1. Deployment overview
Strimzi simplifies the process of running Apache Kafka within a Kubernetes cluster.
This guide provides instructions for deploying and managing Strimzi. Deployment options and steps are covered using the example installation files included with Strimzi. While the guide highlights important configuration considerations, it does not cover all available options. For a deeper understanding of the Kafka component configuration options, refer to the Strimzi Custom Resource API Reference.
In addition to deployment instructions, the guide offers pre- and post-deployment guidance. It covers setting up and securing client access to your Kafka cluster. Furthermore, it explores additional deployment options such as metrics integration, distributed tracing, and cluster management tools like Cruise Control and the Strimzi Drain Cleaner. You’ll also find recommendations on managing Strimzi and fine-tuning Kafka configuration for optimal performance.
Upgrade instructions are provided for both Strimzi and Kafka, to help keep your deployment up to date.
Strimzi is designed to be compatible with all types of Kubernetes clusters, irrespective of their distribution. Whether your deployment involves public or private clouds, or if you are setting up a local development environment, the instructions in this guide are applicable in all cases.
1.1. Strimzi custom resources
The deployment of Kafka components onto a Kubernetes cluster using Strimzi is highly configurable through the use of custom resources. These resources are created as instances of APIs introduced by Custom Resource Definitions (CRDs), which extend Kubernetes resources.
CRDs act as configuration instructions to describe the custom resources in a Kubernetes cluster, and are provided with Strimzi for each Kafka component used in a deployment, as well as users and topics. CRDs and custom resources are defined as YAML files. Example YAML files are provided with the Strimzi distribution.
CRDs also allow Strimzi resources to benefit from native Kubernetes features like CLI accessibility and configuration validation.
1.1.1. Strimzi custom resource example
CRDs require a one-time installation in a cluster to define the schemas used to instantiate and manage Strimzi-specific resources.
After a new custom resource type is added to your cluster by installing a CRD, you can create instances of the resource based on its specification.
Depending on the cluster setup, installation typically requires cluster admin privileges.
Note
|
Access to manage custom resources is limited to Strimzi administrators. For more information, see Designating Strimzi administrators. |
A CRD defines a new kind
of resource, such as kind:Kafka
, within a Kubernetes cluster.
The Kubernetes API server allows custom resources to be created based on the kind
and understands from the CRD how to validate and store the custom resource when it is added to the Kubernetes cluster.
Each Strimzi-specific custom resource conforms to the schema defined by the CRD for the resource’s kind
.
The custom resources for Strimzi components have common configuration properties, which are defined under spec
.
To understand the relationship between a CRD and a custom resource, let’s look at a sample of the CRD for a Kafka topic.
apiVersion: kafka.strimzi.io/v1beta2
kind: CustomResourceDefinition
metadata: (1)
name: kafkatopics.kafka.strimzi.io
labels:
app: strimzi
spec: (2)
group: kafka.strimzi.io
versions:
v1beta2
scope: Namespaced
names:
# ...
singular: kafkatopic
plural: kafkatopics
shortNames:
- kt (3)
additionalPrinterColumns: (4)
# ...
subresources:
status: {} (5)
validation: (6)
openAPIV3Schema:
properties:
spec:
type: object
properties:
partitions:
type: integer
minimum: 1
replicas:
type: integer
minimum: 1
maximum: 32767
# ...
-
The metadata for the topic CRD, its name and a label to identify the CRD.
-
The specification for this CRD, including the group (domain) name, the plural name and the supported schema version, which are used in the URL to access the API of the topic. The other names are used to identify instance resources in the CLI. For example,
kubectl get kafkatopic my-topic
orkubectl get kafkatopics
. -
The shortname can be used in CLI commands. For example,
kubectl get kt
can be used as an abbreviation instead ofkubectl get kafkatopic
. -
The information presented when using a
get
command on the custom resource. -
The current status of the CRD as described in the schema reference for the resource.
-
openAPIV3Schema validation provides validation for the creation of topic custom resources. For example, a topic requires at least one partition and one replica.
Note
|
You can identify the CRD YAML files supplied with the Strimzi installation files, because the file names contain an index number followed by ‘Crd’. |
Here is a corresponding example of a KafkaTopic
custom resource.
apiVersion: kafka.strimzi.io/v1beta2
kind: KafkaTopic (1)
metadata:
name: my-topic
labels:
strimzi.io/cluster: my-cluster (2)
spec: (3)
partitions: 1
replicas: 1
config:
retention.ms: 7200000
segment.bytes: 1073741824
status:
conditions: (4)
lastTransitionTime: "2019-08-20T11:37:00.706Z"
status: "True"
type: Ready
observedGeneration: 1
/ ...
-
The
kind
andapiVersion
identify the CRD of which the custom resource is an instance. -
A label, applicable only to
KafkaTopic
andKafkaUser
resources, that defines the name of the Kafka cluster (which is same as the name of theKafka
resource) to which a topic or user belongs. -
The spec shows the number of partitions and replicas for the topic as well as the configuration parameters for the topic itself. In this example, the retention period for a message to remain in the topic and the segment file size for the log are specified.
-
Status conditions for the
KafkaTopic
resource. Thetype
condition changed toReady
at thelastTransitionTime
.
Custom resources can be applied to a cluster through the platform CLI. When the custom resource is created, it uses the same validation as the built-in resources of the Kubernetes API.
After a KafkaTopic
custom resource is created, the Topic Operator is notified and corresponding Kafka topics are created in Strimzi.
1.1.2. Performing kubectl
operations on custom resources
You can use kubectl
commands to retrieve information and perform other operations on Strimzi custom resources.
Use kubectl
commands, such as get
, describe
, edit
, or delete
, to perform operations on resource types.
For example, kubectl get kafkatopics
retrieves a list of all Kafka topics and kubectl get kafkas
retrieves all deployed Kafka clusters.
When referencing resource types, you can use both singular and plural names:
kubectl get kafkas
gets the same results as kubectl get kafka
.
You can also use the short name of the resource.
Learning short names can save you time when managing Strimzi.
The short name for Kafka
is k
, so you can also run kubectl get k
to list all Kafka clusters.
kubectl get k
NAME DESIRED KAFKA REPLICAS DESIRED ZK REPLICAS
my-cluster 3 3
Strimzi resource | Long name | Short name |
---|---|---|
Kafka |
kafka |
k |
Kafka Node Pool |
kafkanodepool |
knp |
Kafka Topic |
kafkatopic |
kt |
Kafka User |
kafkauser |
ku |
Kafka Connect |
kafkaconnect |
kc |
Kafka Connector |
kafkaconnector |
kctr |
Kafka MirrorMaker |
kafkamirrormaker |
kmm |
Kafka MirrorMaker 2 |
kafkamirrormaker2 |
kmm2 |
Kafka Bridge |
kafkabridge |
kb |
Kafka Rebalance |
kafkarebalance |
kr |
Resource categories
Categories of custom resources can also be used in kubectl
commands.
All Strimzi custom resources belong to the category strimzi
, so you can use strimzi
to get all the Strimzi resources with one command.
For example, running kubectl get strimzi
lists all Strimzi custom resources in a given namespace.
kubectl get strimzi
NAME DESIRED KAFKA REPLICAS DESIRED ZK REPLICAS
kafka.kafka.strimzi.io/my-cluster 3 3
NAME PARTITIONS REPLICATION FACTOR
kafkatopic.kafka.strimzi.io/kafka-apps 3 3
NAME AUTHENTICATION AUTHORIZATION
kafkauser.kafka.strimzi.io/my-user tls simple
The kubectl get strimzi -o name
command returns all resource types and resource names.
The -o name
option fetches the output in the type/name format
kubectl get strimzi -o name
kafka.kafka.strimzi.io/my-cluster
kafkatopic.kafka.strimzi.io/kafka-apps
kafkauser.kafka.strimzi.io/my-user
You can combine this strimzi
command with other commands.
For example, you can pass it into a kubectl delete
command to delete all resources in a single command.
kubectl delete $(kubectl get strimzi -o name)
kafka.kafka.strimzi.io "my-cluster" deleted
kafkatopic.kafka.strimzi.io "kafka-apps" deleted
kafkauser.kafka.strimzi.io "my-user" deleted
Deleting all resources in a single operation might be useful, for example, when you are testing new Strimzi features.
Querying the status of sub-resources
There are other values you can pass to the -o
option.
For example, by using -o yaml
you get the output in YAML format.
Using -o json
will return it as JSON.
You can see all the options in kubectl get --help
.
One of the most useful options is the JSONPath support, which allows you to pass JSONPath expressions to query the Kubernetes API. A JSONPath expression can extract or navigate specific parts of any resource.
For example, you can use the JSONPath expression {.status.listeners[?(@.name=="tls")].bootstrapServers}
to get the bootstrap address from the status of the Kafka custom resource and use it in your Kafka clients.
Here, the command retrieves the bootstrapServers
value of the listener named tls
:
kubectl get kafka my-cluster -o=jsonpath='{.status.listeners[?(@.name=="tls")].bootstrapServers}{"\n"}'
my-cluster-kafka-bootstrap.myproject.svc:9093
By changing the name condition you can also get the address of the other Kafka listeners.
You can use jsonpath
to extract any other property or group of properties from any custom resource.
1.1.3. Strimzi custom resource status information
Status properties provide status information for certain custom resources.
The following table lists the custom resources that provide status information (when deployed) and the schemas that define the status properties.
For more information on the schemas, see the Strimzi Custom Resource API Reference.
Strimzi resource | Schema reference | Publishes status information on… |
---|---|---|
|
|
The Kafka cluster, its listeners, node pools, and any auto-rebalances on scaling |
|
|
The nodes in the node pool, their roles, and the associated Kafka cluster |
|
|
Kafka topics in the Kafka cluster |
|
|
Kafka users in the Kafka cluster |
|
|
The Kafka Connect cluster and connector plugins |
|
|
KafkaConnector resources |
|
|
The Kafka MirrorMaker 2 cluster and internal connectors |
|
|
The Kafka MirrorMaker cluster |
|
|
The Kafka Bridge |
|
|
The status and results of a rebalance |
|
|
The number of pods: being managed, using the current version, and in a ready state |
The status
property of a resource provides information on the state of the resource.
The status.conditions
and status.observedGeneration
properties are common to all resources.
status.conditions
-
Status conditions describe the current state of a resource. Status condition properties are useful for tracking progress related to the resource achieving its desired state, as defined by the configuration specified in its
spec
. Status condition properties provide the time and reason the state of the resource changed, and details of events preventing or delaying the operator from realizing the desired state. status.observedGeneration
-
Last observed generation denotes the latest reconciliation of the resource by the Cluster Operator. If the value of
observedGeneration
is different from the value ofmetadata.generation
(the current version of the deployment), the operator has not yet processed the latest update to the resource. If these values are the same, the status information reflects the most recent changes to the resource.
The status
properties also provide resource-specific information.
For example, KafkaStatus
provides information on listener addresses, and the ID of the Kafka cluster.
KafkaStatus
also provides information on the Kafka and Strimzi versions being used.
You can check the values of operatorLastSuccessfulVersion
and kafkaVersion
to determine whether an upgrade of Strimzi or Kafka has completed
Strimzi creates and maintains the status of custom resources, periodically evaluating the current state of the custom resource and updating its status accordingly.
When performing an update on a custom resource using kubectl edit
, for example, its status
is not editable. Moreover, changing the status
would not affect the configuration of the Kafka cluster.
Here we see the status
properties for a Kafka
custom resource.
apiVersion: kafka.strimzi.io/v1beta2
kind: Kafka
metadata:
spec:
# ...
status:
clusterId: XP9FP2P-RByvEy0W4cOEUA # (1)
conditions: # (2)
- lastTransitionTime: '2023-01-20T17:56:29.396588Z'
status: 'True'
type: Ready # (3)
kafkaMetadataState: KRaft # (4)
kafkaVersion: 3.9.0 # (5)
kafkaNodePools: # (6)
- name: broker
- name: controller
listeners: # (7)
- addresses:
- host: my-cluster-kafka-bootstrap.prm-project.svc
port: 9092
bootstrapServers: 'my-cluster-kafka-bootstrap.prm-project.svc:9092'
name: plain
- addresses:
- host: my-cluster-kafka-bootstrap.prm-project.svc
port: 9093
bootstrapServers: 'my-cluster-kafka-bootstrap.prm-project.svc:9093'
certificates:
- |
-----BEGIN CERTIFICATE-----
-----END CERTIFICATE-----
name: tls
- addresses:
- host: >-
2054284155.us-east-2.elb.amazonaws.com
port: 9095
bootstrapServers: >-
2054284155.us-east-2.elb.amazonaws.com:9095
certificates:
- |
-----BEGIN CERTIFICATE-----
-----END CERTIFICATE-----
name: external3
- addresses:
- host: ip-10-0-172-202.us-east-2.compute.internal
port: 31644
bootstrapServers: 'ip-10-0-172-202.us-east-2.compute.internal:31644'
certificates:
- |
-----BEGIN CERTIFICATE-----
-----END CERTIFICATE-----
name: external4
observedGeneration: 3 # (8)
operatorLastSuccessfulVersion: latest # (9)
-
The Kafka cluster ID.
-
Status
conditions
describe the current state of the Kafka cluster. -
The
Ready
condition indicates that the Cluster Operator considers the Kafka cluster able to handle traffic. -
Kafka metadata state that shows the mechanism used (KRaft or ZooKeeper) to manage Kafka metadata and coordinate operations.
-
The version of Kafka being used by the Kafka cluster.
-
The node pools belonging to the Kafka cluster.
-
The
listeners
describe Kafka bootstrap addresses by type. -
The
observedGeneration
value indicates the last reconciliation of theKafka
custom resource by the Cluster Operator. -
The version of the operator that successfully completed the last reconciliation.
Note
|
The Kafka bootstrap addresses listed in the status do not signify that those endpoints or the Kafka cluster is in a Ready state.
|
1.1.4. Finding the status of a custom resource
Use kubectl
with the status
subresource of a custom resource to retrieve information about the resource.
-
A Kubernetes cluster.
-
The Cluster Operator is running.
-
Specify the custom resource and use the
-o jsonpath
option to apply a standard JSONPath expression to select thestatus
property:kubectl get kafka <kafka_resource_name> -o jsonpath='{.status}' | jq
This expression returns all the status information for the specified custom resource. You can use dot notation, such as
status.listeners
orstatus.observedGeneration
, to fine-tune the status information you wish to see.Using the
jq
command line JSON parser tool makes it easier to read the output.
-
For more information about using JSONPath, see JSONPath support.
1.2. Strimzi operators
Strimzi operators are purpose-built with specialist operational knowledge to effectively manage Kafka on Kubernetes. Each operator performs a distinct function.
- Cluster Operator
-
The Cluster Operator handles the deployment and management of Apache Kafka clusters on Kubernetes. It automates the setup of Kafka brokers, and other Kafka components and resources.
- Topic Operator
-
The Topic Operator manages the creation, configuration, and deletion of topics within Kafka clusters.
- User Operator
-
The User Operator manages Kafka users that require access to Kafka brokers.
When you deploy Strimzi, you first deploy the Cluster Operator. The Cluster Operator is then ready to handle the deployment of Kafka. You can also deploy the Topic Operator and User Operator using the Cluster Operator (recommended) or as standalone operators. You would use a standalone operator with a Kafka cluster that is not managed by the Cluster Operator.
The Topic Operator and User Operator are part of the Entity Operator. The Cluster Operator can deploy one or both operators based on the Entity Operator configuration.
Important
|
To deploy the standalone operators, you need to set environment variables to connect to a Kafka cluster. These environment variables do not need to be set if you are deploying the operators using the Cluster Operator as they will be set by the Cluster Operator. |
1.2.1. Watching Strimzi resources in Kubernetes namespaces
Operators watch and manage Strimzi resources in Kubernetes namespaces. The Cluster Operator can watch a single namespace, multiple namespaces, or all namespaces in a Kubernetes cluster. The Topic Operator and User Operator can watch a single namespace.
-
The Cluster Operator watches for
Kafka
resources -
The Topic Operator watches for
KafkaTopic
resources -
The User Operator watches for
KafkaUser
resources
The Topic Operator and the User Operator can only watch a single Kafka cluster in a namespace. And they can only be connected to a single Kafka cluster.
If multiple Topic Operators watch the same namespace, name collisions and topic deletion can occur.
This is because each Kafka cluster uses Kafka topics that have the same name (such as __consumer_offsets
).
Make sure that only one Topic Operator watches a given namespace.
When using multiple User Operators with a single namespace, a user with a given username can exist in more than one Kafka cluster.
If you deploy the Topic Operator and User Operator using the Cluster Operator, they watch the Kafka cluster deployed by the Cluster Operator by default.
You can also specify a namespace using watchedNamespace
in the operator configuration.
For a standalone deployment of each operator, you specify a namespace and connection to the Kafka cluster to watch in the configuration.
1.2.2. Managing RBAC resources
The Cluster Operator creates and manages role-based access control (RBAC) resources for Strimzi components that need access to Kubernetes resources.
For the Cluster Operator to function, it needs permission within the Kubernetes cluster to interact with Kafka resources, such as Kafka
and KafkaConnect
, as well as managed resources like ConfigMap
, Pod
, Deployment
, and Service
.
Permission is specified through the following Kubernetes RBAC resources:
-
ServiceAccount
-
Role
andClusterRole
-
RoleBinding
andClusterRoleBinding
Delegating privileges to Strimzi components
The Cluster Operator runs under a service account called strimzi-cluster-operator
, which is assigned cluster roles that give it permission to create the necessary RBAC resources for Strimzi components.
Role bindings associate the cluster roles with the service account.
Kubernetes enforces privilege escalation prevention, meaning the Cluster Operator cannot grant privileges it does not possess, nor can it grant such privileges in a namespace it cannot access. Consequently, the Cluster Operator must have the necessary privileges for all the components it orchestrates.
The Cluster Operator must be able to do the following:
-
Enable the Topic Operator to manage
KafkaTopic
resources by creatingRole
andRoleBinding
resources in the relevant namespace. -
Enable the User Operator to manage
KafkaUser
resources by creatingRole
andRoleBinding
resources in the relevant namespace. -
Allow Strimzi to discover the failure domain of a
Node
by creating aClusterRoleBinding
.
When using rack-aware partition assignment, broker pods need to access information about the Node
they are running on, such as the Availability Zone in Amazon AWS.
Similarly, when using NodePort
type listeners, broker pods need to advertise the address of the Node
they are running on.
Since a Node
is a cluster-scoped resource, this access must be granted through a ClusterRoleBinding
, not a namespace-scoped RoleBinding
.
The following sections describe the RBAC resources required by the Cluster Operator.
ClusterRole
resources
The Cluster Operator uses ClusterRole
resources to provide the necessary access to resources.
Depending on the Kubernetes cluster setup, a cluster administrator might be needed to create the cluster roles.
Note
|
Cluster administrator rights are only needed for the creation of ClusterRole resources.
The Cluster Operator will not run under a cluster admin account.
|
The RBAC resources follow the principle of least privilege and contain only those privileges needed by the Cluster Operator to operate the cluster of the Kafka component.
All cluster roles are required by the Cluster Operator in order to delegate privileges.
Name | Description |
---|---|
|
Access rights for namespace-scoped resources used by the Cluster Operator to deploy and manage the operands. |
|
Access rights for cluster-scoped resources used by the Cluster Operator to deploy and manage the operands. |
|
Access rights used by the Cluster Operator for leader election. |
|
Access rights used by the Cluster Operator to watch and manage the Strimzi custom resources. |
|
Access rights to allow Kafka brokers to get the topology labels from Kubernetes worker nodes when rack-awareness is used. |
|
Access rights used by the Topic and User Operators to manage Kafka users and topics. |
|
Access rights to allow Kafka Connect, MirrorMaker (1 and 2), and Kafka Bridge to get the topology labels from Kubernetes worker nodes when rack-awareness is used. |
ClusterRoleBinding
resources
The Cluster Operator uses ClusterRoleBinding
and RoleBinding
resources to associate its ClusterRole
with its ServiceAccount
.
Cluster role bindings are required by cluster roles containing cluster-scoped resources.
Name | Description |
---|---|
|
Grants the Cluster Operator the rights from the |
|
Grants the Cluster Operator the rights from the |
|
Grants the Cluster Operator the rights from the |
Name | Description |
---|---|
|
Grants the Cluster Operator the rights from the |
|
Grants the Cluster Operator the rights from the |
|
Grants the Cluster Operator the rights from the |
|
Grants the Cluster Operator the rights from the |
ServiceAccount
resources
The Cluster Operator runs using the strimzi-cluster-operator
ServiceAccount
.
This service account grants it the privileges it requires to manage the operands.
The Cluster Operator creates additional ClusterRoleBinding
and RoleBinding
resources to delegate some of these RBAC rights to the operands.
Each of the operands uses its own service account created by the Cluster Operator. This allows the Cluster Operator to follow the principle of least privilege and give the operands only the access rights that are really need.
Name | Used by |
---|---|
|
ZooKeeper pods |
|
Kafka broker pods |
|
Entity Operator |
|
Cruise Control pods |
|
Kafka Exporter pods |
|
Kafka Connect pods |
|
MirrorMaker pods |
|
MirrorMaker 2 pods |
|
Kafka Bridge pods |
1.2.3. Managing pod resources
The StrimziPodSet
custom resource is used by Strimzi to create and manage Kafka, Kafka Connect, and MirrorMaker 2 pods.
If you are using ZooKeeper, ZooKeeper pods are also created and managed using StrimziPodSet
resources.
You must not create, update, or delete StrimziPodSet
resources.
The StrimziPodSet
custom resource is used internally and resources are managed solely by the Cluster Operator.
As a consequence, the Cluster Operator must be running properly to avoid the possibility of pods not starting and Kafka clusters not being available.
Note
|
Kubernetes Deployment resources are used for creating and managing the pods of other components: Kafka Bridge, Kafka Exporter, Cruise Control, (deprecated) MirrorMaker 1, User Operator and Topic Operator.
|
1.3. Using the Kafka Bridge to connect with a Kafka cluster
You can use the Kafka Bridge API to create and manage consumers and send and receive records over HTTP rather than the native Kafka protocol.
When you set up the Kafka Bridge you configure HTTP access to the Kafka cluster. You can then use the Kafka Bridge to produce and consume messages from the cluster, as well as performing other operations through its REST interface.
-
For information on installing and using the Kafka Bridge, see Using the Kafka Bridge.
1.4. Seamless FIPS support
Federal Information Processing Standards (FIPS) are standards for computer security and interoperability. When running Strimzi on a FIPS-enabled Kubernetes cluster, the OpenJDK used in Strimzi container images automatically switches to FIPS mode. From version 0.33, Strimzi can run on FIPS-enabled Kubernetes clusters without any changes or special configuration. It uses only the FIPS-compliant security libraries from the OpenJDK.
Important
|
If you are using FIPS-enabled Kubernetes clusters, you may experience higher memory consumption compared to regular Kubernetes clusters. To avoid any issues, we suggest increasing the memory request to at least 512Mi. |
1.4.1. NIST validation
Strimzi is designed to use FIPS-validated cryptographic libraries for secure communication in a FIPS-enabled Kubernetes cluster. However, it’s important to note that while Strimzi can leverage these libraries in a FIPS environment, the underlying Universal Base Images (UBI) used in Strimzi deployments may not inherently include NIST-validated binaries. This means that while Strimzi can leverage cryptographic libraries for FIPS, the specific binaries within the Strimzi container images might not have undergone NIST validation.
For more information about the NIST validation program and validated modules, see Cryptographic Module Validation Program on the NIST website.
1.4.2. Minimum password length
When running in the FIPS mode, SCRAM-SHA-512 passwords need to be at least 32 characters long. From Strimzi 0.33, the default password length in Strimzi User Operator is set to 32 characters as well. If you have a Kafka cluster with custom configuration that uses a password length that is less than 32 characters, you need to update your configuration. If you have any users with passwords shorter than 32 characters, you need to regenerate a password with the required length. You can do that, for example, by deleting the user secret and waiting for the User Operator to create a new password with the appropriate length.
1.5. Document conventions
User-replaced values, also known as replaceables, are shown in with angle brackets (< >).
Underscores ( _ ) are used for multi-word values.
If the value refers to code or commands, monospace
is also used.
For example, the following code shows that <my_namespace>
must be replaced by the correct namespace name:
sed -i 's/namespace: .*/namespace: <my_namespace>/' install/cluster-operator/*RoleBinding*.yaml
1.6. Additional resources
2. Using Kafka in KRaft mode
KRaft (Kafka Raft metadata) mode replaces Kafka’s dependency on ZooKeeper for cluster management. KRaft mode simplifies the deployment and management of Kafka clusters by bringing metadata management and coordination of clusters into Kafka.
Kafka in KRaft mode is designed to offer enhanced reliability, scalability, and throughput. Metadata operations become more efficient as they are directly integrated. And by removing the need to maintain a ZooKeeper cluster, there’s also a reduction in the operational and security overhead.
To deploy a Kafka cluster in KRaft mode, you must use Kafka
and KafkaNodePool
custom resources.
The Kafka
resource using KRaft mode must also have the annotations strimzi.io/kraft: enabled
and strimzi.io/node-pools: enabled
.
For more details and examples, see Deploying a Kafka cluster in KRaft mode.
Through node pool configuration using KafkaNodePool
resources, nodes are assigned the role of broker, controller, or both:
-
Controller nodes operate in the control plane to manage cluster metadata and the state of the cluster using a Raft-based consensus protocol.
-
Broker nodes operate in the data plane to manage the streaming of messages, receiving and storing data in topic partitions.
-
Dual-role nodes fulfill the responsibilities of controllers and brokers.
Controllers use a metadata log, stored as a single-partition topic (__cluster_metadata
) on every node, which records the state of the cluster.
When requests are made to change the cluster configuration, an active (lead) controller manages updates to the metadata log, and follower controllers replicate these updates.
The metadata log stores information on brokers, replicas, topics, and partitions, including the state of in-sync replicas and partition leadership.
Kafka uses this metadata to coordinate changes and manage the cluster effectively.
Broker nodes act as observers, storing the metadata log passively to stay up-to-date with the cluster’s state. Each node fetches updates to the log independently. If you are using JBOD storage, you can change the volume that stores the metadata log.
Note
|
The KRaft metadata version used in the Kafka cluster must be supported by the Kafka version in use.
Both versions are managed through the Kafka resource configuration.
For more information, see Configuring Kafka in KRaft mode.
|
In the following example, a Kafka cluster comprises a quorum of controller and broker nodes for fault tolerance and high availability.
In a typical production environment, use dedicated broker and controller nodes. However, you might want to use nodes in a dual-role configuration for development or testing.
You can use a combination of nodes that combine roles with nodes that perform a single role. In the following example, three nodes perform a dual role and two nodes act only as brokers.
2.1. KRaft limitations
KRaft limitations primarily relate to controller scaling, which impacts cluster operations.
2.1.1. Controller scaling
KRaft mode supports two types of controller quorums:
-
Static controller quorums
In this mode, the number of controllers is fixed, and scaling requires downtime. -
Dynamic controller quorums
This mode enables dynamic scaling of controllers without downtime. New controllers join as observers, replicate the metadata log, and eventually become eligible to join the quorum. If a controller being removed is the active controller, it will step down from the quorum only after the new quorum is confirmed.
Scaling is useful not only for adding or removing controllers, but supports the following operations:
-
Renaming a node pool, which involves adding a new node pool with the desired name and deleting the old one.
-
Changing non-JBOD storage, which requires creating a new node pool with the updated storage configuration and removing the old one.
Dynamic controller quorums provide the flexibility to make these operations significantly easier to perform.
2.1.2. Limitations with static controller quorums
Migration between static and dynamic controller quorums is not currently supported by Apache Kafka, though it is expected to be introduced in a future release. All pre-existing KRaft-based Apache Kafka clusters that use static controller quorums must continue using them. To ensure compatibility with existing KRaft-based clusters, Strimzi continues to use static controller quorums as well.
This limitation means dynamic scaling of controller quorums cannot be used to support the following:
-
Adding or removing node pools with controller roles
-
Adding the controller role to an existing node pool
-
Removing the controller role from an existing node pool
-
Scaling a node pool with the controller role
-
Renaming a node pool with the controller role
Static controller quorums also limit operations that require scaling. For example, changing the storage type for a node pool with a controller role is not possible because it involves scaling the controller quorum. For non-JBOD storage, creating a new node pool with the desired storage type, adding it to the cluster, and removing the old one would require scaling operations, which are not supported. In some cases, workarounds are possible. For instance, when modifying node pool roles to combine controller and broker functions, you can add broker roles to controller nodes instead of adding controller roles to broker nodes to avoid controller scaling. However, this approach would require reassigning more data, which may temporarily affect cluster performance.
Once migration is possible, Strimzi plans to assess introducing support for dynamic quorums.
2.2. Migrating to KRaft mode
If you are using ZooKeeper for metadata management in your Kafka cluster, you can migrate to using Kafka in KRaft mode.
During the migration, you install a quorum of controller nodes as a node pool, which replaces ZooKeeper for management of your cluster.
You enable KRaft migration in the cluster configuration by applying the strimzi.io/kraft="migration"
annotation.
After the migration is complete, you switch the brokers to using KRaft and the controllers out of migration mode using the strimzi.io/kraft="enabled"
annotation.
Before starting the migration, verify that your environment can support Kafka in KRaft mode, as there are a number of limitations. Note also, the following:
-
Migration is only supported on dedicated controller nodes, not on nodes with dual roles as brokers and controllers.
-
Throughout the migration process, ZooKeeper and controller nodes operate in parallel for a period, requiring sufficient compute resources in the cluster.
-
Once KRaft mode is enabled, rollback to ZooKeeper is not possible. Consider this carefully before proceeding with the migration.
-
You must be using Strimzi 0.40 or newer with Kafka 3.7.0 or newer. If you are using an earlier version of Strimzi or Apache Kafka, upgrade before migrating to KRaft mode.
-
The Cluster Operator that manages the Kafka cluster is running.
-
The Kafka cluster deployment uses Kafka node pools.
If your ZooKeeper-based cluster is already using node pools, it is ready to migrate. If not, you can migrate the cluster to use node pools. To migrate when the cluster is not using node pools, brokers must be contained in a
KafkaNodePool
resource configuration that is assigned abroker
role and has the namekafka
. Support for node pools is enabled in theKafka
resource configuration using thestrimzi.io/node-pools: enabled
annotation.
Important
|
Using a single controller with ephemeral storage for migrating to KRaft will not work. During the migration, controller restart will cause loss of metadata synced from ZooKeeper (such as topics and ACLs). In general, migrating an ephemeral-based ZooKeeper cluster to KRaft is not recommended. |
In this procedure, the Kafka cluster name is my-cluster
, which is located in the my-project
namespace.
The name of the controller node pool created is controller
.
The node pool for the brokers is called kafka
.
-
For the Kafka cluster, create a node pool with a
controller
role.The node pool adds a quorum of controller nodes to the cluster.
Example configuration for a controller node poolapiVersion: kafka.strimzi.io/v1beta2 kind: KafkaNodePool metadata: name: controller labels: strimzi.io/cluster: my-cluster spec: replicas: 3 roles: - controller storage: type: jbod volumes: - id: 0 type: persistent-claim size: 20Gi deleteClaim: false resources: requests: memory: 64Gi cpu: "8" limits: memory: 64Gi cpu: "12"
NoteFor the migration, you cannot use a node pool of nodes that share the broker and controller roles. -
Apply the new
KafkaNodePool
resource to create the controllers.Errors related to using controllers in a ZooKeeper-based environment are expected in the Cluster Operator logs. The errors can block reconciliation. To prevent this, perform the next step immediately.
-
Enable KRaft migration in the
Kafka
resource by setting thestrimzi.io/kraft
annotation tomigration
:kubectl annotate kafka my-cluster strimzi.io/kraft="migration" --overwrite
Enabling KRaft migrationapiVersion: kafka.strimzi.io/v1beta2 kind: Kafka metadata: name: my-cluster namespace: my-project annotations: strimzi.io/kraft: migration # ...
Applying the annotation to the
Kafka
resource configuration starts the migration. -
Check the controllers have started and the brokers have rolled:
kubectl get pods -n my-project
Output shows nodes in broker and controller node poolsNAME READY STATUS RESTARTS my-cluster-kafka-0 1/1 Running 0 my-cluster-kafka-1 1/1 Running 0 my-cluster-kafka-2 1/1 Running 0 my-cluster-controller-3 1/1 Running 0 my-cluster-controller-4 1/1 Running 0 my-cluster-controller-5 1/1 Running 0 # ...
-
Check the status of the migration:
kubectl get kafka my-cluster -n my-project -w
Updates to the metadata stateNAME ... METADATA STATE my-cluster ... Zookeeper my-cluster ... KRaftMigration my-cluster ... KRaftDualWriting my-cluster ... KRaftPostMigration
METADATA STATE
shows the mechanism used to manage Kafka metadata and coordinate operations. At the start of the migration this isZooKeeper
.-
ZooKeeper
is the initial state when metadata is only stored in ZooKeeper. -
KRaftMigration
is the state when the migration is in progress. The flag to enable ZooKeeper to KRaft migration (zookeeper.metadata.migration.enable
) is added to the brokers and they are rolled to register with the controllers. The migration can take some time at this point depending on the number of topics and partitions in the cluster. -
KRaftDualWriting
is the state when the Kafka cluster is working as a KRaft cluster, but metadata are being stored in both Kafka and ZooKeeper. Brokers are rolled a second time to remove the flag to enable migration. -
KRaftPostMigration
is the state when KRaft mode is enabled for brokers. Metadata are still being stored in both Kafka and ZooKeeper.
The migration status is also represented in the
status.kafkaMetadataState
property of theKafka
resource.WarningYou can roll back to using ZooKeeper from this point. The next step is to enable KRaft. Rollback cannot be performed after enabling KRaft. -
-
When the metadata state has reached
KRaftPostMigration
, enable KRaft in theKafka
resource configuration by setting thestrimzi.io/kraft
annotation toenabled
:kubectl annotate kafka my-cluster strimzi.io/kraft="enabled" --overwrite
Enabling KRaft migrationapiVersion: kafka.strimzi.io/v1beta2 kind: Kafka metadata: name: my-cluster namespace: my-project annotations: strimzi.io/kraft: enabled # ...
-
Check the status of the move to full KRaft mode:
kubectl get kafka my-cluster -n my-project -w
Updates to the metadata stateNAME ... METADATA STATE my-cluster ... Zookeeper my-cluster ... KRaftMigration my-cluster ... KRaftDualWriting my-cluster ... KRaftPostMigration my-cluster ... PreKRaft my-cluster ... KRaft
-
PreKRaft
is the state when all ZooKeeper-related resources have been automatically deleted. -
KRaft
is the final state (after the controllers have rolled) when the KRaft migration is finalized.
NoteDepending on how deleteClaim
is configured for ZooKeeper, its Persistent Volume Claims (PVCs) and persistent volumes (PVs) may not be deleted.deleteClaim
specifies whether the PVC is deleted when the cluster is uninstalled. The default isfalse
. -
-
Remove any ZooKeeper-related configuration from the
Kafka
resource.Remove the following section:
-
spec.zookeeper
If present, you can also remove the following options from the
.spec.kafka.config
section:-
log.message.format.version
-
inter.broker.protocol.version
Removing
log.message.format.version
andinter.broker.protocol.version
causes the brokers and controllers to roll again. Removing ZooKeeper properties removes any warning messages related to ZooKeeper configuration being present in a KRaft-operated cluster. -
2.2.1. Performing a rollback on the migration
Before the migration is finalized by enabling KRaft in the Kafka
resource, and the state has moved to the KRaft
state, you can perform a rollback operation as follows:
-
Apply the
strimzi.io/kraft="rollback"
annotation to theKafka
resource to roll back the brokers.kubectl annotate kafka my-cluster strimzi.io/kraft="rollback" --overwrite
Rolling back KRaft migrationapiVersion: kafka.strimzi.io/v1beta2 kind: Kafka metadata: name: my-cluster namespace: my-project annotations: strimzi.io/kraft: rollback # ...
The migration process must be in the
KRaftPostMigration
state to do this. The brokers are rolled back so that they can be connected to ZooKeeper again and the state returns toKRaftDualWriting
. -
Delete the controllers node pool:
kubectl delete KafkaNodePool controller -n my-project
-
Apply the
strimzi.io/kraft="disabled"
annotation to theKafka
resource to return the metadata state toZooKeeper
.kubectl annotate kafka my-cluster strimzi.io/kraft="disabled" --overwrite
Switching back to using ZooKeeperapiVersion: kafka.strimzi.io/v1beta2 kind: Kafka metadata: name: my-cluster namespace: my-project annotations: strimzi.io/kraft: disabled # ...
3. Deployment methods
You can deploy Strimzi on Kubernetes 1.25 and later using one of the following methods:
Installation method | Description |
---|---|
Download the deployment files to manually deploy Strimzi components. For the greatest flexibility, choose this method. |
|
Deploy the Strimzi Cluster operator through the OperatorHub.io, then deploy Strimzi components using custom resources. This method provides a standard configuration and allows you to take advantage of automatic updates. |
|
Use a Helm chart to deploy the Cluster Operator, then deploy Strimzi components using custom resources. Helm charts provide a convenient way to manage the installation of applications. |
4. Deployment path
You can configure a deployment where Strimzi manages a single Kafka cluster in the same namespace, suitable for development or testing. Alternatively, Strimzi can manage multiple Kafka clusters across different namespaces in a production environment.
The basic deployment path includes the following steps:
-
Create a Kubernetes namespace for the Cluster Operator.
-
Deploy the Cluster Operator based on your chosen deployment method.
-
Deploy the Kafka cluster, including the Topic Operator and User Operator if desired.
-
Optionally, deploy additional components:
-
The Topic Operator and User Operator as standalone components, if not deployed with the Kafka cluster
-
Kafka Connect
-
Kafka MirrorMaker
-
Kafka Bridge
-
Metrics monitoring components
-
The Cluster Operator creates Kubernetes resources such as Deployment
, Service
, and Pod
for each component.
The resource names are appended with the name of the deployed component.
For example, a Kafka cluster named my-kafka-cluster
will have a service named my-kafka-cluster-kafka
.
5. Downloading deployment files
To deploy Strimzi components using YAML files, download and extract the latest release archive (strimzi-latest.*
) from the GitHub releases page.
The release archive contains sample YAML files for deploying Strimzi components to Kubernetes using kubectl
.
Begin by deploying the Cluster Operator from the install/cluster-operator
directory to watch a single namespace, multiple namespaces, or all namespaces.
In the install
folder, you can also deploy other Strimzi components, including:
-
Strimzi administrator roles (
strimzi-admin
) -
Standalone Topic Operator (
topic-operator
) -
Standalone User Operator (
user-operator
) -
Strimzi Drain Cleaner (
drain-cleaner
)
The examples
folder provides examples of Strimzi custom resources to help you develop your own Kafka configurations.
Note
|
Strimzi container images are available through the Container Registry, but we recommend using the provided YAML files for deployment. |
6. Preparing for your deployment
Prepare for a deployment of Strimzi by completing any necessary pre-deployment tasks. Take the necessary preparatory steps according to your specific requirements, such as the following:
Note
|
To run the commands in this guide, your cluster user must have the rights to manage role-based access control (RBAC) and CRDs. |
6.1. Deployment prerequisites
To deploy Strimzi, you will need the following:
-
A Kubernetes 1.25 and later cluster.
-
The
kubectl
command-line tool is installed and configured to connect to the running cluster.
For more information on the tools available for running Kubernetes, see Install Tools in the Kubernetes documentation.
Note
|
Strimzi supports some features that are specific to OpenShift, where such integration benefits OpenShift users and there is no equivalent implementation using standard Kubernetes. |
oc
and kubectl
commands
The oc
command functions as an alternative to kubectl
.
In almost all cases the example kubectl
commands used in this guide can be done using oc
simply by replacing the command name (options and arguments remain the same).
In other words, instead of using:
kubectl apply -f <your_file>
when using OpenShift you can use:
oc apply -f <your_file>
Note
|
As an exception to this general rule, oc uses oc adm subcommands for cluster management functionality,
whereas kubectl does not make this distinction.
For example, the oc equivalent of kubectl taint is oc adm taint .
|
6.2. Operator deployment best practices
Potential issues can arise from installing more than one Strimzi operator in the same Kubernetes cluster, especially when using different versions. Each Strimzi operator manages a set of resources in a Kubernetes cluster. When you install multiple Strimzi operators, they may attempt to manage the same resources concurrently. This can lead to conflicts and unpredictable behavior within your cluster. Conflicts can still occur even if you deploy Strimzi operators in different namespaces within the same Kubernetes cluster. Although namespaces provide some degree of resource isolation, certain resources managed by the Strimzi operator, such as Custom Resource Definitions (CRDs) and roles, have a cluster-wide scope.
Additionally, installing multiple operators with different versions can result in compatibility issues between the operators and the Kafka clusters they manage. Different versions of Strimzi operators may introduce changes, bug fixes, or improvements that are not backward-compatible.
To avoid the issues associated with installing multiple Strimzi operators in a Kubernetes cluster, the following guidelines are recommended:
-
Install the Strimzi operator in a separate namespace from the Kafka cluster and other Kafka components it manages, to ensure clear separation of resources and configurations.
-
Use a single Strimzi operator to manage all your Kafka instances within a Kubernetes cluster.
-
Update the Strimzi operator and the supported Kafka version as often as possible to reflect the latest features and enhancements.
By following these best practices and ensuring consistent updates for a single Strimzi operator, you can enhance the stability of managing Kafka instances in a Kubernetes cluster. This approach also enables you to make the most of Strimzi’s latest features and capabilities.
6.3. Pushing container images to your own registry
Container images for Strimzi are available in the Container Registry. The installation YAML files provided by Strimzi will pull the images directly from the Container Registry.
If you do not have access to the Container Registry or want to use your own container repository:
-
Pull all container images listed here
-
Push them into your own registry
-
Update the image names in the YAML files used in deployment
Note
|
Each Kafka version supported for the release has a separate image. |
Container image | Namespace/Repository | Description |
---|---|---|
Kafka |
|
Strimzi image for running Kafka, including:
|
Operator |
|
Strimzi image for running the operators:
|
Kafka Bridge |
|
Strimzi image for running the Kafka Bridge |
Strimzi Drain Cleaner |
|
Strimzi image for running the Strimzi Drain Cleaner |
6.4. Designating Strimzi administrators
Strimzi provides custom resources for configuration of your deployment. By default, permission to view, create, edit, and delete these resources is limited to Kubernetes cluster administrators. Strimzi provides two cluster roles that you can use to assign these rights to other users:
-
strimzi-view
allows users to view and list Strimzi resources. -
strimzi-admin
allows users to also create, edit or delete Strimzi resources.
When you install these roles, they will automatically aggregate (add) these rights to the default Kubernetes cluster roles.
strimzi-view
aggregates to the view
role, and strimzi-admin
aggregates to the edit
and admin
roles.
Because of the aggregation, you might not need to assign these roles to users who already have similar rights.
The following procedure shows how to assign a strimzi-admin
role that allows non-cluster administrators to manage Strimzi resources.
A system administrator can designate Strimzi administrators after the Cluster Operator is deployed.
-
The Strimzi admin deployment files, which are included in the Strimzi deployment files.
-
The Strimzi Custom Resource Definitions (CRDs) and role-based access control (RBAC) resources to manage the CRDs have been deployed with the Cluster Operator.
-
Create the
strimzi-view
andstrimzi-admin
cluster roles in Kubernetes.kubectl create -f install/strimzi-admin
-
If needed, assign the roles that provide access rights to users that require them.
kubectl create clusterrolebinding strimzi-admin --clusterrole=strimzi-admin --user=user1 --user=user2
7. Deploying Strimzi using installation files
Download and use the Strimzi deployment files to deploy Strimzi components to a Kubernetes cluster.
You can deploy Strimzi latest on Kubernetes 1.25 and later.
The steps to deploy Strimzi using the installation files are as follows:
-
Use the Cluster Operator to deploy the following:
-
Optionally, deploy the following Kafka components according to your requirements:
Note
|
To run the commands in this guide, a Kubernetes user must have the rights to manage role-based access control (RBAC) and CRDs. |
7.1. Deploying the Cluster Operator
The first step for any deployment of Strimzi is to install the Cluster Operator, which is responsible for deploying and managing Kafka clusters within a Kubernetes cluster.
A single command applies all the installation files in the install/cluster-operator
folder: kubectl apply -f ./install/cluster-operator
.
The command sets up everything you need to be able to create and manage a Kafka deployment, including the following resources:
-
Cluster Operator (
Deployment
,ConfigMap
) -
Strimzi CRDs (
CustomResourceDefinition
) -
RBAC resources (
ClusterRole
,ClusterRoleBinding
,RoleBinding
) -
Service account (
ServiceAccount
)
Cluster-scoped resources like CustomResourceDefinition
, ClusterRole
, and ClusterRoleBinding
require administrator privileges for installation.
Prior to installation, it’s advisable to review the ClusterRole
specifications to ensure they do not grant unnecessary privileges.
After installation, the Cluster Operator runs as a regular Deployment
to watch for updates of Kafka resources.
Any standard (non-admin) Kubernetes user with privileges to access the Deployment
can configure it.
A cluster administrator can also grant standard users the privileges necessary to manage Strimzi custom resources.
By default, a single replica of the Cluster Operator is deployed. You can add replicas with leader election so that additional Cluster Operators are on standby in case of disruption. For more information, see Running multiple Cluster Operator replicas with leader election.
7.1.1. Specifying the namespaces the Cluster Operator watches
The Cluster Operator watches for updates in the namespaces where the Kafka resources are deployed. When you deploy the Cluster Operator, you specify which namespaces to watch in the Kubernetes cluster. You can specify the following namespaces:
-
A single selected namespace (the same namespace containing the Cluster Operator)
Watching multiple selected namespaces has the most impact on performance due to increased processing overhead. To optimize performance for namespace monitoring, it is generally recommended to either watch a single namespace or monitor the entire cluster. Watching a single namespace allows for focused monitoring of namespace-specific resources, while monitoring all namespaces provides a comprehensive view of the cluster’s resources across all namespaces.
The Cluster Operator watches for changes to the following resources:
-
Kafka
for the Kafka cluster. -
KafkaConnect
for the Kafka Connect cluster. -
KafkaConnector
for creating and managing connectors in a Kafka Connect cluster. -
KafkaMirrorMaker
for the Kafka MirrorMaker instance. -
KafkaMirrorMaker2
for the Kafka MirrorMaker 2 instance. -
KafkaBridge
for the Kafka Bridge instance. -
KafkaRebalance
for the Cruise Control optimization requests.
When one of these resources is created in the Kubernetes cluster, the operator gets the cluster description from the resource and starts creating a new cluster for the resource by creating the necessary Kubernetes resources, such as Deployments, Pods, Services and ConfigMaps.
Each time a Kafka resource is updated, the operator performs corresponding updates on the Kubernetes resources that make up the cluster for the resource.
Resources are either patched or deleted, and then recreated in order to make the cluster for the resource reflect the desired state of the cluster. This operation might cause a rolling update that might lead to service disruption.
When a resource is deleted, the operator undeploys the cluster and deletes all related Kubernetes resources.
Note
|
While the Cluster Operator can watch one, multiple, or all namespaces in a Kubernetes cluster,
the Topic Operator and User Operator watch for KafkaTopic and KafkaUser resources in a single namespace.
For more information, see Watching Strimzi resources in Kubernetes namespaces.
|
7.1.2. Deploying the Cluster Operator to watch a single namespace
This procedure shows how to deploy the Cluster Operator to watch Strimzi resources in a single namespace in your Kubernetes cluster.
-
You need an account with permission to create and manage
CustomResourceDefinition
and RBAC (ClusterRole
, andRoleBinding
) resources.
-
Edit the Strimzi installation files to use the namespace the Cluster Operator is going to be installed into.
For example, in this procedure the Cluster Operator is installed into the namespace
my-cluster-operator-namespace
.On Linux, use:
sed -i 's/namespace: .*/namespace: my-cluster-operator-namespace/' install/cluster-operator/*RoleBinding*.yaml
On MacOS, use:
sed -i '' 's/namespace: .*/namespace: my-cluster-operator-namespace/' install/cluster-operator/*RoleBinding*.yaml
-
Deploy the Cluster Operator:
kubectl create -f install/cluster-operator -n my-cluster-operator-namespace
-
Check the status of the deployment:
kubectl get deployments -n my-cluster-operator-namespace
Output shows the deployment name and readinessNAME READY UP-TO-DATE AVAILABLE strimzi-cluster-operator 1/1 1 1
READY
shows the number of replicas that are ready/expected. The deployment is successful when theAVAILABLE
output shows1
.
7.1.3. Deploying the Cluster Operator to watch multiple namespaces
This procedure shows how to deploy the Cluster Operator to watch Strimzi resources across multiple namespaces in your Kubernetes cluster.
-
You need an account with permission to create and manage
CustomResourceDefinition
and RBAC (ClusterRole
, andRoleBinding
) resources.
-
Edit the Strimzi installation files to use the namespace the Cluster Operator is going to be installed into.
For example, in this procedure the Cluster Operator is installed into the namespace
my-cluster-operator-namespace
.On Linux, use:
sed -i 's/namespace: .*/namespace: my-cluster-operator-namespace/' install/cluster-operator/*RoleBinding*.yaml
On MacOS, use:
sed -i '' 's/namespace: .*/namespace: my-cluster-operator-namespace/' install/cluster-operator/*RoleBinding*.yaml
-
Edit the
install/cluster-operator/060-Deployment-strimzi-cluster-operator.yaml
file to add a list of all the namespaces the Cluster Operator will watch to theSTRIMZI_NAMESPACE
environment variable.For example, in this procedure the Cluster Operator will watch the namespaces
watched-namespace-1
,watched-namespace-2
,watched-namespace-3
.apiVersion: apps/v1 kind: Deployment spec: # ... template: spec: serviceAccountName: strimzi-cluster-operator containers: - name: strimzi-cluster-operator image: quay.io/strimzi/operator:latest imagePullPolicy: IfNotPresent env: - name: STRIMZI_NAMESPACE value: watched-namespace-1,watched-namespace-2,watched-namespace-3
-
For each namespace listed, install the
RoleBindings
.In this example, we replace
watched-namespace
in these commands with the namespaces listed in the previous step, repeating them forwatched-namespace-1
,watched-namespace-2
,watched-namespace-3
:kubectl create -f install/cluster-operator/020-RoleBinding-strimzi-cluster-operator.yaml -n <watched_namespace> kubectl create -f install/cluster-operator/023-RoleBinding-strimzi-cluster-operator.yaml -n <watched_namespace> kubectl create -f install/cluster-operator/031-RoleBinding-strimzi-cluster-operator-entity-operator-delegation.yaml -n <watched_namespace>
-
Deploy the Cluster Operator:
kubectl create -f install/cluster-operator -n my-cluster-operator-namespace
-
Check the status of the deployment:
kubectl get deployments -n my-cluster-operator-namespace
Output shows the deployment name and readinessNAME READY UP-TO-DATE AVAILABLE strimzi-cluster-operator 1/1 1 1
READY
shows the number of replicas that are ready/expected. The deployment is successful when theAVAILABLE
output shows1
.
7.1.4. Deploying the Cluster Operator to watch all namespaces
This procedure shows how to deploy the Cluster Operator to watch Strimzi resources across all namespaces in your Kubernetes cluster.
When running in this mode, the Cluster Operator automatically manages clusters in any new namespaces that are created.
-
You need an account with permission to create and manage
CustomResourceDefinition
and RBAC (ClusterRole
, andRoleBinding
) resources.
-
Edit the Strimzi installation files to use the namespace the Cluster Operator is going to be installed into.
For example, in this procedure the Cluster Operator is installed into the namespace
my-cluster-operator-namespace
.On Linux, use:
sed -i 's/namespace: .*/namespace: my-cluster-operator-namespace/' install/cluster-operator/*RoleBinding*.yaml
On MacOS, use:
sed -i '' 's/namespace: .*/namespace: my-cluster-operator-namespace/' install/cluster-operator/*RoleBinding*.yaml
-
Edit the
install/cluster-operator/060-Deployment-strimzi-cluster-operator.yaml
file to set the value of theSTRIMZI_NAMESPACE
environment variable to*
.apiVersion: apps/v1 kind: Deployment spec: # ... template: spec: # ... serviceAccountName: strimzi-cluster-operator containers: - name: strimzi-cluster-operator image: quay.io/strimzi/operator:latest imagePullPolicy: IfNotPresent env: - name: STRIMZI_NAMESPACE value: "*" # ...
-
Create
ClusterRoleBindings
that grant cluster-wide access for all namespaces to the Cluster Operator.kubectl create clusterrolebinding strimzi-cluster-operator-namespaced --clusterrole=strimzi-cluster-operator-namespaced --serviceaccount my-cluster-operator-namespace:strimzi-cluster-operator kubectl create clusterrolebinding strimzi-cluster-operator-watched --clusterrole=strimzi-cluster-operator-watched --serviceaccount my-cluster-operator-namespace:strimzi-cluster-operator kubectl create clusterrolebinding strimzi-cluster-operator-entity-operator-delegation --clusterrole=strimzi-entity-operator --serviceaccount my-cluster-operator-namespace:strimzi-cluster-operator
-
Deploy the Cluster Operator to your Kubernetes cluster.
kubectl create -f install/cluster-operator -n my-cluster-operator-namespace
-
Check the status of the deployment:
kubectl get deployments -n my-cluster-operator-namespace
Output shows the deployment name and readinessNAME READY UP-TO-DATE AVAILABLE strimzi-cluster-operator 1/1 1 1
READY
shows the number of replicas that are ready/expected. The deployment is successful when theAVAILABLE
output shows1
.
7.2. Deploying Kafka
To be able to manage a Kafka cluster with the Cluster Operator, you must deploy it as a Kafka
resource.
Strimzi provides example deployment files to do this.
You can use these files to deploy the Topic Operator and User Operator at the same time.
After you have deployed the Cluster Operator, use a Kafka
resource to deploy the following components:
-
A Kafka cluster that uses KRaft or ZooKeeper:
Node pools are used in the deployment of a Kafka cluster in KRaft (Kafka Raft metadata) mode, and may be used for the deployment of a Kafka cluster with ZooKeeper.
Node pools represent a distinct group of Kafka nodes within the Kafka cluster that share the same configuration.
For each Kafka node in the node pool, any configuration not defined in node pool is inherited from the cluster configuration in the Kafka
resource.
If you haven’t deployed a Kafka cluster as a Kafka
resource, you can’t use the Cluster Operator to manage it.
This applies, for example, to a Kafka cluster running outside of Kubernetes.
However, you can use the Topic Operator and User Operator with a Kafka cluster that is not managed by Strimzi, by deploying them as standalone components.
You can also deploy and use other Kafka components with a Kafka cluster not managed by Strimzi.
7.2.1. Deploying a Kafka cluster in KRaft mode
This procedure shows how to deploy a Kafka cluster in KRaft mode and associated node pools using the Cluster Operator.
The deployment uses a YAML file to provide the specification to create a Kafka
resource and KafkaNodePool
resources.
Strimzi provides the following example deployment files that you can use to create a Kafka cluster that uses node pools:
kafka/kraft/kafka-with-dual-role-nodes.yaml
-
Deploys a Kafka cluster with one pool of nodes that share the broker and controller roles.
kafka/kraft/kafka.yaml
-
Deploys a persistent Kafka cluster with one pool of controller nodes and one pool of broker nodes.
kafka/kraft/kafka-ephemeral.yaml
-
Deploys an ephemeral Kafka cluster with one pool of controller nodes and one pool of broker nodes.
kafka/kraft/kafka-single-node.yaml
-
Deploys a Kafka cluster with a single node.
kafka/kraft/kafka-jbod.yaml
-
Deploys a Kafka cluster with multiple volumes in each broker node.
In this procedure, we use the example deployment file that deploys a Kafka cluster with one pool of nodes that share the broker and controller roles.
The Kafka
resource configuration for each example includes the strimzi.io/node-pools: enabled
annotation, which is required when using node pools.
Kafka
resources using KRaft mode must also have the annotation strimzi.io/kraft: enabled
.
The example YAML files specify the latest supported Kafka version and KRaft metadata version used by the Kafka cluster.
Note
|
You can perform the steps outlined here to deploy a new Kafka cluster with KafkaNodePool resources or migrate your existing Kafka cluster.
|
By default, the example deployment files specify my-cluster
as the Kafka cluster name.
The name cannot be changed after the cluster has been deployed.
To change the cluster name before you deploy the cluster, edit the Kafka.metadata.name
property of the Kafka
resource in the relevant YAML file.
-
Deploy a KRaft-based Kafka cluster.
To deploy a Kafka cluster with a single node pool that uses dual-role nodes:
kubectl apply -f examples/kafka/kraft/kafka-with-dual-role-nodes.yaml
-
Check the status of the deployment:
kubectl get pods -n <my_cluster_operator_namespace>
Output shows the node pool names and readinessNAME READY STATUS RESTARTS my-cluster-entity-operator 3/3 Running 0 my-cluster-pool-a-0 1/1 Running 0 my-cluster-pool-a-1 1/1 Running 0 my-cluster-pool-a-4 1/1 Running 0
-
my-cluster
is the name of the Kafka cluster. -
pool-a
is the name of the node pool.A sequential index number starting with
0
identifies each Kafka pod created. If you are using ZooKeeper, you’ll also see the ZooKeeper pods.READY
shows the number of replicas that are ready/expected. The deployment is successful when theSTATUS
displays asRunning
.Information on the deployment is also shown in the status of the
KafkaNodePool
resource, including a list of IDs for nodes in the pool.NoteNode IDs are assigned sequentially starting at 0 (zero) across all node pools within a cluster. This means that node IDs might not run sequentially within a specific node pool. If there are gaps in the sequence of node IDs across the cluster, the next node to be added is assigned an ID that fills the gap. When scaling down, the node with the highest node ID within a pool is removed.
-
7.2.2. Deploying a ZooKeeper-based Kafka cluster
This procedure shows how to deploy a ZooKeeper-based Kafka cluster to your Kubernetes cluster using the Cluster Operator.
The deployment uses a YAML file to provide the specification to create a Kafka
resource.
Strimzi provides the following example deployment files to create a Kafka cluster that uses ZooKeeper for cluster management:
kafka-persistent.yaml
-
Deploys a persistent cluster with three ZooKeeper and three Kafka nodes.
kafka-jbod.yaml
-
Deploys a persistent cluster with three ZooKeeper and three Kafka nodes (each using multiple persistent volumes).
kafka-persistent-single.yaml
-
Deploys a persistent cluster with a single ZooKeeper node and a single Kafka node.
kafka-ephemeral.yaml
-
Deploys an ephemeral cluster with three ZooKeeper and three Kafka nodes.
kafka-ephemeral-single.yaml
-
Deploys an ephemeral cluster with three ZooKeeper nodes and a single Kafka node.
To deploy a Kafka cluster that uses node pools, the following example YAML file provides the specification to create a Kafka
resource and KafkaNodePool
resources:
kafka/kafka-with-node-pools.yaml
-
Deploys ZooKeeper with 3 nodes, and 2 different pools of Kafka brokers. Each of the pools has 3 brokers. The pools in the example use different storage configuration.
In this procedure, we use the examples for an ephemeral and persistent Kafka cluster deployment.
The example YAML files specify the latest supported Kafka version and inter-broker protocol version.
Note
|
From Kafka 3.0.0, when the inter.broker.protocol.version is set to 3.0 or higher, the log.message.format.version option is ignored and doesn’t need to be set.
|
By default, the example deployment files specify my-cluster
as the Kafka cluster name.
The name cannot be changed after the cluster has been deployed.
To change the cluster name before you deploy the cluster, edit the Kafka.metadata.name
property of the Kafka
resource in the relevant YAML file.
-
Deploy a ZooKeeper-based Kafka cluster.
-
To deploy an ephemeral cluster:
kubectl apply -f examples/kafka/kafka-ephemeral.yaml
-
To deploy a persistent cluster:
kubectl apply -f examples/kafka/kafka-persistent.yaml
-
-
Check the status of the deployment:
kubectl get pods -n <my_cluster_operator_namespace>
Output shows the pod names and readinessNAME READY STATUS RESTARTS my-cluster-entity-operator 3/3 Running 0 my-cluster-kafka-0 1/1 Running 0 my-cluster-kafka-1 1/1 Running 0 my-cluster-kafka-2 1/1 Running 0 my-cluster-zookeeper-0 1/1 Running 0 my-cluster-zookeeper-1 1/1 Running 0 my-cluster-zookeeper-2 1/1 Running 0
my-cluster
is the name of the Kafka cluster.A sequential index number starting with
0
identifies each Kafka and ZooKeeper pod created.With the default deployment, you create an Entity Operator cluster, 3 Kafka pods, and 3 ZooKeeper pods.
READY
shows the number of replicas that are ready/expected. The deployment is successful when theSTATUS
displays asRunning
.
7.2.3. Deploying the Topic Operator using the Cluster Operator
This procedure describes how to deploy the Topic Operator using the Cluster Operator.
You configure the entityOperator
property of the Kafka
resource to include the topicOperator
.
By default, the Topic Operator watches for KafkaTopic
resources in the namespace of the Kafka cluster deployed by the Cluster Operator.
You can also specify a namespace using watchedNamespace
in the Topic Operator spec
.
A single Topic Operator can watch a single namespace.
One namespace should be watched by only one Topic Operator.
If you use Strimzi to deploy multiple Kafka clusters into the same namespace, enable the Topic Operator for only one Kafka cluster or use the watchedNamespace
property to configure the Topic Operators to watch other namespaces.
If you want to use the Topic Operator with a Kafka cluster that is not managed by Strimzi, you must deploy the Topic Operator as a standalone component.
For more information about configuring the entityOperator
and topicOperator
properties,
see Configuring the Entity Operator.
-
Edit the
entityOperator
properties of theKafka
resource to includetopicOperator
:apiVersion: kafka.strimzi.io/v1beta2 kind: Kafka metadata: name: my-cluster spec: #... entityOperator: topicOperator: {} userOperator: {}
-
Configure the Topic Operator
spec
using the properties described in theEntityTopicOperatorSpec
schema reference.Use an empty object (
{}
) if you want all properties to use their default values. -
Create or update the resource:
kubectl apply -f <kafka_configuration_file>
-
Check the status of the deployment:
kubectl get pods -n <my_cluster_operator_namespace>
Output shows the pod name and readinessNAME READY STATUS RESTARTS my-cluster-entity-operator 3/3 Running 0 # ...
my-cluster
is the name of the Kafka cluster.READY
shows the number of replicas that are ready/expected. The deployment is successful when theSTATUS
displays asRunning
.
7.2.4. Deploying the User Operator using the Cluster Operator
This procedure describes how to deploy the User Operator using the Cluster Operator.
You configure the entityOperator
property of the Kafka
resource to include the userOperator
.
By default, the User Operator watches for KafkaUser
resources in the namespace of the Kafka cluster deployment.
You can also specify a namespace using watchedNamespace
in the User Operator spec
.
A single User Operator can watch a single namespace.
One namespace should be watched by only one User Operator.
If you want to use the User Operator with a Kafka cluster that is not managed by Strimzi, you must deploy the User Operator as a standalone component.
For more information about configuring the entityOperator
and userOperator
properties, see Configuring the Entity Operator.
-
Edit the
entityOperator
properties of theKafka
resource to includeuserOperator
:apiVersion: kafka.strimzi.io/v1beta2 kind: Kafka metadata: name: my-cluster spec: #... entityOperator: topicOperator: {} userOperator: {}
-
Configure the User Operator
spec
using the properties described inEntityUserOperatorSpec
schema reference.Use an empty object (
{}
) if you want all properties to use their default values. -
Create or update the resource:
kubectl apply -f <kafka_configuration_file>
-
Check the status of the deployment:
kubectl get pods -n <my_cluster_operator_namespace>
Output shows the pod name and readinessNAME READY STATUS RESTARTS my-cluster-entity-operator 3/3 Running 0 # ...
my-cluster
is the name of the Kafka cluster.READY
shows the number of replicas that are ready/expected. The deployment is successful when theSTATUS
displays asRunning
.
7.2.5. Connecting to ZooKeeper from a terminal
ZooKeeper services are secured with encryption and authentication and are not intended to be used by external applications that are not part of Strimzi.
However, if you want to use CLI tools that require a connection to ZooKeeper, you can use a terminal inside a ZooKeeper pod and connect to localhost:12181
as the ZooKeeper address.
-
A Kubernetes cluster is available.
-
A Kafka cluster is running.
-
The Cluster Operator is running.
-
Open the terminal using the Kubernetes console or run the
exec
command from your CLI.For example:
kubectl exec -ti my-cluster-zookeeper-0 -- bin/zookeeper-shell.sh localhost:12181 ls /
Be sure to use
localhost:12181
.
7.2.6. List of Kafka cluster resources
The following resources are created by the Cluster Operator in the Kubernetes cluster.
<kafka_cluster_name>-cluster-ca
-
Secret with the Cluster CA private key used to encrypt the cluster communication.
<kafka_cluster_name>-cluster-ca-cert
-
Secret with the Cluster CA public key. This key can be used to verify the identity of the Kafka brokers.
<kafka_cluster_name>-clients-ca
-
Secret with the Clients CA private key used to sign user certificates
<kafka_cluster_name>-clients-ca-cert
-
Secret with the Clients CA public key. This key can be used to verify the identity of the Kafka users.
<kafka_cluster_name>-cluster-operator-certs
-
Secret with Cluster operators keys for communication with Kafka and ZooKeeper.
<kafka_cluster_name>-zookeeper
-
Name given to the following ZooKeeper resources:
-
StrimziPodSet for managing the ZooKeeper node pods.
-
Service account used by the ZooKeeper nodes.
-
PodDisruptionBudget configured for the ZooKeeper nodes.
-
<kafka_cluster_name>-zookeeper-<pod_id>
-
Pods created by the StrimziPodSet.
<kafka_cluster_name>-zookeeper-nodes
-
Headless Service needed to have DNS resolve the ZooKeeper pods IP addresses directly.
<kafka_cluster_name>-zookeeper-client
-
Service used by Kafka brokers to connect to ZooKeeper nodes as clients.
<kafka_cluster_name>-zookeeper-config
-
ConfigMap that contains the ZooKeeper ancillary configuration, and is mounted as a volume by the ZooKeeper node pods.
<kafka_cluster_name>-zookeeper-nodes
-
Secret with ZooKeeper node keys.
<kafka_cluster_name>-network-policy-zookeeper
-
Network policy managing access to the ZooKeeper services.
data-<kafka_cluster_name>-zookeeper-<pod_id>
-
Persistent Volume Claim for the volume used for storing data for a specific ZooKeeper node. This resource will be created only if persistent storage is selected for provisioning persistent volumes to store data.
<kafka_cluster_name>-kafka
-
Name given to the following Kafka resources:
-
StrimziPodSet for managing the Kafka broker pods.
-
Service account used by the Kafka pods.
-
PodDisruptionBudget configured for the Kafka brokers.
-
<kafka_cluster_name>-kafka-<pod_id>
-
Name given to the following Kafka resources:
-
Pods created by the StrimziPodSet.
-
ConfigMaps with Kafka broker configuration.
-
<kafka_cluster_name>-kafka-brokers
-
Service needed to have DNS resolve the Kafka broker pods IP addresses directly.
<kafka_cluster_name>-kafka-bootstrap
-
Service can be used as bootstrap servers for Kafka clients connecting from within the Kubernetes cluster.
<kafka_cluster_name>-kafka-external-bootstrap
-
Bootstrap service for clients connecting from outside the Kubernetes cluster. This resource is created only when an external listener is enabled. The old service name will be used for backwards compatibility when the listener name is
external
and port is9094
. <kafka_cluster_name>-kafka-<pod_id>
-
Service used to route traffic from outside the Kubernetes cluster to individual pods. This resource is created only when an external listener is enabled. The old service name will be used for backwards compatibility when the listener name is
external
and port is9094
. <kafka_cluster_name>-kafka-external-bootstrap
-
Bootstrap route for clients connecting from outside the Kubernetes cluster. This resource is created only when an external listener is enabled and set to type
route
. The old route name will be used for backwards compatibility when the listener name isexternal
and port is9094
. <kafka_cluster_name>-kafka-<pod_id>
-
Route for traffic from outside the Kubernetes cluster to individual pods. This resource is created only when an external listener is enabled and set to type
route
. The old route name will be used for backwards compatibility when the listener name isexternal
and port is9094
. <kafka_cluster_name>-kafka-<listener_name>-bootstrap
-
Bootstrap service for clients connecting from outside the Kubernetes cluster. This resource is created only when an external listener is enabled. The new service name will be used for all other external listeners.
<kafka_cluster_name>-kafka-<listener_name>-<pod_id>
-
Service used to route traffic from outside the Kubernetes cluster to individual pods. This resource is created only when an external listener is enabled. The new service name will be used for all other external listeners.
<kafka_cluster_name>-kafka-<listener_name>-bootstrap
-
Bootstrap route for clients connecting from outside the Kubernetes cluster. This resource is created only when an external listener is enabled and set to type
route
. The new route name will be used for all other external listeners. <kafka_cluster_name>-kafka-<listener_name>-<pod_id>
-
Route for traffic from outside the Kubernetes cluster to individual pods. This resource is created only when an external listener is enabled and set to type
route
. The new route name will be used for all other external listeners. <kafka_cluster_name>-kafka-config
-
ConfigMap containing the Kafka ancillary configuration, which is mounted as a volume by the broker pods when the
UseStrimziPodSets
feature gate is disabled. <kafka_cluster_name>-kafka-brokers
-
Secret with Kafka broker keys.
<kafka_cluster_name>-network-policy-kafka
-
Network policy managing access to the Kafka services.
strimzi-namespace-name-<kafka_cluster_name>-kafka-init
-
Cluster role binding used by the Kafka brokers.
<kafka_cluster_name>-jmx
-
Secret with JMX username and password used to secure the Kafka broker port. This resource is created only when JMX is enabled in Kafka.
data-<kafka_cluster_name>-kafka-<pod_id>
-
Persistent Volume Claim for the volume used for storing data for a specific Kafka broker. This resource is created only if persistent storage is selected for provisioning persistent volumes to store data.
data-<id>-<kafka_cluster_name>-kafka-<pod_id>
-
Persistent Volume Claim for the volume
id
used for storing data for a specific Kafka broker. This resource is created only if persistent storage is selected for JBOD volumes when provisioning persistent volumes to store data.
If you are using Kafka node pools, the resources created apply to the nodes managed in the node pools whether they are operating as brokers, controllers, or both.
The naming convention includes the name of the Kafka cluster and the node pool: <kafka_cluster_name>-<pool_name>
.
<kafka_cluster_name>-<pool_name>
-
Name given to the StrimziPodSet for managing the Kafka node pool.
<kafka_cluster_name>-<pool_name>-<pod_id>
-
Name given to the following Kafka node pool resources:
-
Pods created by the StrimziPodSet.
-
ConfigMaps with Kafka node configuration.
-
data-<kafka_cluster_name>-<pool_name>-<pod_id>
-
Persistent Volume Claim for the volume used for storing data for a specific node. This resource is created only if persistent storage is selected for provisioning persistent volumes to store data.
data-<id>-<kafka_cluster_name>-<pool_name>-<pod_id>
-
Persistent Volume Claim for the volume
id
used for storing data for a specific node. This resource is created only if persistent storage is selected for JBOD volumes when provisioning persistent volumes to store data.
These resources are only created if the Entity Operator is deployed using the Cluster Operator.
<kafka_cluster_name>-entity-operator
-
Name given to the following Entity Operator resources:
-
Deployment with Topic and User Operators.
-
Service account used by the Entity Operator.
-
Network policy managing access to the Entity Operator metrics.
-
<kafka_cluster_name>-entity-operator-<random_string>
-
Pod created by the Entity Operator deployment.
<kafka_cluster_name>-entity-topic-operator-config
-
ConfigMap with ancillary configuration for Topic Operators.
<kafka_cluster_name>-entity-user-operator-config
-
ConfigMap with ancillary configuration for User Operators.
<kafka_cluster_name>-entity-topic-operator-certs
-
Secret with Topic Operator keys for communication with Kafka and ZooKeeper.
<kafka_cluster_name>-entity-user-operator-certs
-
Secret with User Operator keys for communication with Kafka and ZooKeeper.
strimzi-<kafka_cluster_name>-entity-topic-operator
-
Role binding used by the Entity Topic Operator.
strimzi-<kafka_cluster_name>-entity-user-operator
-
Role binding used by the Entity User Operator.
These resources are only created if the Kafka Exporter is deployed using the Cluster Operator.
<kafka_cluster_name>-kafka-exporter
-
Name given to the following Kafka Exporter resources:
-
Deployment with Kafka Exporter.
-
Service used to collect consumer lag metrics.
-
Service account used by the Kafka Exporter.
-
Network policy managing access to the Kafka Exporter metrics.
-
<kafka_cluster_name>-kafka-exporter-<random_string>
-
Pod created by the Kafka Exporter deployment.
These resources are only created if Cruise Control was deployed using the Cluster Operator.
<kafka_cluster_name>-cruise-control
-
Name given to the following Cruise Control resources:
-
Deployment with Cruise Control.
-
Service used to communicate with Cruise Control.
-
Service account used by the Cruise Control.
-
<kafka_cluster_name>-cruise-control-<random_string>
-
Pod created by the Cruise Control deployment.
<kafka_cluster_name>-cruise-control-config
-
ConfigMap that contains the Cruise Control ancillary configuration, and is mounted as a volume by the Cruise Control pods.
<kafka_cluster_name>-cruise-control-certs
-
Secret with Cruise Control keys for communication with Kafka and ZooKeeper.
<kafka_cluster_name>-network-policy-cruise-control
-
Network policy managing access to the Cruise Control service.
7.3. Deploying Kafka Connect
Kafka Connect is an integration toolkit for streaming data between Kafka brokers and other systems using connector plugins. Kafka Connect provides a framework for integrating Kafka with an external data source or target, such as a database or messaging system, for import or export of data using connectors. Connectors are plugins that provide the connection configuration needed.
In Strimzi, Kafka Connect is deployed in distributed mode. Kafka Connect can also work in standalone mode, but this is not supported by Strimzi.
Using the concept of connectors, Kafka Connect provides a framework for moving large amounts of data into and out of your Kafka cluster while maintaining scalability and reliability.
The Cluster Operator manages Kafka Connect clusters deployed using the KafkaConnect
resource and connectors created using the KafkaConnector
resource.
In order to use Kafka Connect, you need to do the following.
Note
|
The term connector is used interchangeably to mean a connector instance running within a Kafka Connect cluster, or a connector class. In this guide, the term connector is used when the meaning is clear from the context. |
7.3.1. Deploying Kafka Connect to your Kubernetes cluster
This procedure shows how to deploy a Kafka Connect cluster to your Kubernetes cluster using the Cluster Operator.
A Kafka Connect cluster deployment is implemented with a configurable number of nodes (also called workers) that distribute the workload of connectors as tasks so that the message flow is highly scalable and reliable.
The deployment uses a YAML file to provide the specification to create a KafkaConnect
resource.
Strimzi provides example configuration files. In this procedure, we use the following example file:
-
examples/connect/kafka-connect.yaml
Important
|
If deploying Kafka Connect clusters to run in parallel, each instance must use unique names for internal Kafka Connect topics. To do this, configure each Kafka Connect instance to replace the defaults. |
-
Deploy Kafka Connect to your Kubernetes cluster. Use the
examples/connect/kafka-connect.yaml
file to deploy Kafka Connect.kubectl apply -f examples/connect/kafka-connect.yaml
-
Check the status of the deployment:
kubectl get pods -n <my_cluster_operator_namespace>
Output shows the deployment name and readinessNAME READY STATUS RESTARTS my-connect-cluster-connect-<pod_id> 1/1 Running 0
my-connect-cluster
is the name of the Kafka Connect cluster.A pod ID identifies each pod created.
With the default deployment, you create a single Kafka Connect pod.
READY
shows the number of replicas that are ready/expected. The deployment is successful when theSTATUS
displays asRunning
.
7.3.2. List of Kafka Connect cluster resources
The following resources are created by the Cluster Operator in the Kubernetes cluster:
- <connect_cluster_name>-connect
-
Name given to the following Kafka Connect resources:
-
StrimziPodSet that creates the Kafka Connect worker node pods.
-
Headless service that provides stable DNS names to the Kafka Connect pods.
-
Service account used by the Kafka Connect pods.
-
Pod disruption budget configured for the Kafka Connect worker nodes.
-
Network policy managing access to the Kafka Connect REST API.
-
- <connect_cluster_name>-connect-<pod_id>
-
Pods created by the Kafka Connect StrimziPodSet.
- <connect_cluster_name>-connect-api
-
Service which exposes the REST interface for managing the Kafka Connect cluster.
- <connect_cluster_name>-connect-config
-
ConfigMap which contains the Kafka Connect ancillary configuration and is mounted as a volume by the Kafka Connect pods.
- strimzi-<namespace-name>-<connect_cluster_name>-connect-init
-
Cluster role binding used by the Kafka Connect cluster.
- <connect_cluster_name>-connect-build
-
Pod used to build a new container image with additional connector plugins (only when Kafka Connect Build feature is used).
- <connect_cluster_name>-connect-dockerfile
-
ConfigMap with the Dockerfile generated to build the new container image with additional connector plugins (only when the Kafka Connect build feature is used).
7.4. Adding Kafka Connect connectors
Kafka Connect uses connectors to integrate with other systems to stream data.
A connector is an instance of a Kafka Connector
class, which can be one of the following type:
- Source connector
-
A source connector is a runtime entity that fetches data from an external system and feeds it to Kafka as messages.
- Sink connector
-
A sink connector is a runtime entity that fetches messages from Kafka topics and feeds them to an external system.
Kafka Connect uses a plugin architecture to provide the implementation artifacts for connectors. Plugins allow connections to other systems and provide additional configuration to manipulate data. Plugins include connectors and other components, such as data converters and transforms. A connector operates with a specific type of external system. Each connector defines a schema for its configuration. You supply the configuration to Kafka Connect to create a connector instance within Kafka Connect. Connector instances then define a set of tasks for moving data between systems.
Plugins provide a set of one or more artifacts that define a connector and task implementation for connecting to a given kind of data source. The configuration describes the source input data and target output data to feed into and out of Kafka Connect. The plugins might also contain the libraries and files needed to transform the data.
A Kafka Connect deployment can have one or more plugins, but only one version of each plugin. Plugins for many external systems are available for use with Kafka Connect. You can also create your own plugins.
Add connector plugins to Kafka Connect in one of the following ways:
-
Configure Kafka Connect to build a new container image with plugins automatically
-
Create a Docker image from the base Kafka Connect image (manually or using continuous integration)
After plugins have been added to the container image, you can start, stop, and manage connector instances in the following ways:
You can also create new connector instances using these options.
7.4.1. Building new container images with connector plugins automatically
Configure Kafka Connect so that Strimzi automatically builds a new container image with additional connectors.
You define the connector plugins using the .spec.build.plugins
property of the KafkaConnect
custom resource.
Strimzi automatically downloads and adds the connector plugins into a new container image.
The container is pushed into the container repository specified in .spec.build.output
and automatically used in the Kafka Connect deployment.
-
A container registry.
You need to provide your own container registry where images can be pushed to, stored, and pulled from. Strimzi supports private container registries as well as public registries such as Quay or Docker Hub.
-
Configure the
KafkaConnect
custom resource by specifying the container registry in.spec.build.output
, and additional connectors in.spec.build.plugins
:apiVersion: kafka.strimzi.io/v1beta2 kind: KafkaConnect metadata: name: my-connect-cluster spec: # (1) #... build: output: # (2) type: docker image: my-registry.io/my-org/my-connect-cluster:latest pushSecret: my-registry-credentials plugins: # (3) - name: connector-1 artifacts: - type: tgz url: <url_to_download_connector_1_artifact> sha512sum: <SHA-512_checksum_of_connector_1_artifact> - name: connector-2 artifacts: - type: jar url: <url_to_download_connector_2_artifact> sha512sum: <SHA-512_checksum_of_connector_2_artifact> #...
-
(Required) Configuration of the container registry where new images are pushed.
-
(Required) List of connector plugins and their artifacts to add to the new container image. Each plugin must be configured with at least one
artifact
.
-
Create or update the resource:
$ kubectl apply -f <kafka_connect_configuration_file>
-
Wait for the new container image to build, and for the Kafka Connect cluster to be deployed.
-
Use the Kafka Connect REST API or
KafkaConnector
custom resources to use the connector plugins you added.
A new container image is built automatically when you change the base image (.spec.image
) or change the connector plugin artifacts configuration (.spec.build.plugins
).
To pull an upgraded base image or to download the latest connector plugin artifacts without changing the KafkaConnect
resource, you can trigger a rebuild of the container image associated with the Kafka Connect cluster by applying the annotation strimzi.io/force-rebuild=true
to the Kafka Connect StrimziPodSet
resource.
The annotation triggers the rebuilding process, fetching any new artifacts for plugins specified in the KafkaConnect
custom resource and incorporating them into the container image.
The rebuild includes downloads of new plugin artifacts without versions.
7.4.2. Building new container images with connector plugins from the Kafka Connect base image
Create a custom Docker image with connector plugins from the Kafka Connect base image.
Add the custom image to the /opt/kafka/plugins
directory.
You can use the Kafka container image on Container Registry as a base image for creating your own custom image with additional connector plugins.
At startup, the Strimzi version of Kafka Connect loads any third-party connector plugins contained in the /opt/kafka/plugins
directory.
-
Create a new
Dockerfile
usingquay.io/strimzi/kafka:latest-kafka-3.9.0
as the base image:FROM quay.io/strimzi/kafka:latest-kafka-3.9.0 USER root:root COPY ./my-plugins/ /opt/kafka/plugins/ USER 1001
Example plugins file$ tree ./my-plugins/ ./my-plugins/ ├── debezium-connector-mongodb │  ├── bson-<version>.jar │  ├── CHANGELOG.md │  ├── CONTRIBUTE.md │  ├── COPYRIGHT.txt │  ├── debezium-connector-mongodb-<version>.jar │  ├── debezium-core-<version>.jar │  ├── LICENSE.txt │  ├── mongodb-driver-core-<version>.jar │  ├── README.md │  └── # ... ├── debezium-connector-mysql │  ├── CHANGELOG.md │  ├── CONTRIBUTE.md │  ├── COPYRIGHT.txt │  ├── debezium-connector-mysql-<version>.jar │  ├── debezium-core-<version>.jar │  ├── LICENSE.txt │  ├── mysql-binlog-connector-java-<version>.jar │  ├── mysql-connector-java-<version>.jar │  ├── README.md │  └── # ... └── debezium-connector-postgres ├── CHANGELOG.md ├── CONTRIBUTE.md ├── COPYRIGHT.txt ├── debezium-connector-postgres-<version>.jar ├── debezium-core-<version>.jar ├── LICENSE.txt ├── postgresql-<version>.jar ├── protobuf-java-<version>.jar   ├── README.md └── # ...
The COPY command points to the plugin files to copy to the container image.
This example adds plugins for Debezium connectors (MongoDB, MySQL, and PostgreSQL), though not all files are listed for brevity. Debezium running in Kafka Connect looks the same as any other Kafka Connect task.
-
Build the container image.
-
Push your custom image to your container registry.
-
Point to the new container image.
You can point to the image in one of the following ways:
-
Edit the
KafkaConnect.spec.image
property of theKafkaConnect
custom resource.If set, this property overrides the
STRIMZI_KAFKA_CONNECT_IMAGES
environment variable in the Cluster Operator.apiVersion: kafka.strimzi.io/v1beta2 kind: KafkaConnect metadata: name: my-connect-cluster spec: (1) #... image: my-new-container-image (2) config: (3) #...
-
The docker image for Kafka Connect pods.
-
Configuration of the Kafka Connect workers (not connectors).
-
Edit the
STRIMZI_KAFKA_CONNECT_IMAGES
environment variable in theinstall/cluster-operator/060-Deployment-strimzi-cluster-operator.yaml
file to point to the new container image, and then reinstall the Cluster Operator.
-
7.4.3. Deploying KafkaConnector resources
Deploy KafkaConnector
resources to manage connectors.
The KafkaConnector
custom resource offers a Kubernetes-native approach to management of connectors by the Cluster Operator.
You don’t need to send HTTP requests to manage connectors, as with the Kafka Connect REST API.
You manage a running connector instance by updating its corresponding KafkaConnector
resource, and then applying the updates.
The Cluster Operator updates the configurations of the running connector instances.
You remove a connector by deleting its corresponding KafkaConnector
.
KafkaConnector
resources must be deployed to the same namespace as the Kafka Connect cluster they link to.
In the configuration shown in this procedure, the autoRestart
feature is enabled (enabled: true
) for automatic restarts of failed connectors and tasks.
You can also annotate the KafkaConnector
resource to restart a connector or restart a connector task manually.
You can use your own connectors or try the examples provided by Strimzi. Up until Apache Kafka 3.1.0, example file connector plugins were included with Apache Kafka. Starting from the 3.1.1 and 3.2.0 releases of Apache Kafka, the examples need to be added to the plugin path as any other connector.
Strimzi provides an example KafkaConnector
configuration file (examples/connect/source-connector.yaml
) for the example file connector plugins, which creates the following connector instances as KafkaConnector
resources:
-
A
FileStreamSourceConnector
instance that reads each line from the Kafka license file (the source) and writes the data as messages to a single Kafka topic. -
A
FileStreamSinkConnector
instance that reads messages from the Kafka topic and writes the messages to a temporary file (the sink).
We use the example file to create connectors in this procedure.
Note
|
The example connectors are not intended for use in a production environment. |
-
A Kafka Connect deployment
-
The Cluster Operator is running
-
Add the
FileStreamSourceConnector
andFileStreamSinkConnector
plugins to Kafka Connect in one of the following ways:-
Configure Kafka Connect to build a new container image with plugins automatically
-
Create a Docker image from the base Kafka Connect image (manually or using continuous integration)
-
-
Set the
strimzi.io/use-connector-resources annotation
totrue
in the Kafka Connect configuration.apiVersion: kafka.strimzi.io/v1beta2 kind: KafkaConnect metadata: name: my-connect-cluster annotations: strimzi.io/use-connector-resources: "true" spec: # ...
With the
KafkaConnector
resources enabled, the Cluster Operator watches for them. -
Edit the
examples/connect/source-connector.yaml
file:Example source connector configurationapiVersion: kafka.strimzi.io/v1beta2 kind: KafkaConnector metadata: name: my-source-connector # (1) labels: strimzi.io/cluster: my-connect-cluster # (2) spec: class: org.apache.kafka.connect.file.FileStreamSourceConnector # (3) tasksMax: 2 # (4) autoRestart: # (5) enabled: true config: # (6) file: "/opt/kafka/LICENSE" # (7) topic: my-topic # (8) # ...
-
Name of the
KafkaConnector
resource, which is used as the name of the connector. Use any name that is valid for a Kubernetes resource. -
Name of the Kafka Connect cluster to create the connector instance in. Connectors must be deployed to the same namespace as the Kafka Connect cluster they link to.
-
Full name of the connector class. This should be present in the image being used by the Kafka Connect cluster.
-
Maximum number of Kafka Connect tasks that the connector can create.
-
Enables automatic restarts of failed connectors and tasks. By default, the number of restarts is indefinite, but you can set a maximum on the number of automatic restarts using the
maxRestarts
property. -
Connector configuration as key-value pairs.
-
Location of the external data file. In this example, we’re configuring the
FileStreamSourceConnector
to read from the/opt/kafka/LICENSE
file. -
Kafka topic to publish the source data to.
-
-
Create the source
KafkaConnector
in your Kubernetes cluster:kubectl apply -f examples/connect/source-connector.yaml
-
Create an
examples/connect/sink-connector.yaml
file:touch examples/connect/sink-connector.yaml
-
Paste the following YAML into the
sink-connector.yaml
file:apiVersion: kafka.strimzi.io/v1beta2 kind: KafkaConnector metadata: name: my-sink-connector labels: strimzi.io/cluster: my-connect spec: class: org.apache.kafka.connect.file.FileStreamSinkConnector # (1) tasksMax: 2 config: # (2) file: "/tmp/my-file" # (3) topics: my-topic # (4)
-
Full name or alias of the connector class. This should be present in the image being used by the Kafka Connect cluster.
-
Connector configuration as key-value pairs.
-
Temporary file to publish the source data to.
-
Kafka topic to read the source data from.
-
-
Create the sink
KafkaConnector
in your Kubernetes cluster:kubectl apply -f examples/connect/sink-connector.yaml
-
Check that the connector resources were created:
kubectl get kctr --selector strimzi.io/cluster=<my_connect_cluster> -o name my-source-connector my-sink-connector
Replace <my_connect_cluster> with the name of your Kafka Connect cluster.
-
In the container, execute
kafka-console-consumer.sh
to read the messages that were written to the topic by the source connector:kubectl exec <my_kafka_cluster>-kafka-0 -i -t -- bin/kafka-console-consumer.sh --bootstrap-server <my_kafka_cluster>-kafka-bootstrap.NAMESPACE.svc:9092 --topic my-topic --from-beginning
Replace <my_kafka_cluster> with the name of your Kafka cluster.
Source and sink connector configuration options
The connector configuration is defined in the spec.config
property of the KafkaConnector
resource.
The FileStreamSourceConnector
and FileStreamSinkConnector
classes support the same configuration options as the Kafka Connect REST API.
Other connectors support different configuration options.
Name | Type | Default value | Description |
---|---|---|---|
|
String |
Null |
Source file to write messages to. If not specified, the standard input is used. |
|
List |
Null |
The Kafka topic to publish data to. |
Name | Type | Default value | Description |
---|---|---|---|
|
String |
Null |
Destination file to write messages to. If not specified, the standard output is used. |
|
List |
Null |
One or more Kafka topics to read data from. |
|
String |
Null |
A regular expression matching one or more Kafka topics to read data from. |
7.4.4. Exposing the Kafka Connect API
Use the Kafka Connect REST API as an alternative to using KafkaConnector
resources to manage connectors.
The Kafka Connect REST API is available as a service running on <connect_cluster_name>-connect-api:8083
, where <connect_cluster_name> is the name of your Kafka Connect cluster.
The service is created when you create a Kafka Connect instance.
The operations supported by the Kafka Connect REST API are described in the Apache Kafka Connect API documentation.
Note
|
The strimzi.io/use-connector-resources annotation enables KafkaConnectors.
If you applied the annotation to your KafkaConnect resource configuration, you need to remove it to use the Kafka Connect API.
Otherwise, manual changes made directly using the Kafka Connect REST API are reverted by the Cluster Operator.
|
You can add the connector configuration as a JSON object.
curl -X POST \
http://my-connect-cluster-connect-api:8083/connectors \
-H 'Content-Type: application/json' \
-d '{ "name": "my-source-connector",
"config":
{
"connector.class":"org.apache.kafka.connect.file.FileStreamSourceConnector",
"file": "/opt/kafka/LICENSE",
"topic":"my-topic",
"tasksMax": "4",
"type": "source"
}
}'
The API is only accessible within the Kubernetes cluster. If you want to make the Kafka Connect API accessible to applications running outside of the Kubernetes cluster, you can expose it manually by creating one of the following features:
-
LoadBalancer
orNodePort
type services -
Ingress
resources (Kubernetes only) -
OpenShift routes (OpenShift only)
Note
|
The connection is insecure, so allow external access advisedly. |
If you decide to create services, use the labels from the selector
of the <connect_cluster_name>-connect-api
service to configure the pods to which the service will route the traffic:
# ...
selector:
strimzi.io/cluster: my-connect-cluster (1)
strimzi.io/kind: KafkaConnect
strimzi.io/name: my-connect-cluster-connect (2)
#...
-
Name of the Kafka Connect custom resource in your Kubernetes cluster.
-
Name of the Kafka Connect deployment created by the Cluster Operator.
You must also create a NetworkPolicy
that allows HTTP requests from external clients.
apiVersion: networking.k8s.io/v1
kind: NetworkPolicy
metadata:
name: my-custom-connect-network-policy
spec:
ingress:
- from:
- podSelector: (1)
matchLabels:
app: my-connector-manager
ports:
- port: 8083
protocol: TCP
podSelector:
matchLabels:
strimzi.io/cluster: my-connect-cluster
strimzi.io/kind: KafkaConnect
strimzi.io/name: my-connect-cluster-connect
policyTypes:
- Ingress
-
The label of the pod that is allowed to connect to the API.
To add the connector configuration outside the cluster, use the URL of the resource that exposes the API in the curl command.
7.4.5. Limiting access to the Kafka Connect API
It is crucial to restrict access to the Kafka Connect API only to trusted users to prevent unauthorized actions and potential security issues. The Kafka Connect API provides extensive capabilities for altering connector configurations, which makes it all the more important to take security precautions. Someone with access to the Kafka Connect API could potentially obtain sensitive information that an administrator may assume is secure.
The Kafka Connect REST API can be accessed by anyone who has authenticated access to the Kubernetes cluster and knows the endpoint URL, which includes the hostname/IP address and port number.
For example, suppose an organization uses a Kafka Connect cluster and connectors to stream sensitive data from a customer database to a central database. The administrator uses a configuration provider plugin to store sensitive information related to connecting to the customer database and the central database, such as database connection details and authentication credentials. The configuration provider protects this sensitive information from being exposed to unauthorized users. However, someone who has access to the Kafka Connect API can still obtain access to the customer database without the consent of the administrator. They can do this by setting up a fake database and configuring a connector to connect to it. They then modify the connector configuration to point to the customer database, but instead of sending the data to the central database, they send it to the fake database. By configuring the connector to connect to the fake database, the login details and credentials for connecting to the customer database are intercepted, even though they are stored securely in the configuration provider.
If you are using the KafkaConnector
custom resources, then by default the Kubernetes RBAC rules permit only Kubernetes cluster administrators to make changes to connectors.
You can also designate non-cluster administrators to manage Strimzi resources.
With KafkaConnector
resources enabled in your Kafka Connect configuration, changes made directly using the Kafka Connect REST API are reverted by the Cluster Operator.
If you are not using the KafkaConnector
resource, the default RBAC rules do not limit access to the Kafka Connect API.
If you want to limit direct access to the Kafka Connect REST API using Kubernetes RBAC, you need to enable and use the KafkaConnector
resources.
For improved security, we recommend configuring the following properties for the Kafka Connect API:
org.apache.kafka.disallowed.login.modules
-
(Kafka 3.4 or later) Set the
org.apache.kafka.disallowed.login.modules
Java system property to prevent the use of insecure login modules. For example, specifyingcom.sun.security.auth.module.JndiLoginModule
prevents the use of the KafkaJndiLoginModule
.Example configuration for disallowing login modulesapiVersion: kafka.strimzi.io/v1beta2 kind: KafkaConnect metadata: name: my-connect-cluster annotations: strimzi.io/use-connector-resources: "true" spec: # ... jvmOptions: javaSystemProperties: - name: org.apache.kafka.disallowed.login.modules value: com.sun.security.auth.module.JndiLoginModule, org.apache.kafka.common.security.kerberos.KerberosLoginModule # ...
Only allow trusted login modules and follow the latest advice from Kafka for the version you are using. As a best practice, you should explicitly disallow insecure login modules in your Kafka Connect configuration by using the
org.apache.kafka.disallowed.login.modules
system property. connector.client.config.override.policy
-
Set the
connector.client.config.override.policy
property toNone
to prevent connector configurations from overriding the Kafka Connect configuration and the consumers and producers it uses.Example configuration to specify connector override policyapiVersion: kafka.strimzi.io/v1beta2 kind: KafkaConnect metadata: name: my-connect-cluster annotations: strimzi.io/use-connector-resources: "true" spec: # ... config: connector.client.config.override.policy: None # ...
7.4.6. Switching to using KafkaConnector
custom resources
You can switch from using the Kafka Connect API to using KafkaConnector
custom resources to manage your connectors.
To make the switch, do the following in the order shown:
-
Deploy
KafkaConnector
resources with the configuration to create your connector instances. -
Enable
KafkaConnector
resources in your Kafka Connect configuration by setting thestrimzi.io/use-connector-resources
annotation totrue
.
Warning
|
If you enable KafkaConnector resources before creating them, you delete all connectors.
|
To switch from using KafkaConnector
resources to using the Kafka Connect API, first remove the annotation that enables the KafkaConnector
resources from your Kafka Connect configuration.
Otherwise, manual changes made directly using the Kafka Connect REST API are reverted by the Cluster Operator.
When making the switch, check the status of the KafkaConnect
resource.
The value of metadata.generation
(the current version of the deployment) must match status.observedGeneration
(the latest reconciliation of the resource).
When the Kafka Connect cluster is Ready
, you can delete the KafkaConnector
resources.
7.5. Deploying Kafka MirrorMaker
Kafka MirrorMaker replicates data between two or more Kafka clusters, within or across data centers. This process is called mirroring to avoid confusion with the concept of Kafka partition replication. MirrorMaker consumes messages from a source cluster and republishes those messages to a target cluster.
Data replication across clusters supports scenarios that require the following:
-
Recovery of data in the event of a system failure
-
Consolidation of data from multiple source clusters for centralized analysis
-
Restriction of data access to a specific cluster
-
Provision of data at a specific location to improve latency
7.5.1. Deploying Kafka MirrorMaker to your Kubernetes cluster
This procedure shows how to deploy a Kafka MirrorMaker cluster to your Kubernetes cluster using the Cluster Operator.
The deployment uses a YAML file to provide the specification to create a KafkaMirrorMaker
or KafkaMirrorMaker2
resource depending on the version of MirrorMaker deployed.
MirrorMaker 2 is based on Kafka Connect and uses its configuration properties.
Important
|
Kafka MirrorMaker 1 (referred to as just MirrorMaker in the documentation) has been deprecated in Apache Kafka 3.0.0 and will be removed in Apache Kafka 4.0.0.
As a result, the KafkaMirrorMaker custom resource which is used to deploy Kafka MirrorMaker 1 has been deprecated in Strimzi as well.
The KafkaMirrorMaker resource will be removed from Strimzi when we adopt Apache Kafka 4.0.0.
As a replacement, use the KafkaMirrorMaker2 custom resource with the IdentityReplicationPolicy .
|
Strimzi provides example configuration files. In this procedure, we use the following example files:
-
examples/mirror-maker/kafka-mirror-maker.yaml
-
examples/mirror-maker/kafka-mirror-maker-2.yaml
Important
|
If deploying MirrorMaker 2 clusters to run in parallel, using the same target Kafka cluster, each instance must use unique names for internal Kafka Connect topics. To do this, configure each MirrorMaker 2 instance to replace the defaults. |
-
Deploy Kafka MirrorMaker to your Kubernetes cluster:
For MirrorMaker:
kubectl apply -f examples/mirror-maker/kafka-mirror-maker.yaml
For MirrorMaker 2:
kubectl apply -f examples/mirror-maker/kafka-mirror-maker-2.yaml
-
Check the status of the deployment:
kubectl get pods -n <my_cluster_operator_namespace>
Output shows the deployment name and readinessNAME READY STATUS RESTARTS my-mirror-maker-mirror-maker-<pod_id> 1/1 Running 1 my-mm2-cluster-mirrormaker2-<pod_id> 1/1 Running 1
my-mirror-maker
is the name of the Kafka MirrorMaker cluster.my-mm2-cluster
is the name of the Kafka MirrorMaker 2 cluster.A pod ID identifies each pod created.
With the default deployment, you install a single MirrorMaker or MirrorMaker 2 pod.
READY
shows the number of replicas that are ready/expected. The deployment is successful when theSTATUS
displays asRunning
.
7.5.2. List of Kafka MirrorMaker 2 cluster resources
The following resources are created by the Cluster Operator in the Kubernetes cluster:
- <mirrormaker2_cluster_name>-mirrormaker2
-
Name given to the following MirrorMaker 2 resources:
-
StrimziPodSet that creates the MirrorMaker 2 worker node pods.
-
Headless service that provides stable DNS names to the MirrorMaker 2 pods.
-
Service account used by the MirrorMaker 2 pods.
-
Pod disruption budget configured for the MirrorMaker 2 worker nodes.
-
Network Policy managing access to the MirrorMaker 2 REST API.
-
- <mirrormaker2_cluster_name>-mirrormaker2-<pod_id>
-
Pods created by the MirrorMaker 2 StrimziPodSet.
- <mirrormaker2_cluster_name>-mirrormaker2-api
-
Service which exposes the REST interface for managing the MirrorMaker 2 cluster.
- <mirrormaker2_cluster_name>-mirrormaker2-config
-
ConfigMap which contains the MirrorMaker 2 ancillary configuration and is mounted as a volume by the MirrorMaker 2 pods.
- strimzi-<namespace-name>-<mirrormaker2_cluster_name>-mirrormaker2-init
-
Cluster role binding used by the MirrorMaker 2 cluster.
7.5.3. List of Kafka MirrorMaker cluster resources
The following resources are created by the Cluster Operator in the Kubernetes cluster:
- <mirrormaker_cluster_name>-mirror-maker
-
Name given to the following MirrorMaker resources:
-
Deployment which is responsible for creating the MirrorMaker pods.
-
Service account used by the MirrorMaker nodes.
-
Pod Disruption Budget configured for the MirrorMaker worker nodes.
-
- <mirrormaker_cluster_name>-mirror-maker-config
-
ConfigMap which contains ancillary configuration for MirrorMaker, and is mounted as a volume by the MirrorMaker pods.
7.6. Deploying Kafka Bridge
Kafka Bridge provides an API for integrating HTTP-based clients with a Kafka cluster.
7.6.1. Deploying Kafka Bridge to your Kubernetes cluster
This procedure shows how to deploy a Kafka Bridge cluster to your Kubernetes cluster using the Cluster Operator.
The deployment uses a YAML file to provide the specification to create a KafkaBridge
resource.
Strimzi provides example configuration files. In this procedure, we use the following example file:
-
examples/bridge/kafka-bridge.yaml
-
Deploy Kafka Bridge to your Kubernetes cluster:
kubectl apply -f examples/bridge/kafka-bridge.yaml
-
Check the status of the deployment:
kubectl get pods -n <my_cluster_operator_namespace>
Output shows the deployment name and readinessNAME READY STATUS RESTARTS my-bridge-bridge-<pod_id> 1/1 Running 0
my-bridge
is the name of the Kafka Bridge cluster.A pod ID identifies each pod created.
With the default deployment, you install a single Kafka Bridge pod.
READY
shows the number of replicas that are ready/expected. The deployment is successful when theSTATUS
displays asRunning
.
7.6.2. Exposing the Kafka Bridge service to your local machine
Use port forwarding to expose the Kafka Bridge service to your local machine on http://localhost:8080.
Note
|
Port forwarding is only suitable for development and testing purposes. |
-
List the names of the pods in your Kubernetes cluster:
kubectl get pods -o name pod/kafka-consumer # ... pod/my-bridge-bridge-<pod_id>
-
Connect to the Kafka Bridge pod on port
8080
:kubectl port-forward pod/my-bridge-bridge-<pod_id> 8080:8080 &
NoteIf port 8080 on your local machine is already in use, use an alternative HTTP port, such as 8008
.
API requests are now forwarded from port 8080 on your local machine to port 8080 in the Kafka Bridge pod.
7.6.3. Accessing the Kafka Bridge outside of Kubernetes
After deployment, the Kafka Bridge can only be accessed by applications running in the same Kubernetes cluster.
These applications use the <kafka_bridge_name>-bridge-service
service to access the API.
If you want to make the Kafka Bridge accessible to applications running outside of the Kubernetes cluster, you can expose it manually by creating one of the following features:
-
LoadBalancer
orNodePort
type services -
Ingress
resources (Kubernetes only) -
OpenShift routes (OpenShift only)
If you decide to create Services, use the labels from the selector
of the <kafka_bridge_name>-bridge-service
service to configure the pods to which the service will route the traffic:
# ...
selector:
strimzi.io/cluster: kafka-bridge-name (1)
strimzi.io/kind: KafkaBridge
#...
-
Name of the Kafka Bridge custom resource in your Kubernetes cluster.
7.6.4. List of Kafka Bridge cluster resources
The following resources are created by the Cluster Operator in the Kubernetes cluster:
- <bridge_cluster_name>-bridge
-
Deployment which is in charge to create the Kafka Bridge worker node pods.
- <bridge_cluster_name>-bridge-service
-
Service which exposes the REST interface of the Kafka Bridge cluster.
- <bridge_cluster_name>-bridge-config
-
ConfigMap which contains the Kafka Bridge ancillary configuration and is mounted as a volume by the Kafka broker pods.
- <bridge_cluster_name>-bridge
-
Pod Disruption Budget configured for the Kafka Bridge worker nodes.
7.7. Alternative standalone deployment options for Strimzi operators
You can perform a standalone deployment of the Topic Operator and User Operator. Consider a standalone deployment of these operators if you are using a Kafka cluster that is not managed by the Cluster Operator.
You deploy the operators to Kubernetes. Kafka can be running outside of Kubernetes. For example, you might be using a Kafka as a managed service. You adjust the deployment configuration for the standalone operator to match the address of your Kafka cluster.
7.7.1. Deploying the standalone Topic Operator
This procedure shows how to deploy the Topic Operator as a standalone component for topic management. You can use a standalone Topic Operator with a Kafka cluster that is not managed by the Cluster Operator.
Standalone deployment files are provided with Strimzi.
Use the 05-Deployment-strimzi-topic-operator.yaml
deployment file to deploy the Topic Operator.
Add or set the environment variables needed to make a connection to a Kafka cluster.
The Topic Operator watches for KafkaTopic
resources in a single namespace.
You specify the namespace to watch, and the connection to the Kafka cluster, in the Topic Operator configuration.
A single Topic Operator can watch a single namespace.
One namespace should be watched by only one Topic Operator.
If you want to use more than one Topic Operator, configure each of them to watch different namespaces.
In this way, you can use Topic Operators with multiple Kafka clusters.
-
The standalone Topic Operator deployment files, which are included in the Strimzi deployment files.
-
You are running a Kafka cluster for the Topic Operator to connect to.
As long as the standalone Topic Operator is correctly configured for connection, the Kafka cluster can be running on a bare-metal environment, a virtual machine, or as a managed cloud application service.
-
Edit the
env
properties in theinstall/topic-operator/05-Deployment-strimzi-topic-operator.yaml
standalone deployment file.Example standalone Topic Operator deployment configurationapiVersion: apps/v1 kind: Deployment metadata: name: strimzi-topic-operator labels: app: strimzi spec: # ... template: # ... spec: # ... containers: - name: strimzi-topic-operator # ... env: - name: STRIMZI_NAMESPACE # (1) valueFrom: fieldRef: fieldPath: metadata.namespace - name: STRIMZI_KAFKA_BOOTSTRAP_SERVERS # (2) value: my-kafka-bootstrap-address:9092 - name: STRIMZI_RESOURCE_LABELS # (3) value: "strimzi.io/cluster=my-cluster" - name: STRIMZI_FULL_RECONCILIATION_INTERVAL_MS # (4) value: "120000" - name: STRIMZI_LOG_LEVEL # (5) value: INFO - name: STRIMZI_TLS_ENABLED # (6) value: "false" - name: STRIMZI_JAVA_OPTS # (7) value: "-Xmx=512M -Xms=256M" - name: STRIMZI_JAVA_SYSTEM_PROPERTIES # (8) value: "-Djavax.net.debug=verbose -DpropertyName=value" - name: STRIMZI_PUBLIC_CA # (9) value: "false" - name: STRIMZI_TLS_AUTH_ENABLED # (10) value: "false" - name: STRIMZI_SASL_ENABLED # (11) value: "false" - name: STRIMZI_SASL_USERNAME # (12) value: "admin" - name: STRIMZI_SASL_PASSWORD # (13) value: "password" - name: STRIMZI_SASL_MECHANISM # (14) value: "scram-sha-512" - name: STRIMZI_SECURITY_PROTOCOL # (15) value: "SSL" - name: STRIMZI_USE_FINALIZERS value: "false" # (16)
-
The Kubernetes namespace for the Topic Operator to watch for
KafkaTopic
resources. Specify the namespace of the Kafka cluster. -
The host and port pair of the bootstrap broker address to discover and connect to all brokers in the Kafka cluster. Use a comma-separated list to specify two or three broker addresses in case a server is down.
-
The label to identify the
KafkaTopic
resources managed by the Topic Operator. This does not have to be the name of the Kafka cluster. It can be the label assigned to theKafkaTopic
resource. If you deploy more than one Topic Operator, the labels must be unique for each. That is, the operators cannot manage the same resources. -
The interval between periodic reconciliations, in milliseconds. The default is
120000
(2 minutes). -
The level for printing logging messages. You can set the level to
ERROR
,WARNING
,INFO
,DEBUG
, orTRACE
. -
Enables TLS support for encrypted communication with the Kafka brokers.
-
(Optional) The Java options used by the JVM running the Topic Operator.
-
(Optional) The debugging (
-D
) options set for the Topic Operator. -
(Optional) Skips the generation of trust store certificates if TLS is enabled through
STRIMZI_TLS_ENABLED
. If this environment variable is enabled, the brokers must use a public trusted certificate authority for their TLS certificates. The default isfalse
. -
(Optional) Generates key store certificates for mTLS authentication. Setting this to
false
disables client authentication with mTLS to the Kafka brokers. The default istrue
. -
(Optional) Enables SASL support for client authentication when connecting to Kafka brokers. The default is
false
. -
(Optional) The SASL username for client authentication. Mandatory only if SASL is enabled through
STRIMZI_SASL_ENABLED
. -
(Optional) The SASL password for client authentication. Mandatory only if SASL is enabled through
STRIMZI_SASL_ENABLED
. -
(Optional) The SASL mechanism for client authentication. Mandatory only if SASL is enabled through
STRIMZI_SASL_ENABLED
. You can set the value toplain
,scram-sha-256
, orscram-sha-512
. -
(Optional) The security protocol used for communication with Kafka brokers. The default value is "PLAINTEXT". You can set the value to
PLAINTEXT
,SSL
,SASL_PLAINTEXT
, orSASL_SSL
. -
Set
STRIMZI_USE_FINALIZERS
tofalse
if you do not want to use finalizers to control topic deletion.
-
-
If you want to connect to Kafka brokers that are using certificates from a public certificate authority, set
STRIMZI_PUBLIC_CA
totrue
. Set this property totrue
, for example, if you are using Amazon AWS MSK service. -
If you enabled mTLS with the
STRIMZI_TLS_ENABLED
environment variable, specify the keystore and truststore used to authenticate connection to the Kafka cluster.Example mTLS configuration# .... env: - name: STRIMZI_TRUSTSTORE_LOCATION # (1) value: "/path/to/truststore.p12" - name: STRIMZI_TRUSTSTORE_PASSWORD # (2) value: "TRUSTSTORE-PASSWORD" - name: STRIMZI_KEYSTORE_LOCATION # (3) value: "/path/to/keystore.p12" - name: STRIMZI_KEYSTORE_PASSWORD # (4) value: "KEYSTORE-PASSWORD" # ...
-
The truststore contains the public keys of the Certificate Authorities used to sign the Kafka and ZooKeeper server certificates.
-
The password for accessing the truststore.
-
The keystore contains the private key for mTLS authentication.
-
The password for accessing the keystore.
-
-
If you need to configure custom SASL authentication, you can define the necessary authentication properties using the
STRIMZI_SASL_CUSTOM_CONFIG_JSON
environment variable for the standalone operator. For example, this configuration may be used for accessing a Kafka cluster in a cloud provider with a custom login module like the Amazon MSK Library for AWS Identity and Access Management (aws-msk_iam-auth
).The property
STRIMZI_ALTERABLE_TOPIC_CONFIG
defaults toALL
, allowing all.spec.config
properties to be set in theKafkaTopic
resource. If this setting is not suitable for a managed Kafka service, do as follows:-
If only a subset of properties is configurable, list them as comma-separated values.
-
If no properties are to be configured, use
NONE
, which is equivalent to an empty property list.
NoteOnly Kafka configuration properties starting with sasl.
can be set with theSTRIMZI_SASL_CUSTOM_CONFIG_JSON
environment variable.Example custom SASL configuration# .... env: - name: STRIMZI_SASL_ENABLED value: "true" - name: STRIMZI_SECURITY_PROTOCOL value: SASL_SSL - name: STRIMZI_SKIP_CLUSTER_CONFIG_REVIEW # (1) value: "true" - name: STRIMZI_ALTERABLE_TOPIC_CONFIG # (2) value: compression.type, max.message.bytes, message.timestamp.difference.max.ms, message.timestamp.type, retention.bytes, retention.ms - name: STRIMZI_SASL_CUSTOM_CONFIG_JSON # (3) value: | { "sasl.mechanism": "AWS_MSK_IAM", "sasl.jaas.config": "software.amazon.msk.auth.iam.IAMLoginModule required;", "sasl.client.callback.handler.class": "software.amazon.msk.auth.iam.IAMClientCallbackHandler" } - name: STRIMZI_PUBLIC_CA value: "true" - name: STRIMZI_TRUSTSTORE_LOCATION value: /etc/pki/java/cacerts - name: STRIMZI_TRUSTSTORE_PASSWORD value: changeit - name: STRIMZI_KAFKA_BOOTSTRAP_SERVERS value: my-kafka-cluster-.kafka-serverless.us-east-1.amazonaws.com:9098 # ...
-
Disables cluster configuration lookup for managed Kafka services that don’t allow topic configuration changes.
-
Defines the topic configuration properties that can be updated based on the limitations set by managed Kafka services.
-
Specifies the SASL properties to be set in JSON format. Only properties starting with
sasl.
are allowed.Example Dockerfile with external jarsFROM quay.io/strimzi/operator:latest USER root RUN mkdir -p ${STRIMZI_HOME}/external-libs RUN chmod +rx ${STRIMZI_HOME}/external-libs COPY ./aws-msk-iam-auth-and-dependencies/* ${STRIMZI_HOME}/external-libs/ ENV JAVA_CLASSPATH=${STRIMZI_HOME}/external-libs/* USER 1001
-
-
Apply the changes to the
Deployment
configuration to deploy the Topic Operator. -
Check the status of the deployment:
kubectl get deployments
Output shows the deployment name and readinessNAME READY UP-TO-DATE AVAILABLE strimzi-topic-operator 1/1 1 1
READY
shows the number of replicas that are ready/expected. The deployment is successful when theAVAILABLE
output shows1
.
7.7.2. Deploying the standalone User Operator
This procedure shows how to deploy the User Operator as a standalone component for user management. You can use a standalone User Operator with a Kafka cluster that is not managed by the Cluster Operator.
A standalone deployment can operate with any Kafka cluster.
Standalone deployment files are provided with Strimzi.
Use the 05-Deployment-strimzi-user-operator.yaml
deployment file to deploy the User Operator.
Add or set the environment variables needed to make a connection to a Kafka cluster.
The User Operator watches for KafkaUser
resources in a single namespace.
You specify the namespace to watch, and the connection to the Kafka cluster, in the User Operator configuration.
A single User Operator can watch a single namespace.
One namespace should be watched by only one User Operator.
If you want to use more than one User Operator, configure each of them to watch different namespaces.
In this way, you can use the User Operator with multiple Kafka clusters.
-
The standalone User Operator deployment files, which are included in the Strimzi deployment files.
-
You are running a Kafka cluster for the User Operator to connect to.
As long as the standalone User Operator is correctly configured for connection, the Kafka cluster can be running on a bare-metal environment, a virtual machine, or as a managed cloud application service.
-
Edit the following
env
properties in theinstall/user-operator/05-Deployment-strimzi-user-operator.yaml
standalone deployment file.Example standalone User Operator deployment configurationapiVersion: apps/v1 kind: Deployment metadata: name: strimzi-user-operator labels: app: strimzi spec: # ... template: # ... spec: # ... containers: - name: strimzi-user-operator # ... env: - name: STRIMZI_NAMESPACE (1) valueFrom: fieldRef: fieldPath: metadata.namespace - name: STRIMZI_KAFKA_BOOTSTRAP_SERVERS (2) value: my-kafka-bootstrap-address:9092 - name: STRIMZI_CA_CERT_NAME (3) value: my-cluster-clients-ca-cert - name: STRIMZI_CA_KEY_NAME (4) value: my-cluster-clients-ca - name: STRIMZI_LABELS (5) value: "strimzi.io/cluster=my-cluster" - name: STRIMZI_FULL_RECONCILIATION_INTERVAL_MS (6) value: "120000" - name: STRIMZI_WORK_QUEUE_SIZE (7) value: 10000 - name: STRIMZI_CONTROLLER_THREAD_POOL_SIZE (8) value: 10 - name: STRIMZI_USER_OPERATIONS_THREAD_POOL_SIZE (9) value: 4 - name: STRIMZI_LOG_LEVEL (10) value: INFO - name: STRIMZI_GC_LOG_ENABLED (11) value: "true" - name: STRIMZI_CA_VALIDITY (12) value: "365" - name: STRIMZI_CA_RENEWAL (13) value: "30" - name: STRIMZI_JAVA_OPTS (14) value: "-Xmx=512M -Xms=256M" - name: STRIMZI_JAVA_SYSTEM_PROPERTIES (15) value: "-Djavax.net.debug=verbose -DpropertyName=value" - name: STRIMZI_SECRET_PREFIX (16) value: "kafka-" - name: STRIMZI_ACLS_ADMIN_API_SUPPORTED (17) value: "true" - name: STRIMZI_MAINTENANCE_TIME_WINDOWS (18) value: '* * 8-10 * * ?;* * 14-15 * * ?' - name: STRIMZI_KAFKA_ADMIN_CLIENT_CONFIGURATION (19) value: | default.api.timeout.ms=120000 request.timeout.ms=60000
-
The Kubernetes namespace for the User Operator to watch for
KafkaUser
resources. Only one namespace can be specified. -
The host and port pair of the bootstrap broker address to discover and connect to all brokers in the Kafka cluster. Use a comma-separated list to specify two or three broker addresses in case a server is down.
-
The Kubernetes
Secret
that contains the public key (ca.crt
) value of the CA (certificate authority) that signs new user certificates for mTLS authentication. -
The Kubernetes
Secret
that contains the private key (ca.key
) value of the CA that signs new user certificates for mTLS authentication. -
The label to identify the
KafkaUser
resources managed by the User Operator. This does not have to be the name of the Kafka cluster. It can be the label assigned to theKafkaUser
resource. If you deploy more than one User Operator, the labels must be unique for each. That is, the operators cannot manage the same resources. -
The interval between periodic reconciliations, in milliseconds. The default is
120000
(2 minutes). -
The size of the controller event queue. The size of the queue should be at least as big as the maximal amount of users you expect the User Operator to operate. The default is
1024
. -
The size of the worker pool for reconciling the users. Bigger pool might require more resources, but it will also handle more
KafkaUser
resources The default is50
. -
The size of the worker pool for Kafka Admin API and Kubernetes operations. Bigger pool might require more resources, but it will also handle more
KafkaUser
resources The default is4
. -
The level for printing logging messages. You can set the level to
ERROR
,WARNING
,INFO
,DEBUG
, orTRACE
. -
Enables garbage collection (GC) logging. The default is
true
. -
The validity period for the CA. The default is
365
days. -
The renewal period for the CA. The renewal period is measured backwards from the expiry date of the current certificate. The default is
30
days to initiate certificate renewal before the old certificates expire. -
(Optional) The Java options used by the JVM running the User Operator
-
(Optional) The debugging (
-D
) options set for the User Operator -
(Optional) Prefix for the names of Kubernetes secrets created by the User Operator.
-
(Optional) Indicates whether the Kafka cluster supports management of authorization ACL rules using the Kafka Admin API. When set to
false
, the User Operator will reject all resources withsimple
authorization ACL rules. This helps to avoid unnecessary exceptions in the Kafka cluster logs. The default istrue
. -
(Optional) Semi-colon separated list of Cron Expressions defining the maintenance time windows during which the expiring user certificates will be renewed.
-
(Optional) Configuration options for configuring the Kafka Admin client used by the User Operator in the properties format.
-
-
If you are using mTLS to connect to the Kafka cluster, specify the secrets used to authenticate connection. Otherwise, go to the next step.
Example mTLS configuration# .... env: - name: STRIMZI_CLUSTER_CA_CERT_SECRET_NAME (1) value: my-cluster-cluster-ca-cert - name: STRIMZI_EO_KEY_SECRET_NAME (2) value: my-cluster-entity-operator-certs # ..."
-
The Kubernetes
Secret
that contains the public key (ca.crt
) value of the CA that signs Kafka broker certificates. -
The Kubernetes
Secret
that contains the certificate public key (entity-operator.crt
) and private key (entity-operator.key
) that is used for mTLS authentication against the Kafka cluster.
-
-
Deploy the User Operator.
kubectl create -f install/user-operator
-
Check the status of the deployment:
kubectl get deployments
Output shows the deployment name and readinessNAME READY UP-TO-DATE AVAILABLE strimzi-user-operator 1/1 1 1
READY
shows the number of replicas that are ready/expected. The deployment is successful when theAVAILABLE
output shows1
.
8. Deploying Strimzi from OperatorHub.io
OperatorHub.io is a catalog of Kubernetes operators sourced from multiple providers. It offers you an alternative way to install a stable version of Strimzi.
The Operator Lifecycle Manager is used for the installation and management of all operators published on OperatorHub.io. Operator Lifecycle Manager is a prerequisite for installing the Strimzi Kafka operator
To install Strimzi, locate Strimzi from OperatorHub.io, and follow the instructions provided to deploy the Cluster Operator.
After you have deployed the Cluster Operator, you can deploy Strimzi components using custom resources.
For example, you can deploy the Kafka
custom resource, and the installed Cluster Operator will create a Kafka cluster.
Upgrades between versions might include manual steps. Always read the release notes before upgrading.
For information on upgrades, see Upgrading Strimzi.
Warning
|
Make sure you use the appropriate update channel.
Installing Strimzi from the default stable channel is generally safe.
However, we do not recommend enabling automatic OLM updates on the stable channel.
An automatic upgrade will skip any necessary steps prior to upgrade.
For example, to upgrade from 0.22 or earlier
you must first update custom resources to support the v1beta2 API version.
Use automatic upgrades only on version-specific channels.
|
9. Deploying Strimzi using Helm
Helm charts are used to package, configure, and deploy Kubernetes resources. Strimzi provides a Helm chart to deploy the Cluster Operator.
After you have deployed the Cluster Operator this way, you can deploy Strimzi components using custom resources.
For example, you can deploy the Kafka
custom resource, and the installed Cluster Operator will create a Kafka cluster.
For information on upgrades, see Upgrading Strimzi.
-
The Helm client must be installed on a local machine.
-
Install the Strimzi Cluster Operator using the Helm command line tool:
helm install strimzi-cluster-operator oci://quay.io/strimzi-helm/strimzi-kafka-operator
Alternatively, you can use parameter values to install a specific version of the Cluster Operator or specify any changes to the default configuration.
Example configuration that installs a specific version of the Cluster Operator and changes the number of replicashelm install strimzi-cluster-operator --set replicas=2 --version 0.35.0 oci://quay.io/strimzi-helm/strimzi-kafka-operator
-
Verify that the Cluster Operator has been deployed successfully using the Helm command line tool:
helm ls
-
Deploy Kafka and other Kafka components using custom resources.
10. Feature gates
Strimzi operators use feature gates to enable or disable specific features and functions. Enabling a feature gate alters the behavior of the associated operator, introducing the corresponding feature to your Strimzi deployment.
The purpose of feature gates is to facilitate the trial and testing of a feature before it is fully adopted. The state (enabled or disabled) of a feature gate may vary by default, depending on its maturity level.
As a feature gate graduates and reaches General Availability (GA), it transitions to an enabled state by default and becomes a permanent part of the Strimzi deployment. A feature gate at the GA stage cannot be disabled.
The supported feature gates are applicable to all Strimzi operators. While a particular feature gate might be used by one operator and ignored by the others, it can still be configured in all operators. When deploying the User Operator and Topic Operator within the context of the`Kafka` custom resource, the Cluster Operator automatically propagates the feature gates configuration to them. When the User Operator and Topic Operator are deployed standalone, without a Cluster Operator available to configure the feature gates, they must be directly configured within their deployments.
10.1. Graduated feature gates (GA)
Graduated feature gates have reached General Availability (GA) and are permanently enabled features.
10.1.1. ControlPlaneListener feature gate
The ControlPlaneListener
feature gate separates listeners for data replication and coordination:
-
Connections between the Kafka controller and brokers use an internal control plane listener on port 9090.
-
Replication of data between brokers, as well as internal connections from Strimzi operators, Cruise Control, or the Kafka Exporter use a replication listener on port 9091.
Important
|
With the ControlPlaneListener feature gate permanently enabled, direct upgrades or downgrades between Strimzi 0.22 and earlier and Strimzi 0.32 and newer are not possible.
You must first upgrade or downgrade through one of the Strimzi versions in-between, disable the ControlPlaneListener feature gate, and then downgrade or upgrade (with the feature gate enabled) to the target version.
|
10.1.2. ServiceAccountPatching feature gate
The ServiceAccountPatching
feature gate ensures that the Cluster Operator always reconciles service accounts and updates them when needed.
For example, when you change service account labels or annotations using the template
property of a custom resource, the operator automatically updates them on the existing service account resources.
10.1.3. UseStrimziPodSets feature gate
The UseStrimziPodSets
feature gate introduced the StrimziPodSet
custom resource for managing Kafka and ZooKeeper pods, replacing the use of Kubernetes StatefulSet
resources.
Important
|
With the UseStrimziPodSets feature gate permanently enabled, direct downgrades from Strimzi 0.35 and newer to Strimzi 0.27 or earlier are not possible. You must first downgrade through one of the Strimzi versions in-between, disable the UseStrimziPodSets feature gate, and then downgrade to Strimzi 0.27 or earlier.
|
10.1.4. StableConnectIdentities feature gate
The StableConnectIdentities
feature gate introduced the StrimziPodSet
custom resource for managing Kafka Connect and Kafka MirrorMaker 2 pods, replacing the use of Kubernetes Deployment
resources.
StrimziPodSet
resources give the pods stable names and stable addresses, which do not change during rolling upgrades, replacing the use of Kubernetes Deployment
resources.
Important
|
With the StableConnectIdentities feature gate permanently enabled, direct downgrades from Strimzi 0.39 and newer to Strimzi 0.33 or earlier are not possible.
You must first downgrade through one of the Strimzi versions in-between, disable the StableConnectIdentities feature gate, and then downgrade to Strimzi 0.33 or earlier.
|
10.1.5. KafkaNodePools feature gate
The KafkaNodePools
feature gate introduced a new KafkaNodePool
custom resource that enables the configuration of different pools of Apache Kafka nodes.
A node pool refers to a distinct group of Kafka nodes within a Kafka cluster.
Each pool has its own unique configuration, which includes mandatory settings such as the number of replicas, storage configuration, and a list of assigned roles.
You can assign the controller role, broker role, or both roles to all nodes in the pool using the .spec.roles
property.
When used with a ZooKeeper-based Apache Kafka cluster, it must be set to the broker
role.
When used with a KRaft-based Apache Kafka cluster, it can be set to broker
, controller
, or both.
In addition, a node pool can have its own configuration of resource requests and limits, Java JVM options, and resource templates.
Configuration options not set in the KafkaNodePool
resource are inherited from the Kafka
custom resource.
The KafkaNodePool
resources use a strimzi.io/cluster
label to indicate to which Kafka cluster they belong.
The label must be set to the name of the Kafka
custom resource.
The Kafka
resource configuration must also include the strimzi.io/node-pools: enabled
annotation, which is required when using node pools.
Examples of the KafkaNodePool
resources can be found in the example configuration files provided by Strimzi.
If your cluster already uses KafkaNodePool
custom resources, and you wish to downgrade to an older version of Strimzi that does not support them or with the KafkaNodePools
feature gate disabled, you must first migrate from KafkaNodePool
custom resources to managing Kafka nodes using only Kafka
custom resources. For more information, see the instructions for reversing a migration to node pools.
10.1.6. UnidirectionalTopicOperator feature gate
The UnidirectionalTopicOperator
feature gate introduced a unidirectional topic management mode for creating Kafka topics using the KafkaTopic
resource.
Unidirectional mode is compatible with using KRaft for cluster management.
With unidirectional mode, you create Kafka topics using the KafkaTopic
resource, which are then managed by the Topic Operator.
Any configuration changes to a topic outside the KafkaTopic
resource are reverted.
For more information on topic management, see Topic management.
10.1.7. UseKRaft feature gate
The UseKRaft
feature gate introduced the KRaft (Kafka Raft metadata) mode for running Apache Kafka clusters without ZooKeeper.
ZooKeeper and KRaft are mechanisms used to manage metadata and coordinate operations in Kafka clusters.
KRaft mode eliminates the need for an external coordination service like ZooKeeper.
In KRaft mode, Kafka nodes take on the roles of brokers, controllers, or both.
They collectively manage the metadata, which is replicated across partitions.
Controllers are responsible for coordinating operations and maintaining the cluster’s state.
For more information on using KRraft, see Using Kafka in KRaft mode.
10.2. Stable feature gates (Beta)
Stable feature gates have reached a beta level of maturity, and are generally enabled by default for all users. Stable feature gates are production-ready, but they can still be disabled.
10.2.1. ContinueReconciliationOnManualRollingUpdateFailure feature gate
The ContinueReconciliationOnManualRollingUpdateFailure
feature gate has a default state of enabled.
The ContinueReconciliationOnManualRollingUpdateFailure
feature gate allows the Cluster Operator to continue a reconciliation if the manual rolling update of the operands fails.
It applies to the following operands that support manual rolling updates using the strimzi.io/manual-rolling-update
annotation:
-
ZooKeeper
-
Kafka
-
Kafka Connect
-
Kafka MirrorMaker 2
Continuing the reconciliation after a manual rolling update failure allows the operator to recover from various situations that might prevent the update from succeeding. For example, a missing Persistent Volume Claim (PVC) or Persistent Volume (PV) might cause the manual rolling update to fail. However, the PVCs and PVs are created only in a later stage of the reconciliation. By continuing the reconciliation after this failure, the process can recreate the missing PVC or PV and recover.
The ContinueReconciliationOnManualRollingUpdateFailure
feature gate is used by the Cluster Operator.
It is ignored by the User and Topic Operators.
To disable the ContinueReconciliationOnManualRollingUpdateFailure
feature gate, specify -ContinueReconciliationOnManualRollingUpdateFailure
in the STRIMZI_FEATURE_GATES
environment variable in the Cluster Operator configuration.
10.3. Early access feature gates (Alpha)
Early access feature gates have not yet reached the beta stage, and are disabled by default. An early access feature gate provides an opportunity for assessment before its functionality is permanently incorporated into Strimzi. Currently, there are no alpha level feature gates.
10.4. Enabling feature gates
To modify a feature gate’s default state, use the STRIMZI_FEATURE_GATES
environment variable in the operator’s configuration.
You can modify multiple feature gates using this single environment variable.
Specify a comma-separated list of feature gate names and prefixes.
A +
prefix enables the feature gate and a -
prefix disables it.
FeatureGate1
and disables FeatureGate2
env:
- name: STRIMZI_FEATURE_GATES
value: +FeatureGate1,-FeatureGate2
10.5. Feature gate releases
Feature gates have three stages of maturity:
-
Alpha — typically disabled by default
-
Beta — typically enabled by default
-
General Availability (GA) — typically always enabled
Alpha stage features might be experimental or unstable, subject to change, or not sufficiently tested for production use. Beta stage features are well tested and their functionality is not likely to change. GA stage features are stable and should not change in the future. Alpha and beta stage features are removed if they do not prove to be useful.
-
The
ControlPlaneListener
feature gate moved to GA stage in Strimzi 0.32. It is now permanently enabled and cannot be disabled. -
The
ServiceAccountPatching
feature gate moved to GA stage in Strimzi 0.30. It is now permanently enabled and cannot be disabled. -
The
UseStrimziPodSets
feature gate moved to GA stage in Strimzi 0.35 and the support for StatefulSets is completely removed. It is now permanently enabled and cannot be disabled. -
The
StableConnectIdentities
feature gate moved to GA stage in Strimzi 0.39. It is now permanently enabled and cannot be disabled. -
The
KafkaNodePools
feature gate moved to GA stage in Strimzi 0.41. It is now permanently enabled and cannot be disabled. To useKafkaNodePool
resources, you still need to use thestrimzi.io/node-pools: enabled
annotation on theKafka
custom resources. -
The
UnidirectionalTopicOperator
feature gate moved to GA stage in Strimzi 0.41. It is now permanently enabled and cannot be disabled. -
The
UseKRaft
feature gate moved to GA stage in Strimzi 0.42. It is now permanently enabled and cannot be disabled. To use KRaft (ZooKeeper-less Apache Kafka), you still need to use thestrimzi.io/kraft: enabled
annotation on theKafka
custom resources or migrate from an existing ZooKeeper-based cluster. -
The
ContinueReconciliationOnManualRollingUpdateFailure
feature was introduced in Strimzi 0.41 and moved to beta stage in Strimzi 0.44.0. It is now enabled by default, but can be disabled if needed.
Note
|
Feature gates might be removed when they reach GA. This means that the feature was incorporated into the Strimzi core features and can no longer be disabled. |
Feature gate | Alpha | Beta | GA |
---|---|---|---|
|
0.23 |
0.27 |
0.32 |
|
0.24 |
0.27 |
0.30 |
|
0.28 |
0.30 |
0.35 |
|
0.29 |
0.40 |
0.42 |
|
0.34 |
0.37 |
0.39 |
|
0.36 |
0.39 |
0.41 |
|
0.36 |
0.39 |
0.41 |
|
0.41 |
0.44 |
0.47 (planned) |
If a feature gate is enabled, you may need to disable it before upgrading or downgrading from a specific Strimzi version (or first upgrade / downgrade to a version of Strimzi where it can be disabled). The following table shows which feature gates you need to disable when upgrading or downgrading Strimzi versions.
Disable Feature gate | Upgrading from Strimzi version | Downgrading to Strimzi version |
---|---|---|
|
0.22 and earlier |
0.22 and earlier |
|
- |
0.27 and earlier |
|
- |
0.33 and earlier |
11. Configuring a deployment
Configure and manage a Strimzi deployment to your precise needs using Strimzi custom resources. Strimzi provides example custom resources with each release, allowing you to configure and create instances of supported Kafka components. Fine-tune your deployment by configuring custom resources to include additional features according to your specific requirements.
Use custom resources to configure and create instances of the following components:
-
Kafka clusters
-
Kafka Connect clusters
-
Kafka MirrorMaker
-
Kafka Bridge
-
Cruise Control
You can use configuration to manage your instances or modify your deployment to introduce additional features. New features are sometimes introduced through feature gates, which are controlled through operator configuration.
The Strimzi Custom Resource API Reference describes the properties you can use in your configuration.
Through configuration of the Kafka
resource, you can introduce the following:
-
Data storage
-
Rack awareness
-
Listeners for authenticated client access to the Kafka cluster
-
Topic Operator for managing Kafka topics
-
User Operator for managing Kafka users (clients)
-
Cruise Control for cluster rebalancing
-
Kafka Exporter for collecting lag metrics
Use KafkaNodePool
resources to configure distinct groups of nodes within a Kafka cluster.
Common configuration is configured independently for each component, such as the following:
-
Bootstrap servers for host/port connection to a Kafka cluster
-
Metrics configuration
-
Healthchecks and liveness probes
-
Resource limits and requests (CPU/Memory)
-
Logging frequency
-
JVM options for maximum and minimum memory allocation
-
Adding additional volumes and volume mounts
For specific areas of configuration, namely metrics, logging, and external configuration for Kafka Connect connectors, you can also use ConfigMap
resources.
By using a ConfigMap
resource to incorporate configuration, you centralize maintenance.
You can also use configuration providers to load configuration from external sources, which we recommend for supplying the credentials for Kafka Connect connector configuration.
When deploying Kafka, the Cluster Operator automatically sets up and renews TLS certificates to enable encryption and authentication within your cluster. If required, you can manually renew the cluster and clients CA certificates before their renewal period starts. You can also replace the keys used by the cluster and clients CA certificates. For more information, see Renewing CA certificates manually and Replacing private keys.
You add configuration to a custom resource using spec
properties.
After adding the configuration, you can use kubectl
to apply the changes to a custom resource configuration file:
kubectl apply -f <kafka_configuration_file>
Note
|
Labels applied to a custom resource are also applied to the Kubernetes resources making up its cluster. This provides a convenient mechanism for resources to be labeled as required. |
11.1. Using example configuration files
Further enhance your deployment by incorporating additional supported configuration.
Example configuration files are included in the Strimzi deployment files.
You can also access the example files directly from the
examples
directory.
The example files include only the essential properties and values for custom resources by default.
You can download and apply the examples using the kubectl
command-line tool.
The examples can serve as a starting point when building your own Kafka component configuration for deployment.
Note
|
If you installed Strimzi using the Operator, you can still download the example files and use them to upload configuration. |
The release artifacts include an examples
directory that contains the configuration examples.
examples
├── user (1)
├── topic (2)
├── security (3)
│ ├── tls-auth
│ ├── scram-sha-512-auth
│ └── keycloak-authorization
├── mirror-maker (4)
├── metrics (5)
├── kafka (6)
│ └── kraft (7)
├── cruise-control (8)
├── connect (9)
└── bridge (10)
-
KafkaUser
custom resource configuration, which is managed by the User Operator. -
KafkaTopic
custom resource configuration, which is managed by Topic Operator. -
Authentication and authorization configuration for Kafka components. Includes example configuration for TLS and SCRAM-SHA-512 authentication. The Keycloak example includes
Kafka
custom resource configuration and a Keycloak realm specification. You can use the example to try Keycloak authorization services. There is also an example with enabledoauth
authentication andkeycloak
authorization metrics. -
KafkaMirrorMaker
andKafkaMirrorMaker2
custom resource configurations for a deployment of MirrorMaker. Includes example configuration for replication policy and synchronization frequency. -
Metrics configuration, including Prometheus installation and Grafana dashboard files.
-
Kafka
andKafkaNodePool
custom resource configurations for a deployment of Kafka clusters that use ZooKeeper mode. Includes example configuration for an ephemeral or persistent single or multi-node deployment. -
Kafka
andKafkaNodePool
configurations for a deployment of Kafka clusters that use KRaft (Kafka Raft metadata) mode. -
Kafka
andKafkaRebalance
configurations for deploying and using Cruise Control to manage clusters.Kafka
configuration examples enable auto-rebalancing on scaling events and set default optimization goals.KakaRebalance
configuration examples set proposal-specific optimization goals and generate optimization proposals in various supported modes. -
KafkaConnect
andKafkaConnector
custom resource configuration for a deployment of Kafka Connect. Includes example configurations for a single or multi-node deployment. -
KafkaBridge
custom resource configuration for a deployment of Kafka Bridge.
11.2. Configuring Kafka in KRaft mode
Update the spec
properties of the Kafka
custom resource to configure your deployment of Kafka in KRaft mode.
As well as configuring Kafka, you can add configuration for Strimzi operators.
The KRaft metadata version (.spec.kafka.metadataVersion
) must be a version supported by the Kafka version (spec.kafka.version
).
If the metadata version is not set in the configuration, the Cluster Operator updates the version to the default for the Kafka version used.
Note
|
The oldest supported metadata version is 3.3. Using a metadata version that is older than the Kafka version might cause some features to be disabled. |
Kafka clusters operating in KRaft mode also use node pools. The following must be specified in the node pool configuration:
-
Roles assigned to each node within the Kafka cluster
-
Number of replica nodes used
-
Storage specification for the nodes
Other optional properties may also be set in node pools.
For a deeper understanding of the Kafka cluster configuration options, refer to the Strimzi Custom Resource API Reference.
Kafka
custom resource configuration# Basic configuration (required)
apiVersion: kafka.strimzi.io/v1beta2
kind: Kafka
metadata:
name: my-cluster
# Deployment specifications
spec:
kafka:
# Listener configuration (required)
listeners: # (1)
- name: plain # (2)
port: 9092 # (3)
type: internal # (4)
tls: false # (5)
configuration:
useServiceDnsDomain: true # (6)
- name: tls
port: 9093
type: internal
tls: true
authentication: # (7)
type: tls
- name: external1 # (8)
port: 9094
type: route
tls: true
configuration:
brokerCertChainAndKey: # (9)
secretName: my-secret
certificate: my-certificate.crt
key: my-key.key
# Kafka version (recommended)
version: 3.9.0 # (10)
# KRaft metadata version (recommended)
metadataVersion: 3.9 # (11)
# Kafka configuration (recommended)
config: # (12)
auto.create.topics.enable: "false"
offsets.topic.replication.factor: 3
transaction.state.log.replication.factor: 3
transaction.state.log.min.isr: 2
default.replication.factor: 3
min.insync.replicas: 2
# Resources requests and limits (recommended)
resources: # (13)
requests:
memory: 64Gi
cpu: "8"
limits:
memory: 64Gi
cpu: "12"
# Logging configuration (optional)
logging: # (14)
type: inline
loggers:
kafka.root.logger.level: INFO
# Readiness probe (optional)
readinessProbe: # (15)
initialDelaySeconds: 15
timeoutSeconds: 5
# Liveness probe (optional)
livenessProbe:
initialDelaySeconds: 15
timeoutSeconds: 5
# JVM options (optional)
jvmOptions: # (16)
-Xms: 8192m
-Xmx: 8192m
# Custom image (optional)
image: my-org/my-image:latest # (17)
# Authorization (optional)
authorization: # (18)
type: simple
# Rack awareness (optional)
rack: # (19)
topologyKey: topology.kubernetes.io/zone
# Metrics configuration (optional)
metricsConfig: # (20)
type: jmxPrometheusExporter
valueFrom:
configMapKeyRef: # (21)
name: my-config-map
key: my-key
# Entity Operator (recommended)
entityOperator: # (22)
topicOperator:
watchedNamespace: my-topic-namespace
reconciliationIntervalMs: 60000
# Resources requests and limits (recommended)
resources:
requests:
memory: 512Mi
cpu: "1"
limits:
memory: 512Mi
cpu: "1"
# Logging configuration (optional)
logging: # (23)
type: inline
loggers:
rootLogger.level: INFO
userOperator:
watchedNamespace: my-topic-namespace
reconciliationIntervalMs: 60000
# Resources requests and limits (recommended)
resources:
requests:
memory: 512Mi
cpu: "1"
limits:
memory: 512Mi
cpu: "1"
# Logging configuration (optional)
logging: # (24)
type: inline
loggers:
rootLogger.level: INFO
# Kafka Exporter (optional)
kafkaExporter: # (25)
# ...
# Cruise Control (optional)
cruiseControl: # (26)
# ...
-
Listeners configure how clients connect to the Kafka cluster via bootstrap addresses. Listeners are configured as internal or external listeners for connection from inside or outside the Kubernetes cluster.
-
Name to identify the listener. Must be unique within the Kafka cluster.
-
Port number used by the listener inside Kafka. The port number has to be unique within a given Kafka cluster. Allowed port numbers are 9092 and higher with the exception of ports 9404 and 9999, which are already used for Prometheus and JMX. Depending on the listener type, the port number might not be the same as the port number that connects Kafka clients.
-
Listener type specified as
internal
orcluster-ip
(to expose Kafka using per-brokerClusterIP
services), or for external listeners, asroute
(OpenShift only),loadbalancer
,nodeport
oringress
(Kubernetes only). -
Enables or disables TLS encryption for each listener. For
route
andingress
type listeners, TLS encryption must always be enabled by setting it totrue
. -
Defines whether the fully-qualified DNS names including the cluster service suffix (usually
.cluster.local
) are assigned. -
Listener authentication mechanism specified as mTLS, SCRAM-SHA-512, or token-based OAuth 2.0.
-
External listener configuration specifies how the Kafka cluster is exposed outside Kubernetes, such as through a
route
,loadbalancer
ornodeport
. -
Optional configuration for a Kafka listener certificate managed by an external CA (certificate authority). The
brokerCertChainAndKey
specifies aSecret
that contains a server certificate and a private key. You can configure Kafka listener certificates on any listener with enabled TLS encryption. -
Kafka version, which can be changed to a supported version by following the upgrade procedure.
-
Kafka metadata version, which can be changed to a supported version by following the upgrade procedure.
-
Broker configuration. Standard Apache Kafka configuration may be provided, restricted to those properties not managed directly by Strimzi.
-
Requests for reservation of supported resources, currently
cpu
andmemory
, and limits to specify the maximum resources that can be consumed. -
Kafka loggers and log levels added directly (
inline
) or indirectly (external
) through a ConfigMap. A custom Log4j configuration must be placed under thelog4j.properties
key in the ConfigMap. For the Kafkakafka.root.logger.level
logger, you can set the log level to INFO, ERROR, WARN, TRACE, DEBUG, FATAL or OFF. -
Healthchecks to know when to restart a container (liveness) and when a container can accept traffic (readiness).
-
JVM configuration options to optimize performance for the Virtual Machine (VM) running Kafka.
-
ADVANCED OPTION: Container image configuration, which is recommended only in special situations.
-
Authorization enables simple, OAUTH 2.0, or OPA authorization on the Kafka broker. Simple authorization uses the
AclAuthorizer
andStandardAuthorizer
Kafka plugins. -
Rack awareness configuration to spread replicas across different racks, data centers, or availability zones. The
topologyKey
must match a node label containing the rack ID. The example used in this configuration specifies a zone using the standardtopology.kubernetes.io/zone
label. -
Prometheus metrics enabled. In this example, metrics are configured for the Prometheus JMX Exporter (the default metrics exporter).
-
Rules for exporting metrics in Prometheus format to a Grafana dashboard through the Prometheus JMX Exporter, which are enabled by referencing a ConfigMap containing configuration for the Prometheus JMX exporter. You can enable metrics without further configuration using a reference to a ConfigMap containing an empty file under
metricsConfig.valueFrom.configMapKeyRef.key
. -
Entity Operator configuration, which specifies the configuration for the Topic Operator and User Operator.
-
Specified Topic Operator loggers and log levels. This example uses
inline
logging. -
Specified User Operator loggers and log levels.
-
Kafka Exporter configuration. Kafka Exporter is an optional component for extracting metrics data from Kafka brokers, in particular consumer lag data. For Kafka Exporter to be able to work properly, consumer groups need to be in use.
-
Optional configuration for Cruise Control, which is used to rebalance the Kafka cluster.
11.2.1. Setting throughput and storage limits on brokers
This procedure describes how to set throughput and storage limits on brokers in your Kafka cluster.
Enable a quota plugin and configure limits using quotas
properties in the Kafka
resource.
There are two types of quota plugins available:
-
The
strimzi
type enables the Strimzi Quotas plugin. -
The
kafka
type enables the built-in Kafka plugin.
Only one quota plugin can be enabled at a time.
The built-in kafka
plugin is enabled by default.
Enabling the strimzi
plugin automatically disables the built-in plugin.
strimzi
pluginThe strimzi
plugin provides storage utilization quotas and dynamic distribution of throughput limits.
-
Storage quotas throttle Kafka producers based on disk storage utilization. Limits can be specified in bytes (
minAvailableBytesPerVolume
) or percentage (minAvailableRatioPerVolume
) of available disk space, applying to each disk individually. When any broker in the cluster exceeds the configured disk threshold, clients are throttled to prevent disks from filling up too quickly and exceeding capacity. -
A total throughput limit is distributed dynamically across all clients. For example, if you set a 40 MBps producer byte-rate threshold, the distribution across two producers is not static. If one producer is using 10 MBps, the other can use up to 30 MBps.
-
Specific users (clients) can be excluded from the restrictions.
Note
|
With the strimzi plugin, you see only aggregated quota metrics, not per-client metrics.
|
kafka
pluginThe kafka
plugin applies throughput limits on a per-user, per-broker basis and includes additional CPU and operation rate limits.
-
Limits are applied per user and per broker. For example, setting a 20 MBps producer byte-rate threshold limits each user to 20 MBps on a per-broker basis across all producer connections for that user. There is no total throughput limit as there is in the
strimzi
plugin. Limits can be overridden by user-specific quota configurations. -
CPU utilization limits for each client can be set as a percentage of the network threads and I/O threads on a per-broker basis.
-
The number of concurrent partition creation and deletion operations (mutations) allowed per second can be set on a per-broker basis.
When using the default Kafka quotas plugin, the default quotas (if set) are applied to all users. This includes internal users such as the Topic Operator and Cruise Control, which may impact their operations. To avoid unduly limiting internal users, consider tuning the quotas effectively.
For example, a quota automatically applied to the Topic Operator by the Kafka quotas plugin could constrain the controller mutation rate, potentially throttling topic creation or deletion operations. Therefore, it is important to understand the minimal quotas required by the Topic Operator to function correctly and explicitly set appropriate quotas to avoid such issues. Monitoring relevant controller and broker metrics can help track and optimize the rate of operations on topics. Cruise Control and its metrics reporter also require sufficient produce and fetch rates to conduct rebalances, depending on the scale and configuration of the Kafka cluster. To prevent issues for Cruise Control, you might start with a rate of at least 1 KB/s for its producers and consumers in small clusters, such as three brokers with moderate traffic, and adjust as needed for larger or more active clusters.
-
The Cluster Operator that manages the Kafka cluster is running.
-
Add the plugin configuration to the
quotas
section of theKafka
resource.Examplestrimzi
plugin configurationapiVersion: kafka.strimzi.io/v1beta2 kind: Kafka metadata: name: my-cluster spec: kafka: # ... quotas: type: strimzi producerByteRate: 1000000 # (1) consumerByteRate: 1000000 # (2) minAvailableBytesPerVolume: 500000000000 # (3) excludedPrincipals: # (4) - my-user
-
Sets a producer byte-rate threshold of 1 MBps.
-
Sets a consumer byte-rate threshold of 1 MBps.
-
Sets an available bytes limit for storage of 500 GB.
-
Excludes
my-user
from the restrictions.
minAvailableBytesPerVolume
andminAvailableRatioPerVolume
are mutually exclusive. Only configure one of these parameters.Examplekafka
plugin configurationapiVersion: kafka.strimzi.io/v1beta2 kind: Kafka metadata: name: my-cluster spec: kafka: # ... quotas: type: kafka producerByteRate: 1000000 consumerByteRate: 1000000 requestPercentage: 55 # (1) controllerMutationRate: 50 # (2)
-
Sets the CPU utilization limit to 55%.
-
Sets the controller mutation rate to 50 operations per second.
-
-
Apply the changes to the
Kafka
configuration.
Note
|
Additional options can be configured in the spec.kafka.config section.
The full list of supported options can be found in the plugin documentation.
|
11.2.2. Deleting Kafka nodes using annotations
This procedure describes how to delete an existing Kafka node by using a Kubernetes annotation.
Deleting a Kafka node consists of deleting both the Pod
on which the Kafka broker is running and the related PersistentVolumeClaim
(if the cluster was deployed with persistent storage).
After deletion, the Pod
and its related PersistentVolumeClaim
are recreated automatically.
Warning
|
Deleting a PersistentVolumeClaim can cause permanent data loss and the availability of your cluster cannot be guaranteed.
The following procedure should only be performed if you have encountered storage issues.
|
-
A running Cluster Operator
-
Find the name of the
Pod
that you want to delete.Kafka broker pods are named
<cluster_name>-kafka-<index_number>
, where<index_number>
starts at zero and ends at the total number of replicas minus one. For example,my-cluster-kafka-0
. -
Use
kubectl annotate
to annotate thePod
resource in Kubernetes:kubectl annotate pod <cluster_name>-kafka-<index_number> strimzi.io/delete-pod-and-pvc="true"
-
Wait for the next reconciliation, when the annotated pod with the underlying persistent volume claim will be deleted and then recreated.
11.3. Configuring Kafka with ZooKeeper
Update the spec
properties of the Kafka
custom resource to configure your deployment of Kafka with ZooKeeper.
As well as configuring Kafka, you can add configuration for ZooKeeper and the Strimzi operators. The configuration options for Kafka and the Strimzi operators are the same as when using Kafka in KRaft mode. For descriptions of the properties, see Configuring Kafka in KRaft mode.
The inter-broker protocol version (inter.broker.protocol.version
) must be a version supported by the Kafka version (spec.kafka.version
).
If the inter-broker protocol version is not set in the configuration, the Cluster Operator updates the version to the default for the Kafka version used.
If you are also using node pools, the following must be specified in the node pool configuration:
-
Roles assigned to each node within the Kafka cluster
-
Number of replica nodes used
-
Storage specification for the nodes
If set in the node pool configuration, the equivalent configuration in the Kafka
resource, such as spec.kafka.replicas
, is not required.
Other optional properties may also be set in node pools.
For a deeper understanding of the ZooKeeper cluster configuration options, refer to the Strimzi Custom Resource API Reference.
Kafka
custom resource configuration when using ZooKeeper# Basic configuration (required)
apiVersion: kafka.strimzi.io/v1beta2
kind: Kafka
metadata:
name: my-cluster
# Deployment specifications
spec:
# Kafka configuration (required)
kafka:
# Replicas (required)
replicas: 3
# Listener configuration (required)
listeners:
- name: plain
port: 9092
type: internal
tls: false
configuration:
useServiceDnsDomain: true
- name: tls
port: 9093
type: internal
tls: true
authentication:
type: tls
- name: external1
port: 9094
type: route
tls: true
configuration:
brokerCertChainAndKey:
secretName: my-secret
certificate: my-certificate.crt
key: my-key.key
# Storage configuration (required)
storage:
type: persistent-claim
size: 10000Gi
# Kafka version (recommended)
version: 3.9.0
# Kafka configuration (recommended)
config:
auto.create.topics.enable: "false"
offsets.topic.replication.factor: 3
transaction.state.log.replication.factor: 3
transaction.state.log.min.isr: 2
default.replication.factor: 3
min.insync.replicas: 2
inter.broker.protocol.version: "3.9"
# Resources requests and limits (recommended)
resources:
requests:
memory: 64Gi
cpu: "8"
limits:
memory: 64Gi
cpu: "12"
# Logging configuration (optional)
logging:
type: inline
loggers:
kafka.root.logger.level: INFO
# Readiness probe (optional)
readinessProbe:
initialDelaySeconds: 15
timeoutSeconds: 5
# Liveness probe (optional)
livenessProbe:
initialDelaySeconds: 15
timeoutSeconds: 5
# JVM options (optional)
jvmOptions:
-Xms: 8192m
-Xmx: 8192m
# Custom image (optional)
image: my-org/my-image:latest
# Authorization (optional)
authorization:
type: simple
# Rack awareness (optional)
rack:
topologyKey: topology.kubernetes.io/zone
# Metrics configuration (optional)
metricsConfig:
type: jmxPrometheusExporter
valueFrom:
configMapKeyRef:
name: my-config-map
key: my-key
# ...
# ZooKeeper configuration (required)
zookeeper: # (1)
# Replicas (required)
replicas: 3 # (2)
# Storage configuration (required)
storage: # (3)
type: persistent-claim
size: 1000Gi
# Resources requests and limits (recommended)
resources: # (4)
requests:
memory: 8Gi
cpu: "2"
limits:
memory: 8Gi
cpu: "2"
# Logging configuration (optional)
logging: # (5)
type: inline
loggers:
zookeeper.root.logger: INFO
# JVM options (optional)
jvmOptions: # (6)
-Xms: 4096m
-Xmx: 4096m
# Metrics configuration (optional)
metricsConfig: # (7)
type: jmxPrometheusExporter
valueFrom:
configMapKeyRef: # (8)
name: my-config-map
key: my-key
# ...
# Entity operator (recommended)
entityOperator:
topicOperator:
# Resources requests and limits (recommended)
resources:
requests:
memory: 512Mi
cpu: "1"
limits:
memory: 512Mi
cpu: "1"
# Logging configuration (optional)
logging:
type: inline
loggers:
rootLogger.level: INFO
watchedNamespace: my-topic-namespace
reconciliationIntervalSeconds: 60
userOperator:
# Resources requests and limits (recommended)
resources:
requests:
memory: 512Mi
cpu: "1"
limits:
memory: 512Mi
cpu: "1"
# Logging configuration (optional)
logging:
type: inline
loggers:
rootLogger.level: INFO
watchedNamespace: my-topic-namespace
reconciliationIntervalSeconds: 60
# Kafka Exporter (optional)
kafkaExporter:
# ...
# Cruise Control (optional)
cruiseControl:
# ...
-
ZooKeeper-specific configuration contains properties similar to the Kafka configuration.
-
The number of ZooKeeper nodes. ZooKeeper clusters or ensembles usually run with an odd number of nodes, typically three, five, or seven. The majority of nodes must be available in order to maintain an effective quorum. If the ZooKeeper cluster loses its quorum, it will stop responding to clients and the Kafka brokers will stop working. Having a stable and highly available ZooKeeper cluster is crucial for Strimzi.
-
Storage size for persistent volumes may be increased and additional volumes may be added to JBOD storage.
-
Requests for reservation of supported resources, currently
cpu
andmemory
, and limits to specify the maximum resources that can be consumed. -
ZooKeeper loggers and log levels.
-
JVM configuration options to optimize performance for the Virtual Machine (VM) running ZooKeeper.
-
Prometheus metrics enabled. In this example, metrics are configured for the Prometheus JMX Exporter (the default metrics exporter).
-
Rules for exporting metrics in Prometheus format to a Grafana dashboard through the Prometheus JMX Exporter, which are enabled by referencing a ConfigMap containing configuration for the Prometheus JMX exporter. You can enable metrics without further configuration using a reference to a ConfigMap containing an empty file under
metricsConfig.valueFrom.configMapKeyRef.key
.
11.3.1. Default ZooKeeper configuration values
When deploying ZooKeeper with Strimzi, some of the default configuration set by Strimzi differs from the standard ZooKeeper defaults. This is because Strimzi sets a number of ZooKeeper properties with values that are optimized for running ZooKeeper within a Kubernetes environment.
The default configuration for key ZooKeeper properties in Strimzi is as follows:
Property | Default value | Description |
---|---|---|
|
2000 |
The length of a single tick in milliseconds, which determines the length of a session timeout. |
|
5 |
The maximum number of ticks that a follower is allowed to fall behind the leader in a ZooKeeper cluster. |
|
2 |
The maximum number of ticks that a follower is allowed to be out of sync with the leader in a ZooKeeper cluster. |
|
1 |
Enables the |
|
false |
Flag to disable the ZooKeeper admin server. The admin server is not used by Strimzi. |
Important
|
Modifying these default values as zookeeper.config in the Kafka custom resource may impact the behavior and performance of your ZooKeeper cluster.
|
11.3.2. Deleting ZooKeeper nodes using annotations
This procedure describes how to delete an existing ZooKeeper node by using a Kubernetes annotation.
Deleting a ZooKeeper node consists of deleting both the Pod
on which ZooKeeper is running and the related PersistentVolumeClaim
(if the cluster was deployed with persistent storage).
After deletion, the Pod
and its related PersistentVolumeClaim
are recreated automatically.
Warning
|
Deleting a PersistentVolumeClaim can cause permanent data loss and the availability of your cluster cannot be guaranteed.
The following procedure should only be performed if you have encountered storage issues.
|
-
A running Cluster Operator
-
Find the name of the
Pod
that you want to delete.ZooKeeper pods are named
<cluster_name>-zookeeper-<index_number>
, where<index_number>
starts at zero and ends at the total number of replicas minus one. For example,my-cluster-zookeeper-0
. -
Use
kubectl annotate
to annotate thePod
resource in Kubernetes:kubectl annotate pod <cluster_name>-zookeeper-<index_number> strimzi.io/delete-pod-and-pvc="true"
-
Wait for the next reconciliation, when the annotated pod with the underlying persistent volume claim will be deleted and then recreated.
11.4. Configuring node pools
Update the spec
properties of the KafkaNodePool
custom resource to configure a node pool deployment.
A node pool refers to a distinct group of Kafka nodes within a Kafka cluster. Each pool has its own unique configuration, which includes mandatory settings for the number of replicas, roles, and storage allocation.
Optionally, you can also specify values for the following properties:
-
resources
to specify memory and cpu requests and limits -
template
to specify custom configuration for pods and other Kubernetes resources -
jvmOptions
to specify custom JVM configuration for heap size, runtime and other options
The relationship between Kafka
and KafkaNodePool
resources is as follows:
-
Kafka
resources represent the configuration for all nodes in a Kafka cluster. -
KafkaNodePool
resources represent the configuration for nodes only in the node pool.
If a configuration property is not specified in KafkaNodePool
, it is inherited from the Kafka
resource.
Configuration specified in the KafkaNodePool
resource takes precedence if set in both resources.
For example, if both the node pool and Kafka configuration includes jvmOptions
, the values specified in the node pool configuration are used.
When -Xmx: 1024m
is set in KafkaNodePool.spec.jvmOptions
and -Xms: 512m
is set in Kafka.spec.kafka.jvmOptions
, the node uses the value from its node pool configuration.
Properties from Kafka
and KafkaNodePool
schemas are not combined.
To clarify, if KafkaNodePool.spec.template
includes only podSet.metadata.labels
, and Kafka.spec.kafka.template
includes podSet.metadata.annotations
and pod.metadata.labels
, the template values from the Kafka configuration are ignored since there is a template value in the node pool configuration.
For a deeper understanding of the node pool configuration options, refer to the Strimzi Custom Resource API Reference.
# Basic configuration (required)
apiVersion: kafka.strimzi.io/v1beta2
kind: KafkaNodePool
metadata:
name: kraft-dual-role # (1)
labels:
strimzi.io/cluster: my-cluster # (2)
# Node pool specifications
spec:
# Replicas (required)
replicas: 3 # (3)
# Roles (required)
roles: # (4)
- controller
- broker
# Storage configuration (required)
storage: # (5)
type: jbod
volumes:
- id: 0
type: persistent-claim
size: 100Gi
deleteClaim: false
# Resources requests and limits (recommended)
resources: # (6)
requests:
memory: 64Gi
cpu: "8"
limits:
memory: 64Gi
cpu: "12"
-
Unique name for the node pool.
-
The Kafka cluster the node pool belongs to. A node pool can only belong to a single cluster.
-
Number of replicas for the nodes.
-
Roles for the nodes in the node pool. In this example, the nodes have dual roles as controllers and brokers.
-
Storage specification for the nodes.
-
Requests for reservation of supported resources, currently
cpu
andmemory
, and limits to specify the maximum resources that can be consumed.
Note
|
The configuration for the Kafka resource must be suitable for KRaft mode. Currently, KRaft mode has a number of limitations.
|
apiVersion: kafka.strimzi.io/v1beta2
kind: KafkaNodePool
metadata:
name: pool-a
labels:
strimzi.io/cluster: my-cluster
spec:
replicas: 3
roles:
- broker # (1)
storage:
type: jbod
volumes:
- id: 0
type: persistent-claim
size: 100Gi
deleteClaim: false
resources:
requests:
memory: 64Gi
cpu: "8"
limits:
memory: 64Gi
cpu: "12"
-
Roles for the nodes in the node pool, which can only be
broker
when using Kafka with ZooKeeper.
11.4.1. Assigning IDs to node pools for scaling operations
This procedure describes how to use annotations for advanced node ID handling by the Cluster Operator when performing scaling operations on node pools. You specify the node IDs to use, rather than the Cluster Operator using the next ID in sequence. Management of node IDs in this way gives greater control.
To add a range of IDs, you assign the following annotations to the KafkaNodePool
resource:
-
strimzi.io/next-node-ids
to add a range of IDs that are used for new brokers -
strimzi.io/remove-node-ids
to add a range of IDs for removing existing brokers
You can specify an array of individual node IDs, ID ranges, or a combination of both.
For example, you can specify the following range of IDs: [0, 1, 2, 10-20, 30]
for scaling up the Kafka node pool.
This format allows you to specify a combination of individual node IDs (0
, 1
, 2
, 30
) as well as a range of IDs (10-20
).
In a typical scenario, you might specify a range of IDs for scaling up and a single node ID to remove a specific node when scaling down.
In this procedure, we add the scaling annotations to node pools as follows:
-
pool-a
is assigned a range of IDs for scaling up -
pool-b
is assigned a range of IDs for scaling down
During the scaling operation, IDs are used as follows:
-
Scale up picks up the lowest available ID in the range for the new node.
-
Scale down removes the node with the highest available ID in the range.
If there are gaps in the sequence of node IDs assigned in the node pool, the next node to be added is assigned an ID that fills the gap.
The annotations don’t need to be updated after every scaling operation. Any unused IDs are still valid for the next scaling event.
The Cluster Operator allows you to specify a range of IDs in either ascending or descending order, so you can define them in the order the nodes are scaled.
For example, when scaling up, you can specify a range such as [1000-1999]
, and the new nodes are assigned the next lowest IDs: 1000
, 1001
, 1002
, 1003
, and so on.
Conversely, when scaling down, you can specify a range like [1999-1000]
, ensuring that nodes with the next highest IDs are removed: 1003
, 1002
, 1001
, 1000
, and so on.
If you don’t specify an ID range using the annotations, the Cluster Operator follows its default behavior for handling IDs during scaling operations. Node IDs start at 0 (zero) and run sequentially across the Kafka cluster. The next lowest ID is assigned to a new node. Gaps to node IDs are filled across the cluster. This means that they might not run sequentially within a node pool. The default behavior for scaling up is to add the next lowest available node ID across the cluster; and for scaling down, it is to remove the node in the node pool with the highest available node ID. The default approach is also applied if the assigned range of IDs is misformatted, the scaling up range runs out of IDs, or the scaling down range does not apply to any in-use nodes.
-
(Optional) Use the
reserved.broker-max.id
configuration property to extend the allowable range for node IDs within your node pools.
By default, Apache Kafka restricts node IDs to numbers ranging from 0 to 999.
To use node ID values greater than 999, add the reserved.broker-max.id
configuration property to the Kafka
custom resource and specify the required maximum node ID value.
In this example, the maximum node ID is set at 10000. Node IDs can then be assigned up to that value.
apiVersion: kafka.strimzi.io/v1beta2
kind: Kafka
metadata:
name: my-cluster
spec:
kafka:
config:
reserved.broker.max.id: 10000
# ...
-
Annotate the node pool with the IDs to use when scaling up or scaling down, as shown in the following examples.
IDs for scaling up are assigned to node pool
pool-a
:Assigning IDs for scaling upkubectl annotate kafkanodepool pool-a strimzi.io/next-node-ids="[0,1,2,10-20,30]"
The lowest available ID from this range is used when adding a node to
pool-a
.IDs for scaling down are assigned to node pool
pool-b
:Assigning IDs for scaling downkubectl annotate kafkanodepool pool-b strimzi.io/remove-node-ids="[60-50,9,8,7]"
The highest available ID from this range is removed when scaling down
pool-b
.NoteIf you want to remove a specific node, you can assign a single node ID to the scaling down annotation: kubectl annotate kafkanodepool pool-b strimzi.io/remove-node-ids="[3]"
. -
You can now scale the node pool.
For more information, see the following:
On reconciliation, a warning is given if the annotations are misformatted.
-
After you have performed the scaling operation, you can remove the annotation if it’s no longer needed.
Removing the annotation for scaling upkubectl annotate kafkanodepool pool-a strimzi.io/next-node-ids-
Removing the annotation for scaling downkubectl annotate kafkanodepool pool-b strimzi.io/remove-node-ids-
11.4.2. Impact on racks when moving nodes from node pools
If rack awareness is enabled on a Kafka cluster, replicas can be spread across different racks, data centers, or availability zones. When moving nodes from node pools, consider the implications on the cluster topology, particularly regarding rack awareness. Removing specific pods from node pools, especially out of order, may break the cluster topology or cause an imbalance in distribution across racks. An imbalance can impact both the distribution of nodes themselves and the partition replicas within the cluster. An uneven distribution of nodes and partitions across racks can affect the performance and resilience of the Kafka cluster.
Plan the removal of nodes strategically to maintain the required balance and resilience across racks.
Use the strimzi.io/remove-node-ids
annotation to move nodes with specific IDs with caution.
Ensure that configuration to spread partition replicas across racks and for clients to consume from the closest replicas is not broken.
Tip
|
Use Cruise Control and the KafkaRebalance resource with the RackAwareGoal to make sure that replicas remain distributed across different racks.
|
11.4.3. Adding nodes to a node pool
This procedure describes how to scale up a node pool to add new nodes. Currently, scale up is only possible for broker-only node pools containing nodes that run as dedicated brokers.
In this procedure, we start with three nodes for node pool pool-a
:
NAME READY STATUS RESTARTS
my-cluster-pool-a-0 1/1 Running 0
my-cluster-pool-a-1 1/1 Running 0
my-cluster-pool-a-2 1/1 Running 0
Node IDs are appended to the name of the node on creation.
We add node my-cluster-pool-a-3
, which has a node ID of 3
.
Note
|
During this process, the ID of the node that holds the partition replicas changes. Consider any dependencies that reference the node ID. |
-
(Optional) Auto-rebalancing is enabled.
If auto-rebalancing is enabled, partition reassignment happens automatically during the node scaling process, so you don’t need to manually initiate the reassignment through Cruise Control. -
(Optional) For scale up operations, you can specify the node IDs to use in the operation.
If you have assigned a range of node IDs for the operation, the ID of the node being added is determined by the sequence of nodes given. If you have assigned a single node ID, a node is added with the specified ID. Otherwise, the lowest available node ID across the cluster is used.
-
Create a new node in the node pool.
For example, node pool
pool-a
has three replicas. We add a node by increasing the number of replicas:kubectl scale kafkanodepool pool-a --replicas=4
-
Check the status of the deployment and wait for the pods in the node pool to be created and ready (
1/1
).kubectl get pods -n <my_cluster_operator_namespace>
Output shows four Kafka nodes in the node poolNAME READY STATUS RESTARTS my-cluster-pool-a-0 1/1 Running 0 my-cluster-pool-a-1 1/1 Running 0 my-cluster-pool-a-2 1/1 Running 0 my-cluster-pool-a-3 1/1 Running 0
-
Reassign the partitions after increasing the number of nodes in the node pool.
-
If auto-rebalancing is enabled, partitions are reassigned to new nodes automatically, so you can skip this step.
-
If auto-rebalancing is not enabled, use the Cruise Control
add-brokers
mode to move partition replicas from existing brokers to the newly added brokers.Using Cruise Control to reassign partition replicasapiVersion: kafka.strimzi.io/v1beta2 kind: KafkaRebalance metadata: # ... spec: mode: add-brokers brokers: [3]
We are reassigning partitions to node
my-cluster-pool-a-3
. The reassignment can take some time depending on the number of topics and partitions in the cluster.
-
11.4.4. Removing nodes from a node pool
This procedure describes how to scale down a node pool to remove nodes. Currently, scale down is only possible for broker-only node pools containing nodes that run as dedicated brokers.
In this procedure, we start with four nodes for node pool pool-a
:
NAME READY STATUS RESTARTS
my-cluster-pool-a-0 1/1 Running 0
my-cluster-pool-a-1 1/1 Running 0
my-cluster-pool-a-2 1/1 Running 0
my-cluster-pool-a-3 1/1 Running 0
Node IDs are appended to the name of the node on creation.
We remove node my-cluster-pool-a-3
, which has a node ID of 3
.
Note
|
During this process, the ID of the node that holds the partition replicas changes. Consider any dependencies that reference the node ID. |
-
(Optional) Auto-rebalancing is enabled.
If auto-rebalancing is enabled, partition reassignment happens automatically during the node scaling process, so you don’t need to manually initiate the reassignment through Cruise Control. -
(Optional) For scale down operations, you can specify the node IDs to use in the operation.
If you have assigned a range of node IDs for the operation, the ID of the node being removed is determined by the sequence of nodes given. If you have assigned a single node ID, the node with the specified ID is removed. Otherwise, the node with the highest available ID in the node pool is removed.
-
Reassign the partitions before decreasing the number of nodes in the node pool.
-
If auto-rebalancing is enabled, partitions are moved off brokers that are going to be removed automatically, so you can skip this step.
-
If auto-rebalancing is not enabled, use the Cruise Control
remove-brokers
mode to move partition replicas off the brokers that are going to be removed.Using Cruise Control to reassign partition replicasapiVersion: kafka.strimzi.io/v1beta2 kind: KafkaRebalance metadata: # ... spec: mode: remove-brokers brokers: [3]
We are reassigning partitions from node
my-cluster-pool-a-3
. The reassignment can take some time depending on the number of topics and partitions in the cluster.
-
-
After the reassignment process is complete, and the node being removed has no live partitions, reduce the number of Kafka nodes in the node pool.
For example, node pool
pool-a
has four replicas. We remove a node by decreasing the number of replicas:kubectl scale kafkanodepool pool-a --replicas=3
Output shows three Kafka nodes in the node poolNAME READY STATUS RESTARTS my-cluster-pool-b-kafka-0 1/1 Running 0 my-cluster-pool-b-kafka-1 1/1 Running 0 my-cluster-pool-b-kafka-2 1/1 Running 0
11.4.5. Moving nodes between node pools
This procedure describes how to move nodes between source and target Kafka node pools without downtime. You create a new node on the target node pool and reassign partitions to move data from the old node on the source node pool. When the replicas on the new node are in-sync, you can delete the old node.
In this procedure, we start with two node pools:
-
pool-a
with three replicas is the target node pool -
pool-b
with four replicas is the source node pool
We scale up pool-a
, and reassign partitions and scale down pool-b
, which results in the following:
-
pool-a
with four replicas -
pool-b
with three replicas
Currently, scaling is only possible for broker-only node pools containing nodes that run as dedicated brokers.
Note
|
During this process, the ID of the node that holds the partition replicas changes. Consider any dependencies that reference the node ID. |
-
(Optional) Auto-rebalancing is enabled.
If auto-rebalancing is enabled, partition reassignment happens automatically during the node scaling process, so you don’t need to manually initiate the reassignment through Cruise Control. -
(Optional) For scale up and scale down operations, you can specify the range of node IDs to use.
If you have assigned node IDs for the operation, the ID of the node being added or removed is determined by the sequence of nodes given. Otherwise, the lowest available node ID across the cluster is used when adding nodes; and the node with the highest available ID in the node pool is removed.
-
Create a new node in the target node pool.
For example, node pool
pool-a
has three replicas. We add a node by increasing the number of replicas:kubectl scale kafkanodepool pool-a --replicas=4
-
Check the status of the deployment and wait for the pods in the node pool to be created and ready (
1/1
).kubectl get pods -n <my_cluster_operator_namespace>
Output shows four Kafka nodes in the source and target node poolsNAME READY STATUS RESTARTS my-cluster-pool-a-0 1/1 Running 0 my-cluster-pool-a-1 1/1 Running 0 my-cluster-pool-a-4 1/1 Running 0 my-cluster-pool-a-7 1/1 Running 0 my-cluster-pool-b-2 1/1 Running 0 my-cluster-pool-b-3 1/1 Running 0 my-cluster-pool-b-5 1/1 Running 0 my-cluster-pool-b-6 1/1 Running 0
Node IDs are appended to the name of the node on creation. We add node
my-cluster-pool-a-7
, which has a node ID of7
.If auto-rebalancing is enabled, partitions are reassigned to new nodes and moved off brokers that are going to be removed automatically, so you can skip the next step.
-
If auto-rebalancing is not enabled, reassign partitions before decreasing the number of nodes in the source node pool.
Use the Cruise Control
remove-brokers
mode to move partition replicas off the brokers that are going to be removed.Using Cruise Control to reassign partition replicasapiVersion: kafka.strimzi.io/v1beta2 kind: KafkaRebalance metadata: # ... spec: mode: remove-brokers brokers: [6]
We are reassigning partitions from node
my-cluster-pool-b-6
. The reassignment can take some time depending on the number of topics and partitions in the cluster. -
After the reassignment process is complete, reduce the number of Kafka nodes in the source node pool.
For example, node pool
pool-b
has four replicas. We remove a node by decreasing the number of replicas:kubectl scale kafkanodepool pool-b --replicas=3
The node with the highest ID (
6
) within the pool is removed.Output shows three Kafka nodes in the source node poolNAME READY STATUS RESTARTS my-cluster-pool-b-kafka-2 1/1 Running 0 my-cluster-pool-b-kafka-3 1/1 Running 0 my-cluster-pool-b-kafka-5 1/1 Running 0
11.4.6. Changing node pool roles
Node pools can be used with Kafka clusters that operate in KRaft mode (using Kafka Raft metadata) or use ZooKeeper for metadata management. If you are using KRaft mode, you can specify roles for all nodes in the node pool to operate as brokers, controllers, or both. If you are using ZooKeeper, nodes must be set as brokers only.
In certain circumstances you might want to change the roles assigned to a node pool. For example, you may have a node pool that contains nodes that perform dual broker and controller roles, and then decide to split the roles between two node pools. In this case, you create a new node pool with nodes that act only as brokers, and then reassign partitions from the dual-role nodes to the new brokers. You can then switch the old node pool to a controller-only role.
You can also perform the reverse operation by moving from node pools with controller-only and broker-only roles to a node pool that contains nodes that perform dual broker and controller roles.
In this case, you add the broker
role to the existing controller-only node pool, reassign partitions from the broker-only nodes to the dual-role nodes, and then delete the broker-only node pool.
When removing broker
roles in the node pool configuration, keep in mind that Kafka does not automatically reassign partitions.
Before removing the broker role, ensure that nodes changing to controller-only roles do not have any assigned partitions.
If partitions are assigned, the change is prevented.
No replicas must be left on the node before removing the broker role.
The best way to reassign partitions before changing roles is to apply a Cruise Control optimization proposal in remove-brokers
mode.
For more information, see Generating optimization proposals.
11.4.7. Transitioning to separate broker and controller roles
This procedure describes how to transition to using node pools with separate roles. If your Kafka cluster is using a node pool with combined controller and broker roles, you can transition to using two node pools with separate roles. To do this, rebalance the cluster to move partition replicas to a node pool with a broker-only role, and then switch the old node pool to a controller-only role.
In this procedure, we start with node pool pool-a
, which has controller
and broker
roles:
apiVersion: kafka.strimzi.io/v1beta2
kind: KafkaNodePool
metadata:
name: pool-a
labels:
strimzi.io/cluster: my-cluster
spec:
replicas: 3
roles:
- controller
- broker
storage:
type: jbod
volumes:
- id: 0
type: persistent-claim
size: 20Gi
deleteClaim: false
# ...
The node pool has three nodes:
NAME READY STATUS RESTARTS
my-cluster-pool-a-0 1/1 Running 0
my-cluster-pool-a-1 1/1 Running 0
my-cluster-pool-a-2 1/1 Running 0
Each node performs a combined role of broker and controller.
We create a second node pool called pool-b
, with three nodes that act as brokers only.
Note
|
During this process, the ID of the node that holds the partition replicas changes. Consider any dependencies that reference the node ID. |
-
Create a node pool with a
broker
role.Example node pool configurationapiVersion: kafka.strimzi.io/v1beta2 kind: KafkaNodePool metadata: name: pool-b labels: strimzi.io/cluster: my-cluster spec: replicas: 3 roles: - broker storage: type: jbod volumes: - id: 0 type: persistent-claim size: 100Gi deleteClaim: false # ...
The new node pool also has three nodes. If you already have a broker-only node pool, you can skip this step.
-
Apply the new
KafkaNodePool
resource to create the brokers. -
Check the status of the deployment and wait for the pods in the node pool to be created and ready (
1/1
).kubectl get pods -n <my_cluster_operator_namespace>
Output shows pods running in two node poolsNAME READY STATUS RESTARTS my-cluster-pool-a-0 1/1 Running 0 my-cluster-pool-a-1 1/1 Running 0 my-cluster-pool-a-2 1/1 Running 0 my-cluster-pool-b-3 1/1 Running 0 my-cluster-pool-b-4 1/1 Running 0 my-cluster-pool-b-5 1/1 Running 0
Node IDs are appended to the name of the node on creation.
-
Use the Cruise Control
remove-brokers
mode to reassign partition replicas from the dual-role nodes to the newly added brokers.Using Cruise Control to reassign partition replicasapiVersion: kafka.strimzi.io/v1beta2 kind: KafkaRebalance metadata: # ... spec: mode: remove-brokers brokers: [0, 1, 2]
The reassignment can take some time depending on the number of topics and partitions in the cluster.
NoteIf nodes changing to controller-only roles have any assigned partitions, the change is prevented. The status.conditions
of theKafka
resource provide details of events preventing the change. -
Remove the
broker
role from the node pool that originally had a combined role.Dual-role nodes switched to controllersapiVersion: kafka.strimzi.io/v1beta2 kind: KafkaNodePool metadata: name: pool-a labels: strimzi.io/cluster: my-cluster spec: replicas: 3 roles: - controller storage: type: jbod volumes: - id: 0 type: persistent-claim size: 20Gi deleteClaim: false # ...
-
Apply the configuration change so that the node pool switches to a controller-only role.
11.4.8. Transitioning to dual-role nodes
This procedure describes how to transition from separate node pools with broker-only and controller-only roles to using a dual-role node pool.
If your Kafka cluster is using node pools with dedicated controller and broker nodes, you can transition to using a single node pool with both roles.
To do this, add the broker
role to the controller-only node pool, rebalance the cluster to move partition replicas to the dual-role node pool, and then delete the old broker-only node pool.
In this procedure, we start with two node pools pool-a
, which has only the controller
role and pool-b
which has only the broker
role:
apiVersion: kafka.strimzi.io/v1beta2
kind: KafkaNodePool
metadata:
name: pool-a
labels:
strimzi.io/cluster: my-cluster
spec:
replicas: 3
roles:
- controller
storage:
type: jbod
volumes:
- id: 0
type: persistent-claim
size: 100Gi
deleteClaim: false
# ...
---
apiVersion: kafka.strimzi.io/v1beta2
kind: KafkaNodePool
metadata:
name: pool-b
labels:
strimzi.io/cluster: my-cluster
spec:
replicas: 3
roles:
- broker
storage:
type: jbod
volumes:
- id: 0
type: persistent-claim
size: 100Gi
deleteClaim: false
# ...
The Kafka cluster has six nodes:
NAME READY STATUS RESTARTS
my-cluster-pool-a-0 1/1 Running 0
my-cluster-pool-a-1 1/1 Running 0
my-cluster-pool-a-2 1/1 Running 0
my-cluster-pool-b-3 1/1 Running 0
my-cluster-pool-b-4 1/1 Running 0
my-cluster-pool-b-5 1/1 Running 0
The pool-a
nodes perform the role of controller.
The pool-b
nodes perform the role of broker.
Note
|
During this process, the ID of the node that holds the partition replicas changes. Consider any dependencies that reference the node ID. |
-
Edit the node pool
pool-a
and add thebroker
role to it.Example node pool configurationapiVersion: kafka.strimzi.io/v1beta2 kind: KafkaNodePool metadata: name: pool-a labels: strimzi.io/cluster: my-cluster spec: replicas: 3 roles: - controller - broker storage: type: jbod volumes: - id: 0 type: persistent-claim size: 100Gi deleteClaim: false # ...
-
Check the status and wait for the pods in the node pool to be restarted and ready (
1/1
).kubectl get pods -n <my_cluster_operator_namespace>
Output shows pods running in two node poolsNAME READY STATUS RESTARTS my-cluster-pool-a-0 1/1 Running 0 my-cluster-pool-a-1 1/1 Running 0 my-cluster-pool-a-2 1/1 Running 0 my-cluster-pool-b-3 1/1 Running 0 my-cluster-pool-b-4 1/1 Running 0 my-cluster-pool-b-5 1/1 Running 0
Node IDs are appended to the name of the node on creation.
-
Use the Cruise Control
remove-brokers
mode to reassign partition replicas from the broker-only nodes to the dual-role nodes.Using Cruise Control to reassign partition replicasapiVersion: kafka.strimzi.io/v1beta2 kind: KafkaRebalance metadata: # ... spec: mode: remove-brokers brokers: [3, 4, 5]
The reassignment can take some time depending on the number of topics and partitions in the cluster.
-
Remove the
pool-b
node pool that has the old broker-only nodes.kubectl delete kafkanodepool pool-b -n <my_cluster_operator_namespace>
11.4.9. Migrating existing Kafka clusters to use Kafka node pools
This procedure describes how to migrate existing Kafka clusters to use Kafka node pools. After you have updated the Kafka cluster, you can use the node pools to manage the configuration of nodes within each pool.
Note
|
Currently, replica and storage configuration in the KafkaNodePool resource must also be present in the Kafka resource. The configuration is ignored when node pools are being used.
|
-
Create a new
KafkaNodePool
resource.-
Name the resource
kafka
. -
Point a
strimzi.io/cluster
label to your existingKafka
resource. -
Set the replica count and storage configuration to match your current Kafka cluster.
-
Set the roles to
broker
.
Example configuration for a node pool used in migrating a Kafka clusterapiVersion: kafka.strimzi.io/v1beta2 kind: KafkaNodePool metadata: name: kafka labels: strimzi.io/cluster: my-cluster spec: replicas: 3 roles: - broker storage: type: jbod volumes: - id: 0 type: persistent-claim size: 100Gi deleteClaim: false
WarningTo preserve cluster data and the names of its nodes and resources, the node pool name must be kafka
, and thestrimzi.io/cluster
label matches the Kafka resource name. Otherwise, nodes and resources are created with new names, including the persistent volume storage used by the nodes. Consequently, your previous data may not be available. -
-
Apply the
KafkaNodePool
resource:kubectl apply -f <node_pool_configuration_file>
By applying this resource, you switch Kafka to using node pools.
There is no change or rolling update and resources are identical to how they were before.
-
Enable support for node pools in the
Kafka
resource using thestrimzi.io/node-pools: enabled
annotation.Example configuration for a node pool in a cluster using ZooKeeperapiVersion: kafka.strimzi.io/v1beta2 kind: Kafka metadata: name: my-cluster annotations: strimzi.io/node-pools: enabled spec: kafka: # ... zookeeper: # ...
-
Apply the
Kafka
resource:kubectl apply -f <kafka_configuration_file>
There is no change or rolling update. The resources remain identical to how they were before.
-
Remove the replicated properties from the
Kafka
custom resource. When theKafkaNodePool
resource is in use, you can remove the properties that you copied to theKafkaNodePool
resource, such as the.spec.kafka.replicas
and.spec.kafka.storage
properties.
To revert to managing Kafka nodes using only Kafka
custom resources:
-
If you have multiple node pools, consolidate them into a single
KafkaNodePool
namedkafka
with node IDs from 0 to N (where N is the number of replicas). -
Ensure that the
.spec.kafka
configuration in theKafka
resource matches theKafkaNodePool
configuration, including storage, resources, and replicas. -
Disable support for node pools in the
Kafka
resource using thestrimzi.io/node-pools: disabled
annotation. -
Delete the Kafka node pool named
kafka
.
11.5. Configuring Kafka storage
Strimzi supports different Kafka storage options. You can choose between the following basic types:
- Ephemeral storage
-
Ephemeral storage is temporary and only persists while a pod is running. When a pod is deleted, the data is lost, though data can be recovered in a highly available environment. Due to its transient nature, ephemeral storage is only recommended for development and testing environments.
- Persistent storage
-
Persistent storage retains data across pod restarts and system disruptions, making it ideal for production environments.
JBOD (Just a Bunch of Disks) storage allows you to configure your Kafka cluster to use multiple disks or volumes as ephemeral or persistent storage.
When specifying JBOD storage, you must still decide between using ephemeral or persistent volumes for each disk. Even if you start with only one volume, using JBOD allows for future scaling by adding more volumes as needed, and that is why it is always recommended.
Note
|
Persistent, ephemeral, and JBOD storage types cannot be changed after a Kafka cluster is deployed. However, you can add or remove volumes of different types from the JBOD storage. You can also create and migrate to node pools with new storage specifications. |
Tiered storage, currently available as an early access feature, provides additional flexibility for managing Kafka data by combining different storage types with varying performance and cost characteristics. It allows Kafka to offload older data to cheaper, long-term storage (such as object storage) while keeping recent, frequently accessed data on faster, more expensive storage (such as block storage).
Tiered storage is an add-on capability.
After configuring storage (ephemeral, persistent, or JBOD) for Kafka nodes, you can configure tiered storage at the cluster level and enable it for specific topics using the remote.storage.enable
topic-level configuration.
11.5.1. Storage considerations
Efficient data storage is essential for Strimzi to operate effectively, and block storage is strongly recommended. Strimzi has been tested only with block storage, and file storage solutions like NFS are not guaranteed to work.
Common block storage types supported by Kubernetes include:
-
Cloud-based block storage solutions:
-
Amazon EBS (for AWS)
-
Azure Disk Storage (for Microsoft Azure)
-
Persistent Disk (for Google Cloud)
-
-
Persistent storage (for bare metal deployments) using local persistent volumes
-
Storage Area Network (SAN) volumes accessed by protocols like Fibre Channel or iSCSI
Note
|
Strimzi does not require Kubernetes raw block volumes. |
File systems
Kafka uses a file system for storing messages. Strimzi is compatible with the XFS and ext4 file systems, which are commonly used with Kafka. Consider the underlying architecture and requirements of your deployment when choosing and setting up your file system.
For more information, refer to Filesystem Selection in the Kafka documentation.
Disk usage
Solid-state drives (SSDs), though not essential, can improve the performance of Kafka in large clusters where data is sent to and received from multiple topics asynchronously.
Note
|
Replicated storage is not required, as Kafka provides built-in data replication. |
11.5.2. Configuring Kafka storage in KRaft mode
Use the storage
properties of the KafkaNodePool
custom resource to configure storage for a deployment of Kafka in KRaft mode.
Configuring ephemeral storage
To use ephemeral storage, specify ephemeral
as the storage type.
apiVersion: kafka.strimzi.io/v1beta2
kind: KafkaNodePool
metadata:
name: my-node-pool
labels:
strimzi.io/cluster: my-cluster
spec:
replicas: 3
roles:
- broker
storage:
type: ephemeral
# ...
Ephemeral storage uses emptyDir volumes, which are created when a pod is assigned to a node.
You can limit the size of the emptyDir
volume with the sizeLimit
property.
The ephemeral volume used by Kafka brokers for log directories is mounted at /var/lib/kafka/data/kafka-log<pod_id>
.
Important
|
Ephemeral storage is not suitable for Kafka topics with a replication factor of 1. |
For more information on ephemeral storage configuration options, see the EphemeralStorage
schema reference.
Configuring persistent storage
To use persistent storage, specify one of the following as the storage type:
-
persistent-claim
for a single persistent volume -
jbod
for multiple persistent volumes in a Kafka cluster (Recommended for Kafka in a production environment)
apiVersion: kafka.strimzi.io/v1beta2
kind: KafkaNodePool
metadata:
name: my-node-pool
labels:
strimzi.io/cluster: my-cluster
spec:
replicas: 3
roles:
- broker
storage:
type: persistent-claim
size: 500Gi
deleteClaim: true
# ...
Strimzi uses Persistent Volume Claims (PVCs) to request storage on persistent volumes (PVs). The PVC binds to a PV that meets the requested storage criteria, without needing to know the underlying storage infrastructure.
PVCs created for Kafka pods follow the naming convention data-<kafka_cluster_name>-<pool_name>-<pod_id>
, and the persistent volumes for Kafka logs are mounted at /var/lib/kafka/data/kafka-log<pod_id>
.
You can also specify custom storage classes (StorageClass) and volume selectors in the storage configuration.
# ...
storage:
type: persistent-claim
size: 500Gi
class: my-storage-class
selector:
hdd-type: ssd
deleteClaim: true
# ...
Storage classes define storage profiles and dynamically provision persistent volumes (PVs) based on those profiles. This is useful, for example, when storage classes are restricted to different availability zones or data centers. If a storage class is not specified, the default storage class in the Kubernetes cluster is used. Selectors specify persistent volumes that offer specific features, such as solid-state drive (SSD) volumes.
For more information on persistent storage configuration options, see the PersistentClaimStorage
schema reference.
Resizing persistent volumes
Persistent volumes can be resized by changing the size
storage property without any risk of data loss, as long as the storage infrastructure supports it.
Following a configuration update to change the size of the storage, Strimzi instructs the storage infrastructure to make the change.
Storage expansion is supported in Strimzi clusters that use persistent-claim volumes. Decreasing the size of persistent volumes is not supported in Kubernetes. For more information about resizing persistent volumes in Kubernetes, see Resizing Persistent Volumes using Kubernetes.
After increasing the value of the size
property, Kubernetes increases the capacity of the selected persistent volumes in response to a request from the Cluster Operator.
When the resizing is complete, the Cluster Operator restarts all pods that use the resized persistent volumes.
This happens automatically.
In this example, the volumes are increased to 2000Gi.
2000Gi
apiVersion: kafka.strimzi.io/v1beta2
kind: KafkaNodePool
metadata:
name: my-node-pool
labels:
strimzi.io/cluster: my-cluster
spec:
replicas: 3
roles:
- broker
storage:
type: jbod
volumes:
- id: 0
type: persistent-claim
size: 2000Gi
deleteClaim: false
- id: 1
type: persistent-claim
size: 2000Gi
deleteClaim: false
- id: 2
type: persistent-claim
size: 2000Gi
deleteClaim: false
# ...
Returning information on the PVs verifies the changes:
kubectl get pv
NAME CAPACITY CLAIM
pvc-0ca459ce-... 2000Gi my-project/data-my-cluster-my-node-pool-2
pvc-6e1810be-... 2000Gi my-project/data-my-cluster-my-node-pool-0
pvc-82dc78c9-... 2000Gi my-project/data-my-cluster-my-node-pool-1
The output shows the names of each PVC associated with a broker pod.
Note
|
Storage reduction is only possible when using multiple disks per broker. You can remove a disk after moving all partitions on the disk to other volumes within the same broker (intra-broker) or to other brokers within the same cluster (intra-cluster). |
Configuring JBOD storage
To use JBOD storage, specify jbod
as the storage type and add configuration for the JBOD volumes.
JBOD volumes can be persistent or ephemeral, with the configuration options and constraints applicable to each type.
apiVersion: kafka.strimzi.io/v1beta2
kind: KafkaNodePool
metadata:
name: my-node-pool
labels:
strimzi.io/cluster: my-cluster
spec:
replicas: 3
roles:
- broker
storage:
type: jbod
volumes:
- id: 0
type: persistent-claim
size: 100Gi
deleteClaim: false
- id: 1
type: persistent-claim
size: 100Gi
deleteClaim: false
- id: 2
type: persistent-claim
size: 100Gi
deleteClaim: false
# ...
PVCs are created for the JBOD volumes using the naming convention data-<volume_id>-<kafka_cluster_name>-<pool_name>-<pod_id>
, and the JBOD volumes used for log directories are mounted at /var/lib/kafka/data-<volume_id>/kafka-log<pod_id>
.
Adding or removing volumes from JBOD storage
Volume IDs cannot be changed once JBOD volumes are created, though you can add or remove volumes.
When adding a new volume to the to the volumes
array under an id
which was already used in the past and removed, make sure that the previously used PersistentVolumeClaims
have been deleted.
Use Cruise Control to reassign partitions when adding or removing volumes. For information on intra-broker disk balancing, see Tuning options for rebalances.
Configuring KRaft metadata log storage
In KRaft mode, each node (including brokers and controllers) stores a copy of the Kafka cluster’s metadata log on one of its data volumes.
By default, the log is stored on the volume with the lowest ID, but you can specify a different volume using the kraftMetadata
property.
For controller-only nodes, storage is exclusively for the metadata log. Since the log is always stored on a single volume, using JBOD storage with multiple volumes does not improve performance or increase available disk space.
In contrast, broker nodes or nodes that combine broker and controller roles can share the same volume for both the metadata log and partition replica data, optimizing disk utilization. They can also use JBOD storage, where one volume is shared for the metadata log and partition replica data, while additional volumes are used solely for partition replica data.
Changing the volume that stores the metadata log triggers a rolling update of the cluster nodes, involving the deletion of the old log and the creation of a new one in the specified location.
If kraftMetadata
isn’t specified, adding a new volume with a lower ID also prompts an update and relocation of the metadata log.
apiVersion: kafka.strimzi.io/v1beta2
kind: KafkaNodePool
metadata:
name: pool-a
# ...
spec:
storage:
type: jbod
volumes:
- id: 0
type: persistent-claim
size: 100Gi
deleteClaim: false
- id: 1
type: persistent-claim
size: 100Gi
kraftMetadata: shared
deleteClaim: false
# ...
Managing storage using node pools
Storage management in Strimzi is usually straightforward, and requires little change when set up, but there might be situations where you need to modify your storage configurations. Node pools simplify this process, because you can set up separate node pools that specify your new storage requirements.
In this procedure we create and manage storage for a node pool called pool-a
containing three nodes.
The steps require a scaling operation to add a new node pool.
Currently, scaling is only possible for broker-only node pools containing nodes that run as dedicated brokers.
We show how to change the storage class (volumes.class
) that defines the type of persistent storage it uses.
You can use the same steps to change the storage size (volumes.size
).
This approach is particularly useful if you want to reduce disk sizes.
When increasing disk sizes, you have the option to dynamically resize persistent volumes.
Note
|
We strongly recommend using block storage. Strimzi is only tested for use with block storage. |
-
For storage that uses persistent volume claims for dynamic volume allocation, storage classes are defined and available in the Kubernetes cluster that correspond to the storage solutions you need.
-
Create the node pool with its own storage settings.
For example, node pool
pool-a
uses JBOD storage with persistent volumes:apiVersion: kafka.strimzi.io/v1beta2 kind: KafkaNodePool metadata: name: pool-a labels: strimzi.io/cluster: my-cluster spec: roles: - broker replicas: 3 storage: type: jbod volumes: - id: 0 type: persistent-claim size: 500Gi class: gp2-ebs # ...
Nodes in
pool-a
are configured to use Amazon EBS (Elastic Block Store) GP2 volumes. -
Apply the node pool configuration for
pool-a
. -
Check the status of the deployment and wait for the pods in
pool-a
to be created and ready (1/1
).kubectl get pods -n <my_cluster_operator_namespace>
Output shows three Kafka nodes in the node poolNAME READY STATUS RESTARTS my-cluster-pool-a-0 1/1 Running 0 my-cluster-pool-a-1 1/1 Running 0 my-cluster-pool-a-2 1/1 Running 0
-
To migrate to a new storage class, create a new node pool with the required storage configuration:
apiVersion: kafka.strimzi.io/v1beta2 kind: KafkaNodePool metadata: name: pool-b labels: strimzi.io/cluster: my-cluster spec: roles: - broker replicas: 3 storage: type: jbod volumes: - id: 0 type: persistent-claim size: 1Ti class: gp3-ebs # ...
Nodes in
pool-b
are configured to use Amazon EBS (Elastic Block Store) GP3 volumes. -
Apply the node pool configuration for
pool-b
. -
Check the status of the deployment and wait for the pods in
pool-b
to be created and ready. -
Reassign the partitions from
pool-a
topool-b
.When migrating to a new storage configuration, use the Cruise Control
remove-brokers
mode to move partition replicas off the brokers that are going to be removed.Using Cruise Control to reassign partition replicasapiVersion: kafka.strimzi.io/v1beta2 kind: KafkaRebalance metadata: # ... spec: mode: remove-brokers brokers: [0, 1, 2]
We are reassigning partitions from
pool-a
. The reassignment can take some time depending on the number of topics and partitions in the cluster. -
After the reassignment process is complete, delete the old node pool:
kubectl delete kafkanodepool pool-a
Managing storage affinity using node pools
In situations where storage resources, such as local persistent volumes, are constrained to specific worker nodes, or availability zones, configuring storage affinity helps to schedule pods to use the right nodes.
Node pools allow you to configure affinity independently.
In this procedure, we create and manage storage affinity for two availability zones: zone-1
and zone-2
.
You can configure node pools for separate availability zones, but use the same storage class.
We define an all-zones
persistent storage class representing the storage resources available in each zone.
We also use the .spec.template.pod
properties to configure the node affinity and schedule Kafka pods on zone-1
and zone-2
worker nodes.
The storage class and affinity is specified in node pools representing the nodes in each availability zone:
-
pool-zone-1
-
pool-zone-2
.
-
If you are not familiar with the concepts of affinity, see the Kubernetes node and pod affinity documentation.
-
Define the storage class for use with each availability zone:
apiVersion: storage.k8s.io/v1 kind: StorageClass metadata: name: all-zones provisioner: kubernetes.io/my-storage parameters: type: ssd volumeBindingMode: WaitForFirstConsumer
-
Create node pools representing the two availability zones, specifying the
all-zones
storage class and the affinity for each zone:Node pool configuration for zone-1apiVersion: kafka.strimzi.io/v1beta2 kind: KafkaNodePool metadata: name: pool-zone-1 labels: strimzi.io/cluster: my-cluster spec: replicas: 3 storage: type: jbod volumes: - id: 0 type: persistent-claim size: 500Gi class: all-zones template: pod: affinity: nodeAffinity: requiredDuringSchedulingIgnoredDuringExecution: nodeSelectorTerms: - matchExpressions: - key: topology.kubernetes.io/zone operator: In values: - zone-1 # ...
Node pool configuration for zone-2apiVersion: kafka.strimzi.io/v1beta2 kind: KafkaNodePool metadata: name: pool-zone-2 labels: strimzi.io/cluster: my-cluster spec: replicas: 4 storage: type: jbod volumes: - id: 0 type: persistent-claim size: 500Gi class: all-zones template: pod: affinity: nodeAffinity: requiredDuringSchedulingIgnoredDuringExecution: nodeSelectorTerms: - matchExpressions: - key: topology.kubernetes.io/zone operator: In values: - zone-2 # ...
-
Apply the node pool configuration.
-
Check the status of the deployment and wait for the pods in the node pools to be created and ready (
1/1
).kubectl get pods -n <my_cluster_operator_namespace>
Output shows 3 Kafka nodes inpool-zone-1
and 4 Kafka nodes inpool-zone-2
NAME READY STATUS RESTARTS my-cluster-pool-zone-1-kafka-0 1/1 Running 0 my-cluster-pool-zone-1-kafka-1 1/1 Running 0 my-cluster-pool-zone-1-kafka-2 1/1 Running 0 my-cluster-pool-zone-2-kafka-3 1/1 Running 0 my-cluster-pool-zone-2-kafka-4 1/1 Running 0 my-cluster-pool-zone-2-kafka-5 1/1 Running 0 my-cluster-pool-zone-2-kafka-6 1/1 Running 0
11.5.3. Configuring Kafka storage with ZooKeeper
If you are using ZooKeeper, configure its storage in the Kafka
resource.
Depending on whether the deployment uses node pools, configure storage for the Kafka cluster in Kafka
or KafkaNodePool
resources.
This section focuses only on ZooKeeper storage and Kafka storage configuration in the Kafka
resource.
For detailed information on Kafka storage, refer to the section describing storage configuration using node pools.
The same configuration options for storage are available in the Kafka
resource.
Note
|
Replicated storage is not required for ZooKeeper, as it has built-in data replication. |
Configuring ephemeral storage
To use ephemeral storage, specify ephemeral
as the storage type.
apiVersion: kafka.strimzi.io/v1beta2
kind: Kafka
metadata:
name: my-cluster
spec:
kafka:
storage:
type: ephemeral
zookeeper:
storage:
type: ephemeral
# ...
The ephemeral volume used by Kafka brokers for log directories is mounted at /var/lib/kafka/data/kafka-log<pod_id>
.
Important
|
Ephemeral storage is unsuitable for single-node ZooKeeper clusters or Kafka topics with a replication factor of 1. |
Configuring persistent storage
The same persistent storage configuration options available for node pools can also be specified for Kafka in the Kafka
resource.
For more information, see the section on configuring Kafka storage using node pools.
The size
property can also be adjusted to resize persistent volumes.
The storage type must always be persistent-claim
for ZooKeeper, as it does not support JBOD storage.
apiVersion: kafka.strimzi.io/v1beta2
kind: Kafka
metadata:
name: my-cluster
spec:
kafka:
storage:
type: persistent-claim
size: 500Gi
deleteClaim: true
# ...
zookeeper:
storage:
type: persistent-claim
size: 1000Gi
PVCs created for Kafka pods when storage is configured in the Kafka
resource use the naming convention data-<cluster_name>-kafka-<pod_id>
, and the persistent volumes for Kafka logs are mounted at /var/lib/kafka/data/kafka-log<pod_id>
.
PVCs created for ZooKeeper follow the naming convention data-<cluster_name>-zookeeper-<pod_id>
.
Note
|
As in KRaft mode, you can also specify custom storage classes and volume selectors. |
Configuring JBOD storage
ZooKeeper does not support JBOD storage, but Kafka nodes in a ZooKeeper-based cluster can still be configured to use JBOD storage.
The same JBOD configuration options available for node pools can also be specified for Kafka in the Kafka
resource.
For more information, see the section on configuring Kafka storage using node pools.
The volumes
array can also be adjusted to add or remove volumes.
apiVersion: kafka.strimzi.io/v1beta2
kind: Kafka
metadata:
name: my-cluster
spec:
kafka:
storage:
type: jbod
volumes:
- id: 0
type: persistent-claim
size: 100Gi
deleteClaim: false
- id: 1
type: persistent-claim
size: 100Gi
deleteClaim: false
- id: 2
type: persistent-claim
size: 100Gi
deleteClaim: false
# ...
zookeeper:
storage:
type: persistent-claim
size: 1000Gi
Migrating from storage class overrides (deprecated)
The use of node pools to change the storage classes used by volumes replaces the deprecated overrides
properties previously used for Kafka and ZooKeeper in the Kafka
resource.
apiVersion: kafka.strimzi.io/v1beta2
kind: Kafka
metadata:
labels:
app: my-cluster
name: my-cluster
namespace: myproject
spec:
# ...
kafka:
replicas: 3
storage:
type: jbod
volumes:
- id: 0
type: persistent-claim
size: 100Gi
deleteClaim: false
class: my-storage-class
overrides:
- broker: 0
class: my-storage-class-zone-1a
- broker: 1
class: my-storage-class-zone-1b
- broker: 2
class: my-storage-class-zone-1c
# ...
# ...
zookeeper:
replicas: 3
storage:
deleteClaim: true
size: 100Gi
type: persistent-claim
class: my-storage-class
overrides:
- broker: 0
class: my-storage-class-zone-1a
- broker: 1
class: my-storage-class-zone-1b
- broker: 2
class: my-storage-class-zone-1c
# ...
If you are using storage class overrides for Kafka, we encourage you to transition to using node pools instead. To migrate the existing configuration, follow these steps:
-
Make sure you already use node pools resources. If not, you should migrate the cluster to use node pools first.
-
Create new node pools with storage configuration using the desired storage class without using the overrides.
-
Move all partition replicas from the old broker using the storage class overrides. You can do this using Cruise Control or using the partition reassignment tool.
-
Delete the old node pool with the old brokers using the storage class overrides.
11.5.4. Tiered storage (early access)
Tiered storage introduces a flexible approach to managing Kafka data whereby log segments are moved to a separate storage system. For example, you can combine the use of block storage on brokers for frequently accessed data and offload older or less frequently accessed data from the block storage to more cost-effective, scalable remote storage solutions, such as Amazon S3, without compromising data accessibility and durability.
Warning
|
Tiered storage is an early access Kafka feature, which is also available in Strimzi. Due to its current limitations, it is not recommended for production environments. |
Tiered storage requires an implementation of Kafka’s RemoteStorageManager
interface to handle communication between Kafka and the remote storage system, which is enabled through configuration of the Kafka
resource.
Strimzi uses Kafka’s TopicBasedRemoteLogMetadataManager
for Remote Log Metadata Management (RLMM) when custom tiered storage is enabled.
The RLMM manages the metadata related to remote storage.
To use custom tiered storage, do the following:
-
Include a tiered storage plugin for Kafka in the Strimzi image by building a custom container image. The plugin must provide the necessary functionality for a Kafka cluster managed by Strimzi to interact with the tiered storage solution.
-
Configure Kafka for tiered storage using
tieredStorage
properties in theKafka
resource. Specify the class name and path for the customRemoteStorageManager
implementation, as well as any additional configuration. -
If required, specify RLMM-specific tiered storage configuration.
apiVersion: kafka.strimzi.io/v1beta2
kind: Kafka
metadata:
name: my-cluster
spec:
kafka:
tieredStorage:
type: custom # (1)
remoteStorageManager: # (2)
className: com.example.kafka.tiered.storage.s3.S3RemoteStorageManager
classPath: /opt/kafka/plugins/tiered-storage-s3/*
config:
storage.bucket.name: my-bucket # (3)
# ...
config:
rlmm.config.remote.log.metadata.topic.replication.factor: 1 # (4)
# ...
-
The
type
must be set tocustom
. -
The configuration for the custom
RemoteStorageManager
implementation, including class name and path. -
Configuration to pass to the custom
RemoteStorageManager
implementation, which Strimzi automatically prefixes withrsm.config.
. -
Tiered storage configuration to pass to the RLMM, which requires an
rlmm.config.
prefix. For more information on tiered storage configuration, see the Apache Kafka documentation.
11.6. Configuring the Entity Operator
Use the entityOperator
property in Kafka.spec
to configure the Entity Operator.
The Entity Operator is responsible for managing Kafka-related entities in a running Kafka cluster. It comprises the following operators:
-
Topic Operator to manage Kafka topics
-
User Operator to manage Kafka users
By configuring the Kafka
resource, the Cluster Operator can deploy the Entity Operator, including one or both operators.
Once deployed, the operators are automatically configured to handle the topics and users of the Kafka cluster.
Each operator can only monitor a single namespace. For more information, see Watching Strimzi resources in Kubernetes namespaces.
The entityOperator
property supports several sub-properties:
-
topicOperator
-
userOperator
-
template
The template
property contains the configuration of the Entity Operator pod, such as labels, annotations, affinity, and tolerations.
For more information on configuring templates, see Customizing Kubernetes resources.
The topicOperator
property contains the configuration of the Topic Operator.
When this option is missing, the Entity Operator is deployed without the Topic Operator.
The userOperator
property contains the configuration of the User Operator.
When this option is missing, the Entity Operator is deployed without the User Operator.
For more information on the properties used to configure the Entity Operator, see the EntityOperatorSpec
schema reference.
apiVersion: kafka.strimzi.io/v1beta2
kind: Kafka
metadata:
name: my-cluster
spec:
kafka:
# ...
zookeeper:
# ...
entityOperator:
topicOperator: {}
userOperator: {}
If an empty object ({}
) is used for the topicOperator
and userOperator
, all properties use their default values.
When both topicOperator
and userOperator
properties are missing, the Entity Operator is not deployed.
11.6.1. Configuring the Topic Operator
Use topicOperator
properties in Kafka.spec.entityOperator
to configure the Topic Operator.
The following properties are supported:
watchedNamespace
-
The Kubernetes namespace in which the Topic Operator watches for
KafkaTopic
resources. Default is the namespace where the Kafka cluster is deployed. reconciliationIntervalMs
-
The interval between periodic reconciliations in milliseconds. Default
120000
. image
-
The
image
property can be used to configure the container image which is used. To learn more, refer to the information provided on configuring theimage
property`. resources
-
The
resources
property configures the amount of resources allocated to the Topic Operator. You can specify requests and limits formemory
andcpu
resources. The requests should be enough to ensure a stable performance of the operator. logging
-
The
logging
property configures the logging of the Topic Operator. To learn more, refer to the information provided on Topic Operator logging.
apiVersion: kafka.strimzi.io/v1beta2
kind: Kafka
metadata:
name: my-cluster
spec:
kafka:
# ...
zookeeper:
# ...
entityOperator:
# ...
topicOperator:
watchedNamespace: my-topic-namespace
reconciliationIntervalMs: 60000
resources:
requests:
cpu: "1"
memory: 500Mi
limits:
cpu: "1"
memory: 500Mi
# ...
11.6.2. Configuring the User Operator
Use userOperator
properties in Kafka.spec.entityOperator
to configure the User Operator.
The following properties are supported:
watchedNamespace
-
The Kubernetes namespace in which the User Operator watches for
KafkaUser
resources. Default is the namespace where the Kafka cluster is deployed. reconciliationIntervalMs
-
The interval between periodic reconciliations in milliseconds. Default
120000
. image
-
The
image
property can be used to configure the container image which will be used. To learn more, refer to the information provided on configuring theimage
property`. resources
-
The
resources
property configures the amount of resources allocated to the User Operator. You can specify requests and limits formemory
andcpu
resources. The requests should be enough to ensure a stable performance of the operator. logging
-
The
logging
property configures the logging of the User Operator. To learn more, refer to the information provided on User Operator logging. secretPrefix
-
The
secretPrefix
property adds a prefix to the name of all Secrets created from the KafkaUser resource. For example,secretPrefix: kafka-
would prefix all Secret names withkafka-
. So a KafkaUser namedmy-user
would create a Secret namedkafka-my-user
.
apiVersion: kafka.strimzi.io/v1beta2
kind: Kafka
metadata:
name: my-cluster
spec:
kafka:
# ...
zookeeper:
# ...
entityOperator:
# ...
userOperator:
watchedNamespace: my-user-namespace
reconciliationIntervalMs: 60000
resources:
requests:
cpu: "1"
memory: 500Mi
limits:
cpu: "1"
memory: 500Mi
# ...
11.7. Configuring the Cluster Operator
Use environment variables to configure the Cluster Operator.
Specify the environment variables for the container image of the Cluster Operator in its Deployment
configuration file.
You can use the following environment variables to configure the Cluster Operator.
If you are running Cluster Operator replicas in standby mode, there are additional environment variables for enabling leader election.
Kafka, Kafka Connect, and Kafka MirrorMaker support multiple versions.
Use their STRIMZI_<COMPONENT_NAME>_IMAGES
environment variables to configure the default container images used for each version.
The configuration provides a mapping between a version and an image.
The required syntax is whitespace or comma-separated <version> = <image>
pairs, which determine the image to use for a given version.
For example, 3.9.0=quay.io/strimzi/kafka:latest-kafka-3.9.0.
Theses default images are overridden if image
property values are specified in the configuration of a component.
For more information on image
configuration of components, see the Strimzi Custom Resource API Reference.
Note
|
The Deployment configuration file provided with the Strimzi release artifacts is install/cluster-operator/060-Deployment-strimzi-cluster-operator.yaml .
|
STRIMZI_NAMESPACE
-
A comma-separated list of namespaces that the operator operates in. When not set, set to empty string, or set to
*
, the Cluster Operator operates in all namespaces.The Cluster Operator deployment might use the downward API to set this automatically to the namespace the Cluster Operator is deployed in.
Example configuration for Cluster Operator namespacesenv: - name: STRIMZI_NAMESPACE valueFrom: fieldRef: fieldPath: metadata.namespace
STRIMZI_FULL_RECONCILIATION_INTERVAL_MS
-
Optional, default is 120000 ms. The interval between periodic reconciliations, in milliseconds.
STRIMZI_OPERATION_TIMEOUT_MS
-
Optional, default 300000 ms. The timeout for internal operations, in milliseconds. Increase this value when using Strimzi on clusters where regular Kubernetes operations take longer than usual (due to factors such as prolonged download times for container images, for example).
STRIMZI_ZOOKEEPER_ADMIN_SESSION_TIMEOUT_MS
-
Optional, default 10000 ms. The session timeout for the Cluster Operator’s ZooKeeper admin client, in milliseconds. Increase the value if ZooKeeper requests from the Cluster Operator are regularly failing due to timeout issues. There is a maximum allowed session time set on the ZooKeeper server side via the
maxSessionTimeout
config. By default, the maximum session timeout value is 20 times the defaulttickTime
(whose default is 2000) at 40000 ms. If you require a higher timeout, change themaxSessionTimeout
ZooKeeper server configuration value. STRIMZI_OPERATIONS_THREAD_POOL_SIZE
-
Optional, default 10. The worker thread pool size, which is used for various asynchronous and blocking operations that are run by the Cluster Operator.
STRIMZI_OPERATOR_NAME
-
Optional, defaults to the pod’s hostname. The operator name identifies the Strimzi instance when emitting Kubernetes events.
STRIMZI_OPERATOR_NAMESPACE
-
The name of the namespace where the Cluster Operator is running. Do not configure this variable manually. Use the downward API.
env: - name: STRIMZI_OPERATOR_NAMESPACE valueFrom: fieldRef: fieldPath: metadata.namespace
STRIMZI_OPERATOR_NAMESPACE_LABELS
-
Optional. The labels of the namespace where the Strimzi Cluster Operator is running. Use namespace labels to configure the namespace selector in network policies. Network policies allow the Strimzi Cluster Operator access only to the operands from the namespace with these labels. When not set, the namespace selector in network policies is configured to allow access to the Cluster Operator from any namespace in the Kubernetes cluster.
env: - name: STRIMZI_OPERATOR_NAMESPACE_LABELS value: label1=value1,label2=value2
STRIMZI_POD_DISRUPTION_BUDGET_GENERATION
-
Optional, default
true
. Pod disruption budget for resources. A pod disruption budget with themaxUnavailable
value set to zero prevents Kubernetes from evicting pods automatically.Set this environment variable to
false
to disable pod disruption budget generation. You might do this, for example, if you want to manage the pod disruption budgets yourself, or if you have a development environment where availability is not important. STRIMZI_LABELS_EXCLUSION_PATTERN
-
Optional, default regex pattern is
(app.kubernetes.io/(?!part-of).|kustomize.toolkit.fluxcd.io.)
. The regex exclusion pattern used to filter labels propagation from the main custom resource to its subresources. The labels exclusion filter is not applied to labels in template sections such asspec.kafka.template.pod.metadata.labels
.env: - name: STRIMZI_LABELS_EXCLUSION_PATTERN value: "^key1.*"
STRIMZI_CUSTOM_<COMPONENT_NAME>_LABELS
-
Optional. One or more custom labels to apply to all the pods created by the custom resource of the component. The Cluster Operator labels the pods when the custom resource is created or is next reconciled.
Labels can be applied to the following components:
-
KAFKA
-
KAFKA_CONNECT
-
KAFKA_CONNECT_BUILD
-
ZOOKEEPER
-
ENTITY_OPERATOR
-
KAFKA_MIRROR_MAKER2
-
KAFKA_MIRROR_MAKER
-
CRUISE_CONTROL
-
KAFKA_BRIDGE
-
KAFKA_EXPORTER
-
STRIMZI_CUSTOM_RESOURCE_SELECTOR
-
Optional. The label selector to filter the custom resources handled by the Cluster Operator. The operator will operate only on those custom resources that have the specified labels set. Resources without these labels will not be seen by the operator. The label selector applies to
Kafka
,KafkaConnect
,KafkaBridge
,KafkaMirrorMaker
, andKafkaMirrorMaker2
resources.KafkaRebalance
andKafkaConnector
resources are operated only when their corresponding Kafka and Kafka Connect clusters have the matching labels.env: - name: STRIMZI_CUSTOM_RESOURCE_SELECTOR value: label1=value1,label2=value2
STRIMZI_KAFKA_IMAGES
-
Required. The mapping from the Kafka version to the corresponding image containing a Kafka broker for that version. For example
3.8.0=quay.io/strimzi/kafka:latest-kafka-3.8.0, 3.9.0=quay.io/strimzi/kafka:latest-kafka-3.9.0
. STRIMZI_KAFKA_CONNECT_IMAGES
-
Required. The mapping from the Kafka version to the corresponding image of Kafka Connect for that version. For example
3.8.0=quay.io/strimzi/kafka:latest-kafka-3.8.0, 3.9.0=quay.io/strimzi/kafka:latest-kafka-3.9.0
. STRIMZI_KAFKA_MIRROR_MAKER2_IMAGES
-
Required. The mapping from the Kafka version to the corresponding image of MirrorMaker 2 for that version. For example
3.8.0=quay.io/strimzi/kafka:latest-kafka-3.8.0, 3.9.0=quay.io/strimzi/kafka:latest-kafka-3.9.0
. - (Deprecated)
STRIMZI_KAFKA_MIRROR_MAKER_IMAGES
-
Required. The mapping from the Kafka version to the corresponding image of MirrorMaker for that version. For example
3.8.0=quay.io/strimzi/kafka:latest-kafka-3.8.0, 3.9.0=quay.io/strimzi/kafka:latest-kafka-3.9.0
. STRIMZI_DEFAULT_TOPIC_OPERATOR_IMAGE
-
Optional. The default is
quay.io/strimzi/operator:latest
. The image name to use as the default when deploying the Topic Operator if no image is specified as theKafka.spec.entityOperator.topicOperator.image
in theKafka
resource. STRIMZI_DEFAULT_USER_OPERATOR_IMAGE
-
Optional. The default is
quay.io/strimzi/operator:latest
. The image name to use as the default when deploying the User Operator if no image is specified as theKafka.spec.entityOperator.userOperator.image
in theKafka
resource. STRIMZI_DEFAULT_KAFKA_EXPORTER_IMAGE
-
Optional. The default is
quay.io/strimzi/kafka:latest-kafka-3.9.0
. The image name to use as the default when deploying the Kafka Exporter if no image is specified as theKafka.spec.kafkaExporter.image
in theKafka
resource. STRIMZI_DEFAULT_CRUISE_CONTROL_IMAGE
-
Optional. The default is
quay.io/strimzi/kafka:latest-kafka-3.9.0
. The image name to use as the default when deploying Cruise Control if no image is specified as theKafka.spec.cruiseControl.image
in theKafka
resource. STRIMZI_DEFAULT_KAFKA_BRIDGE_IMAGE
-
Optional. The default is
quay.io/strimzi/kafka-bridge:0.31.1
. The image name to use as the default when deploying the Kafka Bridge if no image is specified as theKafka.spec.kafkaBridge.image
in theKafka
resource. STRIMZI_DEFAULT_KAFKA_INIT_IMAGE
-
Optional. The default is
quay.io/strimzi/operator:latest
. The image name to use as the default for the Kafka initializer container if no image is specified in thebrokerRackInitImage
of theKafka
resource or theclientRackInitImage
of the Kafka Connect resource. The init container is started before the Kafka cluster for initial configuration work, such as rack support. STRIMZI_IMAGE_PULL_POLICY
-
Optional. The
ImagePullPolicy
that is applied to containers in all pods managed by the Cluster Operator. The valid values areAlways
,IfNotPresent
, andNever
. If not specified, the Kubernetes defaults are used. Changing the policy will result in a rolling update of all your Kafka, Kafka Connect, and Kafka MirrorMaker clusters. STRIMZI_IMAGE_PULL_SECRETS
-
Optional. A comma-separated list of
Secret
names. The secrets referenced here contain the credentials to the container registries where the container images are pulled from. The secrets are specified in theimagePullSecrets
property for all pods created by the Cluster Operator. Changing this list results in a rolling update of all your Kafka, Kafka Connect, and Kafka MirrorMaker clusters. STRIMZI_KUBERNETES_VERSION
-
Optional. Overrides the Kubernetes version information detected from the API server.
Example configuration for Kubernetes version overrideenv: - name: STRIMZI_KUBERNETES_VERSION value: | major=1 minor=16 gitVersion=v1.16.2 gitCommit=c97fe5036ef3df2967d086711e6c0c405941e14b gitTreeState=clean buildDate=2019-10-15T19:09:08Z goVersion=go1.12.10 compiler=gc platform=linux/amd64
KUBERNETES_SERVICE_DNS_DOMAIN
-
Optional. Overrides the default Kubernetes DNS domain name suffix.
By default, services assigned in the Kubernetes cluster have a DNS domain name that uses the default suffix
cluster.local
.For example, for broker kafka-0:
<cluster-name>-kafka-0.<cluster-name>-kafka-brokers.<namespace>.svc.cluster.local
The DNS domain name is added to the Kafka broker certificates used for hostname verification.
If you are using a different DNS domain name suffix in your cluster, change the
KUBERNETES_SERVICE_DNS_DOMAIN
environment variable from the default to the one you are using in order to establish a connection with the Kafka brokers. STRIMZI_CONNECT_BUILD_TIMEOUT_MS
-
Optional, default 300000 ms. The timeout for building new Kafka Connect images with additional connectors, in milliseconds. Consider increasing this value when using Strimzi to build container images containing many connectors or using a slow container registry.
STRIMZI_NETWORK_POLICY_GENERATION
-
Optional, default
true
. Network policy for resources. Network policies allow connections between Kafka components.Set this environment variable to
false
to disable network policy generation. You might do this, for example, if you want to use custom network policies. Custom network policies allow more control over maintaining the connections between components. STRIMZI_DNS_CACHE_TTL
-
Optional, default
30
. Number of seconds to cache successful name lookups in local DNS resolver. Any negative value means cache forever. Zero means do not cache, which can be useful for avoiding connection errors due to long caching policies being applied. STRIMZI_POD_SET_RECONCILIATION_ONLY
-
Optional, default
false
. When set totrue
, the Cluster Operator reconciles only theStrimziPodSet
resources and any changes to the other custom resources (Kafka
,KafkaConnect
, and so on) are ignored. This mode is useful for ensuring that your pods are recreated if needed, but no other changes happen to the clusters. STRIMZI_FEATURE_GATES
-
Optional. Enables or disables the features and functionality controlled by feature gates.
STRIMZI_POD_SECURITY_PROVIDER_CLASS
-
Optional. Configuration for the pluggable
PodSecurityProvider
class, which can be used to provide the security context configuration for Pods and containers.
11.7.1. Restricting access to the Cluster Operator using network policy
Use the STRIMZI_OPERATOR_NAMESPACE_LABELS
environment variable to establish network policy for the Cluster Operator using namespace labels.
The Cluster Operator can run in the same namespace as the resources it manages, or in a separate namespace.
By default, the STRIMZI_OPERATOR_NAMESPACE
environment variable is configured to use the downward API to find the namespace the Cluster Operator is running in.
If the Cluster Operator is running in the same namespace as the resources, only local access is required and allowed by Strimzi.
If the Cluster Operator is running in a separate namespace to the resources it manages, any namespace in the Kubernetes cluster is allowed access to the Cluster Operator unless network policy is configured. By adding namespace labels, access to the Cluster Operator is restricted to the namespaces specified.
#...
env:
# ...
- name: STRIMZI_OPERATOR_NAMESPACE_LABELS
value: label1=value1,label2=value2
#...
11.7.2. Setting periodic reconciliation of custom resources
Use the STRIMZI_FULL_RECONCILIATION_INTERVAL_MS
variable to set the time interval for periodic reconciliations by the Cluster Operator.
Replace its value with the required interval in milliseconds.
#...
env:
# ...
- name: STRIMZI_FULL_RECONCILIATION_INTERVAL_MS
value: "120000"
#...
The Cluster Operator reacts to all notifications about applicable cluster resources received from the Kubernetes cluster. If the operator is not running, or if a notification is not received for any reason, resources will get out of sync with the state of the running Kubernetes cluster. In order to handle failovers properly, a periodic reconciliation process is executed by the Cluster Operator so that it can compare the state of the resources with the current cluster deployments in order to have a consistent state across all of them.
11.7.3. Pausing reconciliation of custom resources using annotations
Sometimes it is useful to pause the reconciliation of custom resources managed by Strimzi operators, so that you can perform fixes or make updates. If reconciliations are paused, any changes made to custom resources are ignored by the operators until the pause ends.
If you want to pause reconciliation of a custom resource, set the strimzi.io/pause-reconciliation
annotation to true
in its configuration.
This instructs the appropriate operator to pause reconciliation of the custom resource.
For example, you can apply the annotation to the KafkaConnect
resource so that reconciliation by the Cluster Operator is paused.
You can also create a custom resource with the pause annotation enabled. The custom resource is created, but it is ignored.
-
The Strimzi Operator that manages the custom resource is running.
-
Annotate the custom resource in Kubernetes, setting
pause-reconciliation
totrue
:kubectl annotate <kind_of_custom_resource> <name_of_custom_resource> strimzi.io/pause-reconciliation="true"
For example, for the
KafkaConnect
custom resource:kubectl annotate KafkaConnect my-connect strimzi.io/pause-reconciliation="true"
-
Check that the status conditions of the custom resource show a change to
ReconciliationPaused
:kubectl describe <kind_of_custom_resource> <name_of_custom_resource>
The
type
condition changes toReconciliationPaused
at thelastTransitionTime
.Example custom resource with a paused reconciliation condition typeapiVersion: kafka.strimzi.io/v1beta2 kind: KafkaConnect metadata: annotations: strimzi.io/pause-reconciliation: "true" strimzi.io/use-connector-resources: "true" creationTimestamp: 2021-03-12T10:47:11Z #... spec: # ... status: conditions: - lastTransitionTime: 2021-03-12T10:47:41.689249Z status: "True" type: ReconciliationPaused
-
To resume reconciliation, you can set the annotation to
false
, or remove the annotation.
11.7.4. Running multiple Cluster Operator replicas with leader election
The default Cluster Operator configuration enables leader election to run multiple parallel replicas of the Cluster Operator. One replica is elected as the active leader and operates the deployed resources. The other replicas run in standby mode. When the leader stops or fails, one of the standby replicas is elected as the new leader and starts operating the deployed resources.
By default, Strimzi runs with a single Cluster Operator replica that is always the leader replica. When a single Cluster Operator replica stops or fails, Kubernetes starts a new replica.
Running the Cluster Operator with multiple replicas is not essential. But it’s useful to have replicas on standby in case of large-scale disruptions caused by major failure. For example, suppose multiple worker nodes or an entire availability zone fails. This failure might cause the Cluster Operator pod and many Kafka pods to go down at the same time. If subsequent pod scheduling causes congestion through lack of resources, this can delay operations when running a single Cluster Operator.
Enabling leader election for Cluster Operator replicas
Configure leader election environment variables when running additional Cluster Operator replicas. The following environment variables are supported:
STRIMZI_LEADER_ELECTION_ENABLED
-
Optional, disabled (
false
) by default. Enables or disables leader election, which allows additional Cluster Operator replicas to run on standby.
Note
|
Leader election is disabled by default. It is only enabled when applying this environment variable on installation. |
STRIMZI_LEADER_ELECTION_LEASE_NAME
-
Required when leader election is enabled. The name of the Kubernetes
Lease
resource that is used for the leader election. STRIMZI_LEADER_ELECTION_LEASE_NAMESPACE
-
Required when leader election is enabled. The namespace where the Kubernetes
Lease
resource used for leader election is created. You can use the downward API to configure it to the namespace where the Cluster Operator is deployed.env: - name: STRIMZI_LEADER_ELECTION_LEASE_NAMESPACE valueFrom: fieldRef: fieldPath: metadata.namespace
STRIMZI_LEADER_ELECTION_IDENTITY
-
Required when leader election is enabled. Configures the identity of a given Cluster Operator instance used during the leader election. The identity must be unique for each operator instance. You can use the downward API to configure it to the name of the pod where the Cluster Operator is deployed.
env: - name: STRIMZI_LEADER_ELECTION_IDENTITY valueFrom: fieldRef: fieldPath: metadata.name
STRIMZI_LEADER_ELECTION_LEASE_DURATION_MS
-
Optional, default 15000 ms. Specifies the duration the acquired lease is valid.
STRIMZI_LEADER_ELECTION_RENEW_DEADLINE_MS
-
Optional, default 10000 ms. Specifies the period the leader should try to maintain leadership.
STRIMZI_LEADER_ELECTION_RETRY_PERIOD_MS
-
Optional, default 2000 ms. Specifies the frequency of updates to the lease lock by the leader.
Configuring Cluster Operator replicas
To run additional Cluster Operator replicas in standby mode, you will need to increase the number of replicas and enable leader election. To configure leader election, use the leader election environment variables.
To make the required changes, configure the following Cluster Operator installation files located in install/cluster-operator/
:
-
060-Deployment-strimzi-cluster-operator.yaml
-
022-ClusterRole-strimzi-cluster-operator-role.yaml
-
022-RoleBinding-strimzi-cluster-operator.yaml
Leader election has its own ClusterRole
and RoleBinding
RBAC resources that target the namespace where the Cluster Operator is running, rather than the namespace it is watching.
The default deployment configuration creates a Lease
resource called strimzi-cluster-operator
in the same namespace as the Cluster Operator.
The Cluster Operator uses leases to manage leader election.
The RBAC resources provide the permissions to use the Lease
resource.
If you use a different Lease
name or namespace, update the ClusterRole
and RoleBinding
files accordingly.
-
You need an account with permission to create and manage
CustomResourceDefinition
and RBAC (ClusterRole
, andRoleBinding
) resources.
Edit the Deployment
resource that is used to deploy the Cluster Operator, which is defined in the 060-Deployment-strimzi-cluster-operator.yaml
file.
-
Change the
replicas
property from the default (1) to a value that matches the required number of replicas.Increasing the number of Cluster Operator replicasapiVersion: apps/v1 kind: Deployment metadata: name: strimzi-cluster-operator labels: app: strimzi spec: replicas: 3
-
Check that the leader election
env
properties are set.If they are not set, configure them.
To enable leader election,
STRIMZI_LEADER_ELECTION_ENABLED
must be set totrue
(default).In this example, the name of the lease is changed to
my-strimzi-cluster-operator
.Configuring leader election environment variables for the Cluster Operator# ... spec containers: - name: strimzi-cluster-operator # ... env: - name: STRIMZI_LEADER_ELECTION_ENABLED value: "true" - name: STRIMZI_LEADER_ELECTION_LEASE_NAME value: "my-strimzi-cluster-operator" - name: STRIMZI_LEADER_ELECTION_LEASE_NAMESPACE valueFrom: fieldRef: fieldPath: metadata.namespace - name: STRIMZI_LEADER_ELECTION_IDENTITY valueFrom: fieldRef: fieldPath: metadata.name
For a description of the available environment variables, see Enabling leader election for Cluster Operator replicas.
If you specified a different name or namespace for the
Lease
resource used in leader election, update the RBAC resources. -
(optional) Edit the
ClusterRole
resource in the022-ClusterRole-strimzi-cluster-operator-role.yaml
file.Update
resourceNames
with the name of theLease
resource.Updating the ClusterRole references to the leaseapiVersion: rbac.authorization.k8s.io/v1 kind: ClusterRole metadata: name: strimzi-cluster-operator-leader-election labels: app: strimzi rules: - apiGroups: - coordination.k8s.io resourceNames: - my-strimzi-cluster-operator # ...
-
(optional) Edit the
RoleBinding
resource in the022-RoleBinding-strimzi-cluster-operator.yaml
file.Update
subjects.name
andsubjects.namespace
with the name of theLease
resource and the namespace where it was created.Updating the RoleBinding references to the leaseapiVersion: rbac.authorization.k8s.io/v1 kind: RoleBinding metadata: name: strimzi-cluster-operator-leader-election labels: app: strimzi subjects: - kind: ServiceAccount name: my-strimzi-cluster-operator namespace: myproject # ...
-
Deploy the Cluster Operator:
kubectl create -f install/cluster-operator -n myproject
-
Check the status of the deployment:
kubectl get deployments -n myproject
Output shows the deployment name and readinessNAME READY UP-TO-DATE AVAILABLE strimzi-cluster-operator 3/3 3 3
READY
shows the number of replicas that are ready/expected. The deployment is successful when theAVAILABLE
output shows the correct number of replicas.
11.7.5. Configuring Cluster Operator HTTP proxy settings
If you are running a Kafka cluster behind a HTTP proxy, you can still pass data in and out of the cluster. For example, you can run Kafka Connect with connectors that push and pull data from outside the proxy. Or you can use a proxy to connect with an authorization server.
Configure the Cluster Operator deployment to specify the proxy environment variables.
The Cluster Operator accepts standard proxy configuration (HTTP_PROXY
, HTTPS_PROXY
and NO_PROXY
) as environment variables.
The proxy settings are applied to all Strimzi containers.
The format for a proxy address is http://<ip_address>:<port_number>. To set up a proxy with a name and password, the format is http://<username>:<password>@<ip-address>:<port_number>.
-
You need an account with permission to create and manage
CustomResourceDefinition
and RBAC (ClusterRole
, andRoleBinding
) resources.
-
To add proxy environment variables to the Cluster Operator, update its
Deployment
configuration (install/cluster-operator/060-Deployment-strimzi-cluster-operator.yaml
).Example proxy configuration for the Cluster OperatorapiVersion: apps/v1 kind: Deployment spec: # ... template: spec: serviceAccountName: strimzi-cluster-operator containers: # ... env: # ... - name: "HTTP_PROXY" value: "http://proxy.com" (1) - name: "HTTPS_PROXY" value: "https://proxy.com" (2) - name: "NO_PROXY" value: "internal.com, other.domain.com" (3) # ...
-
Address of the proxy server.
-
Secure address of the proxy server.
-
Addresses for servers that are accessed directly as exceptions to the proxy server. The URLs are comma-separated.
Alternatively, edit the
Deployment
directly:kubectl edit deployment strimzi-cluster-operator
-
-
If you updated the YAML file instead of editing the
Deployment
directly, apply the changes:kubectl create -f install/cluster-operator/060-Deployment-strimzi-cluster-operator.yaml
11.7.6. Disabling FIPS mode using Cluster Operator configuration
Strimzi automatically switches to FIPS mode when running on a FIPS-enabled Kubernetes cluster.
Disable FIPS mode by setting the FIPS_MODE
environment variable to disabled
in the deployment configuration for the Cluster Operator.
With FIPS mode disabled, Strimzi automatically disables FIPS in the OpenJDK for all components.
With FIPS mode disabled, Strimzi is not FIPS compliant.
The Strimzi operators, as well as all operands, run in the same way as if they were running on an Kubernetes cluster without FIPS enabled.
-
To disable the FIPS mode in the Cluster Operator, update its
Deployment
configuration (install/cluster-operator/060-Deployment-strimzi-cluster-operator.yaml
) and add theFIPS_MODE
environment variable.Example FIPS configuration for the Cluster OperatorapiVersion: apps/v1 kind: Deployment spec: # ... template: spec: serviceAccountName: strimzi-cluster-operator containers: # ... env: # ... - name: "FIPS_MODE" value: "disabled" # (1) # ...
-
Disables the FIPS mode.
Alternatively, edit the
Deployment
directly:kubectl edit deployment strimzi-cluster-operator
-
-
If you updated the YAML file instead of editing the
Deployment
directly, apply the changes:kubectl apply -f install/cluster-operator/060-Deployment-strimzi-cluster-operator.yaml
11.8. Configuring Kafka Connect
Update the spec
properties of the KafkaConnect
custom resource to configure your Kafka Connect deployment.
Use Kafka Connect to set up external data connections to your Kafka cluster.
Use the properties of the KafkaConnect
resource to configure your Kafka Connect deployment.
You can also use the KafkaConnect
resource to specify the following:
-
Connector plugin configuration to build a container image that includes the plugins to make connections
-
Configuration for the Kafka Connect worker pods that run connectors
-
An annotation to enable use of the
KafkaConnector
resource to manage connectors
The Cluster Operator manages Kafka Connect clusters deployed using the KafkaConnect
resource and connectors created using the KafkaConnector
resource.
For a deeper understanding of the Kafka Connect cluster configuration options, refer to the Strimzi Custom Resource API Reference.
You can tune the configuration to handle high volumes of messages. For more information, see Handling high volumes of messages.
KafkaConnect
custom resource configuration# Basic configuration (required)
apiVersion: kafka.strimzi.io/v1beta2
kind: KafkaConnect # (1)
metadata:
name: my-connect-cluster
annotations:
strimzi.io/use-connector-resources: "true" # (2)
# Deployment specifications
spec:
# Replicas (required)
replicas: 3 # (3)
# Bootstrap servers (required)
bootstrapServers: my-cluster-kafka-bootstrap:9092 # (4)
# Kafka Connect configuration (recommended)
config: # (5)
group.id: my-connect-cluster
offset.storage.topic: my-connect-cluster-offsets
config.storage.topic: my-connect-cluster-configs
status.storage.topic: my-connect-cluster-status
key.converter: org.apache.kafka.connect.json.JsonConverter
value.converter: org.apache.kafka.connect.json.JsonConverter
key.converter.schemas.enable: true
value.converter.schemas.enable: true
config.storage.replication.factor: 3
offset.storage.replication.factor: 3
status.storage.replication.factor: 3
# Resources requests and limits (recommended)
resources: # (6)
requests:
cpu: "1"
memory: 2Gi
limits:
cpu: "2"
memory: 2Gi
# Authentication (optional)
authentication: # (7)
type: tls
certificateAndKey:
certificate: source.crt
key: source.key
secretName: my-user-source
# TLS configuration (optional)
tls: # (8)
trustedCertificates:
- secretName: my-cluster-cluster-cert
pattern: "*.crt"
- secretName: my-cluster-cluster-cert
pattern: "*.crt"
# Build configuration (optional)
build: # (9)
output: # (10)
type: docker
image: my-registry.io/my-org/my-connect-cluster:latest
pushSecret: my-registry-credentials
plugins: # (11)
- name: connector-1
artifacts:
- type: tgz
url: <url_to_download_connector_1_artifact>
sha512sum: <SHA-512_checksum_of_connector_1_artifact>
- name: connector-2
artifacts:
- type: jar
url: <url_to_download_connector_2_artifact>
sha512sum: <SHA-512_checksum_of_connector_2_artifact>
# Logging configuration (optional)
logging: # (12)
type: inline
loggers:
log4j.rootLogger: INFO
# Readiness probe (optional)
readinessProbe: # (13)
initialDelaySeconds: 15
timeoutSeconds: 5
# Liveness probe (optional)
livenessProbe:
initialDelaySeconds: 15
timeoutSeconds: 5
# Metrics configuration (optional)
metricsConfig: # (14)
type: jmxPrometheusExporter
valueFrom:
configMapKeyRef:
name: my-config-map
key: my-key
# JVM options (optional)
jvmOptions: # (15)
"-Xmx": "1g"
"-Xms": "1g"
# Custom image (optional)
image: my-org/my-image:latest # (16)
# Rack awareness (optional)
rack:
topologyKey: topology.kubernetes.io/zone # (17)
# Pod and container template (optional)
template: # (18)
pod:
affinity:
podAntiAffinity:
requiredDuringSchedulingIgnoredDuringExecution:
- labelSelector:
matchExpressions:
- key: application
operator: In
values:
- postgresql
- mongodb
topologyKey: "kubernetes.io/hostname"
connectContainer: # (19)
env:
- name: OTEL_SERVICE_NAME
value: my-otel-service
- name: OTEL_EXPORTER_OTLP_ENDPOINT
value: "http://otlp-host:4317"
- name: AWS_ACCESS_KEY_ID
valueFrom:
secretKeyRef:
name: aws-creds
key: awsAccessKey
- name: AWS_SECRET_ACCESS_KEY
valueFrom:
secretKeyRef:
name: aws-creds
key: awsSecretAccessKey
# Tracing configuration (optional)
tracing:
type: opentelemetry # (20)
-
Use
KafkaConnect
. -
Enables the use of
KafkaConnector
resources to start, stop, and manage connector instances. -
The number of replica nodes for the workers that run tasks.
-
Bootstrap address for connection to the Kafka cluster. The address takes the format
<cluster_name>-kafka-bootstrap:<port_number>
. The Kafka cluster doesn’t need to be managed by Strimzi or deployed to a Kubernetes cluster. -
Kafka Connect configuration of workers (not connectors) that run connectors and their tasks. Standard Apache Kafka configuration may be provided, restricted to those properties not managed directly by Strimzi. In this example, JSON convertors are specified. A replication factor of 3 is set for the internal topics used by Kafka Connect (minimum requirement for production environment). Changing the replication factor after the topics have been created has no effect.
-
Requests for reservation of supported resources, currently
cpu
andmemory
, and limits to specify the maximum resources that can be consumed. -
Authentication for the Kafka Connect cluster, specified as mTLS, token-based OAuth, SASL-based SCRAM-SHA-256/SCRAM-SHA-512, or PLAIN. By default, Kafka Connect connects to Kafka brokers using a plain text connection.
-
TLS configuration for encrypted connections to the Kafka cluster, with trusted certificates stored in X.509 format within the specified secrets.
-
Build configuration properties for building a container image with connector plugins automatically.
-
(Required) Configuration of the container registry where new images are pushed.
-
(Required) List of connector plugins and their artifacts to add to the new container image. Each plugin must be configured with at least one
artifact
. -
Specified Kafka Connect loggers and log levels added directly (
inline
) or indirectly (external
) through a ConfigMap. A custom Log4j configuration must be placed under thelog4j.properties
orlog4j2.properties
key in the ConfigMap. For the Kafka Connectlog4j.rootLogger
logger, you can set the log level to INFO, ERROR, WARN, TRACE, DEBUG, FATAL or OFF. -
Healthchecks to know when to restart a container (liveness) and when a container can accept traffic (readiness).
-
Prometheus metrics, which are enabled by referencing a ConfigMap containing configuration for the Prometheus JMX exporter in this example. You can enable metrics without further configuration using a reference to a ConfigMap containing an empty file under
metricsConfig.valueFrom.configMapKeyRef.key
. -
JVM configuration options to optimize performance for the Virtual Machine (VM) running Kafka Connect.
-
ADVANCED OPTION: Container image configuration, which is recommended only in special situations.
-
SPECIALIZED OPTION: Rack awareness configuration for the deployment. This is a specialized option intended for a deployment within the same location, not across regions. Use this option if you want connectors to consume from the closest replica rather than the leader replica. In certain cases, consuming from the closest replica can improve network utilization or reduce costs . The
topologyKey
must match a node label containing the rack ID. The example used in this configuration specifies a zone using the standardtopology.kubernetes.io/zone
label. To consume from the closest replica, enable theRackAwareReplicaSelector
in the Kafka broker configuration. -
Template customization. Here a pod is scheduled with anti-affinity, so the pod is not scheduled on nodes with the same hostname.
-
Environment variables are set for distributed tracing and to pass credentials to connectors.
-
Distributed tracing is enabled by using OpenTelemetry.
11.8.1. Configuring Kafka Connect for multiple instances
By default, Strimzi configures the group ID and names of the internal topics used by Kafka Connect.
When running multiple instances of Kafka Connect, you must change these default settings using the following config
properties:
apiVersion: kafka.strimzi.io/v1beta2
kind: KafkaConnect
metadata:
name: my-connect
spec:
config:
group.id: my-connect-cluster # (1)
offset.storage.topic: my-connect-cluster-offsets # (2)
config.storage.topic: my-connect-cluster-configs # (3)
status.storage.topic: my-connect-cluster-status # (4)
# ...
# ...
-
The Kafka Connect cluster group ID within Kafka.
-
Kafka topic that stores connector offsets.
-
Kafka topic that stores connector and task status configurations.
-
Kafka topic that stores connector and task status updates.
Note
|
Values for the three topics must be the same for all instances with the same group.id .
|
Unless you modify these default settings, each instance connecting to the same Kafka cluster is deployed with the same values. In practice, this means all instances form a cluster and use the same internal topics.
Multiple instances attempting to use the same internal topics will cause unexpected errors, so you must change the values of these properties for each instance.
11.8.2. Configuring Kafka Connect user authorization
When using authorization in Kafka, a Kafka Connect user requires read/write access to the cluster group and internal topics of Kafka Connect.
This procedure outlines how access is granted using simple
authorization and ACLs.
Properties for the Kafka Connect cluster group ID and internal topics are configured by Strimzi by default.
Alternatively, you can define them explicitly in the spec
of the KafkaConnect
resource.
This is useful when configuring Kafka Connect for multiple instances, as the values for the group ID and topics must differ when running multiple Kafka Connect instances.
Simple authorization uses ACL rules managed by the Kafka AclAuthorizer
and StandardAuthorizer
plugins to ensure appropriate access levels.
For more information on configuring a KafkaUser
resource to use simple authorization, see the AclRule
schema reference.
-
A Kubernetes cluster
-
A running Cluster Operator
-
Edit the
authorization
property in theKafkaUser
resource to provide access rights to the user.Access rights are configured for the Kafka Connect topics and cluster group using
literal
name values. The following table shows the default names configured for the topics and cluster group ID.Table 12. Names for the access rights configuration Property Name offset.storage.topic
connect-cluster-offsets
status.storage.topic
connect-cluster-status
config.storage.topic
connect-cluster-configs
group
connect-cluster
In this example configuration, the default names are used to specify access rights. If you are using different names for a Kafka Connect instance, use those names in the ACLs configuration.
Example configuration for simple authorizationapiVersion: kafka.strimzi.io/v1beta2 kind: KafkaUser metadata: name: my-user labels: strimzi.io/cluster: my-cluster spec: # ... authorization: type: simple acls: # access to offset.storage.topic - resource: type: topic name: connect-cluster-offsets patternType: literal operations: - Create - Describe - Read - Write host: "*" # access to status.storage.topic - resource: type: topic name: connect-cluster-status patternType: literal operations: - Create - Describe - Read - Write host: "*" # access to config.storage.topic - resource: type: topic name: connect-cluster-configs patternType: literal operations: - Create - Describe - Read - Write host: "*" # cluster group - resource: type: group name: connect-cluster patternType: literal operations: - Read host: "*"
-
Create or update the resource.
kubectl apply -f KAFKA-USER-CONFIG-FILE
11.9. Configuring Kafka Connect connectors
The KafkaConnector
resource provides a Kubernetes-native approach to management of connectors by the Cluster Operator.
To create, delete, or reconfigure connectors with KafkaConnector
resources, you must set the use-connector-resources
annotation to true
in your KafkaConnect
custom resource.
apiVersion: kafka.strimzi.io/v1beta2
kind: KafkaConnect
metadata:
name: my-connect-cluster
annotations:
strimzi.io/use-connector-resources: "true"
# ...
When the use-connector-resources
annotation is enabled in your KafkaConnect
configuration, you must define and manage connectors using KafkaConnector
resources.
Note
|
Alternatively, you can manage connectors using the Kafka Connect REST API instead of KafkaConnector resources.
To use the API, you must remove the strimzi.io/use-connector-resources annotation to use KafkaConnector resources in the KafkaConnect the resource.
|
KafkaConnector
resources provide the configuration needed to create connectors within a Kafka Connect cluster, which interacts with a Kafka cluster as specified in the KafkaConnect
configuration.
The Kafka cluster does not need to be managed by Strimzi or deployed to a Kubernetes cluster.
The configuration also specifies how the connector instances interact with external data systems, including any required authentication methods. Additionally, you must define the data to watch. For example, in a source connector that reads data from a database, the configuration might include the database name. You can also define where this data should be placed in Kafka by specifying the target topic name.
Use the tasksMax
property to specify the maximum number of tasks.
For instance, a source connector with tasksMax: 2
might split the import of source data into two tasks.
apiVersion: kafka.strimzi.io/v1beta2
kind: KafkaConnector
metadata:
name: my-source-connector # (1)
labels:
strimzi.io/cluster: my-connect-cluster # (2)
spec:
class: org.apache.kafka.connect.file.FileStreamSourceConnector # (3)
tasksMax: 2 # (4)
autoRestart: # (5)
enabled: true
config: # (6)
file: "/opt/kafka/LICENSE" # (7)
topic: my-topic # (8)
# ...
-
Name of the
KafkaConnector
resource, which is used as the name of the connector. Use any name that is valid for a Kubernetes resource. -
Name of the Kafka Connect cluster to create the connector instance in. Connectors must be deployed to the same namespace as the Kafka Connect cluster they link to.
-
Full name of the connector class. This should be present in the image being used by the Kafka Connect cluster.
-
Maximum number of Kafka Connect tasks that the connector can create.
-
Enables automatic restarts of failed connectors and tasks. By default, the number of restarts is indefinite, but you can set a maximum on the number of automatic restarts using the
maxRestarts
property. -
Connector configuration as key-value pairs.
-
Location of the external data file. In this example, we’re configuring the
FileStreamSourceConnector
to read from the/opt/kafka/LICENSE
file. -
Kafka topic to publish the source data to.
To include external connector configurations, such as user access credentials stored in a secret, use the template
property of the KafkaConnect
resource.
You can also load values using configuration providers.
11.9.1. Manually stopping or pausing Kafka Connect connectors
If you are using KafkaConnector
resources to configure connectors, use the state
configuration to either stop or pause a connector.
In contrast to the paused state, where the connector and tasks remain instantiated, stopping a connector retains only the configuration, with no active processes.
Stopping a connector from running may be more suitable for longer durations than just pausing.
While a paused connector is quicker to resume, a stopped connector has the advantages of freeing up memory and resources.
Note
|
The state configuration replaces the (deprecated) pause configuration in the KafkaConnectorSpec schema, which allows pauses on connectors.
If you were previously using the pause configuration to pause connectors, we encourage you to transition to using the state configuration only to avoid conflicts.
|
-
The Cluster Operator is running.
-
Find the name of the
KafkaConnector
custom resource that controls the connector you want to pause or stop:kubectl get KafkaConnector
-
Edit the
KafkaConnector
resource to stop or pause the connector.Example configuration for stopping a Kafka Connect connectorapiVersion: kafka.strimzi.io/v1beta2 kind: KafkaConnector metadata: name: my-source-connector labels: strimzi.io/cluster: my-connect-cluster spec: class: org.apache.kafka.connect.file.FileStreamSourceConnector tasksMax: 2 config: file: "/opt/kafka/LICENSE" topic: my-topic state: stopped # ...
Change the
state
configuration tostopped
orpaused
. The default state for the connector when this property is not set isrunning
. -
Apply the changes to the
KafkaConnector
configuration.You can resume the connector by changing
state
torunning
or removing the configuration.
Note
|
Alternatively, you can expose the Kafka Connect API and use the stop and pause endpoints to stop a connector from running.
For example, PUT /connectors/<connector_name>/stop .
You can then use the resume endpoint to restart it.
|
11.9.2. Manually restarting Kafka Connect connectors
If you are using KafkaConnector
resources to manage connectors, use the strimzi.io/restart
annotation to manually trigger a restart of a connector.
-
The Cluster Operator is running.
-
Find the name of the
KafkaConnector
custom resource that controls the Kafka connector you want to restart:kubectl get KafkaConnector
-
Restart the connector by annotating the
KafkaConnector
resource in Kubernetes:kubectl annotate KafkaConnector <kafka_connector_name> strimzi.io/restart="true"
The
restart
annotation is set totrue
. -
Wait for the next reconciliation to occur (every two minutes by default).
The Kafka connector is restarted, as long as the annotation was detected by the reconciliation process. When Kafka Connect accepts the restart request, the annotation is removed from the
KafkaConnector
custom resource.
11.9.3. Manually restarting Kafka Connect connector tasks
If you are using KafkaConnector
resources to manage connectors, use the strimzi.io/restart-task
annotation to manually trigger a restart of a connector task.
-
The Cluster Operator is running.
-
Find the name of the
KafkaConnector
custom resource that controls the Kafka connector task you want to restart:kubectl get KafkaConnector
-
Find the ID of the task to be restarted from the
KafkaConnector
custom resource:kubectl describe KafkaConnector <kafka_connector_name>
Task IDs are non-negative integers, starting from 0.
-
Use the ID to restart the connector task by annotating the
KafkaConnector
resource in Kubernetes:kubectl annotate KafkaConnector <kafka_connector_name> strimzi.io/restart-task="0"
In this example, task
0
is restarted. -
Wait for the next reconciliation to occur (every two minutes by default).
The Kafka connector task is restarted, as long as the annotation was detected by the reconciliation process. When Kafka Connect accepts the restart request, the annotation is removed from the
KafkaConnector
custom resource.
11.9.4. Listing connector offsets
To track connector offsets using KafkaConnector
resources, add the listOffsets
configuration.
The offsets, which keep track of the flow of data, are written to a config map specified in the configuration.
If the config map does not exist, Strimzi creates it.
After the configuration is in place, annotate the KafkaConnector
resource to write the list to the config map.
Sink connectors use Kafka’s standard consumer offset mechanism, while source connectors store offsets in a custom format within a Kafka topic.
-
For sink connectors, the list shows Kafka topic partitions and the last committed offset for each partition.
-
For source connectors, the list shows the source system’s partition and the last offset processed.
-
The Cluster Operator is running.
-
Edit the
KafkaConnector
resource for the connector to include thelistOffsets
configuration.Example configuration to list offsetsapiVersion: kafka.strimzi.io/v1beta2 kind: KafkaConnector metadata: name: my-source-connector labels: strimzi.io/cluster: my-connect-cluster spec: listOffsets: toConfigMap: # (1) name: my-connector-offsets # (2) # ...
-
The reference to the config map where the list of offsets will be written to.
-
The name of the config map, which is named
my-connector-offsets
in this example.
-
-
Run the command to write the list to the config map by annotating the
KafkaConnector
resource:kubectl annotate kafkaconnector my-source-connector strimzi.io/connector-offsets=list -n <namespace>
The annotation remains until either the list operation succeeds or it is manually removed from the resource.
-
After the
KafkaConnector
resource is updated, use the following command to check if the config map with the offsets was created:kubectl get configmap my-connector-offsets -n <namespace>
-
Inspect the contents of the config map to verify the offsets are being listed:
kubectl describe configmap my-connector-offsets -n <namespace>
Strimzi puts the offset information into the
offsets.json
property. This does not overwrite any other properties when updating an existing config map.Example source connector offset listapiVersion: v1 kind: ConfigMap metadata: # ... ownerReferences: # (1) - apiVersion: kafka.strimzi.io/v1beta2 blockOwnerDeletion: false controller: false kind: KafkaConnector name: my-source-connector uid: 637e3be7-bd96-43ab-abde-c55b4c4550e0 resourceVersion: "66951" uid: 641d60a9-36eb-4f29-9895-8f2c1eb9638e data: offsets.json: |- { "offsets" : [ { "partition" : { "filename" : "/data/myfile.txt" # (2) }, "offset" : { "position" : 15295 # (3) } } ] }
-
The owner reference pointing to the
KafkaConnector
resource for the source connector. To provide a custom owner reference, create the config map in advance and set the owner reference. -
The source partition, represented by the filename
/data/myfile.txt
in this example for a file-based connector. -
The last processed offset position in the source partition.
Example sink connector offset listapiVersion: v1 kind: ConfigMap metadata: # ... ownerReferences: # (1) - apiVersion: kafka.strimzi.io/v1beta2 blockOwnerDeletion: false controller: false kind: KafkaConnector name: my-sink-connector uid: 84a29d7f-77e6-43ac-bfbb-719f9b9a4b3b resourceVersion: "79241" uid: 721e30bc-23df-41a2-9b48-fb2b7d9b042c data: offsets.json: |- { "offsets": [ { "partition": { "kafka_topic": "my-topic", # (2) "kafka_partition": 2 # (3) }, "offset": { "kafka_offset": 4 # (4) } } ] }
-
The owner reference pointing to the
KafkaConnector
resource for the sink connector. -
The Kafka topic that the sink connector is consuming from.
-
The partition of the Kafka topic.
-
The last committed Kafka offset for this topic and partition.
-
11.9.5. Altering connector offsets
To alter connector offsets using KafkaConnector
resources, configure the resource to stop the connector and add alterOffsets
configuration to specify the offset changes in a config map.
You can reuse the same config map used to list offsets.
After the connector is stopped and the configuration is in place, annotate the KafkaConnector
resource to apply the offset alteration, then restart the connector.
Altering connector offsets can be useful, for example, to skip a poison record or replay a record.
In this procedure, we alter the offset position for a source connector named my-source-connector
.
-
The Cluster Operator is running.
-
Edit the
KafkaConnector
resource to stop the connector and include thealterOffsets
configuration.Example configuration to stop a connector and alter offsetsapiVersion: kafka.strimzi.io/v1beta2 kind: KafkaConnector metadata: name: my-source-connector labels: strimzi.io/cluster: my-connect-cluster spec: state: stopped # (1) alterOffsets: fromConfigMap: # (2) name: my-connector-offsets # (3) # ...
-
Changes the state of the connector to
stopped
. The default state for the connector when this property is not set isrunning
. -
The reference to the config map that provides the update.
-
The name of the config map, which is named
my-connector-offsets
in this example.
-
-
Edit the config map to make the alteration.
In this example, we’re resetting the offset position for a source connector to 15000.
Example source connector offset list configurationapiVersion: v1 kind: ConfigMap metadata: # ... data: offsets.json: |- # (1) { "offsets" : [ { "partition" : { "filename" : "/data/myfile.txt" }, "offset" : { "position" : 15000 # (2) } } ] }
-
Edits must be made within the
offsets.json
property. -
The updated offset position in the source partition.
-
-
Run the command to update the offset position by annotating the
KafkaConnector
resource:kubectl annotate kafkaconnector my-source-connector strimzi.io/connector-offsets=alter -n <namespace>
The annotation remains until either the update operation succeeds or it is manually removed from the resource.
-
Check the changes by using the procedure to list connector offsets.
-
Restart the connector by changing the state to
running
.Example configuration to start a connectorapiVersion: kafka.strimzi.io/v1beta2 kind: KafkaConnector metadata: name: my-source-connector labels: strimzi.io/cluster: my-connect-cluster spec: state: running # ...
11.9.6. Resetting connector offsets
To reset connector offsets using KafkaConnector
resources, configure the resource to stop the connector.
After the connector is stopped, annotate the KafkaConnector
resource to clear the offsets, then restart the connector.
In this procedure, we reset the offset position for a source connector named my-source-connector
.
-
The Cluster Operator is running.
-
Edit the
KafkaConnector
resource to stop the connector.Example configuration to stop a connectorapiVersion: kafka.strimzi.io/v1beta2 kind: KafkaConnector metadata: name: my-source-connector labels: strimzi.io/cluster: my-connect-cluster spec: # ... state: stopped # (1) # ...
-
Changes the state of the connector to
stopped
. The default state for the connector when this property is not set isrunning
.
-
-
Run the command to reset the offset position by annotating the
KafkaConnector
resource:kubectl annotate kafkaconnector my-source-connector strimzi.io/connector-offsets=reset -n <namespace>
The annotation remains until either the reset operation succeeds or it is manually removed from the resource.
-
Check the changes by using the procedure to list connector offsets.
After resetting, the
offsets.json
property is empty.Example source connector offset listapiVersion: v1 kind: ConfigMap metadata: # ... data: offsets.json: |- { "offsets" : [] }
-
Restart the connector by changing the state to
running
.Example configuration to start a connectorapiVersion: kafka.strimzi.io/v1beta2 kind: KafkaConnector metadata: name: my-source-connector labels: strimzi.io/cluster: my-connect-cluster spec: state: running # ...
11.10. Configuring Kafka MirrorMaker 2
Update the spec
properties of the KafkaMirrorMaker2
custom resource to configure your MirrorMaker 2 deployment.
MirrorMaker 2 uses source cluster configuration for data consumption and target cluster configuration for data output.
MirrorMaker 2 is based on the Kafka Connect framework, connectors managing the transfer of data between clusters.
You configure MirrorMaker 2 to define the Kafka Connect deployment, including the connection details of the source and target clusters, and then run a set of MirrorMaker 2 connectors to make the connection.
MirrorMaker 2 supports topic configuration synchronization between the source and target clusters. You specify source topics in the MirrorMaker 2 configuration. MirrorMaker 2 monitors the source topics. MirrorMaker 2 detects and propagates changes to the source topics to the remote topics. Changes might include automatically creating missing topics and partitions.
Note
|
In most cases you write to local topics and read from remote topics. Though write operations are not prevented on remote topics, they should be avoided. |
The configuration must specify:
-
Each Kafka cluster
-
Connection information for each cluster, including authentication
-
The replication flow and direction
-
Cluster to cluster
-
Topic to topic
-
For a deeper understanding of the Kafka MirrorMaker 2 cluster configuration options, refer to the Strimzi Custom Resource API Reference.
Note
|
MirrorMaker 2 resource configuration differs from the previous version of MirrorMaker, which is now deprecated. There is currently no legacy support, so any resources must be manually converted into the new format. |
MirrorMaker 2 provides default configuration values for properties such as replication factors. A minimal configuration, with defaults left unchanged, would be something like this example:
apiVersion: kafka.strimzi.io/v1beta2
kind: KafkaMirrorMaker2
metadata:
name: my-mirror-maker2
spec:
version: 3.9.0
connectCluster: "my-cluster-target"
clusters:
- alias: "my-cluster-source"
bootstrapServers: my-cluster-source-kafka-bootstrap:9092
- alias: "my-cluster-target"
bootstrapServers: my-cluster-target-kafka-bootstrap:9092
mirrors:
- sourceCluster: "my-cluster-source"
targetCluster: "my-cluster-target"
sourceConnector: {}
You can configure access control for source and target clusters using mTLS or SASL authentication. This procedure shows a configuration that uses TLS encryption and mTLS authentication for the source and target cluster.
You can specify the topics and consumer groups you wish to replicate from a source cluster in the KafkaMirrorMaker2
resource.
You use the topicsPattern
and groupsPattern
properties to do this.
You can provide a list of names or use a regular expression.
By default, all topics and consumer groups are replicated if you do not set the topicsPattern
and groupsPattern
properties.
You can also replicate all topics and consumer groups by using ".*"
as a regular expression.
However, try to specify only the topics and consumer groups you need to avoid causing any unnecessary extra load on the cluster.
You can tune the configuration to handle high volumes of messages. For more information, see Handling high volumes of messages.
KafkaMirrorMaker2
custom resource configuration# Basic configuration (required)
apiVersion: kafka.strimzi.io/v1beta2
kind: KafkaMirrorMaker2
metadata:
name: my-mirror-maker2
# Deployment specifications
spec:
# Replicas (required)
replicas: 3 # (1)
# Connect cluster name (required)
connectCluster: "my-cluster-target" # (2)
# Cluster configurations (required)
clusters: # (3)
- alias: "my-cluster-source" # (4)
# Authentication (optional)
authentication: # (5)
certificateAndKey:
certificate: source.crt
key: source.key
secretName: my-user-source
type: tls
bootstrapServers: my-cluster-source-kafka-bootstrap:9092 # (6)
# TLS configuration (optional)
tls: # (7)
trustedCertificates:
- pattern: "*.crt"
secretName: my-cluster-source-cluster-ca-cert
- alias: "my-cluster-target" # (8)
# Authentication (optional)
authentication: # (9)
certificateAndKey:
certificate: target.crt
key: target.key
secretName: my-user-target
type: tls
bootstrapServers: my-cluster-target-kafka-bootstrap:9092 # (10)
# Kafka Connect configuration (optional)
config: # (11)
config.storage.replication.factor: 1
offset.storage.replication.factor: 1
status.storage.replication.factor: 1
# TLS configuration (optional)
tls: # (12)
trustedCertificates:
- pattern: "*.crt"
secretName: my-cluster-target-cluster-ca-cert
# Mirroring configurations (required)
mirrors: # (13)
- sourceCluster: "my-cluster-source" # (14)
targetCluster: "my-cluster-target" # (15)
# Topic and group patterns (required)
topicsPattern: "topic1|topic2|topic3" # (16)
groupsPattern: "group1|group2|group3" # (17)
# Source connector configuration (required)
sourceConnector: # (18)
tasksMax: 10 # (19)
autoRestart: # (20)
enabled: true
config:
replication.factor: 1 # (21)
offset-syncs.topic.replication.factor: 1 # (22)
sync.topic.acls.enabled: "false" # (23)
refresh.topics.interval.seconds: 60 # (24)
replication.policy.class: "org.apache.kafka.connect.mirror.IdentityReplicationPolicy" # (25)
# Heartbeat connector configuration (optional)
heartbeatConnector: # (26)
autoRestart:
enabled: true
config:
heartbeats.topic.replication.factor: 1 # (27)
replication.policy.class: "org.apache.kafka.connect.mirror.IdentityReplicationPolicy"
# Checkpoint connector configuration (optional)
checkpointConnector: # (28)
autoRestart:
enabled: true
config:
checkpoints.topic.replication.factor: 1 # (29)
refresh.groups.interval.seconds: 600 # (30)
sync.group.offsets.enabled: true # (31)
sync.group.offsets.interval.seconds: 60 # (32)
emit.checkpoints.interval.seconds: 60 # (33)
replication.policy.class: "org.apache.kafka.connect.mirror.IdentityReplicationPolicy"
# Kafka version (recommended)
version: 3.9.0 # (34)
# Resources requests and limits (recommended)
resources: # (35)
requests:
cpu: "1"
memory: 2Gi
limits:
cpu: "2"
memory: 2Gi
# Logging configuration (optional)
logging: # (36)
type: inline
loggers:
connect.root.logger.level: INFO
# Readiness probe (optional)
readinessProbe: # (37)
initialDelaySeconds: 15
timeoutSeconds: 5
# Liveness probe (optional)
livenessProbe:
initialDelaySeconds: 15
timeoutSeconds: 5
# JVM options (optional)
jvmOptions: # (38)
"-Xmx": "1g"
"-Xms": "1g"
# Custom image (optional)
image: my-org/my-image:latest # (39)
# Rack awareness (optional)
rack:
topologyKey: topology.kubernetes.io/zone # (40)
# Pod template (optional)
template: # (41)
pod:
affinity:
podAntiAffinity:
requiredDuringSchedulingIgnoredDuringExecution:
- labelSelector:
matchExpressions:
- key: application
operator: In
values:
- postgresql
- mongodb
topologyKey: "kubernetes.io/hostname"
connectContainer: # (42)
env:
- name: OTEL_SERVICE_NAME
value: my-otel-service
- name: OTEL_EXPORTER_OTLP_ENDPOINT
value: "http://otlp-host:4317"
# Tracing configuration (optional)
tracing:
type: opentelemetry # (43)
-
The number of replica nodes for the workers that run tasks.
-
Kafka cluster alias for Kafka Connect, which must specify the target Kafka cluster. The Kafka cluster is used by Kafka Connect for its internal topics.
-
Specification for the Kafka clusters being synchronized.
-
Cluster alias for the source Kafka cluster.
-
Authentication for the source cluster, specified as mTLS, token-based OAuth, SASL-based SCRAM-SHA-256/SCRAM-SHA-512, or PLAIN.
-
Bootstrap address for connection to the source Kafka cluster. The address takes the format
<cluster_name>-kafka-bootstrap:<port_number>
. The Kafka cluster doesn’t need to be managed by Strimzi or deployed to a Kubernetes cluster. -
TLS configuration for encrypted connections to the Kafka cluster, with trusted certificates stored in X.509 format within the specified secrets.
-
Cluster alias for the target Kafka cluster.
-
Authentication for the target Kafka cluster is configured in the same way as for the source Kafka cluster.
-
Bootstrap address for connection to the target Kafka cluster. The address takes the format
<cluster_name>-kafka-bootstrap:<port_number>
. The Kafka cluster doesn’t need to be managed by Strimzi or deployed to a Kubernetes cluster. -
Kafka Connect configuration. Standard Apache Kafka configuration may be provided, restricted to those properties not managed directly by Strimzi.
-
TLS encryption for the target Kafka cluster is configured in the same way as for the source Kafka cluster.
-
MirrorMaker 2 connectors.
-
Cluster alias for the source cluster used by the MirrorMaker 2 connectors.
-
Cluster alias for the target cluster used by the MirrorMaker 2 connectors.
-
Topic replication from the source cluster defined as a comma-separated list or regular expression pattern. The source connector replicates the specified topics. The checkpoint connector tracks offsets for the specified topics. Here we request three topics by name.
-
Consumer group replication from the source cluster defined as a comma-separated list or regular expression pattern. The checkpoint connector replicates the specified consumer groups. Here we request three consumer groups by name.
-
Configuration for the
MirrorSourceConnector
that creates remote topics. Theconfig
overrides the default configuration options. -
The maximum number of tasks that the connector may create. Tasks handle the data replication and run in parallel. If the infrastructure supports the processing overhead, increasing this value can improve throughput. Kafka Connect distributes the tasks between members of the cluster. If there are more tasks than workers, workers are assigned multiple tasks. For sink connectors, aim to have one task for each topic partition consumed. For source connectors, the number of tasks that can run in parallel may also depend on the external system. The connector creates fewer than the maximum number of tasks if it cannot achieve the parallelism.
-
Enables automatic restarts of failed connectors and tasks. By default, the number of restarts is indefinite, but you can set a maximum on the number of automatic restarts using the
maxRestarts
property. -
Replication factor for mirrored topics created at the target cluster.
-
Replication factor for the
MirrorSourceConnector
offset-syncs
internal topic that maps the offsets of the source and target clusters. -
When ACL rules synchronization is enabled, ACLs are applied to synchronized topics. The default is
true
. This feature is not compatible with the User Operator. If you are using the User Operator, set this property tofalse
. -
Optional setting to change the frequency of checks for new topics. The default is for a check every 10 minutes.
-
Adds a policy that overrides the automatic renaming of remote topics. Instead of prepending the name with the name of the source cluster, the topic retains its original name. This optional setting is useful for active/passive backups and data migration. The property must be specified for all connectors. For bidirectional (active/active) replication, use the
DefaultReplicationPolicy
class to automatically rename remote topics and specify thereplication.policy.separator
property for all connectors to add a custom separator. -
Configuration for the
MirrorHeartbeatConnector
that performs connectivity checks. Theconfig
overrides the default configuration options. -
Replication factor for the heartbeat topic created at the target cluster.
-
Configuration for the
MirrorCheckpointConnector
that tracks offsets. Theconfig
overrides the default configuration options. -
Replication factor for the checkpoints topic created at the target cluster.
-
Optional setting to change the frequency of checks for new consumer groups. The default is for a check every 10 minutes.
-
Optional setting to synchronize consumer group offsets, which is useful for recovery in an active/passive configuration. Synchronization is not enabled by default.
-
If the synchronization of consumer group offsets is enabled, you can adjust the frequency of the synchronization.
-
Adjusts the frequency of checks for offset tracking. If you change the frequency of offset synchronization, you might also need to adjust the frequency of these checks.
-
The Kafka Connect and MirrorMaker 2 version, which will always be the same.
-
Requests for reservation of supported resources, currently
cpu
andmemory
, and limits to specify the maximum resources that can be consumed. -
Specified Kafka Connect loggers and log levels added directly (
inline
) or indirectly (external
) through a ConfigMap. A custom Log4j configuration must be placed under thelog4j.properties
orlog4j2.properties
key in the ConfigMap. For the Kafka Connectlog4j.rootLogger
logger, you can set the log level to INFO, ERROR, WARN, TRACE, DEBUG, FATAL or OFF. -
Healthchecks to know when to restart a container (liveness) and when a container can accept traffic (readiness).
-
JVM configuration options to optimize performance for the Virtual Machine (VM) running Kafka MirrorMaker.
-
ADVANCED OPTION: Container image configuration, which is recommended only in special situations.
-
SPECIALIZED OPTION: Rack awareness configuration for the deployment. This is a specialized option intended for a deployment within the same location, not across regions. Use this option if you want connectors to consume from the closest replica rather than the leader replica. In certain cases, consuming from the closest replica can improve network utilization or reduce costs . The
topologyKey
must match a node label containing the rack ID. The example used in this configuration specifies a zone using the standardtopology.kubernetes.io/zone
label. To consume from the closest replica, enable theRackAwareReplicaSelector
in the Kafka broker configuration. -
Template customization. Here a pod is scheduled with anti-affinity, so the pod is not scheduled on nodes with the same hostname.
-
Environment variables are set for distributed tracing.
-
Distributed tracing is enabled by using OpenTelemetry.
11.10.1. Configuring active/active or active/passive modes
You can use MirrorMaker 2 in active/passive or active/active cluster configurations.
- active/active cluster configuration
-
An active/active configuration has two active clusters replicating data bidirectionally. Applications can use either cluster. Each cluster can provide the same data. In this way, you can make the same data available in different geographical locations. As consumer groups are active in both clusters, consumer offsets for replicated topics are not synchronized back to the source cluster.
- active/passive cluster configuration
-
An active/passive configuration has an active cluster replicating data to a passive cluster. The passive cluster remains on standby. You might use the passive cluster for data recovery in the event of system failure.
The expectation is that producers and consumers connect to active clusters only. A MirrorMaker 2 cluster is required at each target destination.
Bidirectional replication (active/active)
The MirrorMaker 2 architecture supports bidirectional replication in an active/active cluster configuration.
Each cluster replicates the data of the other cluster using the concept of source and remote topics. As the same topics are stored in each cluster, remote topics are automatically renamed by MirrorMaker 2 to represent the source cluster. The name of the originating cluster is prepended to the name of the topic.
By flagging the originating cluster, topics are not replicated back to that cluster.
The concept of replication through remote topics is useful when configuring an architecture that requires data aggregation. Consumers can subscribe to source and remote topics within the same cluster, without the need for a separate aggregation cluster.
Unidirectional replication (active/passive)
The MirrorMaker 2 architecture supports unidirectional replication in an active/passive cluster configuration.
You can use an active/passive cluster configuration to make backups or migrate data to another cluster. In this situation, you might not want automatic renaming of remote topics.
You can override automatic renaming by adding IdentityReplicationPolicy
to the source connector configuration.
With this configuration applied, topics retain their original names.
11.10.2. Configuring MirrorMaker 2 for multiple instances
By default, Strimzi configures the group ID and names of the internal topics used by the Kafka Connect framework that MirrorMaker 2 runs on.
When running multiple instances of MirrorMaker 2, and they share the same connectCluster
value, you must change these default settings using the following config
properties:
apiVersion: kafka.strimzi.io/v1beta2
kind: KafkaMirrorMaker2
metadata:
name: my-mirror-maker2
spec:
connectCluster: "my-cluster-target"
clusters:
- alias: "my-cluster-target"
config:
group.id: my-connect-cluster # (1)
offset.storage.topic: my-connect-cluster-offsets # (2)
config.storage.topic: my-connect-cluster-configs # (3)
status.storage.topic: my-connect-cluster-status # (4)
# ...
# ...
-
The Kafka Connect cluster group ID within Kafka.
-
Kafka topic that stores connector offsets.
-
Kafka topic that stores connector and task status configurations.
-
Kafka topic that stores connector and task status updates.
Note
|
Values for the three topics must be the same for all instances with the same group.id .
|
The connectCluster
setting specifies the alias of the target Kafka cluster used by Kafka Connect for its internal topics.
As a result, modifications to the connectCluster
, group ID, and internal topic naming configuration are specific to the target Kafka cluster.
You don’t need to make changes if two MirrorMaker 2 instances are using the same source Kafka cluster or in an active-active mode where each MirrorMaker 2 instance has a different connectCluster
setting and target cluster.
However, if multiple MirrorMaker 2 instances share the same connectCluster
, each instance connecting to the same target Kafka cluster is deployed with the same values.
In practice, this means all instances form a cluster and use the same internal topics.
Multiple instances attempting to use the same internal topics will cause unexpected errors, so you must change the values of these properties for each instance.
11.10.3. Configuring MirrorMaker 2 connectors
Use MirrorMaker 2 connector configuration for the internal connectors that orchestrate the synchronization of data between Kafka clusters.
MirrorMaker 2 consists of the following connectors:
MirrorSourceConnector
-
The source connector replicates topics from a source cluster to a target cluster. It also replicates ACLs and is necessary for the
MirrorCheckpointConnector
to run. MirrorCheckpointConnector
-
The checkpoint connector periodically tracks offsets. If enabled, it also synchronizes consumer group offsets between the source and target cluster.
MirrorHeartbeatConnector
-
The heartbeat connector periodically checks connectivity between the source and target cluster.
The following table describes connector properties and the connectors you configure to use them.
Property | sourceConnector | checkpointConnector | heartbeatConnector |
---|---|---|---|
|
✓ |
✓ |
✓ |
|
✓ |
✓ |
✓ |
|
✓ |
✓ |
✓ |
|
✓ |
✓ |
|
|
✓ |
✓ |
|
|
✓ |
✓ |
|
|
✓ |
||
|
✓ |
||
|
✓ |
||
|
✓ |
||
|
✓ |
||
|
✓ |
||
|
✓ |
||
|
✓ |
||
|
✓ |
||
|
✓ |
||
|
✓ |
||
|
✓ |
||
|
✓ |
||
|
✓ |
||
|
✓ |
||
|
✓ |
||
|
✓ |
||
|
✓ |
||
|
✓ |
||
|
✓ |
||
|
✓ |
||
|
✓ |
Changing the location of the consumer group offsets topic
MirrorMaker 2 tracks offsets for consumer groups using internal topics.
offset-syncs
topic-
The
offset-syncs
topic maps the source and target offsets for replicated topic partitions from record metadata. checkpoints
topic-
The
checkpoints
topic maps the last committed offset in the source and target cluster for replicated topic partitions in each consumer group.
As they are used internally by MirrorMaker 2, you do not interact directly with these topics.
MirrorCheckpointConnector
emits checkpoints for offset tracking.
Offsets for the checkpoints
topic are tracked at predetermined intervals through configuration.
Both topics enable replication to be fully restored from the correct offset position on failover.
The location of the offset-syncs
topic is the source
cluster by default.
You can use the offset-syncs.topic.location
connector configuration to change this to the target
cluster.
You need read/write access to the cluster that contains the topic.
Using the target cluster as the location of the offset-syncs
topic allows you to use MirrorMaker 2 even if you have only read access to the source cluster.
Synchronizing consumer group offsets
The __consumer_offsets
topic stores information on committed offsets for each consumer group.
Offset synchronization periodically transfers the consumer offsets for the consumer groups of a source cluster into the consumer offsets topic of a target cluster.
Offset synchronization is particularly useful in an active/passive configuration. If the active cluster goes down, consumer applications can switch to the passive (standby) cluster and pick up from the last transferred offset position.
To use topic offset synchronization, enable the synchronization by adding sync.group.offsets.enabled
to the checkpoint connector configuration, and setting the property to true
.
Synchronization is disabled by default.
When using the IdentityReplicationPolicy
in the source connector, it also has to be configured in the checkpoint connector configuration.
This ensures that the mirrored consumer offsets will be applied for the correct topics.
Consumer offsets are only synchronized for consumer groups that are not active in the target cluster.
If the consumer groups are in the target cluster, the synchronization cannot be performed and an UNKNOWN_MEMBER_ID
error is returned.
If enabled, the synchronization of offsets from the source cluster is made periodically.
You can change the frequency by adding sync.group.offsets.interval.seconds
and emit.checkpoints.interval.seconds
to the checkpoint connector configuration.
The properties specify the frequency in seconds that the consumer group offsets are synchronized, and the frequency of checkpoints emitted for offset tracking.
The default for both properties is 60 seconds.
You can also change the frequency of checks for new consumer groups using the refresh.groups.interval.seconds
property, which is performed every 10 minutes by default.
Because the synchronization is time-based, any switchover by consumers to a passive cluster will likely result in some duplication of messages.
Note
|
If you have an application written in Java, you can use the RemoteClusterUtils.java utility to synchronize offsets through the application. The utility fetches remote offsets for a consumer group from the checkpoints topic.
|
Deciding when to use the heartbeat connector
The heartbeat connector emits heartbeats to check connectivity between source and target Kafka clusters.
An internal heartbeat
topic is replicated from the source cluster, which means that the heartbeat connector must be connected to the source cluster.
The heartbeat
topic is located on the target cluster, which allows it to do the following:
-
Identify all source clusters it is mirroring data from
-
Verify the liveness and latency of the mirroring process
This helps to make sure that the process is not stuck or has stopped for any reason. While the heartbeat connector can be a valuable tool for monitoring the mirroring processes between Kafka clusters, it’s not always necessary to use it. For example, if your deployment has low network latency or a small number of topics, you might prefer to monitor the mirroring process using log messages or other monitoring tools. If you decide not to use the heartbeat connector, simply omit it from your MirrorMaker 2 configuration.
Aligning the configuration of MirrorMaker 2 connectors
To ensure that MirrorMaker 2 connectors work properly, make sure to align certain configuration settings across connectors. Specifically, ensure that the following properties have the same value across all applicable connectors:
-
replication.policy.class
-
replication.policy.separator
-
offset-syncs.topic.location
-
topic.filter.class
For example, the value for replication.policy.class
must be the same for the source, checkpoint, and heartbeat connectors.
Mismatched or missing settings cause issues with data replication or offset syncing, so it’s essential to keep all relevant connectors configured with the same settings.
Listing the offsets of MirrorMaker 2 connectors
To list the offset positions of the internal MirrorMaker 2 connectors, use the same configuration that’s used to manage Kafka Connect connectors. For more information on setting up the configuration and listing offsets, see Listing connector offsets.
In this example, the sourceConnector
configuration is updated to return the connector offset position.
The offset information is written to a specified config map.
apiVersion: kafka.strimzi.io/v1beta2
kind: KafkaMirrorMaker2
metadata:
name: my-mirror-maker2
spec:
version: 3.9.0
# ...
clusters:
- alias: "my-cluster-source"
bootstrapServers: my-cluster-source-kafka-bootstrap:9092
- alias: "my-cluster-target"
bootstrapServers: my-cluster-target-kafka-bootstrap:9092
mirrors:
- sourceCluster: "my-cluster-source"
targetCluster: "my-cluster-target"
sourceConnector:
listOffsets:
toConfigMap:
name: my-connector-offsets
# ...
You must apply the following annotations to the KafkaMirrorMaker2
resource be able to manage connector offsets:
-
strimzi.io/connector-offsets
-
strimzi.io/mirrormaker-connector
The strimzi.io/mirrormaker-connector
annotation must be set to the name of the connector.
These annotations remain until the operation succeeds or they are manually removed from the resource.
MirrorMaker 2 connectors are named using the aliases of the source and target clusters, followed by the connector type: <source_alias>-><target_alias>.<connector_type>
.
In the following example, the annotations are applied for a connector named my-cluster-source->my-cluster-target.MirrorSourceConnector
.
kubectl annotate kafkamirrormaker2 my-mirror-maker-2 strimzi.io/connector-offsets=list strimzi.io/mirrormaker-connector="my-cluster-source->my-cluster-target.MirrorSourceConnector" -n kafka
The offsets are listed in the specified config map.
Strimzi puts the offset information into a .json
property named after the connector.
This does not overwrite any other properties when updating an existing config map.
apiVersion: v1
kind: ConfigMap
metadata:
# ...
ownerReferences: # (1)
- apiVersion: kafka.strimzi.io/v1beta2
blockOwnerDeletion: false
controller: false
kind: KafkaMirrorMaker2
name: my-mirror-maker2
uid: 637e3be7-bd96-43ab-abde-c55b4c4550e0
data:
my-cluster-source--my-cluster-target.MirrorSourceConnector.json: |- # (2)
{
"offsets": [
{
"partition": {
"cluster": "east-kafka",
"partition": 0,
"topic": "mirrormaker2-cluster-configs"
},
"offset": {
"offset": 0
}
}
]
}
-
The owner reference pointing to the
KafkaMirrorMaker2
resource. To provide a custom owner reference, create the config map in advance and set the owner reference. -
The
.json
property uses the connector name. Since->
characters are not allowed in config map keys,->
is changed to--
in the connector name.
11.10.4. Configuring MirrorMaker 2 connector producers and consumers
MirrorMaker 2 connectors use internal producers and consumers. If needed, you can configure these producers and consumers to override the default settings.
For example, you can increase the batch.size
for the source producer that sends topics to the target Kafka cluster to better accommodate large volumes of messages.
Important
|
Producer and consumer configuration options depend on the MirrorMaker 2 implementation, and may be subject to change. |
The following tables describe the producers and consumers for each of the connectors and where you can add configuration.
Type | Description | Configuration |
---|---|---|
Producer |
Sends topic messages to the target Kafka cluster. Consider tuning the configuration of this producer when it is handling large volumes of data. |
|
Producer |
Writes to the |
|
Consumer |
Retrieves topic messages from the source Kafka cluster. |
|
Type | Description | Configuration |
---|---|---|
Producer |
Emits consumer offset checkpoints. |
|
Consumer |
Loads the |
|
Note
|
You can set offset-syncs.topic.location to target to use the target Kafka cluster as the location of the offset-syncs topic.
|
Type | Description | Configuration |
---|---|---|
Producer |
Emits heartbeats. |
|
The following example shows how you configure the producers and consumers.
apiVersion: kafka.strimzi.io/v1beta2
kind: KafkaMirrorMaker2
metadata:
name: my-mirror-maker2
spec:
version: 3.9.0
# ...
mirrors:
- sourceCluster: "my-cluster-source"
targetCluster: "my-cluster-target"
sourceConnector:
tasksMax: 5
config:
producer.override.batch.size: 327680
producer.override.linger.ms: 100
producer.request.timeout.ms: 30000
consumer.fetch.max.bytes: 52428800
# ...
checkpointConnector:
config:
producer.override.request.timeout.ms: 30000
consumer.max.poll.interval.ms: 300000
# ...
heartbeatConnector:
config:
producer.override.request.timeout.ms: 30000
# ...
11.10.5. Specifying a maximum number of data replication tasks
Connectors create the tasks that are responsible for moving data in and out of Kafka. Each connector comprises one or more tasks that are distributed across a group of worker pods that run the tasks. Increasing the number of tasks can help with performance issues when replicating a large number of partitions or synchronizing the offsets of a large number of consumer groups.
Tasks run in parallel. Workers are assigned one or more tasks. A single task is handled by one worker pod, so you don’t need more worker pods than tasks. If there are more tasks than workers, workers handle multiple tasks.
You can specify the maximum number of connector tasks in your MirrorMaker configuration using the tasksMax
property.
Without specifying a maximum number of tasks, the default setting is a single task.
The heartbeat connector always uses a single task.
The number of tasks that are started for the source and checkpoint connectors is the lower value between the maximum number of possible tasks and the value for tasksMax
.
For the source connector, the maximum number of tasks possible is one for each partition being replicated from the source cluster.
For the checkpoint connector, the maximum number of tasks possible is one for each consumer group being replicated from the source cluster.
When setting a maximum number of tasks, consider the number of partitions and the hardware resources that support the process.
If the infrastructure supports the processing overhead, increasing the number of tasks can improve throughput and latency. For example, adding more tasks reduces the time taken to poll the source cluster when there is a high number of partitions or consumer groups.
Increasing the number of tasks for the source connector is useful when you have a large number of partitions.
apiVersion: kafka.strimzi.io/v1beta2
kind: KafkaMirrorMaker2
metadata:
name: my-mirror-maker2
spec:
# ...
mirrors:
- sourceCluster: "my-cluster-source"
targetCluster: "my-cluster-target"
sourceConnector:
tasksMax: 10
# ...
Increasing the number of tasks for the checkpoint connector is useful when you have a large number of consumer groups.
apiVersion: kafka.strimzi.io/v1beta2
kind: KafkaMirrorMaker2
metadata:
name: my-mirror-maker2
spec:
# ...
mirrors:
- sourceCluster: "my-cluster-source"
targetCluster: "my-cluster-target"
checkpointConnector:
tasksMax: 10
# ...
By default, MirrorMaker 2 checks for new consumer groups every 10 minutes.
You can adjust the refresh.groups.interval.seconds
configuration to change the frequency.
Take care when adjusting lower.
More frequent checks can have a negative impact on performance.
Checking connector task operations
If you are using Prometheus and Grafana to monitor your deployment, you can check MirrorMaker 2 performance. The example MirrorMaker 2 Grafana dashboard provided with Strimzi shows the following metrics related to tasks and latency.
-
The number of tasks
-
Replication latency
-
Offset synchronization latency
11.10.6. Synchronizing ACL rules for remote topics
When using MirrorMaker 2 with Strimzi, it is possible to synchronize ACL rules for remote topics. However, this feature is only available if you are not using the User Operator.
If you are using type: simple
authorization without the User Operator, the ACL rules that manage access to brokers also apply to remote topics.
This means that users who have read access to a source topic can also read its remote equivalent.
Note
|
OAuth 2.0 authorization does not support access to remote topics in this way. |
11.10.7. Securing a Kafka MirrorMaker 2 deployment
This procedure describes in outline the configuration required to secure a MirrorMaker 2 deployment.
You need separate configuration for the source Kafka cluster and the target Kafka cluster. You also need separate user configuration to provide the credentials required for MirrorMaker to connect to the source and target Kafka clusters.
For the Kafka clusters, you specify internal listeners for secure connections within a Kubernetes cluster and external listeners for connections outside the Kubernetes cluster.
You can configure authentication and authorization mechanisms. The security options implemented for the source and target Kafka clusters must be compatible with the security options implemented for MirrorMaker 2.
After you have created the cluster and user authentication credentials, you specify them in your MirrorMaker configuration for secure connections.
Note
|
In this procedure, the certificates generated by the Cluster Operator are used, but you can replace them by installing your own certificates. You can also configure your listener to use a Kafka listener certificate managed by an external CA (certificate authority). |
Before starting this procedure, take a look at the example configuration files provided by Strimzi. They include examples for securing a deployment of MirrorMaker 2 using mTLS or SCRAM-SHA-512 authentication. The examples specify internal listeners for connecting within a Kubernetes cluster.
The examples also provide the configuration for full authorization, including the ACLs that allow user operations on the source and target Kafka clusters.
When configuring user access to source and target Kafka clusters, ACLs must grant access rights to internal MirrorMaker 2 connectors and read/write access to the cluster group and internal topics used by the underlying Kafka Connect framework in the target cluster. If you’ve renamed the cluster group or internal topics, such as when configuring MirrorMaker 2 for multiple instances, use those names in the ACLs configuration.
Simple authorization uses ACL rules managed by the Kafka AclAuthorizer
and StandardAuthorizer
plugins to ensure appropriate access levels.
For more information on configuring a KafkaUser
resource to use simple authorization, see the AclRule
schema reference.
-
Strimzi is running
-
Separate namespaces for source and target clusters
The procedure assumes that the source and target Kafka clusters are installed to separate namespaces. If you want to use the Topic Operator, you’ll need to do this. The Topic Operator only watches a single cluster in a specified namespace.
By separating the clusters into namespaces, you will need to copy the cluster secrets so they can be accessed outside the namespace. You need to reference the secrets in the MirrorMaker configuration.
-
Configure two
Kafka
resources, one to secure the source Kafka cluster and one to secure the target Kafka cluster.You can add listener configuration for authentication and enable authorization.
In this example, an internal listener is configured for a Kafka cluster with TLS encryption and mTLS authentication. Kafka
simple
authorization is enabled.Example source Kafka cluster configuration with TLS encryption and mTLS authenticationapiVersion: kafka.strimzi.io/v1beta2 kind: Kafka metadata: name: my-source-cluster spec: kafka: version: 3.9.0 replicas: 1 listeners: - name: tls port: 9093 type: internal tls: true authentication: type: tls authorization: type: simple config: offsets.topic.replication.factor: 1 transaction.state.log.replication.factor: 1 transaction.state.log.min.isr: 1 default.replication.factor: 1 min.insync.replicas: 1 inter.broker.protocol.version: "3.9" storage: type: jbod volumes: - id: 0 type: persistent-claim size: 100Gi deleteClaim: false zookeeper: replicas: 1 storage: type: persistent-claim size: 100Gi deleteClaim: false entityOperator: topicOperator: {} userOperator: {}
Example target Kafka cluster configuration with TLS encryption and mTLS authenticationapiVersion: kafka.strimzi.io/v1beta2 kind: Kafka metadata: name: my-target-cluster spec: kafka: version: 3.9.0 replicas: 1 listeners: - name: tls port: 9093 type: internal tls: true authentication: type: tls authorization: type: simple config: offsets.topic.replication.factor: 1 transaction.state.log.replication.factor: 1 transaction.state.log.min.isr: 1 default.replication.factor: 1 min.insync.replicas: 1 inter.broker.protocol.version: "3.9" storage: type: jbod volumes: - id: 0 type: persistent-claim size: 100Gi deleteClaim: false zookeeper: replicas: 1 storage: type: persistent-claim size: 100Gi deleteClaim: false entityOperator: topicOperator: {} userOperator: {}
-
Create or update the
Kafka
resources in separate namespaces.kubectl apply -f <kafka_configuration_file> -n <namespace>
The Cluster Operator creates the listeners and sets up the cluster and client certificate authority (CA) certificates to enable authentication within the Kafka cluster.
The certificates are created in the secret
<cluster_name>-cluster-ca-cert
. -
Configure two
KafkaUser
resources, one for a user of the source Kafka cluster and one for a user of the target Kafka cluster.-
Configure the same authentication and authorization types as the corresponding source and target Kafka cluster. For example, if you used
tls
authentication and thesimple
authorization type in theKafka
configuration for the source Kafka cluster, use the same in theKafkaUser
configuration. -
Configure the ACLs needed by MirrorMaker 2 to allow operations on the source and target Kafka clusters.
Example source user configuration for mTLS authenticationapiVersion: kafka.strimzi.io/v1beta2 kind: KafkaUser metadata: name: my-source-user labels: strimzi.io/cluster: my-source-cluster spec: authentication: type: tls authorization: type: simple acls: # MirrorSourceConnector - resource: # Not needed if offset-syncs.topic.location=target type: topic name: mm2-offset-syncs.my-target-cluster.internal operations: - Create - DescribeConfigs - Read - Write - resource: # Needed for every topic which is mirrored type: topic name: "*" operations: - DescribeConfigs - Read # MirrorCheckpointConnector - resource: type: cluster operations: - Describe - resource: # Needed for every group for which offsets are synced type: group name: "*" operations: - Describe - resource: # Not needed if offset-syncs.topic.location=target type: topic name: mm2-offset-syncs.my-target-cluster.internal operations: - Read
Example target user configuration for mTLS authenticationapiVersion: kafka.strimzi.io/v1beta2 kind: KafkaUser metadata: name: my-target-user labels: strimzi.io/cluster: my-target-cluster spec: authentication: type: tls authorization: type: simple acls: # cluster group - resource: type: group name: mirrormaker2-cluster operations: - Read # access to config.storage.topic - resource: type: topic name: mirrormaker2-cluster-configs operations: - Create - Describe - DescribeConfigs - Read - Write # access to status.storage.topic - resource: type: topic name: mirrormaker2-cluster-status operations: - Create - Describe - DescribeConfigs - Read - Write # access to offset.storage.topic - resource: type: topic name: mirrormaker2-cluster-offsets operations: - Create - Describe - DescribeConfigs - Read - Write # MirrorSourceConnector - resource: # Needed for every topic which is mirrored type: topic name: "*" operations: - Create - Alter - AlterConfigs - Write # MirrorCheckpointConnector - resource: type: cluster operations: - Describe - resource: type: topic name: my-source-cluster.checkpoints.internal operations: - Create - Describe - Read - Write - resource: # Needed for every group for which the offset is synced type: group name: "*" operations: - Read - Describe # MirrorHeartbeatConnector - resource: type: topic name: heartbeats operations: - Create - Describe - Write
NoteYou can use a certificate issued outside the User Operator by setting type
totls-external
. For more information, see theKafkaUserSpec
schema reference. -
-
Create or update a
KafkaUser
resource in each of the namespaces you created for the source and target Kafka clusters.kubectl apply -f <kafka_user_configuration_file> -n <namespace>
The User Operator creates the users representing the client (MirrorMaker), and the security credentials used for client authentication, based on the chosen authentication type.
The User Operator creates a new secret with the same name as the
KafkaUser
resource. The secret contains a private and public key for mTLS authentication. The public key is contained in a user certificate, which is signed by the clients CA. -
Configure a
KafkaMirrorMaker2
resource with the authentication details to connect to the source and target Kafka clusters.Example MirrorMaker 2 configuration with TLS encryption and mTLS authenticationapiVersion: kafka.strimzi.io/v1beta2 kind: KafkaMirrorMaker2 metadata: name: my-mirror-maker-2 spec: version: 3.9.0 replicas: 1 connectCluster: "my-target-cluster" clusters: - alias: "my-source-cluster" bootstrapServers: my-source-cluster-kafka-bootstrap:9093 tls: # (1) trustedCertificates: - secretName: my-source-cluster-cluster-ca-cert pattern: "*.crt" authentication: # (2) type: tls certificateAndKey: secretName: my-source-user certificate: user.crt key: user.key - alias: "my-target-cluster" bootstrapServers: my-target-cluster-kafka-bootstrap:9093 tls: # (3) trustedCertificates: - secretName: my-target-cluster-cluster-ca-cert pattern: "*.crt" authentication: # (4) type: tls certificateAndKey: secretName: my-target-user certificate: user.crt key: user.key config: # -1 means it will use the default replication factor configured in the broker config.storage.replication.factor: -1 offset.storage.replication.factor: -1 status.storage.replication.factor: -1 mirrors: - sourceCluster: "my-source-cluster" targetCluster: "my-target-cluster" sourceConnector: config: replication.factor: 1 offset-syncs.topic.replication.factor: 1 sync.topic.acls.enabled: "false" heartbeatConnector: config: heartbeats.topic.replication.factor: 1 checkpointConnector: config: checkpoints.topic.replication.factor: 1 sync.group.offsets.enabled: "true" topicsPattern: "topic1|topic2|topic3" groupsPattern: "group1|group2|group3"
-
The TLS certificates for the source Kafka cluster. If they are in a separate namespace, copy the cluster secrets from the namespace of the Kafka cluster.
-
The user authentication for accessing the source Kafka cluster using the TLS mechanism.
-
The TLS certificates for the target Kafka cluster.
-
The user authentication for accessing the target Kafka cluster.
-
-
Create or update the
KafkaMirrorMaker2
resource in the same namespace as the target Kafka cluster.kubectl apply -f <mirrormaker2_configuration_file> -n <namespace_of_target_cluster>
11.10.8. Manually stopping or pausing MirrorMaker 2 connectors
If you are using KafkaMirrorMaker2
resources to configure internal MirrorMaker connectors, use the state
configuration to either stop or pause a connector.
In contrast to the paused state, where the connector and tasks remain instantiated, stopping a connector retains only the configuration, with no active processes.
Stopping a connector from running may be more suitable for longer durations than just pausing.
While a paused connector is quicker to resume, a stopped connector has the advantages of freeing up memory and resources.
Note
|
The state configuration replaces the (deprecated) pause configuration in the KafkaMirrorMaker2ConnectorSpec schema, which allows pauses on connectors.
If you were previously using the pause configuration to pause connectors, we encourage you to transition to using the state configuration only to avoid conflicts.
|
-
The Cluster Operator is running.
-
Find the name of the
KafkaMirrorMaker2
custom resource that controls the MirrorMaker 2 connector you want to pause or stop:kubectl get KafkaMirrorMaker2
-
Edit the
KafkaMirrorMaker2
resource to stop or pause the connector.Example configuration for stopping a MirrorMaker 2 connectorapiVersion: kafka.strimzi.io/v1beta2 kind: KafkaMirrorMaker2 metadata: name: my-mirror-maker2 spec: version: 3.9.0 replicas: 3 connectCluster: "my-cluster-target" clusters: # ... mirrors: - sourceCluster: "my-cluster-source" targetCluster: "my-cluster-target" sourceConnector: tasksMax: 10 autoRestart: enabled: true state: stopped # ...
Change the
state
configuration tostopped
orpaused
. The default state for the connector when this property is not set isrunning
. -
Apply the changes to the
KafkaMirrorMaker2
configuration.You can resume the connector by changing
state
torunning
or removing the configuration.
Note
|
Alternatively, you can expose the Kafka Connect API and use the stop and pause endpoints to stop a connector from running.
For example, PUT /connectors/<connector_name>/stop .
You can then use the resume endpoint to restart it.
|
11.10.9. Manually restarting MirrorMaker 2 connectors
Use the strimzi.io/restart-connector
annotation to manually trigger a restart of a MirrorMaker 2 connector.
-
The Cluster Operator is running.
-
Find the name of the
KafkaMirrorMaker2
custom resource that controls the Kafka MirrorMaker 2 connector you want to restart:kubectl get KafkaMirrorMaker2
-
Find the name of the Kafka MirrorMaker 2 connector to be restarted from the
KafkaMirrorMaker2
custom resource:kubectl describe KafkaMirrorMaker2 <mirrormaker_cluster_name>
-
Use the name of the connector to restart the connector by annotating the
KafkaMirrorMaker2
resource in Kubernetes:kubectl annotate KafkaMirrorMaker2 <mirrormaker_cluster_name> "strimzi.io/restart-connector=<mirrormaker_connector_name>"
In this example, connector
my-connector
in themy-mirror-maker-2
cluster is restarted:kubectl annotate KafkaMirrorMaker2 my-mirror-maker-2 "strimzi.io/restart-connector=my-connector"
-
Wait for the next reconciliation to occur (every two minutes by default).
The MirrorMaker 2 connector is restarted, as long as the annotation was detected by the reconciliation process. When MirrorMaker 2 accepts the request, the annotation is removed from the
KafkaMirrorMaker2
custom resource.
11.10.10. Manually restarting MirrorMaker 2 connector tasks
Use the strimzi.io/restart-connector-task
annotation to manually trigger a restart of a MirrorMaker 2 connector.
-
The Cluster Operator is running.
-
Find the name of the
KafkaMirrorMaker2
custom resource that controls the MirrorMaker 2 connector task you want to restart:kubectl get KafkaMirrorMaker2
-
Find the name of the connector and the ID of the task to be restarted from the
KafkaMirrorMaker2
custom resource:kubectl describe KafkaMirrorMaker2 <mirrormaker_cluster_name>
Task IDs are non-negative integers, starting from 0.
-
Use the name and ID to restart the connector task by annotating the
KafkaMirrorMaker2
resource in Kubernetes:kubectl annotate KafkaMirrorMaker2 <mirrormaker_cluster_name> "strimzi.io/restart-connector-task=<mirrormaker_connector_name>:<task_id>"
In this example, task
0
for connectormy-connector
in themy-mirror-maker-2
cluster is restarted:kubectl annotate KafkaMirrorMaker2 my-mirror-maker-2 "strimzi.io/restart-connector-task=my-connector:0"
-
Wait for the next reconciliation to occur (every two minutes by default).
The MirrorMaker 2 connector task is restarted, as long as the annotation was detected by the reconciliation process. When MirrorMaker 2 accepts the request, the annotation is removed from the
KafkaMirrorMaker2
custom resource.
11.11. Configuring Kafka MirrorMaker (deprecated)
Update the spec
properties of the KafkaMirrorMaker
custom resource to configure your Kafka MirrorMaker deployment.
You can configure access control for producers and consumers using TLS or SASL authentication. This procedure shows a configuration that uses TLS encryption and mTLS authentication on the consumer and producer side.
For a deeper understanding of the Kafka MirrorMaker cluster configuration options, refer to the Strimzi Custom Resource API Reference.
Important
|
Kafka MirrorMaker 1 (referred to as just MirrorMaker in the documentation) has been deprecated in Apache Kafka 3.0.0 and will be removed in Apache Kafka 4.0.0.
As a result, the KafkaMirrorMaker custom resource which is used to deploy Kafka MirrorMaker 1 has been deprecated in Strimzi as well.
The KafkaMirrorMaker resource will be removed from Strimzi when we adopt Apache Kafka 4.0.0.
As a replacement, use the KafkaMirrorMaker2 custom resource with the IdentityReplicationPolicy .
|
KafkaMirrorMaker
custom resource configurationapiVersion: kafka.strimzi.io/v1beta2
kind: KafkaMirrorMaker
metadata:
name: my-mirror-maker
spec:
replicas: 3 # (1)
consumer:
bootstrapServers: my-source-cluster-kafka-bootstrap:9092 # (2)
groupId: "my-group" # (3)
numStreams: 2 # (4)
offsetCommitInterval: 120000 # (5)
tls: # (6)
trustedCertificates:
- secretName: my-source-cluster-ca-cert
pattern: "*.crt"
authentication: # (7)
type: tls
certificateAndKey:
secretName: my-source-secret
certificate: public.crt
key: private.key
config: # (8)
max.poll.records: 100
receive.buffer.bytes: 32768
producer:
bootstrapServers: my-target-cluster-kafka-bootstrap:9092
abortOnSendFailure: false # (9)
tls:
trustedCertificates:
- secretName: my-target-cluster-ca-cert
pattern: "*.crt"
authentication:
type: tls
certificateAndKey:
secretName: my-target-secret
certificate: public.crt
key: private.key
config:
compression.type: gzip
batch.size: 8192
include: "my-topic|other-topic" # (10)
resources: # (11)
requests:
cpu: "1"
memory: 2Gi
limits:
cpu: "2"
memory: 2Gi
logging: # (12)
type: inline
loggers:
mirrormaker.root.logger: INFO
readinessProbe: # (13)
initialDelaySeconds: 15
timeoutSeconds: 5
livenessProbe:
initialDelaySeconds: 15
timeoutSeconds: 5
metricsConfig: # (14)
type: jmxPrometheusExporter
valueFrom:
configMapKeyRef:
name: my-config-map
key: my-key
jvmOptions: # (15)
"-Xmx": "1g"
"-Xms": "1g"
image: my-org/my-image:latest # (16)
template: # (17)
pod:
affinity:
podAntiAffinity:
requiredDuringSchedulingIgnoredDuringExecution:
- labelSelector:
matchExpressions:
- key: application
operator: In
values:
- postgresql
- mongodb
topologyKey: "kubernetes.io/hostname"
mirrorMakerContainer: # (18)
env:
- name: OTEL_SERVICE_NAME
value: my-otel-service
- name: OTEL_EXPORTER_OTLP_ENDPOINT
value: "http://otlp-host:4317"
tracing: # (19)
type: opentelemetry
-
The number of replica nodes.
-
Bootstrap servers for consumer and producer.
-
Group ID for the consumer.
-
The number of consumer streams.
-
The offset auto-commit interval in milliseconds.
-
TLS configuration for encrypted connections to the Kafka cluster, with trusted certificates stored in X.509 format within the specified secrets.
-
Authentication for consumer or producer, specified as mTLS, token-based OAuth, SASL-based SCRAM-SHA-256/SCRAM-SHA-512, or PLAIN.
-
Kafka configuration options for consumer and producer.
-
If the
abortOnSendFailure
property is set totrue
, Kafka MirrorMaker will exit and the container will restart following a send failure for a message. -
A list of included topics mirrored from source to target Kafka cluster.
-
Requests for reservation of supported resources, currently
cpu
andmemory
, and limits to specify the maximum resources that can be consumed. -
Specified loggers and log levels added directly (
inline
) or indirectly (external
) through a ConfigMap. A custom Log4j configuration must be placed under thelog4j.properties
orlog4j2.properties
key in the ConfigMap. MirrorMaker has a single logger calledmirrormaker.root.logger
. You can set the log level to INFO, ERROR, WARN, TRACE, DEBUG, FATAL or OFF. -
Healthchecks to know when to restart a container (liveness) and when a container can accept traffic (readiness).
-
Prometheus metrics, which are enabled by referencing a ConfigMap containing configuration for the Prometheus JMX exporter in this example. You can enable metrics without further configuration using a reference to a ConfigMap containing an empty file under
metricsConfig.valueFrom.configMapKeyRef.key
. -
JVM configuration options to optimize performance for the Virtual Machine (VM) running Kafka MirrorMaker.
-
ADVANCED OPTION: Container image configuration, which is recommended only in special situations.
-
Template customization. Here a pod is scheduled with anti-affinity, so the pod is not scheduled on nodes with the same hostname.
-
Environment variables are set for distributed tracing.
-
Distributed tracing is enabled by using OpenTelemetry.
WarningWith the abortOnSendFailure
property set tofalse
, the producer attempts to send the next message in a topic. The original message might be lost, as there is no attempt to resend a failed message.
11.12. Configuring the Kafka Bridge
Update the spec
properties of the KafkaBridge
custom resource to configure your Kafka Bridge deployment.
In order to prevent issues arising when client consumer requests are processed by different Kafka Bridge instances, address-based routing must be employed to ensure that requests are routed to the right Kafka Bridge instance. Additionally, each independent Kafka Bridge instance must have a replica. A Kafka Bridge instance has its own state which is not shared with another instances.
For a deeper understanding of the Kafka Bridge and its cluster configuration options, refer to the Using the Kafka Bridge guide and the Strimzi Custom Resource API Reference.
KafkaBridge
custom resource configuration# Basic configuration (required)
apiVersion: kafka.strimzi.io/v1beta2
kind: KafkaBridge
metadata:
name: my-bridge
spec:
# Replicas (required)
replicas: 3 # (1)
# Kafka bootstrap servers (required)
bootstrapServers: <cluster_name>-cluster-kafka-bootstrap:9092 # (2)
# HTTP configuration (required)
http: # (3)
port: 8080
# CORS configuration (optional)
cors: # (4)
allowedOrigins: "https://strimzi.io"
allowedMethods: "GET,POST,PUT,DELETE,OPTIONS,PATCH"
# Resources requests and limits (recommended)
resources: # (5)
requests:
cpu: "1"
memory: 2Gi
limits:
cpu: "2"
memory: 2Gi
# TLS configuration (optional)
tls: # (6)
trustedCertificates:
- secretName: my-cluster-cluster-cert
pattern: "*.crt"
- secretName: my-cluster-cluster-cert
certificate: ca2.crt
# Authentication (optional)
authentication: # (7)
type: tls
certificateAndKey:
secretName: my-secret
certificate: public.crt
key: private.key
# Consumer configuration (optional)
consumer: # (8)
config:
auto.offset.reset: earliest
# Producer configuration (optional)
producer: # (9)
config:
delivery.timeout.ms: 300000
# Logging configuration (optional)
logging: # (10)
type: inline
loggers:
logger.bridge.level: INFO
# Enabling DEBUG just for send operation
logger.send.name: "http.openapi.operation.send"
logger.send.level: DEBUG
# JVM options (optional)
jvmOptions: # (11)
"-Xmx": "1g"
"-Xms": "1g"
# Readiness probe (optional)
readinessProbe: # (12)
initialDelaySeconds: 15
timeoutSeconds: 5
# Liveness probe (optional)
livenessProbe:
initialDelaySeconds: 15
timeoutSeconds: 5
# Custom image (optional)
image: my-org/my-image:latest # (13)
# Pod template (optional)
template: # (14)
pod:
affinity:
podAntiAffinity:
requiredDuringSchedulingIgnoredDuringExecution:
- labelSelector:
matchExpressions:
- key: application
operator: In
values:
- postgresql
- mongodb
topologyKey: "kubernetes.io/hostname"
bridgeContainer: # (15)
env:
- name: OTEL_SERVICE_NAME
value: my-otel-service
- name: OTEL_EXPORTER_OTLP_ENDPOINT
value: "http://otlp-host:4317"
# Tracing configuration (optional)
tracing:
type: opentelemetry # (16)
-
The number of replica nodes.
-
Bootstrap address for connection to the target Kafka cluster. The address takes the format
<cluster_name>-kafka-bootstrap:<port_number>
. The Kafka cluster doesn’t need to be managed by Strimzi or deployed to a Kubernetes cluster. -
HTTP access to Kafka brokers.
-
CORS access specifying selected resources and access methods. Additional HTTP headers in requests describe the origins that are permitted access to the Kafka cluster.
-
Requests for reservation of supported resources, currently
cpu
andmemory
, and limits to specify the maximum resources that can be consumed. -
TLS configuration for encrypted connections to the Kafka cluster, with trusted certificates stored in X.509 format within the specified secrets.
-
Authentication for the Kafka Bridge cluster, specified as mTLS, token-based OAuth, SASL-based SCRAM-SHA-256/SCRAM-SHA-512, or PLAIN. By default, the Kafka Bridge connects to Kafka brokers without authentication.
-
Consumer configuration options.
-
Producer configuration options.
-
Specified Kafka Bridge loggers and log levels added directly (
inline
) or indirectly (external
) through a ConfigMap. A custom Log4j configuration must be placed under thelog4j.properties
orlog4j2.properties
key in the ConfigMap. For the Kafka Bridge loggers, you can set the log level to INFO, ERROR, WARN, TRACE, DEBUG, FATAL or OFF. -
JVM configuration options to optimize performance for the Virtual Machine (VM) running the Kafka Bridge.
-
Healthchecks to know when to restart a container (liveness) and when a container can accept traffic (readiness).
-
Optional: Container image configuration, which is recommended only in special situations.
-
Template customization. Here a pod is scheduled with anti-affinity, so the pod is not scheduled on nodes with the same hostname.
-
Environment variables are set for distributed tracing.
-
Distributed tracing is enabled by using OpenTelemetry.
11.13. Configuring CPU and memory resource limits and requests
By default, the Strimzi Cluster Operator does not specify CPU and memory resource requests and limits for its deployed operands. Ensuring an adequate allocation of resources is crucial for maintaining stability and achieving optimal performance in Kafka. The ideal resource allocation depends on your specific requirements and use cases.
It is recommended to configure CPU and memory resources for each container by setting appropriate requests and limits.
11.14. Restrictions on Kubernetes labels
Kubernetes labels make it easier to organize, manage, and discover Kubernetes resources within your applications.
The Cluster Operator is responsible for applying the following Kubernetes labels to the operands it deploys.
These labels cannot be overridden through template
configuration of Strimzi resources:
-
app.kubernetes.io/name
: Identifies the component type within Strimzi, such askafka
,zookeeper
, and`cruise-control`. -
app.kubernetes.io/instance
: Represents the name of the custom resource to which the operand belongs to. For instance, if a Kafka custom resource is namedmy-cluster
, this label will bear that name on the associated pods. -
app.kubernetes.io/part-of
: Similar toapp.kubernetes.io/instance
, but prefixed withstrimzi-
. -
app.kubernetes.io/managed-by
: Defines the application responsible for managing the operand, such asstrimzi-cluster-operator
orstrimzi-user-operator
.
Kafka
custom resource named my-cluster
apiVersion: kafka.strimzi.io/v1beta2
kind: Pod
metadata:
name: my-cluster-kafka-0
labels:
app.kubernetes.io/instance: my-cluster
app.kubernetes.io/managed-by: strimzi-cluster-operator
app.kubernetes.io/name: kafka
app.kubernetes.io/part-of: strimzi-my-cluster
spec:
# ...
11.15. Configuring pod scheduling
To avoid performance degradation caused by resource conflicts between applications scheduled on the same Kubernetes node, you can schedule Kafka pods separately from critical workloads. This can be achieved by either selecting specific nodes or dedicating a set of nodes exclusively for Kafka.
11.15.1. Specifying affinity, tolerations, and topology spread constraints
Use affinity, tolerations and topology spread constraints to schedule the pods of kafka resources onto nodes.
Affinity, tolerations and topology spread constraints are configured using the affinity
, tolerations
, and topologySpreadConstraint
properties in following resources:
-
Kafka.spec.kafka.template.pod
-
Kafka.spec.zookeeper.template.pod
-
Kafka.spec.entityOperator.template.pod
-
KafkaConnect.spec.template.pod
-
KafkaBridge.spec.template.pod
-
KafkaMirrorMaker.spec.template.pod
-
KafkaMirrorMaker2.spec.template.pod
The format of the affinity
, tolerations
, and topologySpreadConstraint
properties follows the Kubernetes specification.
The affinity configuration can include different types of affinity:
-
Pod affinity and anti-affinity
-
Node affinity
Use pod anti-affinity to avoid critical applications sharing nodes
Use pod anti-affinity to ensure that critical applications are never scheduled on the same disk. When running a Kafka cluster, it is recommended to use pod anti-affinity to ensure that the Kafka brokers do not share nodes with other workloads, such as databases.
Use node affinity to schedule workloads onto specific nodes
The Kubernetes cluster usually consists of many different types of worker nodes. Some are optimized for CPU heavy workloads, some for memory, while other might be optimized for storage (fast local SSDs) or network. Using different nodes helps to optimize both costs and performance. To achieve the best possible performance, it is important to allow scheduling of Strimzi components to use the right nodes.
Kubernetes uses node affinity to schedule workloads onto specific nodes.
Node affinity allows you to create a scheduling constraint for the node on which the pod will be scheduled.
The constraint is specified as a label selector.
You can specify the label using either the built-in node label like beta.kubernetes.io/instance-type
or custom labels to select the right node.
Use node affinity and tolerations for dedicated nodes
Use taints to create dedicated nodes, then schedule Kafka pods on the dedicated nodes by configuring node affinity and tolerations.
Cluster administrators can mark selected Kubernetes nodes as tainted. Nodes with taints are excluded from regular scheduling and normal pods will not be scheduled to run on them. Only services which can tolerate the taint set on the node can be scheduled on it. The only other services running on such nodes will be system services such as log collectors or software defined networks.
Running Kafka and its components on dedicated nodes can have many advantages. There will be no other applications running on the same nodes which could cause disturbance or consume the resources needed for Kafka. That can lead to improved performance and stability.
11.15.2. Configuring pod anti-affinity to schedule each Kafka broker on a different worker node
Many Kafka brokers or ZooKeeper nodes can run on the same Kubernetes worker node.
If the worker node fails, they will all become unavailable at the same time.
To improve reliability, you can use podAntiAffinity
configuration to schedule each Kafka broker or ZooKeeper node on a different Kubernetes worker node.
-
A Kubernetes cluster
-
A running Cluster Operator
-
Edit the
affinity
property in the resource specifying the cluster deployment. To make sure that no worker nodes are shared by Kafka brokers or ZooKeeper nodes, use thestrimzi.io/name
label. Set thetopologyKey
tokubernetes.io/hostname
to specify that the selected pods are not scheduled on nodes with the same hostname. This will still allow the same worker node to be shared by a single Kafka broker and a single ZooKeeper node. For example:apiVersion: kafka.strimzi.io/v1beta2 kind: Kafka spec: kafka: # ... template: pod: affinity: podAntiAffinity: requiredDuringSchedulingIgnoredDuringExecution: - labelSelector: matchExpressions: - key: strimzi.io/name operator: In values: - CLUSTER-NAME-kafka topologyKey: "kubernetes.io/hostname" # ... zookeeper: # ... template: pod: affinity: podAntiAffinity: requiredDuringSchedulingIgnoredDuringExecution: - labelSelector: matchExpressions: - key: strimzi.io/name operator: In values: - CLUSTER-NAME-zookeeper topologyKey: "kubernetes.io/hostname" # ...
Where
CLUSTER-NAME
is the name of your Kafka custom resource. -
If you even want to make sure that a Kafka broker and ZooKeeper node do not share the same worker node, use the
strimzi.io/cluster
label. For example:apiVersion: kafka.strimzi.io/v1beta2 kind: Kafka spec: kafka: # ... template: pod: affinity: podAntiAffinity: requiredDuringSchedulingIgnoredDuringExecution: - labelSelector: matchExpressions: - key: strimzi.io/cluster operator: In values: - CLUSTER-NAME topologyKey: "kubernetes.io/hostname" # ... zookeeper: # ... template: pod: affinity: podAntiAffinity: requiredDuringSchedulingIgnoredDuringExecution: - labelSelector: matchExpressions: - key: strimzi.io/cluster operator: In values: - CLUSTER-NAME topologyKey: "kubernetes.io/hostname" # ...
Where
CLUSTER-NAME
is the name of your Kafka custom resource. -
Create or update the resource.
kubectl apply -f <kafka_configuration_file>
11.15.3. Configuring pod anti-affinity in Kafka components
Pod anti-affinity configuration helps with the stability and performance of Kafka brokers. By using podAntiAffinity
, Kubernetes will not schedule Kafka brokers on the same nodes as other workloads.
Typically, you want to avoid Kafka running on the same worker node as other network or storage intensive applications such as databases, storage or other messaging platforms.
-
A Kubernetes cluster
-
A running Cluster Operator
-
Edit the
affinity
property in the resource specifying the cluster deployment. Use labels to specify the pods which should not be scheduled on the same nodes. ThetopologyKey
should be set tokubernetes.io/hostname
to specify that the selected pods should not be scheduled on nodes with the same hostname. For example:apiVersion: kafka.strimzi.io/v1beta2 kind: Kafka spec: kafka: # ... template: pod: affinity: podAntiAffinity: requiredDuringSchedulingIgnoredDuringExecution: - labelSelector: matchExpressions: - key: application operator: In values: - postgresql - mongodb topologyKey: "kubernetes.io/hostname" # ... zookeeper: # ...
-
Create or update the resource.
This can be done using
kubectl apply
:kubectl apply -f <kafka_configuration_file>
11.15.4. Configuring node affinity in Kafka components
-
A Kubernetes cluster
-
A running Cluster Operator
-
Label the nodes where Strimzi components should be scheduled.
This can be done using
kubectl label
:kubectl label node NAME-OF-NODE node-type=fast-network
Alternatively, some of the existing labels might be reused.
-
Edit the
affinity
property in the resource specifying the cluster deployment. For example:apiVersion: kafka.strimzi.io/v1beta2 kind: Kafka spec: kafka: # ... template: pod: affinity: nodeAffinity: requiredDuringSchedulingIgnoredDuringExecution: nodeSelectorTerms: - matchExpressions: - key: node-type operator: In values: - fast-network # ... zookeeper: # ...
-
Create or update the resource.
This can be done using
kubectl apply
:kubectl apply -f <kafka_configuration_file>
11.15.5. Setting up dedicated nodes and scheduling pods on them
-
A Kubernetes cluster
-
A running Cluster Operator
-
Select the nodes which should be used as dedicated.
-
Make sure there are no workloads scheduled on these nodes.
-
Set the taints on the selected nodes:
This can be done using
kubectl taint
:kubectl taint node NAME-OF-NODE dedicated=Kafka:NoSchedule
-
Additionally, add a label to the selected nodes as well.
This can be done using
kubectl label
:kubectl label node NAME-OF-NODE dedicated=Kafka
-
Edit the
affinity
andtolerations
properties in the resource specifying the cluster deployment.For example:
apiVersion: kafka.strimzi.io/v1beta2 kind: Kafka spec: kafka: # ... template: pod: tolerations: - key: "dedicated" operator: "Equal" value: "Kafka" effect: "NoSchedule" affinity: nodeAffinity: requiredDuringSchedulingIgnoredDuringExecution: nodeSelectorTerms: - matchExpressions: - key: dedicated operator: In values: - Kafka # ... zookeeper: # ...
-
Create or update the resource.
This can be done using
kubectl apply
:kubectl apply -f <kafka_configuration_file>
11.16. Disabling pod disruption budget generation
Strimzi generates pod disruption budget resources for Kafka, Kafka Connect worker, MirrorMaker2 worker, and Kafka Bridge worker nodes.
If you want to use custom pod disruption budget resources, you can set the STRIMZI_POD_DISRUPTION_BUDGET_GENERATION
environment variable to false
in the Cluster Operator configuration.
For more information, see Configuring the Cluster Operator.
11.17. Configuring logging levels
Configure logging levels in the custom resources of Kafka components and Strimzi operators.
You can specify the logging levels directly in the spec.logging
property of the custom resource.
Or you can define the logging properties in a ConfigMap that’s referenced in the custom resource using the configMapKeyRef
property.
The advantages of using a ConfigMap are that the logging properties are maintained in one place and are accessible to more than one resource. You can also reuse the ConfigMap for more than one resource. If you are using a ConfigMap to specify loggers for Strimzi Operators, you can also append the logging specification to add filters.
You specify a logging type
in your logging specification:
-
inline
when specifying logging levels directly -
external
when referencing a ConfigMap
inline
logging configuration# ...
logging:
type: inline
loggers:
kafka.root.logger.level: INFO
# ...
external
logging configuration# ...
logging:
type: external
valueFrom:
configMapKeyRef:
name: my-config-map
key: my-config-map-key
# ...
Values for the name
and key
of the ConfigMap are mandatory.
Default logging is used if the name
or key
is not set.
11.17.1. Logging options for Kafka components and operators
For more information on configuring logging for specific Kafka components or operators, see the following sections.
11.17.2. Creating a ConfigMap for logging
To use a ConfigMap to define logging properties, you create the ConfigMap and then reference it as part of the logging definition in the spec
of a resource.
The ConfigMap must contain the appropriate logging configuration.
-
log4j.properties
for Kafka components, ZooKeeper, and the Kafka Bridge -
log4j2.properties
for the Topic Operator and User Operator
The configuration must be placed under these properties.
In this procedure a ConfigMap defines a root logger for a Kafka resource.
-
Create the ConfigMap.
You can create the ConfigMap as a YAML file or from a properties file.
ConfigMap example with a root logger definition for Kafka:
kind: ConfigMap apiVersion: v1 metadata: name: logging-configmap data: log4j.properties: kafka.root.logger.level="INFO"
If you are using a properties file, specify the file at the command line:
kubectl create configmap logging-configmap --from-file=log4j.properties
The properties file defines the logging configuration:
# Define the logger kafka.root.logger.level="INFO" # ...
-
Define external logging in the
spec
of the resource, setting thelogging.valueFrom.configMapKeyRef.name
to the name of the ConfigMap andlogging.valueFrom.configMapKeyRef.key
to the key in this ConfigMap.# ... logging: type: external valueFrom: configMapKeyRef: name: logging-configmap key: log4j.properties # ...
-
Create or update the resource.
kubectl apply -f <kafka_configuration_file>
11.17.3. Configuring Cluster Operator logging
Cluster Operator logging is configured through a ConfigMap
named strimzi-cluster-operator
.
A ConfigMap
containing logging configuration is created when installing the Cluster Operator.
This ConfigMap
is described in the file install/cluster-operator/050-ConfigMap-strimzi-cluster-operator.yaml
.
You configure Cluster Operator logging by changing the data.log4j2.properties
values in this ConfigMap
.
To update the logging configuration, you can edit the 050-ConfigMap-strimzi-cluster-operator.yaml
file and then run the following command:
kubectl create -f install/cluster-operator/050-ConfigMap-strimzi-cluster-operator.yaml
Alternatively, edit the ConfigMap
directly:
kubectl edit configmap strimzi-cluster-operator
With this ConfigMap, you can control various aspects of logging, including the root logger level, log output format, and log levels for different components.
The monitorInterval
setting, determines how often the logging configuration is reloaded.
You can also control the logging levels for the Kafka AdminClient
, ZooKeeper ZKTrustManager
, Netty, and the OkHttp client.
Netty is a framework used in Strimzi for network communication, and OkHttp is a library used for making HTTP requests.
If the ConfigMap
is missing when the Cluster Operator is deployed, the default logging values are used.
If the ConfigMap
is accidentally deleted after the Cluster Operator is deployed, the most recently loaded logging configuration is used.
Create a new ConfigMap
to load a new logging configuration.
Note
|
Do not remove the monitorInterval option from the ConfigMap .
|
11.17.4. Adding logging filters to Strimzi operators
If you are using a ConfigMap to configure the (log4j2) logging levels for Strimzi operators, you can also define logging filters to limit what’s returned in the log.
Logging filters are useful when you have a large number of logging messages.
Suppose you set the log level for the logger as DEBUG (rootLogger.level="DEBUG"
).
Logging filters reduce the number of logs returned for the logger at that level, so you can focus on a specific resource.
When the filter is set, only log messages matching the filter are logged.
Filters use markers to specify what to include in the log.
You specify a kind, namespace and name for the marker.
For example, if a Kafka cluster is failing, you can isolate the logs by specifying the kind as Kafka
, and use the namespace and name of the failing cluster.
This example shows a marker filter for a Kafka cluster named my-kafka-cluster
.
rootLogger.level="INFO"
appender.console.filter.filter1.type=MarkerFilter (1)
appender.console.filter.filter1.onMatch=ACCEPT (2)
appender.console.filter.filter1.onMismatch=DENY (3)
appender.console.filter.filter1.marker=Kafka(my-namespace/my-kafka-cluster) (4)
-
The
MarkerFilter
type compares a specified marker for filtering. -
The
onMatch
property accepts the log if the marker matches. -
The
onMismatch
property rejects the log if the marker does not match. -
The marker used for filtering is in the format KIND(NAMESPACE/NAME-OF-RESOURCE).
You can create one or more filters. Here, the log is filtered for two Kafka clusters.
appender.console.filter.filter1.type=MarkerFilter
appender.console.filter.filter1.onMatch=ACCEPT
appender.console.filter.filter1.onMismatch=DENY
appender.console.filter.filter1.marker=Kafka(my-namespace/my-kafka-cluster-1)
appender.console.filter.filter2.type=MarkerFilter
appender.console.filter.filter2.onMatch=ACCEPT
appender.console.filter.filter2.onMismatch=DENY
appender.console.filter.filter2.marker=Kafka(my-namespace/my-kafka-cluster-2)
To add filters to the Cluster Operator, update its logging ConfigMap YAML file (install/cluster-operator/050-ConfigMap-strimzi-cluster-operator.yaml
).
-
Update the
050-ConfigMap-strimzi-cluster-operator.yaml
file to add the filter properties to the ConfigMap.In this example, the filter properties return logs only for the
my-kafka-cluster
Kafka cluster:kind: ConfigMap apiVersion: v1 metadata: name: strimzi-cluster-operator data: log4j2.properties: #... appender.console.filter.filter1.type=MarkerFilter appender.console.filter.filter1.onMatch=ACCEPT appender.console.filter.filter1.onMismatch=DENY appender.console.filter.filter1.marker=Kafka(my-namespace/my-kafka-cluster)
Alternatively, edit the
ConfigMap
directly:kubectl edit configmap strimzi-cluster-operator
-
If you updated the YAML file instead of editing the
ConfigMap
directly, apply the changes by deploying the ConfigMap:kubectl create -f install/cluster-operator/050-ConfigMap-strimzi-cluster-operator.yaml
To add filters to the Topic Operator or User Operator, create or edit a logging ConfigMap.
In this procedure a logging ConfigMap is created with filters for the Topic Operator. The same approach is used for the User Operator.
-
Create the ConfigMap.
You can create the ConfigMap as a YAML file or from a properties file.
In this example, the filter properties return logs only for the
my-topic
topic:kind: ConfigMap apiVersion: v1 metadata: name: logging-configmap data: log4j2.properties: rootLogger.level="INFO" appender.console.filter.filter1.type=MarkerFilter appender.console.filter.filter1.onMatch=ACCEPT appender.console.filter.filter1.onMismatch=DENY appender.console.filter.filter1.marker=KafkaTopic(my-namespace/my-topic)
If you are using a properties file, specify the file at the command line:
kubectl create configmap logging-configmap --from-file=log4j2.properties
The properties file defines the logging configuration:
# Define the logger rootLogger.level="INFO" # Set the filters appender.console.filter.filter1.type=MarkerFilter appender.console.filter.filter1.onMatch=ACCEPT appender.console.filter.filter1.onMismatch=DENY appender.console.filter.filter1.marker=KafkaTopic(my-namespace/my-topic) # ...
-
Define external logging in the
spec
of the resource, setting thelogging.valueFrom.configMapKeyRef.name
to the name of the ConfigMap andlogging.valueFrom.configMapKeyRef.key
to the key in this ConfigMap.For the Topic Operator, logging is specified in the
topicOperator
configuration of theKafka
resource.spec: # ... entityOperator: topicOperator: logging: type: external valueFrom: configMapKeyRef: name: logging-configmap key: log4j2.properties
-
Apply the changes by deploying the Cluster Operator:
create -f install/cluster-operator -n my-cluster-operator-namespace
11.17.5. Lock acquisition warnings for cluster operations
The Cluster Operator ensures that only one operation runs at a time for each cluster by using locks. If another operation attempts to start while a lock is held, it waits until the current operation completes.
Operations such as cluster creation, rolling updates, scaling down, and scaling up are managed by the Cluster Operator.
If acquiring a lock takes longer than the configured timeout (STRIMZI_OPERATION_TIMEOUT_MS
), a DEBUG message is logged:
DEBUG AbstractOperator:406 - Reconciliation #55(timer) Kafka(myproject/my-cluster): Failed to acquire lock lock::myproject::Kafka::my-cluster within 10000ms.
Timed-out operations are retried during the next periodic reconciliation in intervals defined by STRIMZI_FULL_RECONCILIATION_INTERVAL_MS
(by default 120 seconds).
If an INFO message continues to appear with the same same reconciliation number, it might indicate a lock release error:
INFO AbstractOperator:399 - Reconciliation #1(watch) Kafka(myproject/my-cluster): Reconciliation is in progress
Restarting the Cluster Operator can resolve such issues.
11.18. Using ConfigMaps to add configuration
Add specific configuration to your Strimzi deployment using ConfigMap
resources.
ConfigMaps use key-value pairs to store non-confidential data.
Configuration data added to ConfigMaps is maintained in one place and can be reused amongst components.
ConfigMaps can only store the following types of configuration data:
-
Logging configuration
-
Metrics configuration
-
External configuration for Kafka Connect connectors
You can’t use ConfigMaps for other areas of configuration.
When you configure a component, you can add a reference to a ConfigMap using the configMapKeyRef
property.
For example, you can use configMapKeyRef
to reference a ConfigMap that provides configuration for logging.
You might use a ConfigMap to pass a Log4j configuration file.
You add the reference to the logging
configuration.
# ...
logging:
type: external
valueFrom:
configMapKeyRef:
name: my-config-map
key: my-config-map-key
# ...
To use a ConfigMap for metrics configuration, you add a reference to the metricsConfig
configuration of the component in the same way.
template
properties allow data from a ConfigMap
or Secret
to be mounted in a pod as environment variables or volumes.
You can use external configuration data for the connectors used by Kafka Connect.
The data might be related to an external data source, providing the values needed for the connector to communicate with that data source.
For example, you can use the configMapKeyRef
property to pass configuration data from a ConfigMap as an environment variable.
apiVersion: kafka.strimzi.io/v1beta2
kind: KafkaConnect
metadata:
name: my-connect
spec:
# ...
template:
connectContainer:
env:
- name: MY_ENVIRONMENT_VARIABLE
valueFrom:
configMapKeyRef:
name: my-config-map
key: my-key
If you are using ConfigMaps that are managed externally, use configuration providers to load the data in the ConfigMaps.
11.18.1. Naming custom ConfigMaps
Strimzi creates its own ConfigMaps and other resources when it is deployed to Kubernetes. The ConfigMaps contain data necessary for running components. The ConfigMaps created by Strimzi must not be edited.
Make sure that any custom ConfigMaps you create do not have the same name as these default ConfigMaps. If they have the same name, they will be overwritten. For example, if your ConfigMap has the same name as the ConfigMap for the Kafka cluster, it will be overwritten when there is an update to the Kafka cluster.
11.19. Loading configuration values from external sources
Use configuration providers to load configuration data from external sources. The providers operate independently of Strimzi. You can use them to load configuration data for all Kafka components, including producers and consumers. You reference the external source in the configuration of the component and provide access rights. The provider loads data without needing to restart the Kafka component or extracting files, even when referencing a new external source. For example, use providers to supply the credentials for the Kafka Connect connector configuration. The configuration must include any access rights to the external source.
11.19.1. Enabling configuration providers
You can enable one or more configuration providers using the config.providers
properties in the spec
configuration of a component.
apiVersion: kafka.strimzi.io/v1beta2
kind: KafkaConnect
metadata:
name: my-connect
annotations:
strimzi.io/use-connector-resources: "true"
spec:
# ...
config:
# ...
config.providers: env
config.providers.env.class: org.apache.kafka.common.config.provider.EnvVarConfigProvider
# ...
- KubernetesSecretConfigProvider
-
Loads configuration data from Kubernetes secrets. You specify the name of the secret and the key within the secret where the configuration data is stored. This provider is useful for storing sensitive configuration data like passwords or other user credentials.
- KubernetesConfigMapConfigProvider
-
Loads configuration data from Kubernetes config maps. You specify the name of the config map and the key within the config map where the configuration data is stored. This provider is useful for storing non-sensitive configuration data.
- EnvVarConfigProvider
-
Loads configuration data from environment variables. You specify the name of the environment variable where the configuration data is stored. This provider is useful for configuring applications running in containers, for example, to load certificates or JAAS configuration from environment variables mapped from secrets.
- FileConfigProvider
-
Loads configuration data from a file. You specify the path to the file where the configuration data is stored. This provider is useful for loading configuration data from files that are mounted into containers.
- DirectoryConfigProvider
-
Loads configuration data from files within a directory. You specify the path to the directory where the configuration files are stored. This provider is useful for loading multiple configuration files and for organizing configuration data into separate files.
To use KubernetesSecretConfigProvider
and KubernetesConfigMapConfigProvider
, which are part of the Kubernetes Configuration Provider plugin, you must set up access rights to the namespace that contains the configuration file.
You can use the other providers without setting up access rights. You can supply connector configuration for Kafka Connect or MirrorMaker 2 in this way by doing the following:
-
Mount config maps or secrets into the Kafka Connect pod as environment variables or volumes
-
Enable
EnvVarConfigProvider
,FileConfigProvider
, orDirectoryConfigProvider
in the Kafka Connect or MirrorMaker 2 configuration -
Pass connector configuration using the
template
property in thespec
of theKafkaConnect
orKafkaMirrorMaker2
resource
Using providers help prevent the passing of restricted information through the Kafka Connect REST interface. You can use this approach in the following scenarios:
-
Mounting environment variables with the values a connector uses to connect and communicate with a data source
-
Mounting a properties file with values that are used to configure Kafka Connect connectors
-
Mounting files in a directory that contains values for the TLS truststore and keystore used by a connector
Note
|
A restart is required when using a new Secret or ConfigMap for a connector, which can disrupt other connectors.
|
11.19.2. Loading configuration values from secrets or config maps
Use the KubernetesSecretConfigProvider
to provide configuration properties from a secret or the KubernetesConfigMapConfigProvider
to provide configuration properties from a config map.
In this procedure, a config map provides configuration properties for a connector. The properties are specified as key values of the config map. The config map is mounted into the Kafka Connect pod as a volume.
-
A Kafka cluster is running.
-
The Cluster Operator is running.
-
You have a config map containing the connector configuration.
apiVersion: v1
kind: ConfigMap
metadata:
name: my-connector-configuration
data:
option1: value1
option2: value2
-
Configure the
KafkaConnect
resource.-
Enable the
KubernetesConfigMapConfigProvider
The specification shown here can support loading values from config maps and secrets.
Example Kafka Connect configuration to use config maps and secretsapiVersion: kafka.strimzi.io/v1beta2 kind: KafkaConnect metadata: name: my-connect annotations: strimzi.io/use-connector-resources: "true" spec: # ... config: # ... config.providers: secrets,configmaps # (1) config.providers.configmaps.class: io.strimzi.kafka.KubernetesConfigMapConfigProvider # (2) config.providers.secrets.class: io.strimzi.kafka.KubernetesSecretConfigProvider # (3) # ...
-
The alias for the configuration provider is used to define other configuration parameters. The provider parameters use the alias from
config.providers
, taking the formconfig.providers.${alias}.class
. -
KubernetesConfigMapConfigProvider
provides values from config maps. -
KubernetesSecretConfigProvider
provides values from secrets.
-
-
Create or update the resource to enable the provider.
kubectl apply -f <kafka_connect_configuration_file>
-
Create a role that permits access to the values in the external config map.
Example role to access values from a config mapapiVersion: rbac.authorization.k8s.io/v1 kind: Role metadata: name: connector-configuration-role rules: - apiGroups: [""] resources: ["configmaps"] resourceNames: ["my-connector-configuration"] verbs: ["get"] # ...
The rule gives the role permission to access the
my-connector-configuration
config map. -
Create a role binding to permit access to the namespace that contains the config map.
Example role binding to access the namespace that contains the config mapapiVersion: rbac.authorization.k8s.io/v1 kind: RoleBinding metadata: name: connector-configuration-role-binding subjects: - kind: ServiceAccount name: my-connect-connect namespace: my-project roleRef: kind: Role name: connector-configuration-role apiGroup: rbac.authorization.k8s.io # ...
The role binding gives the role permission to access the
my-project
namespace.The service account must be the same one used by the Kafka Connect deployment. The service account name format is
<cluster_name>-connect
, where<cluster_name>
is the name of theKafkaConnect
custom resource. -
Reference the config map in the connector configuration.
Example connector configuration referencing the config mapapiVersion: kafka.strimzi.io/v1beta2 kind: KafkaConnector metadata: name: my-connector labels: strimzi.io/cluster: my-connect spec: # ... config: option: ${configmaps:my-project/my-connector-configuration:option1} # ... # ...
The placeholder structure is
configmaps:<path_and_file_name>:<property>
.KubernetesConfigMapConfigProvider
reads and extracts theoption1
property value from the external config map.
11.19.3. Loading configuration values from environment variables
Use the EnvVarConfigProvider
to provide configuration properties as environment variables.
Environment variables can contain values from config maps or secrets.
In this procedure, environment variables provide configuration properties for a connector to communicate with Amazon AWS.
The connector must be able to read the AWS_ACCESS_KEY_ID
and AWS_SECRET_ACCESS_KEY
.
The values of the environment variables are derived from a secret mounted into the Kafka Connect pod.
Note
|
The names of user-defined environment variables cannot start with KAFKA_ or STRIMZI_ .
|
-
A Kafka cluster is running.
-
The Cluster Operator is running.
-
You have a secret containing the connector configuration.
apiVersion: v1
kind: Secret
metadata:
name: aws-creds
type: Opaque
data:
awsAccessKey: QUtJQVhYWFhYWFhYWFhYWFg=
awsSecretAccessKey: Ylhsd1lYTnpkMjl5WkE=
-
Configure the
KafkaConnect
resource.-
Enable the
EnvVarConfigProvider
-
Specify the environment variables using the
template
property.
Example Kafka Connect configuration to use external environment variablesapiVersion: kafka.strimzi.io/v1beta2 kind: KafkaConnect metadata: name: my-connect annotations: strimzi.io/use-connector-resources: "true" spec: # ... config: # ... config.providers: env # (1) config.providers.env.class: org.apache.kafka.common.config.provider.EnvVarConfigProvider # (2) # ... template: connectContainer: env: - name: AWS_ACCESS_KEY_ID # (3) valueFrom: secretKeyRef: name: aws-creds # (4) key: awsAccessKey # (5) - name: AWS_SECRET_ACCESS_KEY valueFrom: secretKeyRef: name: aws-creds key: awsSecretAccessKey # ...
-
The alias for the configuration provider is used to define other configuration parameters. The provider parameters use the alias from
config.providers
, taking the formconfig.providers.${alias}.class
. -
EnvVarConfigProvider
provides values from environment variables. -
The environment variable takes a value from the secret.
-
The name of the secret containing the environment variable.
-
The name of the key stored in the secret.
NoteThe secretKeyRef
property references keys in a secret. If you are using a config map instead of a secret, use theconfigMapKeyRef
property.
-
-
Create or update the resource to enable the provider.
kubectl apply -f <kafka_connect_configuration_file>
-
Reference the environment variable in the connector configuration.
Example connector configuration referencing the environment variableapiVersion: kafka.strimzi.io/v1beta2 kind: KafkaConnector metadata: name: my-connector labels: strimzi.io/cluster: my-connect spec: # ... config: option: ${env:AWS_ACCESS_KEY_ID} option: ${env:AWS_SECRET_ACCESS_KEY} # ... # ...
The placeholder structure is
env:<environment_variable_name>
.EnvVarConfigProvider
reads and extracts the environment variable values from the mounted secret.
11.19.4. Loading configuration values from a file within a directory
Use the FileConfigProvider
to provide configuration properties from a file within a directory.
Files can be stored in config maps or secrets.
In this procedure, a file provides configuration properties for a connector.
A database name and password are specified as properties of a secret.
The secret is mounted to the Kafka Connect pod as a volume.
Volumes are mounted on the path /mnt/<volume-name>
.
-
A Kafka cluster is running.
-
The Cluster Operator is running.
-
You have a secret containing the connector configuration.
apiVersion: v1
kind: Secret
metadata:
name: mysecret
type: Opaque
stringData:
connector.properties: |- # (1)
dbUsername: my-username # (2)
dbPassword: my-password
-
The connector configuration in properties file format.
-
Database username and password properties used in the configuration.
-
Configure the
KafkaConnect
resource.-
Enable the
FileConfigProvider
-
Specify the additional volume using the
template
property.
Example Kafka Connect configuration to use an external property fileapiVersion: kafka.strimzi.io/v1beta2 kind: KafkaConnect metadata: name: my-connect spec: # ... config: config.providers: file # (1) config.providers.file.class: org.apache.kafka.common.config.provider.FileConfigProvider # (2) #... template: pod: volumes: - name: connector-config-volume # (3) secret: secretName: mysecret # (4) connectContainer: volumeMounts: - name: connector-config-volume # (5) mountPath: /mnt/mysecret # (6)
-
The alias for the configuration provider is used to define other configuration parameters.
-
FileConfigProvider
provides values from properties files. The parameter uses the alias fromconfig.providers
, taking the formconfig.providers.${alias}.class
. -
The name of the volume containing the secret.
-
The name of the secret.
-
The name of the mounted volume, which must match the volume name in the
volumes
list. -
The path where the secret is mounted, which must start with
/mnt/
.
-
-
Create or update the resource to enable the provider.
kubectl apply -f <kafka_connect_configuration_file>
-
Reference the file properties in the connector configuration as placeholders.
Example connector configuration referencing the fileapiVersion: kafka.strimzi.io/v1beta2 kind: KafkaConnector metadata: name: my-source-connector labels: strimzi.io/cluster: my-connect-cluster spec: class: io.debezium.connector.mysql.MySqlConnector tasksMax: 2 config: database.hostname: 192.168.99.1 database.port: "3306" database.user: "${file:/mnt/mysecret/connector.properties:dbUsername}" database.password: "${file:/mnt/mysecret/connector.properties:dbPassword}" database.server.id: "184054" #...
The placeholder structure is
file:<path_and_file_name>:<property>
.FileConfigProvider
reads and extracts the database username and password property values from the mounted secret.
11.19.5. Loading configuration values from multiple files within a directory
Use the DirectoryConfigProvider
to provide configuration properties from multiple files within a directory.
Files can be config maps or secrets.
In this procedure, a secret provides the TLS keystore and truststore user credentials for a connector.
The credentials are in separate files.
The secrets are mounted into the Kafka Connect pod as volumes.
Volumes are mounted on the path /mnt/<volume-name>
.
-
A Kafka cluster is running.
-
The Cluster Operator is running.
-
You have a secret containing the user credentials.
apiVersion: v1
kind: Secret
metadata:
name: my-user
labels:
strimzi.io/kind: KafkaUser
strimzi.io/cluster: my-cluster
type: Opaque
data:
ca.crt: <public_key> # Public key of the clients CA used to sign this user certificate
user.crt: <user_certificate> # Public key of the user
user.key: <user_private_key> # Private key of the user
user.p12: <store> # PKCS #12 store for user certificates and keys
user.password: <password_for_store> # Protects the PKCS #12 store
The my-user
secret provides the keystore credentials (user.crt
and user.key
) for the connector.
The <cluster_name>-cluster-ca-cert
secret generated when deploying the Kafka cluster provides the cluster CA certificate as truststore credentials (ca.crt
).
-
Configure the
KafkaConnect
resource.-
Enable the
DirectoryConfigProvider
-
Specify the additional volume using the
template
property.
Example Kafka Connect configuration to use external property filesapiVersion: kafka.strimzi.io/v1beta2 kind: KafkaConnect metadata: name: my-connect spec: # ... config: config.providers: directory # (1) config.providers.directory.class: org.apache.kafka.common.config.provider.DirectoryConfigProvider # (2) #... template: pod: volumes: - name: my-user-volume # (3) secret: secretName: my-user # (4) - name: cluster-ca-volume secret: secretName: my-cluster-cluster-ca-cert connectContainer: volumeMounts: - name: my-user-volume # (5) mountPath: /mnt/my-user # (6) - name: cluster-ca-volume mountPath: /mnt/cluster-ca
-
The alias for the configuration provider is used to define other configuration parameters.
-
DirectoryConfigProvider
provides values from files in a directory. The parameter uses the alias fromconfig.providers
, taking the formconfig.providers.${alias}.class
. -
The name of the volume containing the secret.
-
The name of the secret.
-
The name of the mounted volume, which must match the volume name in the
volumes
list. -
The path where the secret is mounted, which must start with
/mnt/
.
-
-
Create or update the resource to enable the provider.
kubectl apply -f <kafka_connect_configuration_file>
-
Reference the file properties in the connector configuration as placeholders.
Example connector configuration referencing the filesapiVersion: kafka.strimzi.io/v1beta2 kind: KafkaConnector metadata: name: my-source-connector labels: strimzi.io/cluster: my-connect-cluster spec: class: io.debezium.connector.mysql.MySqlConnector tasksMax: 2 config: # ... database.history.producer.security.protocol: SSL database.history.producer.ssl.truststore.type: PEM database.history.producer.ssl.truststore.certificates: "${directory:/mtn/cluster-ca:ca.crt}" database.history.producer.ssl.keystore.type: PEM database.history.producer.ssl.keystore.certificate.chain: "${directory:/mnt/my-user:user.crt}" database.history.producer.ssl.keystore.key: "${directory:/mnt/my-user:user.key}" #...
The placeholder structure is
directory:<path>:<file_name>
.DirectoryConfigProvider
reads and extracts the credentials from the mounted secrets.
11.20. Customizing Kubernetes resources
A Strimzi deployment creates Kubernetes resources, such as Deployment
, Pod
, and Service
resources.
These resources are managed by Strimzi operators.
Only the operator that is responsible for managing a particular Kubernetes resource can change that resource.
If you try to manually change an operator-managed Kubernetes resource, the operator will revert your changes back.
Changing an operator-managed Kubernetes resource can be useful if you want to perform certain tasks, such as the following:
-
Adding custom labels or annotations that control how
Pods
are treated by Istio or other services -
Managing how
Loadbalancer
-type Services are created by the cluster
To make the changes to a Kubernetes resource, you can use the template
property within the spec
section of various Strimzi custom resources.
Here is a list of the custom resources where you can apply the changes:
-
Kafka.spec.kafka
-
Kafka.spec.zookeeper
-
Kafka.spec.entityOperator
-
Kafka.spec.kafkaExporter
-
Kafka.spec.cruiseControl
-
KafkaNodePool.spec
-
KafkaConnect.spec
-
KafkaMirrorMaker.spec
-
KafkaMirrorMaker2.spec
-
KafkaBridge.spec
-
KafkaUser.spec
For more information about these properties, see the Strimzi Custom Resource API Reference.
The Strimzi Custom Resource API Reference provides more details about the customizable fields.
In the following example, the template
property is used to modify the labels in a Kafka broker’s pod.
apiVersion: kafka.strimzi.io/v1beta2
kind: Kafka
metadata:
name: my-cluster
labels:
app: my-cluster
spec:
kafka:
# ...
template:
pod:
metadata:
labels:
mylabel: myvalue
# ...
11.20.1. Customizing the image pull policy
Strimzi allows you to customize the image pull policy for containers in all pods deployed by the Cluster Operator.
The image pull policy is configured using the environment variable STRIMZI_IMAGE_PULL_POLICY
in the Cluster Operator deployment.
The STRIMZI_IMAGE_PULL_POLICY
environment variable can be set to three different values:
Always
-
Container images are pulled from the registry every time the pod is started or restarted.
IfNotPresent
-
Container images are pulled from the registry only when they were not pulled before.
Never
-
Container images are never pulled from the registry.
Currently, the image pull policy can only be customized for all Kafka, Kafka Connect, and Kafka MirrorMaker clusters at once. Changing the policy will result in a rolling update of all your Kafka, Kafka Connect, and Kafka MirrorMaker clusters.
11.20.2. Applying a termination grace period
Apply a termination grace period to give a Kafka cluster enough time to shut down cleanly.
Specify the time using the terminationGracePeriodSeconds
property.
Add the property to the template.pod
configuration of the Kafka
custom resource.
The time you add will depend on the size of your Kafka cluster. The Kubernetes default for the termination grace period is 30 seconds. If you observe that your clusters are not shutting down cleanly, you can increase the termination grace period.
A termination grace period is applied every time a pod is restarted. The period begins when Kubernetes sends a term (termination) signal to the processes running in the pod. The period should reflect the amount of time required to transfer the processes of the terminating pod to another pod before they are stopped. After the period ends, a kill signal stops any processes still running in the pod.
The following example adds a termination grace period of 120 seconds to the Kafka
custom resource.
You can also specify the configuration in the custom resources of other Kafka components.
apiVersion: kafka.strimzi.io/v1beta2
kind: Kafka
metadata:
name: my-cluster
spec:
kafka:
# ...
template:
pod:
terminationGracePeriodSeconds: 120
# ...
# ...
12. Using the Topic Operator to manage Kafka topics
The KafkaTopic
resource configures topics, including partition and replication factor settings.
When you create, modify, or delete a topic using KafkaTopic
, the Topic Operator ensures that these changes are reflected in the Kafka cluster.
For more information on the KafkaTopic
resource, see the KafkaTopic
schema reference.
12.1. Topic management
The KafkaTopic
resource is responsible for managing a single topic within a Kafka cluster.
The Topic Operator operates as follows:
-
When a
KafkaTopic
is created, deleted, or changed, the Topic Operator performs the corresponding operation on the Kafka topic.
If a topic is created, deleted, or modified directly within the Kafka cluster, without the presence of a corresponding KafkaTopic
resource, the Topic Operator does not manage that topic.
The Topic Operator will only manage Kafka topics associated with KafkaTopic
resources and does not interfere with topics managed independently within the Kafka cluster.
If a KafkaTopic
does exist for a Kafka topic, any configuration changes made outside the resource are reverted.
The Topic Operator can detect cases where where multiple KafkaTopic
resources are attempting to manage a Kafka topic using the same .spec.topicName
.
Only the oldest resource is reconciled, while the other resources fail with a resource conflict error.
12.2. Topic naming conventions
A KafkaTopic
resource includes a name for the topic and a label that identifies the name of the Kafka cluster it belongs to.
apiVersion: kafka.strimzi.io/v1beta2
kind: KafkaTopic
metadata:
name: topic-name-1
labels:
strimzi.io/cluster: my-cluster
spec:
topicName: topic-name-1
The label provides the cluster name of the Kafka
resource.
The Topic Operator uses the label as a mechanism for determining which KafkaTopic
resources to manage.
If the label does not match the Kafka cluster, the Topic Operator cannot see the KafkaTopic
, and the topic is not created.
Kafka and Kubernetes have their own naming validation rules, and a Kafka topic name might not be a valid resource name in Kubernetes. If possible, try and stick to a naming convention that works for both.
Consider the following guidelines:
-
Use topic names that reflect the nature of the topic
-
Be concise and keep the name under 63 characters
-
Use all lower case and hyphens
-
Avoid special characters, spaces or symbols
The KafkaTopic
resource allows you to specify the Kafka topic name using the metadata.name
field.
However, if the desired Kafka topic name is not a valid Kubernetes resource name, you can use the spec.topicName
property to specify the actual name.
The spec.topicName
field is optional, and when it’s absent, the Kafka topic name defaults to the metadata.name
of the topic.
When a topic is created, the topic name cannot be changed later.
apiVersion: kafka.strimzi.io/v1beta2
kind: KafkaTopic
metadata:
name: my-topic-1 # (1)
spec:
topicName: My.Topic.1 # (2)
# ...
-
A valid topic name that works in Kubernetes.
-
A Kafka topic name that uses upper case and periods, which are invalid in Kubernetes.
If more than one KafkaTopic
resource refers to the same Kafka topic, the resource that was created first is considered to be the one managing the topic.
The status of the newer resources is updated to indicate a conflict, and their Ready
status is changed to False
.
A Kafka client application, such as Kafka Streams, can automatically create topics with invalid Kubernetes resource names.
If you want to manage these topics, you must create KafkaTopic
resources with a different .metadata.name
, as shown in the previous example.
Note
|
For more information on the requirements for identifiers and names in a cluster, refer to the Kubernetes documentation Object Names and IDs. |
12.3. Handling changes to topics
Configuration changes only go in one direction: from the KafkaTopic
resource to the Kafka topic.
Any changes to a Kafka topic managed outside the KafkaTopic
resource are reverted.
12.3.1. Downgrading to a Strimzi version that uses internal topics to store topic metadata
If you are reverting back to a version of Strimzi earlier than 0.41, which uses internal topics for the storage of topic metadata, you still downgrade your Cluster Operator to the previous version, then downgrade Kafka brokers and client applications to the previous Kafka version as standard.
12.3.2. Downgrading to a Strimzi version that uses ZooKeeper to store topic metadata
If you are reverting back to a version of Strimzi earlier than 0.22, which uses ZooKeeper for the storage of topic metadata, you still downgrade your Cluster Operator to the previous version, then downgrade Kafka brokers and client applications to the previous Kafka version as standard.
However, you must also delete the topics that were created for the topic store using a kafka-topics
command, specifying the bootstrap address of the Kafka cluster.
For example:
kubectl run kafka-admin -ti --image=quay.io/strimzi/kafka:latest-kafka-3.9.0 --rm=true --restart=Never -- ./bin/kafka-topics.sh --bootstrap-server localhost:9092 --topic __strimzi-topic-operator-kstreams-topic-store-changelog --delete && ./bin/kafka-topics.sh --bootstrap-server localhost:9092 --topic __strimzi_store_topic --delete
The command must correspond to the type of listener and authentication used to access the Kafka cluster.
The Topic Operator will reconstruct the ZooKeeper topic metadata from the state of the topics in Kafka.
12.3.3. Automatic creation of topics
Applications can trigger the automatic creation of topics in the Kafka cluster.
By default, the Kafka broker configuration auto.create.topics.enable
is set to true
, allowing the broker to create topics automatically when an application attempts to produce or consume from a non-existing topic.
Applications might also use the Kafka AdminClient
to automatically create topics.
When an application is deployed along with its KafkaTopic
resources, it is possible that automatic topic creation in the cluster happens before the Topic Operator can react to the KafkaTopic
.
The topics created for an application deployment are initially created with default topic configuration.
If the Topic Operator attempts to reconfigure the topics based on KafkaTopic
resource specifications included with the application deployment, the operation might fail because the required change to the configuration is not allowed.
For example, if the change means lowering the number of topic partitions.
For this reason, it is recommended to disable auto.create.topics.enable
in the Kafka cluster configuration.
12.4. Configuring Kafka topics
Use the properties of the KafkaTopic
resource to configure Kafka topics.
Changes made to topic configuration in the KafkaTopic
are propagated to Kafka.
You can use kubectl apply
to create or modify topics, and kubectl delete
to delete existing topics.
For example:
-
kubectl apply -f <topic_config_file>
-
kubectl delete KafkaTopic <topic_name>
To be able to delete topics, delete.topic.enable
must be set to true
(default) in the spec.kafka.config
of the Kafka resource.
This procedure shows how to create a topic with 10 partitions and 2 replicas.
The KafkaTopic resource does not allow the following changes:
-
Renaming the topic defined in
spec.topicName
. A mismatch betweenspec.topicName
andstatus.topicName
will be detected. -
Decreasing the number of partitions using
spec.partitions
(not supported by Kafka). -
Modifying the number of replicas specified in
spec.replicas
.
Warning
|
Increasing spec.partitions for topics with keys will alter the partitioning of records, which can cause issues, especially when the topic uses semantic partitioning.
|
-
A running Kafka cluster configured with a Kafka broker listener using mTLS authentication and TLS encryption.
-
A running Topic Operator (typically deployed with the Entity Operator).
-
For deleting a topic,
delete.topic.enable=true
(default) in thespec.kafka.config
of theKafka
resource.
-
Configure the
KafkaTopic
resource.Example Kafka topic configurationapiVersion: kafka.strimzi.io/v1beta2 kind: KafkaTopic metadata: name: my-topic-1 labels: strimzi.io/cluster: my-cluster spec: partitions: 10 replicas: 2
TipWhen modifying a topic, you can get the current version of the resource using kubectl get kafkatopic my-topic-1 -o yaml
. -
Create the
KafkaTopic
resource in Kubernetes.kubectl apply -f <topic_config_file>
-
Wait for the ready status of the topic to change to
True
:kubectl get kafkatopics -o wide -w -n <namespace>
Kafka topic statusNAME CLUSTER PARTITIONS REPLICATION FACTOR READY my-topic-1 my-cluster 10 3 True my-topic-2 my-cluster 10 3 my-topic-3 my-cluster 10 3 True
Topic creation is successful when the
READY
output showsTrue
. -
If the
READY
column stays blank, get more details on the status from the resource YAML or from the Topic Operator logs.Status messages provide details on the reason for the current status.
oc get kafkatopics my-topic-2 -o yaml
Details on a topic with aNotReady
status# ... status: conditions: - lastTransitionTime: "2022-06-13T10:14:43.351550Z" message: Number of partitions cannot be decreased reason: PartitionDecreaseException status: "True" type: NotReady
In this example, the reason the topic is not ready is because the original number of partitions was reduced in the
KafkaTopic
configuration. Kafka does not support this.After resetting the topic configuration, the status shows the topic is ready.
kubectl get kafkatopics my-topic-2 -o wide -w -n <namespace>
Status update of the topicNAME CLUSTER PARTITIONS REPLICATION FACTOR READY my-topic-2 my-cluster 10 3 True
Fetching the details shows no messages
kubectl get kafkatopics my-topic-2 -o yaml
Details on a topic with aREADY
status# ... status: conditions: - lastTransitionTime: '2022-06-13T10:15:03.761084Z' status: 'True' type: Ready
12.5. Configuring topics for replication and number of partitions
The recommended configuration for topics managed by the Topic Operator is a topic replication factor of 3, and a minimum of 2 in-sync replicas.
apiVersion: kafka.strimzi.io/v1beta2
kind: KafkaTopic
metadata:
name: my-topic
labels:
strimzi.io/cluster: my-cluster
spec:
partitions: 10 # (1)
replicas: 3 # (2)
config:
min.insync.replicas: 2 # (3)
#...
-
The number of partitions for the topic.
-
The number of replica topic partitions. Changing the number of replicas in the topic configuration requires a deployment of Cruise Control. For more information, see Using Cruise Control to modify topic replication factor.
-
The minimum number of replica partitions that a message must be successfully written to, or an exception is raised.
Note
|
In-sync replicas are used in conjunction with the acks configuration for producer applications.
The acks configuration determines the number of follower partitions a message must be replicated to before the message is acknowledged as successfully received.
Replicas need to be reassigned when adding or removing brokers (see Scaling clusters by adding or removing brokers).
|
12.6. Managing KafkaTopic resources without impacting Kafka topics
This procedure describes how to convert Kafka topics that are currently managed through the KafkaTopic
resource into unmanaged topics.
This capability can be useful in various scenarios.
For instance, you might want to update the metadata.name
of a KafkaTopic
resource.
You can only do that by deleting the original KafkaTopic
resource and recreating a new one.
By annotating a KafkaTopic
resource with strimzi.io/managed=false
, you indicate that the Topic Operator should no longer manage that particular topic.
This allows you to retain the Kafka topic while making changes to the resource’s configuration or other administrative tasks.
-
Annotate the
KafkaTopic
resource in Kubernetes, settingstrimzi.io/managed
tofalse
:kubectl annotate kafkatopic my-topic-1 strimzi.io/managed="false" --overwrite
Specify the
metadata.name
of the topic in yourKafkaTopic
resource, which ismy-topic-1
in this example. -
Check the status of the
KafkaTopic
resource to make sure the request was successful:kubectl get kafkatopic my-topic-1 -o yaml
Example topic with anUnmanaged
statusapiVersion: kafka.strimzi.io/v1beta2 kind: KafkaTopic metadata: generation: 124 name: my-topic-1 finalizer: - strimzi.io/topic-operator labels: strimzi.io/cluster: my-cluster annotations: strimzi.io/managed: "false" spec: partitions: 10 replicas: 2 config: retention.ms: 432000000 status: observedGeneration: 124 # (1) conditions: - lastTransitionTime: "2024-08-22T06:07:57.671085635Z" status: "True" type: Unmanaged # (2)
-
The value of
metadata.generation
mustmatch status.observedGeneration
. -
The
Unmanaged
condition means that theKafkaTopic
is no longer reconciled.
-
-
You can now make changes to the
KafkaTopic
resource without it affecting the Kafka topic it was managing.For example, to change the
metadata.name
, do as follows:-
Delete the original
KafkTopic
resource:kubectl delete kafkatopic <kafka_topic_name>
-
Recreate the
KafkTopic
resource with a differentmetadata.name
, but usespec.topicName
to refer to the same topic that was managed by the original
-
-
If you haven’t deleted the original
KafkaTopic
resource, and you wish to resume management of the Kafka topic again, set thestrimzi.io/managed
annotation totrue
or remove the annotation.
12.7. Enabling topic management for existing Kafka topics
This procedure describes how to enable topic management for topics that are not currently managed through the KafkaTopic
resource.
You do this by creating a matching KafkaTopic
resource.
-
Create a
KafkaTopic
resource with ametadata.name
that is the same as the Kafka topic.Or use
spec.topicName
if the name of the topic in Kafka would not be a legal Kubernetes resource name.Example Kafka topic configurationapiVersion: kafka.strimzi.io/v1beta2 kind: KafkaTopic metadata: name: my-topic-1 labels: strimzi.io/cluster: my-cluster spec: partitions: 10 replicas: 2
In this example, the Kafka topic is named
my-topic-1
.The Topic Operator checks whether the topic is managed by another
KafkaTopic
resource. If it is, the older resource takes precedence and a resource conflict error is returned in the status of the new resource. -
Apply the
KafkaTopic
resource:kubectl apply -f <topic_configuration_file>
-
Wait for the operator to update the topic in Kafka.
The operator updates the Kafka topic with the
spec
of theKafkaTopic
that has the same name. -
Check the status of the
KafkaTopic
resource to make sure the request was successful:oc get kafkatopics my-topic-1 -o yaml
Example topic with aReady
statusapiVersion: kafka.strimzi.io/v1beta2 kind: KafkaTopic metadata: generation: 1 name: my-topic-1 labels: strimzi.io/cluster: my-cluster spec: partitions: 10 replicas: 2 # ... status: observedGeneration: 1 # (1) topicName: my-topic-1 conditions: - type: Ready status: True lastTransitionTime: 20230301T103000Z
-
Successful reconciliation of the resource means the topic is now managed.
The value of
metadata.generation
(the current version of the deployment) mustmatch status.observedGeneration
(the latest reconciliation of the resource).
-
12.8. Deleting managed topics
The Topic Operator supports the deletion of topics managed through the KafkaTopic
resource with or without Kubernetes finalizers.
This is determined by the STRIMZI_USE_FINALIZERS
Topic Operator environment variable.
By default, this is set to true
, though it can be set to false
in the Topic Operator env
configuration if you do not want the Topic Operator to add finalizers.
Finalizers ensure orderly and controlled deletion of KafkaTopic
resources.
A finalizer for the Topic Operator is added to the metadata of the KafkaTopic
resource:
apiVersion: kafka.strimzi.io/v1beta2
kind: KafkaTopic
metadata:
generation: 1
name: my-topic-1
finalizers:
- strimzi.io/topic-operator
labels:
strimzi.io/cluster: my-cluster
In this example, the finalizer is added for topic my-topic-1
.
The finalizer prevents the topic from being fully deleted until the finalization process is complete.
If you then delete the topic using kubectl delete kafkatopic my-topic-1
, a timestamp is added to the metadata:
apiVersion: kafka.strimzi.io/v1beta2
kind: KafkaTopic
metadata:
generation: 1
name: my-topic-1
finalizers:
- strimzi.io/topic-operator
labels:
strimzi.io/cluster: my-cluster
deletionTimestamp: 20230301T000000.000
The resource is still present. If the deletion fails, it is shown in the status of the resource.
When the finalization tasks are successfully executed, the finalizer is removed from the metadata, and the resource is fully deleted.
Finalizers also serve to prevent related resources from being deleted.
If the Topic Operator is not running, it won’t be able to remove its finalizer from the metadata.finalizers
.
And any attempt to directly delete the KafkaTopic
resources or the namespace will fail or timeout, leaving the namespace in a stuck terminating state.
If this happens, you can bypass the finalization process by removing the finalizers on topics.
12.9. Removing finalizers on topics
If the Topic Operator is not running, and you want to bypass the finalization process when deleting managed topics, you must remove the finalizers. You can do this manually by editing the resources directly or by using a command.
To remove finalizers on all topics, use the following command:
kubectl get kt -o=json | jq '.items[].metadata.finalizers = null' | kubectl apply -f -
The command uses the jq
command line JSON parser tool to modify the KafkaTopic
(kt
) resources by setting the finalizers to null
.
You can also use the command for a specific topic:
kubectl get kt <topic_name> -o=json | jq '.metadata.finalizers = null' | kubectl apply -f -
After running the command, you can go ahead and delete the topics. Alternatively, if the topics were already being deleted but were blocked due to outstanding finalizers then their deletion should complete.
Warning
|
Be careful when removing finalizers, as any cleanup operations associated with the finalization process are not performed if the Topic Operator is not running.
For example, if you remove the finalizer from a KafkaTopic resource and subsequently delete the resource, the related Kafka topic won’t be deleted.
|
12.10. Considerations when disabling topic deletion
When the delete.topic.enable
configuration in Kafka is set to false
, topics cannot be deleted.
This might be required in certain scenarios, but it introduces a consideration when using the Topic Operator.
As topics cannot be deleted, finalizers added to the metadata of a KafkaTopic
resource to control topic deletion are never removed by the Topic Operator (though they can be removed manually).
Similarly, any Custom Resource Definitions (CRDs) or namespaces associated with topics cannot be deleted.
Before configuring delete.topic.enable=false
, assess these implications to ensure it aligns with your specific requirements.
Note
|
To avoid using finalizers, you can set the STRIMZI_USE_FINALIZERS Topic Operator environment variable to false .
|
12.11. Tuning request batches for topic operations
The Topic Operator uses the request batching capabilities of the Kafka Admin API for operations on topic resources. You can fine-tune the batching mechanism using the following operator configuration properties:
-
STRIMZI_MAX_QUEUE_SIZE
to set the maximum size of the topic event queue. The default value is 1024. -
STRIMZI_MAX_BATCH_SIZE
to set the maximum number of topic events allowed in a single batch. The default value is 100. -
MAX_BATCH_LINGER_MS
to specify the maximum time to wait for a batch to accumulate items before processing. The default is 100 milliseconds.
If the maximum size of the request batching queue is exceeded, the Topic Operator shuts down and is restarted.
To prevent frequent restarts, consider adjusting the STRIMZI_MAX_QUEUE_SIZE
property to accommodate the typical load.
13. Using the User Operator to manage Kafka users
When you create, modify or delete a user using the KafkaUser
resource,
the User Operator ensures that these changes are reflected in the Kafka cluster.
For more information on the KafkaUser
resource, see the KafkaUser
schema reference.
13.1. Configuring Kafka users
Use the properties of the KafkaUser
resource to configure Kafka users.
You can use kubectl apply
to create or modify users, and kubectl delete
to delete existing users.
For example:
-
kubectl apply -f <user_config_file>
-
kubectl delete KafkaUser <user_name>
Users represent Kafka clients.
When you configure Kafka users, you enable the user authentication and authorization mechanisms required by clients to access Kafka.
The mechanism used must match the equivalent Kafka
configuration.
For more information on using Kafka
and KafkaUser
resources to secure access to Kafka brokers, see Securing access to a Kafka cluster.
-
A running Kafka cluster configured with a Kafka broker listener using mTLS authentication and TLS encryption.
-
A running User Operator (typically deployed with the Entity Operator).
-
Configure the
KafkaUser
resource.This example specifies mTLS authentication and simple authorization using ACLs.
Example Kafka user configurationapiVersion: kafka.strimzi.io/v1beta2 kind: KafkaUser metadata: name: my-user-1 labels: strimzi.io/cluster: my-cluster spec: authentication: type: tls authorization: type: simple acls: # Example consumer Acls for topic my-topic using consumer group my-group - resource: type: topic name: my-topic patternType: literal operations: - Describe - Read host: "*" - resource: type: group name: my-group patternType: literal operations: - Read host: "*" # Example Producer Acls for topic my-topic - resource: type: topic name: my-topic patternType: literal operations: - Create - Describe - Write host: "*"
-
Create the
KafkaUser
resource in Kubernetes.kubectl apply -f <user_config_file>
-
Wait for the ready status of the user to change to
True
:kubectl get kafkausers -o wide -w -n <namespace>
Kafka user statusNAME CLUSTER AUTHENTICATION AUTHORIZATION READY my-user-1 my-cluster tls simple True my-user-2 my-cluster tls simple my-user-3 my-cluster tls simple True
User creation is successful when the
READY
output showsTrue
. -
If the
READY
column stays blank, get more details on the status from the resource YAML or User Operator logs.Messages provide details on the reason for the current status.
kubectl get kafkausers my-user-2 -o yaml
Details on a user with aNotReady
status# ... status: conditions: - lastTransitionTime: "2022-06-10T10:07:37.238065Z" message: Simple authorization ACL rules are configured but not supported in the Kafka cluster configuration. reason: InvalidResourceException status: "True" type: NotReady
In this example, the reason the user is not ready is because simple authorization is not enabled in the
Kafka
configuration.Kafka configuration for simple authorizationapiVersion: kafka.strimzi.io/v1beta2 kind: Kafka metadata: name: my-cluster spec: kafka: # ... authorization: type: simple
After updating the Kafka configuration, the status shows the user is ready.
kubectl get kafkausers my-user-2 -o wide -w -n <namespace>
Status update of the userNAME CLUSTER AUTHENTICATION AUTHORIZATION READY my-user-2 my-cluster tls simple True
Fetching the details shows no messages.
kubectl get kafkausers my-user-2 -o yaml
Details on a user with aREADY
status# ... status: conditions: - lastTransitionTime: "2022-06-10T10:33:40.166846Z" status: "True" type: Ready
14. Setting up client access to a Kafka cluster
After you have deployed Strimzi, you can set up client access to your Kafka cluster. To verify the deployment, you can deploy example producer and consumer clients. Otherwise, create listeners that provide client access within or outside the Kubernetes cluster.
14.1. Deploying example clients
Send and receive messages from a Kafka cluster installed on Kubernetes.
This procedure describes how to deploy Kafka clients to the Kubernetes cluster, then produce and consume messages to test your installation. The clients are deployed using the Kafka container image.
-
The Kafka cluster is available for the clients.
-
Deploy a Kafka producer.
This example deploys a Kafka producer that connects to the Kafka cluster
my-cluster
.A topic named
my-topic
is created.Deploying a Kafka producer to Kuberneteskubectl run kafka-producer -ti --image=quay.io/strimzi/kafka:latest-kafka-3.9.0 --rm=true --restart=Never -- bin/kafka-console-producer.sh --bootstrap-server my-cluster-kafka-bootstrap:9092 --topic my-topic
-
Type a message into the console where the producer is running.
-
Press Enter to send the message.
-
Deploy a Kafka consumer.
The consumer should consume messages produced to
my-topic
in the Kafka clustermy-cluster
.Deploying a Kafka consumer to Kuberneteskubectl run kafka-consumer -ti --image=quay.io/strimzi/kafka:latest-kafka-3.9.0 --rm=true --restart=Never -- bin/kafka-console-consumer.sh --bootstrap-server my-cluster-kafka-bootstrap:9092 --topic my-topic --from-beginning
-
Confirm that you see the incoming messages in the consumer console.
14.2. Configuring listeners to connect to Kafka
Use listeners to enable client connections to Kafka.
Strimzi provides a generic GenericKafkaListener
schema with properties to configure listeners through the Kafka
resource.
When configuring a Kafka cluster, you specify a listener type
based on your requirements, environment, and infrastructure.
Services, routes, load balancers, and ingresses for clients to connect to a cluster are created according to the listener type.
Internal and external listener types are supported.
- Internal listeners
-
Use internal listener types to connect clients within a kubernetes cluster.
-
internal
to connect within the same Kubernetes cluster -
cluster-ip
to expose Kafka using per-brokerClusterIP
servicesInternal listeners use a headless service and the DNS names assigned to the broker pods. By default, they do not use the Kubernetes service DNS domain (typically
.cluster.local
). However, you can customize this configuration using theuseServiceDnsDomain
property. Consider using acluster-ip
type listener if routing through the headless service isn’t feasible or if you require a custom access mechanism, such as when integrating with specific Ingress controllers or the Kubernetes Gateway API.
-
- External listeners
-
Use external listener types to connect clients outside a Kubernetes cluster.
-
nodeport
to use ports on Kubernetes nodes -
loadbalancer
to use loadbalancer services -
ingress
to use KubernetesIngress
and the Ingress NGINX Controller for Kubernetes (Kubernetes only) -
route
to use OpenShiftRoute
and the default HAProxy router (OpenShift only)External listeners handle access to a Kafka cluster from networks that require different authentication mechanisms. For example, loadbalancers might not be suitable for certain infrastructure, such as bare metal, where node ports provide a better option.
-
Important
|
Do not use the built-in ingress controller on OpenShift, use the route type instead. The Ingress NGINX Controller is only intended for use on Kubernetes. The route type is only supported on OpenShift.
|
Each listener is defined as an array in the Kafka
resource.
apiVersion: kafka.strimzi.io/v1beta2
kind: Kafka
metadata:
name: my-cluster
spec:
kafka:
# ...
listeners:
- name: plain
port: 9092
type: internal
tls: false
configuration:
useServiceDnsDomain: true
- name: tls
port: 9093
type: internal
tls: true
authentication:
type: tls
- name: external1
port: 9094
type: route
tls: true
configuration:
brokerCertChainAndKey:
secretName: my-secret
certificate: my-certificate.crt
key: my-key.key
# ...
You can configure as many listeners as required, as long as their names and ports are unique. You can also configure listeners for secure connection using authentication.
Note
|
If you scale your Kafka cluster while using external listeners, it might trigger a rolling update of all Kafka brokers. This depends on the configuration. |
14.3. Listener naming conventions
From the listener configuration, the resulting listener bootstrap and per-broker service names are structured according to the following naming conventions:
Listener type | Bootstrap service name | Per-Broker service name |
---|---|---|
|
<cluster_name>-kafka-bootstrap |
Not applicable |
|
<cluster_name>-kafka-<listener-name>-bootstrap |
<cluster_name>-kafka-<listener-name>-<idx> |
For example, my-cluster-kafka-bootstrap
, my-cluster-kafka-external1-bootstrap
, and my-cluster-kafka-external1-0
.
The names are assigned to the services, routes, load balancers, and ingresses created through the listener configuration.
You can use certain backwards compatible names and port numbers to transition listeners initially configured under the retired KafkaListeners
schema.
The resulting external listener naming convention varies slightly.
The specific combinations of listener name and port configuration values in the following table are backwards compatible.
Listener name | Port | Bootstrap service name | Per-Broker service name |
---|---|---|---|
|
|
<cluster_name>-kafka-bootstrap |
Not applicable |
|
|
<cluster-name>-kafka-bootstrap |
Not applicable |
|
|
<cluster_name>-kafka-bootstrap |
<cluster_name>-kafka-bootstrap-<idx> |
14.4. Accessing Kafka using node ports
Use node ports to access a Kafka cluster from an external client outside the Kubernetes cluster.
To connect to a broker, you specify a hostname and port number for the Kafka bootstrap address, as well as the certificate used for TLS encryption.
The procedure shows basic nodeport
listener configuration.
You can use listener properties to enable TLS encryption (tls
) and specify a client authentication mechanism (authentication
).
Add additional configuration using configuration
properties.
For example, you can use the following configuration properties with nodeport
listeners:
preferredNodePortAddressType
-
Specifies the first address type that’s checked as the node address.
externalTrafficPolicy
-
Specifies whether the service routes external traffic to node-local or cluster-wide endpoints.
nodePort
-
Overrides the assigned node port numbers for the bootstrap and broker services.
For more information on listener configuration, see the GenericKafkaListener
schema reference.
-
A running Cluster Operator
In this procedure, the Kafka cluster name is my-cluster
.
The name of the listener is external4
.
-
Configure a
Kafka
resource with an external listener set to thenodeport
type.For example:
apiVersion: kafka.strimzi.io/v1beta2 kind: Kafka metadata: labels: app: my-cluster name: my-cluster namespace: myproject spec: kafka: # ... listeners: - name: external4 port: 9094 type: nodeport tls: true authentication: type: tls # ... # ... zookeeper: # ...
-
Create or update the resource.
kubectl apply -f <kafka_configuration_file>
A cluster CA certificate to verify the identity of the kafka brokers is created in the secret
my-cluster-cluster-ca-cert
.NodePort
type services are created for each Kafka broker, as well as an external bootstrap service.Node port services created for the bootstrap and brokersNAME TYPE CLUSTER-IP PORT(S) my-cluster-kafka-external4-0 NodePort 172.30.55.13 9094:31789/TCP my-cluster-kafka-external4-1 NodePort 172.30.250.248 9094:30028/TCP my-cluster-kafka-external4-2 NodePort 172.30.115.81 9094:32650/TCP my-cluster-kafka-external4-bootstrap NodePort 172.30.30.23 9094:32650/TCP
The bootstrap address used for client connection is propagated to the
status
of theKafka
resource.Example status for the bootstrap addressstatus: clusterId: Y_RJQDGKRXmNF7fEcWldJQ conditions: - lastTransitionTime: '2023-01-31T14:59:37.113630Z' status: 'True' type: Ready kafkaVersion: 3.9.0 listeners: # ... - addresses: - host: ip-10-0-224-199.us-west-2.compute.internal port: 32650 bootstrapServers: 'ip-10-0-224-199.us-west-2.compute.internal:32650' certificates: - | -----BEGIN CERTIFICATE----- -----END CERTIFICATE----- name: external4 observedGeneration: 2 operatorLastSuccessfulVersion: latest # ...
-
Retrieve the bootstrap address you can use to access the Kafka cluster from the status of the
Kafka
resource.kubectl get kafka my-cluster -o=jsonpath='{.status.listeners[?(@.name=="external4")].bootstrapServers}{"\n"}' ip-10-0-224-199.us-west-2.compute.internal:32650
-
Extract the cluster CA certificate.
kubectl get secret my-cluster-cluster-ca-cert -o jsonpath='{.data.ca\.crt}' | base64 -d > ca.crt
-
Configure your client to connect to the brokers.
-
Specify the bootstrap host and port in your Kafka client as the bootstrap address to connect to the Kafka cluster. For example,
ip-10-0-224-199.us-west-2.compute.internal:32650
. -
Add the extracted certificate to the truststore of your Kafka client to configure a TLS connection.
If you enabled a client authentication mechanism, you will also need to configure it in your client.
-
Note
|
If you are using your own listener certificates, check whether you need to add the CA certificate to the client’s truststore configuration. If it is a public (external) CA, you usually won’t need to add it. |
14.5. Accessing Kafka using loadbalancers
Use loadbalancers to access a Kafka cluster from an external client outside the Kubernetes cluster.
To connect to a broker, you specify a hostname and port number for the Kafka bootstrap address, as well as the certificate used for TLS encryption.
The procedure shows basic loadbalancer
listener configuration.
You can use listener properties to enable TLS encryption (tls
) and specify a client authentication mechanism (authentication
).
Add additional configuration using configuration
properties.
For example, you can use the following configuration properties with loadbalancer
listeners:
loadBalancerSourceRanges
-
Restricts traffic to a specified list of CIDR (Classless Inter-Domain Routing) ranges.
externalTrafficPolicy
-
Specifies whether the service routes external traffic to node-local or cluster-wide endpoints.
loadBalancerIP
-
Requests a specific IP address when creating a loadbalancer.
For more information on listener configuration, see the GenericKafkaListener
schema reference.
-
A running Cluster Operator
In this procedure, the Kafka cluster name is my-cluster
.
The name of the listener is external3
.
-
Configure a
Kafka
resource with an external listener set to theloadbalancer
type.For example:
apiVersion: kafka.strimzi.io/v1beta2 kind: Kafka metadata: labels: app: my-cluster name: my-cluster namespace: myproject spec: kafka: # ... listeners: - name: external3 port: 9094 type: loadbalancer tls: true authentication: type: tls # ... # ... zookeeper: # ...
-
Create or update the resource.
kubectl apply -f <kafka_configuration_file>
A cluster CA certificate to verify the identity of the kafka brokers is also created in the secret
my-cluster-cluster-ca-cert
.loadbalancer
type services and loadbalancers are created for each Kafka broker, as well as an external bootstrap service.Loadbalancer services and loadbalancers created for the bootstraps and brokersNAME TYPE CLUSTER-IP PORT(S) my-cluster-kafka-external3-0 LoadBalancer 172.30.204.234 9094:30011/TCP my-cluster-kafka-external3-1 LoadBalancer 172.30.164.89 9094:32544/TCP my-cluster-kafka-external3-2 LoadBalancer 172.30.73.151 9094:32504/TCP my-cluster-kafka-external3-bootstrap LoadBalancer 172.30.30.228 9094:30371/TCP NAME EXTERNAL-IP (loadbalancer) my-cluster-kafka-external3-0 a8a519e464b924000b6c0f0a05e19f0d-1132975133.us-west-2.elb.amazonaws.com my-cluster-kafka-external3-1 ab6adc22b556343afb0db5ea05d07347-611832211.us-west-2.elb.amazonaws.com my-cluster-kafka-external3-2 a9173e8ccb1914778aeb17eca98713c0-777597560.us-west-2.elb.amazonaws.com my-cluster-kafka-external3-bootstrap a8d4a6fb363bf447fb6e475fc3040176-36312313.us-west-2.elb.amazonaws.com
The bootstrap address used for client connection is propagated to the
status
of theKafka
resource.Example status for the bootstrap addressstatus: clusterId: Y_RJQDGKRXmNF7fEcWldJQ conditions: - lastTransitionTime: '2023-01-31T14:59:37.113630Z' status: 'True' type: Ready kafkaVersion: 3.9.0 listeners: # ... - addresses: - host: >- a8d4a6fb363bf447fb6e475fc3040176-36312313.us-west-2.elb.amazonaws.com port: 9094 bootstrapServers: >- a8d4a6fb363bf447fb6e475fc3040176-36312313.us-west-2.elb.amazonaws.com:9094 certificates: - | -----BEGIN CERTIFICATE----- -----END CERTIFICATE----- name: external3 observedGeneration: 2 operatorLastSuccessfulVersion: latest # ...
The DNS addresses used for client connection are propagated to the
status
of each loadbalancer service.Example status for the bootstrap loadbalancerstatus: loadBalancer: ingress: - hostname: >- a8d4a6fb363bf447fb6e475fc3040176-36312313.us-west-2.elb.amazonaws.com # ...
-
Retrieve the bootstrap address you can use to access the Kafka cluster from the status of the
Kafka
resource.kubectl get kafka my-cluster -o=jsonpath='{.status.listeners[?(@.name=="external3")].bootstrapServers}{"\n"}' a8d4a6fb363bf447fb6e475fc3040176-36312313.us-west-2.elb.amazonaws.com:9094
-
Extract the cluster CA certificate.
kubectl get secret my-cluster-cluster-ca-cert -o jsonpath='{.data.ca\.crt}' | base64 -d > ca.crt
-
Configure your client to connect to the brokers.
-
Specify the bootstrap host and port in your Kafka client as the bootstrap address to connect to the Kafka cluster. For example,
a8d4a6fb363bf447fb6e475fc3040176-36312313.us-west-2.elb.amazonaws.com:9094
. -
Add the extracted certificate to the truststore of your Kafka client to configure a TLS connection.
If you enabled a client authentication mechanism, you will also need to configure it in your client.
-
Note
|
If you are using your own listener certificates, check whether you need to add the CA certificate to the client’s truststore configuration. If it is a public (external) CA, you usually won’t need to add it. |
14.6. Accessing Kafka using an Ingress NGINX Controller for Kubernetes
Use an Ingress NGINX Controller for Kubernetes to access a Kafka cluster from clients outside the Kubernetes cluster.
To be able to use an Ingress NGINX Controller for Kubernetes, add configuration for an ingress
type listener in the Kafka
custom resource.
When applied, the configuration creates a dedicated ingress and service for an external bootstrap and each broker in the cluster.
Clients connect to the bootstrap ingress, which routes them through the bootstrap service to connect to a broker.
Per-broker connections are then established using DNS names, which route traffic from the client to the broker through the broker-specific ingresses and services.
To connect to a broker, you specify a hostname for the ingress bootstrap address, as well as the certificate used for TLS encryption. For access using an ingress, the port used in the Kafka client is typically 443.
The procedure shows basic ingress
listener configuration.
TLS encryption (tls
) must be enabled.
You can also specify a client authentication mechanism (authentication
).
Add additional configuration using configuration
properties.
For example, you can use the class
configuration property with ingress
listeners to specify the ingress controller used.
For more information on listener configuration, see the GenericKafkaListener
schema reference.
Make sure that you enable TLS passthrough in your Ingress NGINX Controller for Kubernetes deployment. Kafka uses a binary protocol over TCP, but the Ingress NGINX Controller for Kubernetes is designed to work with a HTTP protocol. To be able to route TCP traffic through ingresses, Strimzi uses TLS passthrough with Server Name Indication (SNI).
SNI helps with identifying and passing connection to Kafka brokers.
In passthrough mode, TLS encryption is always used.
Because the connection passes to the brokers, the listeners use the TLS certificates signed by the internal cluster CA and not the ingress certificates.
To configure listeners to use your own listener certificates, use the brokerCertChainAndKey
property.
For more information about enabling TLS passthrough, see the TLS passthrough documentation.
-
An Ingress NGINX Controller for Kubernetes is running with TLS passthrough enabled
-
A running Cluster Operator
In this procedure, the Kafka cluster name is my-cluster
.
The name of the listener is external2
.
-
Configure a
Kafka
resource with an external listener set to theingress
type.Specify an ingress hostname for the bootstrap service and for the Kafka brokers in the Kafka cluster.
For example:
apiVersion: kafka.strimzi.io/v1beta2 kind: Kafka metadata: labels: app: my-cluster name: my-cluster namespace: myproject spec: kafka: # ... listeners: - name: external2 port: 9094 type: ingress tls: true # (1) authentication: type: tls configuration: class: nginx # (2) hostTemplate: broker-{nodeId}.myingress.com # (3) bootstrap: host: bootstrap.myingress.com # (4) # ... zookeeper: # ...
-
For
ingress
type listeners, TLS encryption must be enabled (true
). -
(Optional) Class that specifies the ingress controller to use. You might need to add a class if you have not set up a default and a class name is missing in the ingresses created.
-
The host template used to generate the hostnames for the per-broker Ingress resources.
-
The host used as the hostnames for the bootstrap Ingress resource.
-
-
Create or update the resource.
kubectl apply -f <kafka_configuration_file>
A cluster CA certificate to verify the identity of the kafka brokers is created in the secret
my-cluster-cluster-ca-cert
.ClusterIP
type services are created for each Kafka broker, as well as an external bootstrap service.An
ingress
is also created for each service, with a DNS address to expose them using the Ingress NGINX Controller for Kubernetes.Ingresses created for the bootstrap and brokersNAME CLASS HOSTS ADDRESS PORTS my-cluster-kafka-external2-0 nginx broker-0.myingress.com 192.168.49.2 80,443 my-cluster-kafka-external2-1 nginx broker-1.myingress.com 192.168.49.2 80,443 my-cluster-kafka-external2-2 nginx broker-2.myingress.com 192.168.49.2 80,443 my-cluster-kafka-external2-bootstrap nginx bootstrap.myingress.com 192.168.49.2 80,443
The DNS addresses used for client connection are propagated to the
status
of each ingress.Status for the bootstrap ingressstatus: loadBalancer: ingress: - ip: 192.168.49.2 # ...
-
Use a target broker to check the client-server TLS connection on port 443 using the OpenSSL
s_client
.openssl s_client -connect broker-0.myingress.com:443 -servername broker-0.myingress.com -showcerts
The server name is the SNI for passing the connection to the broker.
If the connection is successful, the certificates for the broker are returned.
Certificates for the brokerCertificate chain 0 s:O = io.strimzi, CN = my-cluster-kafka i:O = io.strimzi, CN = cluster-ca v0
-
Extract the cluster CA certificate.
kubectl get secret my-cluster-cluster-ca-cert -o jsonpath='{.data.ca\.crt}' | base64 -d > ca.crt
-
Configure your client to connect to the brokers.
-
Specify the bootstrap host (from the listener
configuration
) and port 443 in your Kafka client as the bootstrap address to connect to the Kafka cluster. For example,bootstrap.myingress.com:443
. -
Add the extracted certificate to the truststore of your Kafka client to configure a TLS connection.
If you enabled a client authentication mechanism, you will also need to configure it in your client.
-
Note
|
If you are using your own listener certificates, check whether you need to add the CA certificate to the client’s truststore configuration. If it is a public (external) CA, you usually won’t need to add it. |
14.7. Accessing Kafka using OpenShift routes
Use OpenShift routes to access a Kafka cluster from clients outside the OpenShift cluster.
To be able to use routes, add configuration for a route
type listener in the Kafka
custom resource.
When applied, the configuration creates a dedicated route and service for an external bootstrap and each broker in the cluster.
Clients connect to the bootstrap route, which routes them through the bootstrap service to connect to a broker.
Per-broker connections are then established using DNS names, which route traffic from the client to the broker through the broker-specific routes and services.
To connect to a broker, you specify a hostname for the route bootstrap address, as well as the certificate used for TLS encryption. For access using routes, the port is always 443.
Warning
|
An OpenShift route address comprises the Kafka cluster name, the listener name, the project name, and the domain of the router.
For example, my-cluster-kafka-external1-bootstrap-my-project.domain.com (<cluster_name>-kafka-<listener_name>-bootstrap-<namespace>.<domain>).
Each DNS label (between periods “.”) must not exceed 63 characters, and the total length of the address must not exceed 255 characters.
|
The procedure shows basic listener configuration.
TLS encryption (tls
) must be enabled.
You can also specify a client authentication mechanism (authentication
).
Add additional configuration using configuration
properties.
For example, you can use the host
configuration property with route
listeners to specify the hostnames used by the bootstrap and per-broker services.
For more information on listener configuration, see the GenericKafkaListener
schema reference.
TLS passthrough is enabled for routes created by Strimzi. Kafka uses a binary protocol over TCP, but routes are designed to work with a HTTP protocol. To be able to route TCP traffic through routes, Strimzi uses TLS passthrough with Server Name Indication (SNI).
SNI helps with identifying and passing connection to Kafka brokers.
In passthrough mode, TLS encryption is always used.
Because the connection passes to the brokers, the listeners use TLS certificates signed by the internal cluster CA and not the ingress certificates.
To configure listeners to use your own listener certificates, use the brokerCertChainAndKey
property.
-
A running Cluster Operator
In this procedure, the Kafka cluster name is my-cluster
.
The name of the listener is external1
.
-
Configure a
Kafka
resource with an external listener set to theroute
type.For example:
apiVersion: kafka.strimzi.io/v1beta2 kind: Kafka metadata: labels: app: my-cluster name: my-cluster namespace: myproject spec: kafka: # ... listeners: - name: external1 port: 9094 type: route tls: true # (1) authentication: type: tls # ... # ... zookeeper: # ...
-
For
route
type listeners, TLS encryption must be enabled (true
).
-
-
Create or update the resource.
kubectl apply -f <kafka_configuration_file>
A cluster CA certificate to verify the identity of the kafka brokers is created in the secret
my-cluster-cluster-ca-cert
.ClusterIP
type services are created for each Kafka broker, as well as an external bootstrap service.A
route
is also created for each service, with a DNS address (host/port) to expose them using the default OpenShift HAProxy router.The routes are preconfigured with TLS passthrough.
Routes created for the bootstraps and brokersNAME HOST/PORT SERVICES PORT TERMINATION my-cluster-kafka-external1-0 my-cluster-kafka-external1-0-my-project.router.com my-cluster-kafka-external1-0 9094 passthrough my-cluster-kafka-external1-1 my-cluster-kafka-external1-1-my-project.router.com my-cluster-kafka-external1-1 9094 passthrough my-cluster-kafka-external1-2 my-cluster-kafka-external1-2-my-project.router.com my-cluster-kafka-external1-2 9094 passthrough my-cluster-kafka-external1-bootstrap my-cluster-kafka-external1-bootstrap-my-project.router.com my-cluster-kafka-external1-bootstrap 9094 passthrough
The DNS addresses used for client connection are propagated to the
status
of each route.Example status for the bootstrap routestatus: ingress: - host: >- my-cluster-kafka-external1-bootstrap-my-project.router.com # ...
-
Use a target broker to check the client-server TLS connection on port 443 using the OpenSSL
s_client
.openssl s_client -connect my-cluster-kafka-external1-0-my-project.router.com:443 -servername my-cluster-kafka-external1-0-my-project.router.com -showcerts
The server name is the Server Name Indication (SNI) for passing the connection to the broker.
If the connection is successful, the certificates for the broker are returned.
Certificates for the brokerCertificate chain 0 s:O = io.strimzi, CN = my-cluster-kafka i:O = io.strimzi, CN = cluster-ca v0
-
Retrieve the address of the bootstrap service from the status of the
Kafka
resource.kubectl get kafka my-cluster -o=jsonpath='{.status.listeners[?(@.name=="external1")].bootstrapServers}{"\n"}' my-cluster-kafka-external1-bootstrap-my-project.router.com:443
The address comprises the Kafka cluster name, the listener name, the project name and the domain of the router (
router.com
in this example). -
Extract the cluster CA certificate.
kubectl get secret my-cluster-cluster-ca-cert -o jsonpath='{.data.ca\.crt}' | base64 -d > ca.crt
-
Configure your client to connect to the brokers.
-
Specify the address for the bootstrap service and port 443 in your Kafka client as the bootstrap address to connect to the Kafka cluster.
-
Add the extracted certificate to the truststore of your Kafka client to configure a TLS connection.
If you enabled a client authentication mechanism, you will also need to configure it in your client.
-
Note
|
If you are using your own listener certificates, check whether you need to add the CA certificate to the client’s truststore configuration. If it is a public (external) CA, you usually won’t need to add it. |
14.8. Discovering connection details for clients
Service discovery makes it easier for client applications running in the same Kubernetes cluster as Strimzi to interact with a Kafka cluster.
A service discovery label and annotation are created for the following services:
-
Internal Kafka bootstrap service
-
Kafka Bridge service
- Service discovery label
-
The service discovery label,
strimzi.io/discovery
, is set totrue
forService
resources to make them discoverable for client connections. - Service discovery annotation
-
The service discovery annotation provides connection details in JSON format for each service for client applications to use to establish connections.
apiVersion: v1
kind: Service
metadata:
annotations:
strimzi.io/discovery: |-
[ {
"port" : 9092,
"tls" : false,
"protocol" : "kafka",
"auth" : "scram-sha-512"
}, {
"port" : 9093,
"tls" : true,
"protocol" : "kafka",
"auth" : "tls"
} ]
labels:
strimzi.io/cluster: my-cluster
strimzi.io/discovery: "true"
strimzi.io/kind: Kafka
strimzi.io/name: my-cluster-kafka-bootstrap
name: my-cluster-kafka-bootstrap
spec:
#...
apiVersion: v1
kind: Service
metadata:
annotations:
strimzi.io/discovery: |-
[ {
"port" : 8080,
"tls" : false,
"auth" : "none",
"protocol" : "http"
} ]
labels:
strimzi.io/cluster: my-bridge
strimzi.io/discovery: "true"
strimzi.io/kind: KafkaBridge
strimzi.io/name: my-bridge-bridge-service
Find services by specifying the discovery label when fetching services from the command line or a corresponding API call.
kubectl get service -l strimzi.io/discovery=true
Connection details are returned when retrieving the service discovery label.
15. Securing access to a Kafka cluster
Secure connections by configuring Kafka and Kafka users. Through configuration, you can implement encryption, authentication, and authorization mechanisms.
To establish secure access to Kafka, configure the Kafka
resource to set up the following configurations based on your specific requirements:
-
Listeners with specified authentication types to define how clients authenticate
-
TLS encryption for communication between Kafka and clients
-
Supported TLS versions and cipher suites for additional security
-
-
Authorization for the entire Kafka cluster
-
Network policies for restricting access
-
Super users for unconstrained access to brokers
Authentication is configured independently for each listener, while authorization is set up for the whole Kafka cluster.
For more information on access configuration for Kafka, see the Kafka
schema reference and GenericKafkaListener
schema reference.
To enable secure client access to Kafka, configure KafkaUser
resources.
These resources represent clients and determine how they authenticate and authorize with the Kafka cluster.
Configure the KafkaUser
resource to set up the following configurations based on your specific requirements:
-
Authentication that must match the enabled listener authentication
-
Supported TLS versions and cipher suites that must match the Kafka configuration
-
-
Simple authorization to apply Access Control List (ACL) rules
-
ACLs for fine-grained control over user access to topics and actions
-
-
Quotas to limit client access based on byte rates or CPU utilization
The User Operator creates the user representing the client and the security credentials used for client authentication, based on the chosen authentication type.
For more information on access configuration for users, see the KafkaUser
schema reference.
15.1. Configuring client authentication on listeners
Configure client authentication for Kafka brokers when creating listeners.
Specify the listener authentication type using the Kafka.spec.kafka.listeners.authentication
property in the Kafka
resource.
For clients inside the Kubernetes cluster, you can create plain
(without encryption) or tls
internal listeners.
The internal
listener type use a headless service and the DNS names given to the broker pods.
As an alternative to the headless service, you can also create a cluster-ip
type of internal listener to expose Kafka using per-broker ClusterIP
services.
For clients outside the Kubernetes cluster, you create external listeners and specify a connection mechanism,
which can be nodeport
, loadbalancer
, ingress
(Kubernetes only), or route
(OpenShift only).
For more information on the configuration options for connecting an external client, see Setting up client access to a Kafka cluster.
Supported authentication options:
-
mTLS authentication (only on the listeners with TLS enabled encryption)
-
SCRAM-SHA-512 authentication
If you’re using OAuth 2.0 for client access management, user authentication and authorization credentials are handled through the authorization server.
The authentication option you choose depends on how you wish to authenticate client access to Kafka brokers.
Note
|
Try exploring the standard authentication options before using custom authentication. Custom authentication allows for any type of Kafka-supported authentication. It can provide more flexibility, but also adds complexity. |
The listener authentication
property is used to specify an authentication mechanism specific to that listener.
If no authentication
property is specified then the listener does not authenticate clients which connect through that listener.
The listener will accept all connections without authentication.
Authentication must be configured when using the User Operator to manage KafkaUsers
.
The following example shows:
-
A
plain
listener configured for SCRAM-SHA-512 authentication -
A
tls
listener with mTLS authentication -
An
external
listener with mTLS authentication
Each listener is configured with a unique name and port within a Kafka cluster.
Important
|
When configuring listeners for client access to brokers, you can use port 9092 or higher (9093, 9094, and so on), but with a few exceptions. The listeners cannot be configured to use the ports reserved for interbroker communication (9090 and 9091), Prometheus metrics (9404), and JMX (Java Management Extensions) monitoring (9999). |
apiVersion: kafka.strimzi.io/v1beta2
kind: Kafka
metadata:
name: my-cluster
namespace: myproject
spec:
kafka:
# ...
listeners:
- name: plain
port: 9092
type: internal
tls: true
authentication:
type: scram-sha-512
- name: tls
port: 9093
type: internal
tls: true
authentication:
type: tls
- name: external3
port: 9094
type: loadbalancer
tls: true
authentication:
type: tls
# ...
15.1.1. mTLS authentication
mTLS authentication is always used for the communication between Kafka brokers and ZooKeeper pods.
Strimzi can configure Kafka to use TLS (Transport Layer Security) to provide encrypted communication between Kafka brokers and clients either with or without mutual authentication. For mutual, or two-way, authentication, both the server and the client present certificates. When you configure mTLS authentication, the broker authenticates the client (client authentication) and the client authenticates the broker (server authentication).
mTLS listener configuration in the Kafka
resource requires the following:
-
tls: true
to specify TLS encryption and server authentication -
authentication.type: tls
to specify the client authentication
When a Kafka cluster is created by the Cluster Operator, it creates a new secret with the name <cluster_name>-cluster-ca-cert
.
The secret contains a CA certificate.
The CA certificate is in PEM and PKCS #12 format.
To verify a Kafka cluster, add the CA certificate to the truststore in your client configuration.
To verify a client, add a user certificate and key to the keystore in your client configuration.
For more information on configuring a client for mTLS, see Configuring user authentication.
Note
|
TLS authentication is more commonly one-way, with one party authenticating the identity of another. For example, when HTTPS is used between a web browser and a web server, the browser obtains proof of the identity of the web server. |
15.1.2. SCRAM-SHA-512 authentication
SCRAM (Salted Challenge Response Authentication Mechanism) is an authentication protocol that can establish mutual authentication using passwords. Strimzi can configure Kafka to use SASL (Simple Authentication and Security Layer) SCRAM-SHA-512 to provide authentication on both unencrypted and encrypted client connections.
When SCRAM-SHA-512 authentication is used with a TLS connection, the TLS protocol provides the encryption, but is not used for authentication.
The following properties of SCRAM make it safe to use SCRAM-SHA-512 even on unencrypted connections:
-
The passwords are not sent in the clear over the communication channel. Instead the client and the server are each challenged by the other to offer proof that they know the password of the authenticating user.
-
The server and client each generate a new challenge for each authentication exchange. This means that the exchange is resilient against replay attacks.
When KafkaUser.spec.authentication.type
is configured with scram-sha-512
the User Operator will generate a random 32-character password consisting of upper and lowercase ASCII letters and numbers.
15.1.3. Restricting access to listeners with network policies
Control listener access by configuring the networkPolicyPeers
property in the Kafka
resource.
By default, Strimzi automatically creates a NetworkPolicy
resource for every enabled Kafka listener, allowing connections from all namespaces.
To restrict listener access to specific applications or namespaces at the network level, configure the networkPolicyPeers
property.
Each listener can have its own networkPolicyPeers
configuration.
For more information on network policy peers, refer to the NetworkPolicyPeer API reference.
If you want to use custom network policies, you can set the STRIMZI_NETWORK_POLICY_GENERATION
environment variable to false
in the Cluster Operator configuration.
For more information, see Configuring the Cluster Operator.
Note
|
Your configuration of Kubernetes must support ingress NetworkPolicies in order to use network policies.
|
-
A Kubernetes cluster with support for Ingress NetworkPolicies.
-
The Cluster Operator is running.
-
Configure the
networkPolicyPeers
property to define the application pods or namespaces allowed to access the Kafka cluster.This example shows configuration for a
tls
listener to allow connections only from application pods with the labelapp
set tokafka-client
:apiVersion: kafka.strimzi.io/v1beta2 kind: Kafka spec: kafka: # ... listeners: - name: tls port: 9093 type: internal tls: true authentication: type: tls networkPolicyPeers: - podSelector: matchLabels: app: kafka-client # ... zookeeper: # ...
-
Apply the changes to the
Kafka
resource configuration.
15.1.4. Using custom listener certificates for TLS encryption
This procedure shows how to configure custom server certificates for TLS listeners or external listeners which have TLS encryption enabled.
By default, Kafka listeners use certificates signed by Strimzi’s internal CA (certificate authority). The Cluster Operator automatically generates a CA certificate when creating a Kafka cluster. To configure a client for TLS, the CA certificate is included in its truststore configuration to authenticate the Kafka cluster. Alternatively, you have the option to install and use your own CA certificates.
However, if you prefer more granular control by using your own custom certificates at the listener-level, you can configure listeners using brokerCertChainAndKey
properties.
You create a secret with your own private key and server certificate, then specify them in the brokerCertChainAndKey
configuration.
User-provided certificates allow you to leverage existing security infrastructure. You can use a certificate signed by a public (external) CA or a private CA. Kafka clients need to trust the CA which was used to sign the listener certificate. If signed by a public CA, you usually won’t need to add it to a client’s truststore configuration.
Custom certificates are not managed by Strimzi, so you need to renew them manually.
Note
|
Listener certificates are used for TLS encryption and server authentication only. They are not used for TLS client authentication. If you want to use your own certificate for TLS client authentication as well, you must install and use your own clients CA. |
-
The Cluster Operator is running.
-
Each listener requires the following:
-
A compatible server certificate signed by an external CA. (Provide an X.509 certificate in PEM format.)
You can use one listener certificate for multiple listeners.
-
Subject Alternative Names (SANs) are specified in the certificate for each listener. For more information, see Specifying SANs for custom listener certificates.
-
If you are not using a self-signed certificate, you can provide a certificate that includes the whole CA chain in the certificate.
You can only use the brokerCertChainAndKey
properties if TLS encryption (tls: true
) is configured for the listener.
Note
|
Strimzi does not support the use of encrypted private keys for TLS. The private key stored in the secret must be unencrypted for this to work. |
-
Create a
Secret
containing your private key and server certificate:kubectl create secret generic <my_secret> --from-file=<my_listener_key.key> --from-file=<my_listener_certificate.crt>
-
Edit the
Kafka
resource for your cluster.Configure the listener to use your
Secret
, certificate file, and private key file in theconfiguration.brokerCertChainAndKey
property.Example configuration for aloadbalancer
external listener with TLS encryption enabled# ... listeners: - name: plain port: 9092 type: internal tls: false - name: external3 port: 9094 type: loadbalancer tls: true configuration: brokerCertChainAndKey: secretName: my-secret certificate: my-listener-certificate.crt key: my-listener-key.key # ...
Example configuration for a TLS listener# ... listeners: - name: plain port: 9092 type: internal tls: false - name: tls port: 9093 type: internal tls: true configuration: brokerCertChainAndKey: secretName: my-secret certificate: my-listener-certificate.crt key: my-listener-key.key # ...
-
Apply the changes to the
Kafka
resource configuration.The Cluster Operator starts a rolling update of the Kafka cluster, which updates the configuration of the listeners.
NoteA rolling update is also started if you update a Kafka listener certificate in a Secret
that is already used by a listener.
15.1.5. Specifying SANs for custom listener certificates
In order to use TLS hostname verification with custom Kafka listener certificates, you must specify the correct Subject Alternative Names (SANs) for each listener.
The certificate SANs must specify hostnames for the following:
-
All of the Kafka brokers in your cluster
-
The Kafka cluster bootstrap service
You can use wildcard certificates if they are supported by your CA.
Examples of SANs for internal listeners
Use the following examples to help you specify hostnames of the SANs in your certificates for your internal listeners.
Replace <cluster-name>
with the name of the Kafka cluster and <namespace>
with the Kubernetes namespace where the cluster is running.
type: internal
listener//Kafka brokers
*.<cluster_name>-kafka-brokers
*.<cluster_name>-kafka-brokers.<namespace>.svc
// Bootstrap service
<cluster_name>-kafka-bootstrap
<cluster_name>-kafka-bootstrap.<namespace>.svc
type: internal
listener// Kafka brokers
<cluster_name>-kafka-0.<cluster_name>-kafka-brokers
<cluster_name>-kafka-0.<cluster_name>-kafka-brokers.<namespace>.svc
<cluster_name>-kafka-1.<cluster_name>-kafka-brokers
<cluster_name>-kafka-1.<cluster_name>-kafka-brokers.<namespace>.svc
# ...
// Bootstrap service
<cluster_name>-kafka-bootstrap
<cluster_name>-kafka-bootstrap.<namespace>.svc
type: cluster-ip
listener// Kafka brokers
<cluster_name>-kafka-<listener-name>-0
<cluster_name>-kafka-<listener-name>-0.<namespace>.svc
<cluster_name>-kafka-_listener-name>-1
<cluster_name>-kafka-<listener-name>-1.<namespace>.svc
# ...
// Bootstrap service
<cluster_name>-kafka-<listener-name>-bootstrap
<cluster_name>-kafka-<listener-name>-bootstrap.<namespace>.svc
Examples of SANs for external listeners
For external listeners which have TLS encryption enabled, the hostnames you need to specify in certificates depends on the external listener type
.
External listener type | In the SANs, specify… |
---|---|
|
Addresses of all Kafka broker You can use a matching wildcard name. |
|
Addresses of all Kafka broker You can use a matching wildcard name. |
|
Addresses of all Kafka broker You can use a matching wildcard name. |
|
Addresses of all Kubernetes worker nodes that the Kafka broker pods might be scheduled to. You can use a matching wildcard name. |
15.2. Configuring authorized access to Kafka
Configure authorized access to a Kafka cluster using the Kafka.spec.kafka.authorization
property in the Kafka
resource.
If the authorization
property is missing, no authorization is enabled and clients have no restrictions.
When enabled, authorization is applied to all enabled listeners.
The authorization method is defined in the type
field.
Supported authorization options:
-
OAuth 2.0 authorization (if you are using OAuth 2.0 token based authentication)
15.2.1. Designating super users
Super users can access all resources in your Kafka cluster regardless of any access restrictions, and are supported by all authorization mechanisms.
To designate super users for a Kafka cluster, add a list of user principals to the superUsers
property.
If a user uses mTLS authentication, the username is the common name from the TLS certificate subject prefixed with CN=
.
If you are not using the User Operator and using your own certificates for mTLS, the username is the full certificate subject.
A full certificate subject can include the following fields:
-
CN=<common_name>
-
OU=<organizational_unit>
-
O=<organization>
-
L=<locality>
-
ST=<state>
-
C=<country_code>
Omit any fields that are not applicable.
apiVersion: kafka.strimzi.io/v1beta2
kind: Kafka
metadata:
name: my-cluster
namespace: myproject
spec:
kafka:
# ...
authorization:
type: simple
superUsers:
- CN=user-1
- user-2
- CN=user-3
- CN=user-4,OU=my-ou,O=my-org,L=my-location,ST=my-state,C=US
- CN=user-5,OU=my-ou,O=my-org,C=GB
- CN=user-6,O=my-org
# ...
15.3. Configuring user (client-side) security mechanisms
When configuring security mechanisms in clients, the clients are represented as users.
Use the KafkaUser
resource to configure the authentication, authorization, and access rights for Kafka clients.
Authentication permits user access, and authorization constrains user access to permissible actions. You can also create super users that have unconstrained access to Kafka brokers.
The authentication and authorization mechanisms must match the specification for the listener used to access the Kafka brokers.
For more information on configuring a KafkaUser
resource to access Kafka brokers securely, see Example: Setting up secure client access.
15.3.1. Associating users with Kafka clusters
A KafkaUser
resource includes a label that defines the appropriate name of the Kafka cluster (derived from the name of the Kafka
resource) to which it belongs.
apiVersion: kafka.strimzi.io/v1beta2
kind: KafkaUser
metadata:
name: my-user
labels:
strimzi.io/cluster: my-cluster
The label enables the User Operator to identify the KafkaUser
resource and create and manager the user.
If the label does not match the Kafka cluster, the User Operator cannot identify the KafkaUser
, and the user is not created.
If the status of the KafkaUser
resource remains empty, check your label configuration.
15.3.2. Configuring user authentication
Use the KafkaUser
custom resource to configure authentication credentials for users (clients) that require access to a Kafka cluster.
Configure the credentials using the authentication
property in KafkaUser.spec
.
By specifying a type
, you control what credentials are generated.
Supported authentication types:
-
tls
for mTLS authentication -
tls-external
for mTLS authentication using external certificates -
scram-sha-512
for SCRAM-SHA-512 authentication
If tls
or scram-sha-512
is specified, the User Operator creates authentication credentials when it creates the user.
If tls-external
is specified, the user still uses mTLS, but no authentication credentials are created.
Use this option when you’re providing your own certificates.
When no authentication type is specified, the User Operator does not create the user or its credentials.
You can use tls-external
to authenticate with mTLS using a certificate issued outside the User Operator.
The User Operator does not generate a TLS certificate or a secret.
You can still manage ACL rules and quotas through the User Operator in the same way as when you’re using the tls
mechanism.
This means that you use the CN=USER-NAME
format when specifying ACL rules and quotas.
USER-NAME is the common name given in a TLS certificate.
mTLS authentication
To use mTLS authentication, you set the type
field in the KafkaUser
resource to tls
.
apiVersion: kafka.strimzi.io/v1beta2
kind: KafkaUser
metadata:
name: my-user
labels:
strimzi.io/cluster: my-cluster
spec:
authentication:
type: tls
# ...
The authentication type must match the equivalent configuration for the Kafka
listener used to access the Kafka cluster.
When the user is created by the User Operator, it creates a new secret with the same name as the KafkaUser
resource.
The secret contains a private and public key for mTLS.
The public key is contained in a user certificate, which is signed by a clients CA (certificate authority) when it is created.
All keys are in X.509 format.
Note
|
If you are using the clients CA generated by the Cluster Operator, the user certificates generated by the User Operator are also renewed when the clients CA is renewed by the Cluster Operator. |
The user secret provides keys and certificates in PEM and PKCS #12 formats.
apiVersion: v1
kind: Secret
metadata:
name: my-user
labels:
strimzi.io/kind: KafkaUser
strimzi.io/cluster: my-cluster
type: Opaque
data:
ca.crt: <public_key> # Public key of the clients CA used to sign this user certificate
user.crt: <user_certificate> # Public key of the user
user.key: <user_private_key> # Private key of the user
user.p12: <store> # PKCS #12 store for user certificates and keys
user.password: <password_for_store> # Protects the PKCS #12 store
When you configure a client, you specify the following:
-
Truststore properties for the public cluster CA certificate to verify the identity of the Kafka cluster
-
Keystore properties for the user authentication credentials to verify the client
The configuration depends on the file format (PEM or PKCS #12). This example uses PKCS #12 stores, and the passwords required to access the credentials in the stores.
bootstrap.servers=<kafka_cluster_name>-kafka-bootstrap:9093 # (1)
security.protocol=SSL # (2)
ssl.truststore.location=/tmp/ca.p12 # (3)
ssl.truststore.password=<truststore_password> # (4)
ssl.keystore.location=/tmp/user.p12 # (5)
ssl.keystore.password=<keystore_password> # (6)
-
The bootstrap server address to connect to the Kafka cluster.
-
The security protocol option when using TLS for encryption.
-
The truststore location contains the public key certificate (
ca.p12
) for the Kafka cluster. A cluster CA certificate and password is generated by the Cluster Operator in the<cluster_name>-cluster-ca-cert
secret when the Kafka cluster is created. -
The password (
ca.password
) for accessing the truststore. -
The keystore location contains the public key certificate (
user.p12
) for the Kafka user. -
The password (
user.password
) for accessing the keystore.
mTLS authentication using a certificate issued outside the User Operator
To use mTLS authentication using a certificate issued outside the User Operator, you set the type
field in the KafkaUser
resource to tls-external
.
A secret and credentials are not created for the user.
apiVersion: kafka.strimzi.io/v1beta2
kind: KafkaUser
metadata:
name: my-user
labels:
strimzi.io/cluster: my-cluster
spec:
authentication:
type: tls-external
# ...
SCRAM-SHA-512 authentication
To use the SCRAM-SHA-512 authentication mechanism, you set the type
field in the KafkaUser
resource to scram-sha-512
.
apiVersion: kafka.strimzi.io/v1beta2
kind: KafkaUser
metadata:
name: my-user
labels:
strimzi.io/cluster: my-cluster
spec:
authentication:
type: scram-sha-512
# ...
When the user is created by the User Operator, it creates a new secret with the same name as the KafkaUser
resource.
The secret contains the generated password in the password
key, which is encoded with base64.
In order to use the password, it must be decoded.
apiVersion: v1
kind: Secret
metadata:
name: my-user
labels:
strimzi.io/kind: KafkaUser
strimzi.io/cluster: my-cluster
type: Opaque
data:
password: Z2VuZXJhdGVkcGFzc3dvcmQ= (1)
sasl.jaas.config: b3JnLmFwYWNoZS5rYWZrYS5jb21tb24uc2VjdXJpdHkuc2NyYW0uU2NyYW1Mb2dpbk1vZHVsZSByZXF1aXJlZCB1c2VybmFtZT0ibXktdXNlciIgcGFzc3dvcmQ9ImdlbmVyYXRlZHBhc3N3b3JkIjsK (2)
-
The generated password, base64 encoded.
-
The JAAS configuration string for SASL SCRAM-SHA-512 authentication, base64 encoded.
Decoding the generated password:
echo "Z2VuZXJhdGVkcGFzc3dvcmQ=" | base64 --decode
Custom password configuration
When a user is created, Strimzi generates a random password.
You can use your own password instead of the one generated by Strimzi. To do so, create a secret with the password and reference it in the KafkaUser
resource.
apiVersion: kafka.strimzi.io/v1beta2
kind: KafkaUser
metadata:
name: my-user
labels:
strimzi.io/cluster: my-cluster
spec:
authentication:
type: scram-sha-512
password:
valueFrom:
secretKeyRef:
name: my-secret # (1)
key: my-password # (2)
# ...
-
The name of the secret containing the predefined password.
-
The key for the password stored inside the secret.
15.3.3. Configuring user authorization
Use the KafkaUser
custom resource to configure authorization rules for users (clients) that require access to a Kafka cluster.
Configure the rules using the authorization
property in KafkaUser.spec
.
By specifying a type
, you control what rules are used.
To use simple authorization, you set the type
property to simple
in KafkaUser.spec.authorization
.
The simple authorization uses the Kafka Admin API to manage the ACL rules inside your Kafka cluster.
Whether ACL management in the User Operator is enabled or not depends on your authorization configuration in the Kafka cluster.
-
For simple authorization, ACL management is always enabled.
-
For OPA authorization, ACL management is always disabled. Authorization rules are configured in the OPA server.
-
For Keycloak authorization, you can manage the ACL rules directly in Keycloak. You can also delegate authorization to the simple authorizer as a fallback option in the configuration. When delegation to the simple authorizer is enabled, the User Operator will enable management of ACL rules as well.
-
For custom authorization using a custom authorization plugin, use the
supportsAdminApi
property in the.spec.kafka.authorization
configuration of theKafka
custom resource to enable or disable the support.
Authorization is cluster-wide.
The authorization type must match the equivalent configuration in the Kafka
custom resource.
If ACL management is not enabled, Strimzi rejects a resource if it contains any ACL rules.
If you’re using a standalone deployment of the User Operator, ACL management is enabled by default.
You can disable it using the STRIMZI_ACLS_ADMIN_API_SUPPORTED
environment variable.
If no authorization is specified, the User Operator does not provision any access rights for the user.
Whether such a KafkaUser
can still access resources depends on the authorizer being used.
For example, for simple
authorization, this is determined by the allow.everyone.if.no.acl.found
configuration in the Kafka cluster.
ACL rules
simple
authorization uses ACL rules to manage access to Kafka brokers.
ACL rules grant access rights to the user, which you specify in the acls
property.
For more information about the AclRule
object, see the AclRule
schema reference.
Super user access to Kafka brokers
If a user is added to a list of super users in a Kafka broker configuration,
the user is allowed unlimited access to the cluster regardless of any authorization constraints defined in ACLs in KafkaUser
.
For more information on configuring super user access to brokers, see Kafka authorization.
15.3.4. Configuring user quotas
Configure the spec
for the KafkaUser
resource to enforce quotas so that a user does not overload Kafka brokers.
Set size-based network usage and time-based CPU utilization thresholds.
Partition mutations occur in response to the following types of user requests:
-
Creating partitions for a new topic
-
Adding partitions to an existing topic
-
Deleting partitions from a topic
You can also add a partition mutation quota to control the rate at which requests to change partitions are accepted.
KafkaUser
with user quotasapiVersion: kafka.strimzi.io/v1beta2
kind: KafkaUser
metadata:
name: my-user
labels:
strimzi.io/cluster: my-cluster
spec:
# ...
quotas:
producerByteRate: 1048576 # (1)
consumerByteRate: 2097152 # (2)
requestPercentage: 55 # (3)
controllerMutationRate: 10 # (4)
-
Byte-per-second quota on the amount of data the user can push to a Kafka broker.
-
Byte-per-second quota on the amount of data the user can fetch from a Kafka broker.
-
CPU utilization limit as a percentage of time for a client group.
-
Number of concurrent partition creation and deletion operations (mutations) allowed per second.
Using quotas for Kafka clients might be useful in a number of situations. Consider a wrongly configured Kafka producer which is sending requests at too high a rate. Such misconfiguration can cause a denial of service to other clients, so the problematic client ought to be blocked. By using a network limiting quota, it is possible to prevent this situation from significantly impacting other clients.
Note
|
Strimzi supports user-level quotas, but not client-level quotas. |
15.4. Example: Setting up secure client access
This procedure shows how to configure client access to a Kafka cluster from outside Kubernetes or from another Kubernetes cluster. It’s split into two parts:
-
Securing Kafka brokers
-
Securing user access to Kafka
Client access to the Kafka cluster is secured with the following configuration:
-
An external listener is configured with TLS encryption and mutual TLS (mTLS) authentication in the
Kafka
resource, as well assimple
authorization. -
A
KafkaUser
is created for the client, utilizing mTLS authentication, and Access Control Lists (ACLs) are defined forsimple
authorization.
At least one listener supporting the desired authentication must be configured for the KafkaUser
.
Listeners can be configured for mutual TLS
, SCRAM-SHA-512
, or OAuth
authentication.
While mTLS always uses encryption, it’s also recommended when using SCRAM-SHA-512 and OAuth 2.0 authentication.
Authorization options for Kafka include simple
, OAuth
, OPA
, or custom
.
When enabled, authorization is applied to all enabled listeners.
To ensure compatibility between Kafka and clients, configuration of the following authentication and authorization mechanisms must align:
-
For
type: tls
andtype: scram-sha-512
authentication types,Kafka.spec.kafka.listeners[*].authentication
must matchKafkaUser.spec.authentication
-
For
type: simple
authorization,Kafka.spec.kafka.authorization
must matchKafkaUser.spec.authorization
For example, mTLS authentication for a user is only possible if it’s also enabled in the Kafka configuration.
Strimzi operators automate the configuration process and create the certificates required for authentication:
-
The Cluster Operator creates the listeners and sets up the cluster and client certificate authority (CA) certificates to enable authentication within the Kafka cluster.
-
The User Operator creates the user representing the client and the security credentials used for client authentication, based on the chosen authentication type.
You add the certificates to your client configuration.
In this procedure, the CA certificates generated by the Cluster Operator are used. Alternatively, you can replace them by installing your own custom CA certificates. You can also configure listeners to use Kafka listener certificates managed by an external CA.
Certificates are available in PEM (.crt) and PKCS #12 (.p12) formats. This procedure uses PEM certificates. Use PEM certificates with clients that support the X.509 certificate format.
Note
|
For internal clients in the same Kubernetes cluster and namespace, you can mount the cluster CA certificate in the pod specification. For more information, see Configuring internal clients to trust the cluster CA. |
-
The Kafka cluster is available for connection by a client running outside the Kubernetes cluster
-
The Cluster Operator and User Operator are running in the cluster
15.4.1. Securing Kafka brokers
-
Configure the Kafka cluster with a Kafka listener.
-
Define the authentication required to access the Kafka broker through the listener.
-
Enable authorization on the Kafka broker.
Example listener configurationapiVersion: kafka.strimzi.io/v1beta2 kind: Kafka metadata: name: my-cluster namespace: myproject spec: kafka: # ... listeners: # (1) - name: external1 # (2) port: 9094 # (3) type: <listener_type> # (4) tls: true # (5) authentication: type: tls # (6) configuration: # (7) #... authorization: # (8) type: simple superUsers: - super-user-name # (9) # ...
-
Configuration options for enabling external listeners are described in the Generic Kafka listener schema reference.
-
Name to identify the listener. Must be unique within the Kafka cluster.
-
Port number used by the listener inside Kafka. The port number has to be unique within a given Kafka cluster. Allowed port numbers are 9092 and higher with the exception of ports 9404 and 9999, which are already used for Prometheus and JMX. Depending on the listener type, the port number might not be the same as the port number that connects Kafka clients.
-
External listener type specified as
route
(OpenShift only),loadbalancer
,nodeport
oringress
(Kubernetes only). An internal listener is specified asinternal
orcluster-ip
. -
Required. TLS encryption on the listener. For
route
andingress
type listeners it must be set totrue
. For mTLS authentication, also use theauthentication
property. -
Client authentication mechanism on the listener. For server and client authentication using mTLS, you specify
tls: true
andauthentication.type: tls
. -
(Optional) Depending on the requirements of the listener type, you can specify additional listener configuration.
-
Authorization specified as
simple
, which uses theAclAuthorizer
andStandardAuthorizer
Kafka plugins. -
(Optional) Super users can access all brokers regardless of any access restrictions defined in ACLs.
WarningAn OpenShift route address comprises the Kafka cluster name, the listener name, the project name, and the domain of the router. For example, my-cluster-kafka-external1-bootstrap-my-project.domain.com
(<cluster_name>-kafka-<listener_name>-bootstrap-<namespace>.<domain>). Each DNS label (between periods “.”) must not exceed 63 characters, and the total length of the address must not exceed 255 characters.
-
-
-
Apply the changes to the
Kafka
resource configuration.The Kafka cluster is configured with a Kafka broker listener using mTLS authentication.
A service is created for each Kafka broker pod.
A service is created to serve as the bootstrap address for connection to the Kafka cluster.
A service is also created as the external bootstrap address for external connection to the Kafka cluster using
nodeport
listeners.The cluster CA certificate to verify the identity of the kafka brokers is also created in the secret
<cluster_name>-cluster-ca-cert
.NoteIf you scale your Kafka cluster while using external listeners, it might trigger a rolling update of all Kafka brokers. This depends on the configuration. -
Retrieve the bootstrap address you can use to access the Kafka cluster from the status of the
Kafka
resource.kubectl get kafka <kafka_cluster_name> -o=jsonpath='{.status.listeners[?(@.name=="<listener_name>")].bootstrapServers}{"\n"}'
For example:
kubectl get kafka my-cluster -o=jsonpath='{.status.listeners[?(@.name=="external")].bootstrapServers}{"\n"}'
Use the bootstrap address in your Kafka client to connect to the Kafka cluster.
15.4.2. Securing user access to Kafka
-
Create or modify a user representing the client that requires access to the Kafka cluster.
-
Specify the same authentication type as the
Kafka
listener. -
Specify the authorization ACLs for
simple
authorization.Example user configurationapiVersion: kafka.strimzi.io/v1beta2 kind: KafkaUser metadata: name: my-user labels: strimzi.io/cluster: my-cluster # (1) spec: authentication: type: tls # (2) authorization: type: simple acls: # (3) - resource: type: topic name: my-topic patternType: literal operations: - Describe - Read - resource: type: group name: my-group patternType: literal operations: - Read
-
The label must match the label of the Kafka cluster.
-
Authentication specified as mutual
tls
. -
Simple authorization requires an accompanying list of ACL rules to apply to the user. The rules define the operations allowed on Kafka resources based on the username (
my-user
).
-
-
-
Apply the changes to the
KafkaUser
resource configuration.The user is created, as well as a secret with the same name as the
KafkaUser
resource. The secret contains a public and private key for mTLS authentication.Example secret with user credentialsapiVersion: v1 kind: Secret metadata: name: my-user labels: strimzi.io/kind: KafkaUser strimzi.io/cluster: my-cluster type: Opaque data: ca.crt: <public_key> # Public key of the clients CA used to sign this user certificate user.crt: <user_certificate> # Public key of the user user.key: <user_private_key> # Private key of the user user.p12: <store> # PKCS #12 store for user certificates and keys user.password: <password_for_store> # Protects the PKCS #12 store
-
Extract the cluster CA certificate from the
<cluster_name>-cluster-ca-cert
secret of the Kafka cluster.kubectl get secret <cluster_name>-cluster-ca-cert -o jsonpath='{.data.ca\.crt}' | base64 -d > ca.crt
-
Extract the user CA certificate from the
<user_name>
secret.kubectl get secret <user_name> -o jsonpath='{.data.user\.crt}' | base64 -d > user.crt
-
Extract the private key of the user from the
<user_name>
secret.kubectl get secret <user_name> -o jsonpath='{.data.user\.key}' | base64 -d > user.key
-
Configure your client with the bootstrap address hostname and port for connecting to the Kafka cluster:
props.put(ConsumerConfig.BOOTSTRAP_SERVERS_CONFIG, "<hostname>:<port>");
-
Configure your client with the truststore credentials to verify the identity of the Kafka cluster.
Specify the public cluster CA certificate.
Example truststore configurationprops.put(CommonClientConfigs.SECURITY_PROTOCOL_CONFIG, "SSL"); props.put(SslConfigs.SSL_TRUSTSTORE_TYPE_CONFIG, "PEM"); props.put(SslConfigs.SSL_TRUSTSTORE_CERTIFICATES_CONFIG, "<ca.crt_file_content>");
SSL is the specified security protocol for mTLS authentication. Specify
SASL_SSL
for SCRAM-SHA-512 authentication over TLS. PEM is the file format of the truststore. -
Configure your client with the keystore credentials to verify the user when connecting to the Kafka cluster.
Specify the public certificate and private key.
Example keystore configurationprops.put(CommonClientConfigs.SECURITY_PROTOCOL_CONFIG, "SSL"); props.put(SslConfigs.SSL_KEYSTORE_TYPE_CONFIG, "PEM"); props.put(SslConfigs.SSL_KEYSTORE_CERTIFICATE_CHAIN_CONFIG, "<user.crt_file_content>"); props.put(SslConfigs.SSL_KEYSTORE_KEY_CONFIG, "<user.key_file_content>");
Add the keystore certificate and the private key directly to the configuration. Add as a single-line format. Between the
BEGIN CERTIFICATE
andEND CERTIFICATE
delimiters, start with a newline character (\n
). End each line from the original certificate with\n
too.Example keystore configurationprops.put(SslConfigs.SSL_KEYSTORE_CERTIFICATE_CHAIN_CONFIG, "-----BEGIN CERTIFICATE----- \n<user_certificate_content_line_1>\n<user_certificate_content_line_n>\n-----END CERTIFICATE---"); props.put(SslConfigs.SSL_KEYSTORE_KEY_CONFIG, "----BEGIN PRIVATE KEY-----\n<user_key_content_line_1>\n<user_key_content_line_n>\n-----END PRIVATE KEY-----");
15.5. Troubleshooting TLS hostname verification with node ports
Off-cluster access using node ports with TLS encryption enabled does not support TLS hostname verification. This is because Strimzi does not know the address of the node where the broker pod is scheduled and cannot include it in the broker certificate. Consequently, clients that perform hostname verification will fail to connect.
For example, a Java client will fail with the following exception:
Caused by: java.security.cert.CertificateException: No subject alternative names matching IP address 168.72.15.231 found
...
To connect, you must disable hostname verification.
In the Java client, set the ssl.endpoint.identification.algorithm
configuration option to an empty string.
When configuring the client using a properties file, you can do it this way:
ssl.endpoint.identification.algorithm=
When configuring the client directly in Java, set the configuration option to an empty string:
props.put("ssl.endpoint.identification.algorithm", "");
Alternatively, if you know the addresses of the worker nodes where the brokers are scheduled, you can add them as additional SANs (Subject Alternative Names) to the broker certificates manually.
For example, this might apply if your cluster is running on a bare metal deployment with a limited number of available worker nodes.
Use the alternativeNames
property to specify additional SANS.
16. Enabling OAuth 2.0 token-based access
Strimzi supports OAuth 2.0 for securing Kafka clusters by integrating with an OAUth 2.0 authorization server. Kafka brokers and clients both need to be configured to use OAuth 2.0.
OAuth 2.0 enables standardized token-based authentication and authorization between applications, using a central authorization server to issue tokens that grant limited access to resources. You can define specific scopes for fine-grained access control. Scopes correspond to different levels of access to Kafka topics or operations within the cluster.
OAuth 2.0 also supports single sign-on and integration with identity providers.
For more information on using OAUth 2.0, see the Strimzi OAuth 2.0 for Apache Kafka project.
16.1. Configuring an OAuth 2.0 authorization server
Before you can use OAuth 2.0 token-based access, you must configure an authorization server for integration with Strimzi. The steps are dependent on the chosen authorization server. Consult the product documentation for the authorization server for information on how to set up OAuth 2.0 access.
Prepare the authorization server to work with Strimzi by defining OAUth 2.0 clients for Kafka and each Kafka client component of your application. In relation to the authorization server, the Kafka cluster and Kafka clients are both regarded as OAuth 2.0 clients.
In general, configure OAuth 2.0 clients in the authorization server with the following client credentials enabled:
-
Client ID (for example,
kafka
for the Kafka cluster) -
Client ID and secret as the authentication mechanism
Note
|
You only need to use a client ID and secret when using a non-public introspection endpoint of the authorization server. The credentials are not typically required when using public authorization server endpoints, as with fast local JWT token validation. |
16.2. Using OAuth 2.0 token-based authentication
Strimzi supports the use of OAuth 2.0 for token-based authentication. An OAuth 2.0 authorization server handles the granting of access and inquiries about access. Kafka clients authenticate to Kafka brokers. Brokers and clients communicate with the authorization server, as necessary, to obtain or validate access tokens.
For a deployment of Strimzi, OAuth 2.0 integration provides the following support:
-
Server-side OAuth 2.0 authentication for Kafka brokers
-
Client-side OAuth 2.0 authentication for Kafka MirrorMaker, Kafka Connect, and the Kafka Bridge
16.2.1. Configuring OAuth 2.0 authentication on listeners
To secure Kafka brokers with OAuth 2.0 authentication, configure a listener in the Kafka
resource to use OAUth 2.0 authentication and a client authentication mechanism, and add further configuration depending on the authentication mechanism and type of token validation used in the authentication.
oauth
authenticationSpecify a listener in the Kafka
resource with an oauth
authentication type.
You can configure internal and external listeners.
We recommend using OAuth 2.0 authentication together with TLS encryption (tls: true
).
Without encryption, the connection is vulnerable to network eavesdropping and unauthorized access through token theft.
apiVersion: kafka.strimzi.io/v1beta2
kind: Kafka
spec:
kafka:
# ...
listeners:
- name: tls
port: 9093
type: internal
tls: true
authentication:
type: oauth
- name: external3
port: 9094
type: loadbalancer
tls: true
authentication:
type: oauth
#...
Use one or both of the following SASL mechanisms for clients to exchange credentials and establish authenticated sessions with Kafka.
OAUTHBEARER
-
Using the
OAUTHBEARER
authentication mechanism, credentials exchange uses a bearer token provided by an OAuth callback handler. Token provision can be configured to use the following methods:-
Client ID and secret (using the OAuth 2.0 client credentials mechanism)
-
Client ID and client assertion
-
Long-lived access token or Service account token
-
Long-lived refresh token obtained manually
OAUTHBEARER
is recommended as it provides a higher level of security thanPLAIN
, though it can only be used by Kafka clients that support theOAUTHBEARER
mechanism at the protocol level. Client credentials are never shared with Kafka. -
PLAIN
-
PLAIN
is a simple authentication mechanism used by all Kafka client tools. Consider usingPLAIN
only with Kafka clients that do not supportOAUTHBEARER
. Using thePLAIN
authentication mechanism, credentials exchange can be configured to use the following methods:-
Client ID and secret (using the OAuth 2.0 client credentials mechanism)
-
Long-lived access token
Regardless of the method used, the client must provideusername
andpassword
properties to Kafka.
Credentials are handled centrally behind a compliant authorization server, similar to how
OAUTHBEARER
authentication is used. The username extraction process depends on the authorization server configuration. -
OAUTHBEARER
is automatically enabled in the oauth
listener configuration for the Kafka broker.
To use the PLAIN
mechanism, you must set the enablePlain
property to true
.
In the following example, the PLAIN
mechanism is enabled, and the OAUTHBEARER
mechanism is disabled on a listener using the enableOauthBearer
property.
PLAIN
mechanismapiVersion: kafka.strimzi.io/v1beta2
kind: Kafka
spec:
kafka:
# ...
listeners:
- name: tls
port: 9093
type: internal
tls: true
authentication:
type: oauth
- name: external3
port: 9094
type: loadbalancer
tls: true
authentication:
type: oauth
enablePlain: true
enableOauthBearer: false
#...
When you have defined the type of authentication as OAuth 2.0, you add configuration based on the type of validation, either as fast local JWT validation or token validation using an introspection endpoint.
Fast local JWT token validation involves checking a JWT token signature locally to ensure that the token meets the following criteria:
-
Contains a
typ
(type) ortoken_type
header claim value ofBearer
to indicate it is an access token -
Is currently valid and not expired
-
Has an issuer that matches a
validIssuerURI
You specify a validIssuerURI
attribute when you configure the listener, so that any tokens not issued by the authorization server are rejected.
The authorization server does not need to be contacted during fast local JWT token validation.
You activate fast local JWT token validation by specifying a jwksEndpointUri
attribute, the endpoint exposed by the OAuth 2.0 authorization server.
The endpoint contains the public keys used to validate signed JWT tokens, which are sent as credentials by Kafka clients.
All communication with the authorization server should be performed using TLS encryption.
You can configure a certificate truststore as a Kubernetes Secret
in your Strimzi project namespace, and use the tlsTrustedCertificates
property to point to the Kubernetes secret containing the truststore file.
You might want to configure a userNameClaim
to properly extract a username from the JWT token.
If required, you can use a JsonPath expression like "['user.info'].['user.id']"
to retrieve the username from nested JSON attributes within a token.
If you want to use Kafka ACL authorization, identify the user by their username during authentication. (The sub
claim in JWT tokens is typically a unique ID, not a username.)
#...
- name: external3
port: 9094
type: loadbalancer
tls: true
authentication:
type: oauth # (1)
validIssuerUri: https://<auth_server_address>/<issuer-context> # (2)
jwksEndpointUri: https://<auth_server_address>/<path_to_jwks_endpoint> # (3)
userNameClaim: preferred_username # (4)
maxSecondsWithoutReauthentication: 3600 # (5)
tlsTrustedCertificates: # (6)
- secretName: oauth-server-cert
pattern: "*.crt"
disableTlsHostnameVerification: true # (7)
jwksExpirySeconds: 360 # (8)
jwksRefreshSeconds: 300 # (9)
jwksMinRefreshPauseSeconds: 1 # (10)
-
Listener type set to
oauth
. -
URI of the token issuer used for authentication.
-
URI of the JWKS certificate endpoint used for local JWT validation.
-
The token claim (or key) that contains the actual username used to identify the user. Its value depends on the authorization server. If necessary, a JsonPath expression like
"['user.info'].['user.id']"
can be used to retrieve the username from nested JSON attributes within a token. -
(Optional) Activates the Kafka re-authentication mechanism that enforces session expiry to the same length of time as the access token. If the specified value is less than the time left for the access token to expire, then the client will have to re-authenticate before the actual token expiry. By default, the session does not expire when the access token expires, and the client does not attempt re-authentication.
-
(Optional) Certificates stored in X.509 format within the specified secrets for TLS connection to the authorization server.
-
(Optional) Disable TLS hostname verification. Default is
false
. -
The duration the JWKS certificates are considered valid before they expire. Default is
360
seconds. If you specify a longer time, consider the risk of allowing access to revoked certificates. -
The period between refreshes of JWKS certificates. The interval must be at least 60 seconds shorter than the expiry interval. Default is
300
seconds. -
The minimum pause in seconds between consecutive attempts to refresh JWKS public keys. When an unknown signing key is encountered, the JWKS keys refresh is scheduled outside the regular periodic schedule with at least the specified pause since the last refresh attempt. The refreshing of keys follows the rule of exponential backoff, retrying on unsuccessful refreshes with ever increasing pause, until it reaches
jwksRefreshSeconds
. The default value is 1.
To configure the listener for Kubernetes service accounts, the Kubernetes API server must be used as the authorization server.
#...
- name: external3
port: 9094
type: loadbalancer
tls: true
authentication:
type: oauth
validIssuerUri: https://kubernetes.default.svc.cluster.local # (1)
jwksEndpointUri: https://kubernetes.default.svc.cluster.local/openid/v1/jwks # (2)
serverBearerTokenLocation: /var/run/secrets/kubernetes.io/serviceaccount/token # (3)
checkAccessTokenType: false # (4)
includeAcceptHeader: false # (5)
tlsTrustedCertificates: # (6)
- secretName: oauth-server-cert
pattern: "*.crt"
maxSecondsWithoutReauthentication: 3600
customClaimCheck: "@.['kubernetes.io'] && @.['kubernetes.io'].['namespace'] in ['myproject']" # (7)
-
URI of the token issuer used for authentication. Must use FQDN, including the
.cluster.local
extension, which may vary based on the Kubernetes cluster configuration. -
URI of the JWKS certificate endpoint used for local JWT validation. Must use FQDN, including the
.cluster.local
extension, which may vary based on the Kubernetes cluster configuration. -
Location to the access token used by the Kafka broker to authenticate to the Kubernetes API server in order to access the
jwksEndpointUri
. -
Skip the access token type check, as the claim for this is not present in service account tokens.
-
Skip sending
Accept
header in HTTP requests to the JWKS endpoint, as the Kubernetes API server does not support it. -
Trusted certificates to connect to authorization server. This should point to a manually created Secret that contains the Kubernetes API server public certificate, which is mounted to the running pods under
/var/run/secrets/kubernetes.io/serviceaccount/ca.crt
. You can use the following command to create the Secret:kubectl get cm kube-root-ca.crt -o jsonpath="{['data']['ca\.crt']}" > /tmp/ca.crt kubectl create secret generic oauth-server-cert --from-file=ca.crt=/tmp/ca.crt
-
(Optional) Additional constraints that JWT token has to fulfill in order to be accepted, expressed as JsonPath filter query. In this example the service account has to belong to
myproject
namespace in order to be allowed to authenticate.
The above configuration uses the sub
claim from the service account JWT token as the user ID. For example, the default service account for pods deployed in the myproject
namespace has the username: system:serviceaccount:myproject:default
.
When configuring ACLs the general form of how to refer to the ServiceAccount user should in that case be: User:system:serviceaccount:<Namespace>:<ServiceAccount-name>
Token validation using an OAuth 2.0 introspection endpoint treats a received access token as opaque. The Kafka broker sends an access token to the introspection endpoint, which responds with the token information necessary for validation. Importantly, it returns up-to-date information if the specific access token is valid, and also information about when the token expires.
To configure OAuth 2.0 introspection-based validation, you specify an introspectionEndpointUri attribute rather than the jwksEndpointUri
attribute specified for fast local JWT token validation.
Depending on the authorization server, you typically have to specify a clientId
and clientSecret
, because the introspection endpoint is usually protected.
- name: external3
port: 9094
type: loadbalancer
tls: true
authentication:
type: oauth
validIssuerUri: https://<auth_server_address>/<issuer-context>
introspectionEndpointUri: https://<auth_server_address>/<path_to_introspection_endpoint> # (1)
clientId: kafka-broker # (2)
clientSecret: # (3)
secretName: my-cluster-oauth
key: clientSecret
userNameClaim: preferred_username # (4)
maxSecondsWithoutReauthentication: 3600 # (5)
tlsTrustedCertificates:
- secretName: oauth-server-cert
pattern: "*.crt"
-
URI of the token introspection endpoint.
-
Client ID to identify the client.
-
Client Secret and client ID is used for authentication.
-
The token claim (or key) that contains the actual username used to identify the user. Its value depends on the authorization server. If necessary, a JsonPath expression like
"['user.info'].['user.id']"
can be used to retrieve the username from nested JSON attributes within a token. -
(Optional) Activates the Kafka re-authentication mechanism that enforces session expiry to the same length of time as the access token. If the specified value is less than the time left for the access token to expire, then the client will have to re-authenticate before the actual token expiry. By default, the session does not expire when the access token expires, and the client does not attempt re-authentication.
Usually, the certificates endpoint of the authorization server (jwksEndpointUri
) is publicly accessible, while the introspection endpoint (introspectionEndpointUri
) is protected.
However, this may vary depending on the authorization server configuration.
The Kafka broker can authenticate to the authorization server’s protected endpoints in one of two ways using HTTP authentication schemes:
-
HTTP Basic authentication uses a client ID and secret.
-
HTTP Bearer authentication uses a bearer token.
To configure HTTP Basic authentication, set the following properties:
-
clientId
-
clientSecret
For HTTP Bearer authentication, set the following property:
-
serverBearerTokenLocation
to specify the file path on disk containing the bearer token.
Specify additional settings depending on the authentication requirements and the authorization server you are using. Some of these properties apply only to certain authentication mechanisms or when used in combination with other properties.
For example, when using OAUth over PLAIN
, access tokens are passed as password
property values with or without an $accessToken:
prefix.
-
If you configure a token endpoint (
tokenEndpointUri
) in the listener configuration, you need the prefix. -
If you don’t configure a token endpoint in the listener configuration, you don’t need the prefix. The Kafka broker interprets the password as a raw access token.
If the password
is set as the access token, the username
must be set to the same principal name that the Kafka broker obtains from the access token.
You can specify username extraction options in your listener using the userNameClaim
, usernamePrefix
, fallbackUserNameClaim
, fallbackUsernamePrefix
, and userInfoEndpointUri
properties.
The username extraction process also depends on your authorization server; in particular, how it maps client IDs to account names.
Note
|
The PLAIN mechanism does not support password grant authentication.
Use either client credentials (client ID + secret) or an access token for authentication.
|
# ...
authentication:
type: oauth
# ...
checkIssuer: false # (1)
checkAudience: true # (2)
usernamePrefix: user- # (3)
fallbackUserNameClaim: client_id # (4)
fallbackUserNamePrefix: client-account- # (5)
serverBearerTokenLocation: path/to/access/token # (6)
validTokenType: bearer # (7)
userInfoEndpointUri: https://<auth_server_address>/<path_to_userinfo_endpoint> # (8)
enableOauthBearer: false # (9)
enablePlain: true # (10)
tokenEndpointUri: https://<auth_server_address>/<path_to_token_endpoint> # (11)
customClaimCheck: "@.custom == 'custom-value'" # (12)
clientAudience: audience # (13)
clientScope: scope # (14)
connectTimeoutSeconds: 60 # (15)
readTimeoutSeconds: 60 # (16)
httpRetries: 2 # (17)
httpRetryPauseMs: 300 # (18)
groupsClaim: "$.groups" # (19)
groupsClaimDelimiter: "," # (20)
includeAcceptHeader: false # (21)
-
If your authorization server does not provide an
iss
claim, it is not possible to perform an issuer check. In this situation, setcheckIssuer
tofalse
and do not specify avalidIssuerUri
. Default istrue
. -
If your authorization server provides an
aud
(audience) claim, and you want to enforce an audience check, setcheckAudience
totrue
. Audience checks identify the intended recipients of tokens. As a result, the Kafka broker will reject tokens that do not have itsclientId
in theiraud
claim. Default isfalse
. -
The prefix used when constructing the user ID. This only takes effect if
userNameClaim
is configured. -
An authorization server may not provide a single attribute to identify both regular users and clients. When a client authenticates in its own name, the server might provide a client ID. When a user authenticates using a username and password to obtain a refresh token or an access token, the server might provide a username attribute in addition to a client ID. Use this fallback option to specify the username claim (attribute) to use if a primary user ID attribute is not available. If necessary, a JsonPath expression like
"['client.info'].['client.id']"
can be used to retrieve the fallback username to retrieve the username from nested JSON attributes within a token. -
In situations where
fallbackUserNameClaim
is applicable, it may also be necessary to prevent name collisions between the values of the username claim, and those of the fallback username claim. Consider a situation where a client calledproducer
exists, but also a regular user calledproducer
exists. In order to differentiate between the two, you can use this property to add a prefix to the user ID of the client. -
The location of the access token used by the Kafka broker to authenticate to the Kubernetes API server for accessing protected endpoints. The authorization server must support
OAUTHBEARER
authentication. This is an alternative to specifyingclientId
andclientSecret
, which usesPLAIN
authentication. -
(Only applicable when using
introspectionEndpointUri
) Depending on the authorization server you are using, the introspection endpoint may or may not return the token type attribute, or it may contain different values. You can specify a valid token type value that the response from the introspection endpoint has to contain. -
(Only applicable when using
introspectionEndpointUri
) The authorization server may be configured or implemented in such a way to not provide any identifiable information in an introspection endpoint response. In order to obtain the user ID, you can configure the URI of theuserinfo
endpoint as a fallback. TheuserNameClaim
,fallbackUserNameClaim
, andfallbackUserNamePrefix
settings are applied to the response ofuserinfo
endpoint. -
Set this to
false
to disable theOAUTHBEARER
mechanism on the listener. At least one ofPLAIN
orOAUTHBEARER
has to be enabled. Default istrue
. -
Set to
true
to enablePLAIN
authentication on the listener, which is supported for clients on all platforms. -
Additional configuration for the
PLAIN
mechanism. If specified, clients can authenticate overPLAIN
by passing an access token as thepassword
using an$accessToken:
prefix. For production, always usehttps://
urls. -
Additional custom rules can be imposed on the JWT access token during validation by setting this to a JsonPath filter query. If the access token does not contain the necessary data, it is rejected. When using the
introspectionEndpointUri
, the custom check is applied to the introspection endpoint response JSON. -
An
audience
parameter passed to the token endpoint. An audience is used when obtaining an access token for inter-broker authentication. It is also used in the name of a client for OAuth 2.0 overPLAIN
client authentication using aclientId
andsecret
. This only affects the ability to obtain the token, and the content of the token, depending on the authorization server. It does not affect token validation rules by the listener. -
A
scope
parameter passed to the token endpoint. A scope is used when obtaining an access token for inter-broker authentication. It is also used in the name of a client for OAuth 2.0 overPLAIN
client authentication using aclientId
andsecret
. This only affects the ability to obtain the token, and the content of the token, depending on the authorization server. It does not affect token validation rules by the listener. -
The connect timeout in seconds when connecting to the authorization server. The default value is 60.
-
The read timeout in seconds when connecting to the authorization server. The default value is 60.
-
The maximum number of times to retry a failed HTTP request to the authorization server. The default value is
0
, meaning that no retries are performed. To use this option effectively, consider reducing the timeout times for theconnectTimeoutSeconds
andreadTimeoutSeconds
options. However, note that retries may prevent the current worker thread from being available to other requests, and if too many requests stall, it could make the Kafka broker unresponsive. -
The time to wait before attempting another retry of a failed HTTP request to the authorization server. By default, this time is set to zero, meaning that no pause is applied. This is because many issues that cause failed requests are per-request network glitches or proxy issues that can be resolved quickly. However, if your authorization server is under stress or experiencing high traffic, you may want to set this option to a value of 100 ms or more to reduce the load on the server and increase the likelihood of successful retries.
-
A JsonPath query that is used to extract groups information from either the JWT token or the introspection endpoint response. This option is not set by default. By configuring this option, a custom authorizer can make authorization decisions based on user groups.
-
A delimiter used to parse groups information when it is returned as a single delimited string. The default value is ',' (comma).
-
Some authorization servers have issues with client sending
Accept: application/json
header. By settingincludeAcceptHeader: false
the header will not be sent. Default istrue
.
16.2.2. Configuring OAuth 2.0 on client applications
To configure OAuth 2.0 on client applications, you must specify the following:
-
SASL (Simple Authentication and Security Layer) security protocols
-
SASL mechanisms
-
A JAAS (Java Authentication and Authorization Service) module
-
Authentication properties to access the authorization server
Specify SASL protocols in the client configuration:
-
SASL_SSL
for authentication over TLS encrypted connections -
SASL_PLAINTEXT
for authentication over unencrypted connections
Use SASL_SSL
for production and SASL_PLAINTEXT
for local development only.
When using SASL_SSL
, additional ssl.truststore
configuration is needed.
The truststore configuration is required for secure connection (https://
) to the OAuth 2.0 authorization server.
To verify the OAuth 2.0 authorization server, add the CA certificate for the authorization server to the truststore in your client configuration.
You can configure a truststore in PEM or PKCS #12 format.
Specify SASL mechanisms in the client configuration:
-
OAUTHBEARER
for credentials exchange using a bearer token -
PLAIN
to pass client credentials (clientId + secret) or an access token
Specify a JAAS module that implements the SASL authentication mechanism as a sasl.jaas.config
property value:
-
org.apache.kafka.common.security.oauthbearer.OAuthBearerLoginModule
implements theOAUTHBEARER
mechanism -
org.apache.kafka.common.security.plain.PlainLoginModule
implements thePLAIN
mechanism
Note
|
For the OAUTHBEARER mechanism, Strimzi provides a callback handler for clients that use Kafka Client Java libraries to enable credentials exchange.
For clients in other languages, custom code may be required to obtain the access token.
For the PLAIN mechanism, Strimzi provides server-side callbacks to enable credentials exchange.
|
To be able to use the OAUTHBEARER
mechanism, you must also add the custom io.strimzi.kafka.oauth.client.JaasClientOauthLoginCallbackHandler
class as the callback handler.
JaasClientOauthLoginCallbackHandler
handles OAuth callbacks to the authorization server for access tokens during client login.
This enables automatic token renewal, ensuring continuous authentication without user intervention.
Additionally, it handles login credentials for clients using the OAuth 2.0 password grant method.
Configure the client to use credentials or access tokens for OAuth 2.0 authentication.
- Using client credentials
-
Using client credentials involves configuring the client with the necessary credentials (client ID and secret, or client ID and client assertion) to obtain a valid access token from an authorization server. This is the simplest mechanism.
- Using access tokens
-
Using access tokens, the client is configured with a valid long-lived access token or refresh token obtained from an authorization server. Using access tokens adds more complexity because there is an additional dependency on authorization server tools. If you are using long-lived access tokens, you may need to configure the client in the authorization server to increase the maximum lifetime of the token.
The only information ever sent to Kafka is the access token. The credentials used to obtain the token are never sent to Kafka. When a client obtains an access token, no further communication with the authorization server is needed.
SASL authentication properties support the following authentication methods:
-
OAuth 2.0 client credentials
-
Access token or Service account token
-
Refresh token
-
OAuth 2.0 password grant (deprecated)
Add the authentication properties as JAAS configuration (sasl.jaas.config
and sasl.login.callback.handler.class
).
If the client application is not configured with an access token directly, the client exchanges one of the following sets of credentials for an access token during Kafka session initiation:
-
Client ID and secret
-
Client ID and client assertion
-
Client ID, refresh token, and (optionally) a secret
-
Username and password, with client ID and (optionally) a secret
Note
|
You can also specify authentication properties as environment variables, or as Java system properties.
For Java system properties, you can set them using setProperty and pass them on the command line using the -D option.
|
security.protocol=SASL_SSL # (1)
sasl.mechanism=OAUTHBEARER # (2)
ssl.truststore.location=/tmp/truststore.p12 (3)
ssl.truststore.password=$STOREPASS
ssl.truststore.type=PKCS12
sasl.jaas.config=org.apache.kafka.common.security.oauthbearer.OAuthBearerLoginModule required \
oauth.token.endpoint.uri="<token_endpoint_url>" \ # (4)
oauth.client.id="<client_id>" \ # (5)
oauth.client.secret="<client_secret>" \ # (6)
oauth.ssl.truststore.location="/tmp/oauth-truststore.p12" \ # (7)
oauth.ssl.truststore.password="$STOREPASS" \ # (8)
oauth.ssl.truststore.type="PKCS12" \ # (9)
oauth.scope="<scope>" \ # (10)
oauth.audience="<audience>" ; # (11)
sasl.login.callback.handler.class=io.strimzi.kafka.oauth.client.JaasClientOauthLoginCallbackHandler
-
SASL_SSL
security protocol for TLS-encrypted connections. UseSASL_PLAINTEXT
over unencrypted connections for local development only. -
The SASL mechanism specified as
OAUTHBEARER
orPLAIN
. -
The truststore configuration for secure access to the Kafka cluster.
-
URI of the authorization server token endpoint.
-
Client ID, which is the name used when creating the client in the authorization server.
-
Client secret created when creating the client in the authorization server.
-
The location contains the public key certificate (
truststore.p12
) for the authorization server. -
The password for accessing the truststore.
-
The truststore type.
-
(Optional) The
scope
for requesting the token from the token endpoint. An authorization server may require a client to specify the scope. -
(Optional) The
audience
for requesting the token from the token endpoint. An authorization server may require a client to specify the audience.
security.protocol=SASL_SSL
sasl.mechanism=OAUTHBEARER
ssl.truststore.location=/tmp/truststore.p12
ssl.truststore.password=$STOREPASS
ssl.truststore.type=PKCS12
sasl.jaas.config=org.apache.kafka.common.security.oauthbearer.OAuthBearerLoginModule required \
oauth.token.endpoint.uri="<token_endpoint_url>" \
oauth.client.id="<client_id>" \
oauth.client.assertion.location="<path_to_client_assertion_token_file>" \ # (1)
oauth.client.assertion.type="urn:ietf:params:oauth:client-assertion-type:jwt-bearer" \ # (2)
oauth.ssl.truststore.location="/tmp/oauth-truststore.p12" \
oauth.ssl.truststore.password="$STOREPASS" \
oauth.ssl.truststore.type="PKCS12" \
oauth.scope="<scope>" \
oauth.audience="<audience>" ;
sasl.login.callback.handler.class=io.strimzi.kafka.oauth.client.JaasClientOauthLoginCallbackHandler
-
Path to the client assertion file used for authenticating the client. This file is a private key file as an alternative to the client secret. Alternatively, use the
oauth.client.assertion
option to specify the client assertion value in clear text. -
(Optional) Sometimes you may need to specify the client assertion type. In not specified, the default value is
urn:ietf:params:oauth:client-assertion-type:jwt-bearer
.
security.protocol=SASL_SSL
sasl.mechanism=OAUTHBEARER
ssl.truststore.location=/tmp/truststore.p12
ssl.truststore.password=$STOREPASS
ssl.truststore.type=PKCS12
sasl.jaas.config=org.apache.kafka.common.security.oauthbearer.OAuthBearerLoginModule required \
oauth.token.endpoint.uri="<token_endpoint_url>" \
oauth.client.id="<client_id>" \ # (1)
oauth.client.secret="<client_secret>" \ # (2)
oauth.password.grant.username="<username>" \ # (3)
oauth.password.grant.password="<password>" \ # (4)
oauth.ssl.truststore.location="/tmp/oauth-truststore.p12" \
oauth.ssl.truststore.password="$STOREPASS" \
oauth.ssl.truststore.type="PKCS12" \
oauth.scope="<scope>" \
oauth.audience="<audience>" ;
sasl.login.callback.handler.class=io.strimzi.kafka.oauth.client.JaasClientOauthLoginCallbackHandler
-
Client ID, which is the name used when creating the client in the authorization server.
-
(Optional) Client secret created when creating the client in the authorization server.
-
Username for password grant authentication. OAuth password grant configuration (username and password) uses the OAuth 2.0 password grant method. To use password grants, create a user account for a client on your authorization server with limited permissions. The account should act like a service account. Use in environments where user accounts are required for authentication, but consider using a refresh token first.
-
Password for password grant authentication.
NoteSASL PLAIN
does not support passing a username and password (password grants) using the OAuth 2.0 password grant method.
security.protocol=SASL_SSL
sasl.mechanism=OAUTHBEARER
ssl.truststore.location=/tmp/truststore.p12
ssl.truststore.password=$STOREPASS
ssl.truststore.type=PKCS12
sasl.jaas.config=org.apache.kafka.common.security.oauthbearer.OAuthBearerLoginModule required \
oauth.access.token="<access_token>" ; # (1)
sasl.login.callback.handler.class=io.strimzi.kafka.oauth.client.JaasClientOauthLoginCallbackHandler
-
Long-lived access token for Kafka clients. Alternatively,
oauth.access.token.location
can be used to specify the file that contains the access token.
security.protocol=SASL_SSL
sasl.mechanism=OAUTHBEARER
ssl.truststore.location=/tmp/truststore.p12
ssl.truststore.password=$STOREPASS
ssl.truststore.type=PKCS12
sasl.jaas.config=org.apache.kafka.common.security.oauthbearer.OAuthBearerLoginModule required \
oauth.access.token.location="/var/run/secrets/kubernetes.io/serviceaccount/token"; # (1)
sasl.login.callback.handler.class=io.strimzi.kafka.oauth.client.JaasClientOauthLoginCallbackHandler
-
Location to the service account token on the filesystem (assuming that the client is deployed as a Kubernetes pod)
security.protocol=SASL_SSL
sasl.mechanism=OAUTHBEARER
ssl.truststore.location=/tmp/truststore.p12
ssl.truststore.password=$STOREPASS
ssl.truststore.type=PKCS12
sasl.jaas.config=org.apache.kafka.common.security.oauthbearer.OAuthBearerLoginModule required \
oauth.token.endpoint.uri="<token_endpoint_url>" \
oauth.client.id="<client_id>" \ # (1)
oauth.client.secret="<client_secret>" \ # (2)
oauth.refresh.token="<refresh_token>" \ # (3)
oauth.ssl.truststore.location="/tmp/oauth-truststore.p12" \
oauth.ssl.truststore.password="$STOREPASS" \
oauth.ssl.truststore.type="PKCS12" ;
sasl.login.callback.handler.class=io.strimzi.kafka.oauth.client.JaasClientOauthLoginCallbackHandler
-
Client ID, which is the name used when creating the client in the authorization server.
-
(Optional) Client secret created when creating the client in the authorization server.
-
Long-lived refresh token for Kafka clients.
OAUTHBEARER
implementationsIf your Kafka broker uses a custom OAUTHBEARER
implementation, you may need to pass additional SASL extension options.
These extensions can include attributes or information required as client context by the authorization server.
The options are passed as key-value pairs and are sent to the Kafka broker when a new session is started.
Pass SASL extension values using oauth.sasl.extension.
as a key prefix.
oauth.sasl.extension.key1="value1"
oauth.sasl.extension.key2="value2"
16.2.3. OAuth 2.0 client authentication flows
OAuth 2.0 authentication flows depend on the underlying Kafka client and Kafka broker configuration. The flows must also be supported by the authorization server used.
The Kafka broker listener configuration determines how clients authenticate using an access token. The client can pass a client ID and secret to request an access token.
If a listener is configured to use PLAIN
authentication, the client can authenticate with a client ID and secret or username and access token.
These values are passed as the username
and password
properties of the PLAIN
mechanism.
Listener configuration supports the following token validation options:
-
You can use fast local token validation based on JWT signature checking and local token introspection, without contacting an authorization server. The authorization server provides a JWKS endpoint with public certificates that are used to validate signatures on the tokens.
-
You can use a call to a token introspection endpoint provided by an authorization server. Each time a new Kafka broker connection is established, the broker passes the access token received from the client to the authorization server. The Kafka broker checks the response to confirm whether the token is valid.
Note
|
An authorization server might only allow the use of opaque access tokens, which means that local token validation is not possible. |
Kafka client credentials can also be configured for the following types of authentication:
-
Direct local access using a previously generated long-lived access token
-
Contact with the authorization server for a new access token to be issued (using a client ID and credentials, or a refresh token, or a username and a password)
Example client authentication flows using the SASL OAUTHBEARER
mechanism
You can use the following communication flows for Kafka authentication using the SASL OAUTHBEARER
mechanism.
-
The Kafka client requests an access token from the authorization server using a client ID and credentials, and optionally a refresh token. Alternatively, the client may authenticate using a username and a password.
-
The authorization server generates a new access token.
-
The Kafka client authenticates with the Kafka broker using the SASL
OAUTHBEARER
mechanism to pass the access token. -
The Kafka broker validates the access token by calling a token introspection endpoint on the authorization server using its own client ID and secret.
-
A Kafka client session is established if the token is valid.
-
The Kafka client authenticates with the authorization server from the token endpoint, using a client ID and credentials, and optionally a refresh token. Alternatively, the client may authenticate using a username and a password.
-
The authorization server generates a new access token.
-
The Kafka client authenticates with the Kafka broker using the SASL
OAUTHBEARER
mechanism to pass the access token. -
The Kafka broker validates the access token locally using a JWT token signature check, and local token introspection.
-
The Kafka client authenticates with the Kafka broker using the SASL
OAUTHBEARER
mechanism to pass the long-lived access token. -
The Kafka broker validates the access token by calling a token introspection endpoint on the authorization server, using its own client ID and secret.
-
A Kafka client session is established if the token is valid.
-
The Kafka client authenticates with the Kafka broker using the SASL
OAUTHBEARER
mechanism to pass the long-lived access token. -
The Kafka broker validates the access token locally using a JWT token signature check and local token introspection.
Warning
|
Fast local JWT token signature validation is suitable only for short-lived tokens as there is no check with the authorization server if a token has been revoked. Token expiration is written into the token, but revocation can happen at any time, so cannot be accounted for without contacting the authorization server. Any issued token would be considered valid until it expires. |
Example client authentication flows using the SASL PLAIN
mechanism
You can use the following communication flows for Kafka authentication using the OAuth PLAIN
mechanism.
-
The Kafka client passes a
clientId
as a username and asecret
as a password. -
The Kafka broker uses a token endpoint to pass the
clientId
andsecret
to the authorization server. -
The authorization server returns a fresh access token or an error if the client credentials are not valid.
-
The Kafka broker validates the token in one of the following ways:
-
If a token introspection endpoint is specified, the Kafka broker validates the access token by calling the endpoint on the authorization server. A session is established if the token validation is successful.
-
If local token introspection is used, a request is not made to the authorization server. The Kafka broker validates the access token locally using a JWT token signature check.
-
-
The Kafka client passes a username and password. The password provides the value of an access token that was obtained manually and configured before running the client.
-
The password is passed with or without an
$accessToken:
string prefix depending on whether or not the Kafka broker listener is configured with a token endpoint for authentication.-
If the token endpoint is configured, the password should be prefixed by
$accessToken:
to let the broker know that the password parameter contains an access token rather than a client secret. The Kafka broker interprets the username as the account username. -
If the token endpoint is not configured on the Kafka broker listener (enforcing a
no-client-credentials mode
), the password should provide the access token without the prefix. The Kafka broker interprets the username as the account username. In this mode, the client doesn’t use a client ID and secret, and thepassword
parameter is always interpreted as a raw access token.
-
-
The Kafka broker validates the token in one of the following ways:
-
If a token introspection endpoint is specified, the Kafka broker validates the access token by calling the endpoint on the authorization server. A session is established if token validation is successful.
-
If local token introspection is used, there is no request made to the authorization server. Kafka broker validates the access token locally using a JWT token signature check.
-
16.2.4. Re-authenticating sessions
Configure oauth
listeners to use Kafka session re-authentication for OAuth 2.0 sessions between Kafka clients and Kafka.
This mechanism enforces the expiry of an authenticated session between the client and the broker after a defined period of time.
When a session expires, the client immediately starts a new session by reusing the existing connection rather than dropping it.
Session re-authentication is disabled by default.
To enable it, you set a time value for maxSecondsWithoutReauthentication
in the oauth
listener configuration.
The same property is used to configure session re-authentication for OAUTHBEARER
and PLAIN
authentication.
For an example configuration, see Configuring OAuth 2.0 authentication on listeners.
Session re-authentication must be supported by the Kafka client libraries used by the client.
Session re-authentication can be used with fast local JWT or introspection endpoint token validation.
When the broker’s authenticated session expires, the client must re-authenticate to the existing session by sending a new, valid access token to the broker, without dropping the connection.
If token validation is successful, a new client session is started using the existing connection. If the client fails to re-authenticate, the broker will close the connection if further attempts are made to send or receive messages. Java clients that use Kafka client library 2.2 or later automatically re-authenticate if the re-authentication mechanism is enabled on the broker.
Session re-authentication also applies to refresh tokens, if used. When the session expires, the client refreshes the access token by using its refresh token. The client then uses the new access token to re-authenticate to the existing session.
When session re-authentication is configured, session expiry works differently for OAUTHBEARER
and PLAIN
authentication.
For OAUTHBEARER
and PLAIN
, using the client ID and secret method:
-
The broker’s authenticated session will expire at the configured
maxSecondsWithoutReauthentication
. -
The session will expire earlier if the access token expires before the configured time.
For PLAIN
using the long-lived access token method:
-
The broker’s authenticated session will expire at the configured
maxSecondsWithoutReauthentication
. -
Re-authentication will fail if the access token expires before the configured time. Although session re-authentication is attempted,
PLAIN
has no mechanism for refreshing tokens.
If maxSecondsWithoutReauthentication
is not configured, OAUTHBEARER
and PLAIN
clients can remain connected to brokers indefinitely, without needing to re-authenticate.
Authenticated sessions do not end with access token expiry.
However, this can be considered when configuring authorization, for example, by using keycloak
authorization or installing a custom authorizer.
16.2.5. Example: Enabling OAuth 2.0 authentication
This example shows how to configure client access to a Kafka cluster using OAUth 2.0 authentication. The procedures describe the configuration required to set up OAuth 2.0 authentication on Kafka listeners, Kafka Java clients, and Kafka components.
Setting up OAuth 2.0 authentication on listeners
Configure Kafka listeners so that they are enabled to use OAuth 2.0 authentication using an authorization server.
We advise using OAuth 2.0 over an encrypted interface through through a listener with tls: true
.
Plain listeners are not recommended.
If the authorization server is using certificates signed by the trusted CA and matching the OAuth 2.0 server hostname, TLS connection works using the default settings. Otherwise, you may need to configure the truststore with proper certificates or disable the certificate hostname validation.
For more information on the configuration of OAuth 2.0 authentication for Kafka broker listeners, see the KafkaListenerAuthenticationOAuth
schema reference.
-
Strimzi and Kafka are running
-
An OAuth 2.0 authorization server is deployed
-
Specify a listener in the
Kafka
resource with anoauth
authentication type.Example listener configuration with OAuth 2.0 authenticationapiVersion: kafka.strimzi.io/v1beta2 kind: Kafka spec: kafka: # ... listeners: - name: tls port: 9093 type: internal tls: true authentication: type: oauth - name: external3 port: 9094 type: loadbalancer tls: true authentication: type: oauth #...
-
Configure the OAuth listener depending on the authorization server and validation type:
-
Apply the changes to the
Kafka
configuration. -
Check the update in the logs or by watching the pod state transitions:
kubectl logs -f ${POD_NAME} -c ${CONTAINER_NAME} kubectl get pod -w
The rolling update configures the brokers to use OAuth 2.0 authentication.
Setting up OAuth 2.0 on Kafka Java clients
Configure Kafka producer and consumer APIs to use OAuth 2.0 for interaction with Kafka brokers.
Add a callback plugin to your client pom.xml
file, then configure your client for OAuth 2.0.
How you configure the authentication properties depends on the authentication method you are using to access the OAuth 2.0 authorization server. In this procedure, the properties are specified in a properties file, then loaded into the client configuration.
-
Strimzi and Kafka are running
-
An OAuth 2.0 authorization server is deployed and configured for OAuth access to Kafka brokers
-
Kafka brokers are configured for OAuth 2.0
-
Add the client library with OAuth 2.0 support to the
pom.xml
file for the Kafka client:<dependency> <groupId>io.strimzi</groupId> <artifactId>kafka-oauth-client</artifactId> <version>0.15.0</version> </dependency>
-
Configure the client depending on the OAuth 2.0 authentication method:
For example, specify the properties for the authentication method in a
client.properties
file. -
Input the client properties for OAUTH 2.0 authentication into the Java client code.
Example showing input of client propertiesProperties props = new Properties(); try (FileReader reader = new FileReader("client.properties", StandardCharsets.UTF_8)) { props.load(reader); }
-
Verify that the Kafka client can access the Kafka brokers.
Setting up OAuth 2.0 on Kafka components
This procedure describes how to set up Kafka components to use OAuth 2.0 authentication using an authorization server.
You can configure OAuth 2.0 authentication for the following components:
-
Kafka Connect
-
Kafka MirrorMaker
-
Kafka Bridge
In this scenario, the Kafka component and the authorization server are running in the same cluster.
For more information on the configuration of OAuth 2.0 authentication for Kafka components, see the KafkaClientAuthenticationOAuth
schema reference.
The schema reference includes examples of configuration options.
-
Strimzi and Kafka are running
-
An OAuth 2.0 authorization server is deployed and configured for OAuth access to Kafka brokers
-
Kafka brokers are configured for OAuth 2.0
-
Create a client secret and mount it to the component as an environment variable.
For example, here we are creating a client
Secret
for the Kafka Bridge:apiVersion: kafka.strimzi.io/v1beta2 kind: Secret metadata: name: my-bridge-oauth type: Opaque data: clientSecret: MGQ1OTRmMzYtZTllZS00MDY2LWI5OGEtMTM5MzM2NjdlZjQw # (1)
-
The
clientSecret
key must be in base64 format.
-
-
Create or edit the resource for the Kafka component so that OAuth 2.0 authentication is configured for the authentication property.
For OAuth 2.0 authentication, you can use the following options:
-
Client ID and secret
-
Client ID and client assertion
-
Client ID and refresh token
-
Access token
-
Username and password
-
TLS
For example, here OAuth 2.0 is assigned to the Kafka Bridge client using a client ID and secret, and TLS:
Example OAuth 2.0 authentication configuration using the client secretapiVersion: kafka.strimzi.io/v1beta2 kind: KafkaBridge metadata: name: my-bridge spec: # ... authentication: type: oauth # (1) tokenEndpointUri: https://<auth_server_address>/<path_to_token_endpoint> # (2) clientId: kafka-bridge clientSecret: secretName: my-bridge-oauth key: clientSecret tlsTrustedCertificates: # (3) - secretName: oauth-server-cert pattern: "*.crt"
-
Authentication type set to
oauth
. -
URI of the token endpoint for authentication.
-
Certificates stored in X.509 format within the specified secrets for TLS connection to the authorization server.
In this example, OAuth 2.0 is assigned to the Kafka Bridge client using a client ID and the location of a client assertion file, with TLS to connect to the authorization server:
Example OAuth 2.0 authentication configuration using client assertionapiVersion: kafka.strimzi.io/v1beta2 kind: KafkaBridge metadata: name: my-bridge spec: # ... authentication: type: oauth tokenEndpointUri: https://<auth_server_address>/<path_to_token_endpoint> clientId: kafka-bridge clientAssertionLocation: /var/run/secrets/sso/assertion # (1) tlsTrustedCertificates: - secretName: oauth-server-cert pattern: "*.crt"
-
Filesystem path to the client assertion file used for authenticating the client. This file is typically added to the deployed pod by an external operator service. Alternatively, use
clientAssertion
to refer to a secret containing the client assertion value.
Here, OAuth 2.0 is assigned to the Kafka Bridge client using a service account token:
Example OAuth 2.0 authentication configuration using the service account tokenapiVersion: kafka.strimzi.io/v1beta2 kind: KafkaBridge metadata: name: my-bridge spec: # ... authentication: type: oauth accessTokenLocation: /var/run/secrets/kubernetes.io/serviceaccount/token # (1)
-
Path to the service account token file location.
Depending on how you apply OAuth 2.0 authentication, and the type of authorization server, there are additional configuration options you can use:
Additional configuration options# ... spec: # ... authentication: # ... disableTlsHostnameVerification: true # (1) accessTokenIsJwt: false # (2) scope: any # (3) audience: kafka # (4) connectTimeoutSeconds: 60 # (5) readTimeoutSeconds: 60 # (6) httpRetries: 2 # (7) httpRetryPauseMs: 300 # (8) includeAcceptHeader: false # (9)
-
(Optional) Disable TLS hostname verification. Default is
false
. -
If you are using opaque tokens, you can apply
accessTokenIsJwt: false
so that access tokens are not treated as JWT tokens. -
(Optional) The
scope
for requesting the token from the token endpoint. An authorization server may require a client to specify the scope. In this case it isany
. -
(Optional) The
audience
for requesting the token from the token endpoint. An authorization server may require a client to specify the audience. In this case it iskafka
. -
(Optional) The connect timeout in seconds when connecting to the authorization server. The default value is 60.
-
(Optional) The read timeout in seconds when connecting to the authorization server. The default value is 60.
-
(Optional) The maximum number of times to retry a failed HTTP request to the authorization server. The default value is
0
, meaning that no retries are performed. To use this option effectively, consider reducing the timeout times for theconnectTimeoutSeconds
andreadTimeoutSeconds
options. However, note that retries may prevent the current worker thread from being available to other requests, and if too many requests stall, it could make the Kafka broker unresponsive. -
(Optional) The time to wait before attempting another retry of a failed HTTP request to the authorization server. By default, this time is set to zero, meaning that no pause is applied. This is because many issues that cause failed requests are per-request network glitches or proxy issues that can be resolved quickly. However, if your authorization server is under stress or experiencing high traffic, you may want to set this option to a value of 100 ms or more to reduce the load on the server and increase the likelihood of successful retries.
-
(Optional) Some authorization servers have issues with client sending
Accept: application/json
header. By settingincludeAcceptHeader: false
the header will not be sent. Default istrue
.
-
-
Apply the changes to the resource configuration of the component.
-
Check the update in the logs or by watching the pod state transitions:
kubectl logs -f ${POD_NAME} -c ${CONTAINER_NAME} kubectl get pod -w
The rolling updates configure the component for interaction with Kafka brokers using OAuth 2.0 authentication.
16.3. Using OAuth 2.0 token-based authorization
Strimzi supports the use of OAuth 2.0 token-based authorization through Keycloak Authorization Services, which lets you manage security policies and permissions centrally.
Security policies and permissions defined in Keycloak grant access to Kafka resources. Users and clients are matched against policies that permit access to perform specific actions on Kafka brokers.
Kafka allows all users full access to brokers by default, but also provides the AclAuthorizer
and StandardAuthorizer
plugins to configure authorization based on Access Control Lists (ACLs).
The ACL rules managed by these plugins are used to grant or deny access to resources based on username, and these rules are stored within the Kafka cluster itself.
However, OAuth 2.0 token-based authorization with Keycloak offers far greater flexibility on how you wish to implement access control to Kafka brokers. In addition, you can configure your Kafka brokers to use OAuth 2.0 authorization and ACLs.
16.3.1. Example: Enabling OAuth 2.0 authorization
This example procedure shows how to configure Kafka to use OAuth 2.0 authorization using Keycloak Authorization Services.
To enable OAuth 2.0 authorization using Keycloak, configure the Kafka
resource to use keycloak
authorization and specify the properties required to access the authorization server and Keycloak Authorization Services.
Keycloak server Authorization Services REST endpoints extend token-based authentication with Keycloak by applying defined security policies on a particular user, and providing a list of permissions granted on different resources for that user. Policies use roles and groups to match permissions to users. OAuth 2.0 authorization enforces permissions locally based on the received list of grants for the user from Keycloak Authorization Services.
A Keycloak authorizer (KeycloakAuthorizer
) is provided with Strimzi.
The authorizer fetches a list of granted permissions from the authorization server as needed,
and enforces authorization locally on Kafka, making rapid authorization decisions for each client request.
Consider the access you require or want to limit for certain users. You can use a combination of Keycloak groups, roles, clients, and users to configure access in Keycloak.
Typically, groups are used to match users based on organizational departments or geographical locations. And roles are used to match users based on their function.
With Keycloak, you can store users and groups in LDAP, whereas clients and roles cannot be stored this way. Storage and access to user data may be a factor in how you choose to configure authorization policies.
Note
|
Super users always have unconstrained access to Kafka regardless of the authorization implemented. |
-
Strimzi must be configured to use OAuth 2.0 with Keycloak for token-based authentication. You use the same Keycloak server endpoint when you set up authorization.
-
OAuth 2.0 authentication must be configured with the
maxSecondsWithoutReauthentication
option to enable re-authentication.
-
Access the Keycloak Admin Console or use the Keycloak Admin CLI to enable Authorization Services for the OAuth 2.0 client for Kafka you created when setting up OAuth 2.0 authentication.
-
Use Authorization Services to define resources, authorization scopes, policies, and permissions for the client.
-
Bind the permissions to users and clients by assigning them roles and groups.
-
Configure the
kafka
resource to usekeycloak
authorization, and to be able to access the authorization server and Authorization Services.Example OAuth 2.0 authorization configurationapiVersion: kafka.strimzi.io/v1beta2 kind: Kafka metadata: name: my-cluster spec: kafka: # ... authorization: type: keycloak (1) tokenEndpointUri: <https://<auth-server-address>/realms/external/protocol/openid-connect/token> (2) clientId: kafka (3) delegateToKafkaAcls: false (4) disableTlsHostnameVerification: false (5) superUsers: (6) - CN=user-1 - user-2 - CN=user-3 tlsTrustedCertificates: (7) - secretName: oauth-server-cert pattern: "*.crt" grantsRefreshPeriodSeconds: 60 (8) grantsRefreshPoolSize: 5 (9) grantsMaxIdleSeconds: 300 (10) grantsGcPeriodSeconds: 300 (11) grantsAlwaysLatest: false (12) connectTimeoutSeconds: 60 (13) readTimeoutSeconds: 60 (14) httpRetries: 2 (15) enableMetrics: false (16) includeAcceptHeader: false (17) #...
-
Type
keycloak
enables Keycloak authorization. -
URI of the Keycloak token endpoint. For production, always use
https://
urls. When you configure token-basedoauth
authentication, you specify ajwksEndpointUri
as the URI for local JWT validation. The hostname for thetokenEndpointUri
URI must be the same. -
The client ID of the OAuth 2.0 client definition in Keycloak that has Authorization Services enabled. Typically,
kafka
is used as the ID. -
(Optional) Delegate authorization to Kafka
AclAuthorizer
andStandardAuthorizer
if access is denied by Keycloak Authorization Services policies. Default isfalse
. -
(Optional) Disable TLS hostname verification. Default is
false
. -
(Optional) Designated super users.
-
(Optional) Certificates stored in X.509 format within the specified secrets for TLS connection to the authorization server.
-
(Optional) The time between two consecutive grants refresh runs. That is the maximum time for active sessions to detect any permissions changes for the user on Keycloak. The default value is 60.
-
(Optional) The number of threads to use to refresh (in parallel) the grants for the active sessions. The default value is 5.
-
(Optional) The time, in seconds, after which an idle grant in the cache can be evicted. The default value is 300.
-
(Optional) The time, in seconds, between consecutive runs of a job that cleans stale grants from the cache. The default value is 300.
-
(Optional) Controls whether the latest grants are fetched for a new session. When enabled, grants are retrieved from Keycloak and cached for the user. The default value is
false
. -
(Optional) The connect timeout in seconds when connecting to the Keycloak token endpoint. The default value is 60.
-
(Optional) The read timeout in seconds when connecting to the Keycloak token endpoint. The default value is 60.
-
(Optional) The maximum number of times to retry (without pausing) a failed HTTP request to the authorization server. The default value is
0
, meaning that no retries are performed. To use this option effectively, consider reducing the timeout times for theconnectTimeoutSeconds
andreadTimeoutSeconds
options. However, note that retries may prevent the current worker thread from being available to other requests, and if too many requests stall, it could make Kafka unresponsive. -
(Optional) Enable or disable OAuth metrics. The default value is
false
. -
(Optional) Some authorization servers have issues with client sending
Accept: application/json
header. By settingincludeAcceptHeader: false
the header will not be sent. Default istrue
.
-
-
Apply the changes to the
Kafka
configuration. -
Check the update in the logs or by watching the pod state transitions:
kubectl logs -f ${POD_NAME} -c kafka kubectl get pod -w
The rolling update configures the brokers to use OAuth 2.0 authorization.
-
Verify the configured permissions by accessing Kafka brokers as clients or users with specific roles, ensuring they have the necessary access and do not have unauthorized access.
16.4. Setting up permissions in Keycloak
When using Keycloak as the OAuth 2.0 authorization server, Kafka permissions are granted to user accounts or service accounts using authorization permissions. To grant permissions to access Kafka, create an OAuth client specification in Keycloak that maps the authorization models of Keycloak Authorization Services and Kafka.
16.4.1. Kafka and Keycloak authorization models
Kafka and Keycloak use different authorization models.
Kafka’s authorization model uses resource types and operations to describe ACLs for a user.
When a Kafka client performs an action on a broker, the broker uses the configured KeycloakAuthorizer
to check the client’s permissions, based on the action and resource type.
Each resource type has a set of available permissions for operations.
For example, the Topic
resource type has Create
and Write
permissions among others.
Refer to the Kafka authorization model in the Kafka documentation for the full list of resources and permissions.
Keycloak’s authorization services model has four concepts for defining and granting permissions:
-
Resources
-
Scopes
-
Policies
-
Permissions
For information on these concepts, see the guide to Keycloak Authorization Services.
16.4.2. Mapping authorization models
The Kafka authorization model is used as a basis for defining the Keycloak roles and resources that control access to Kafka.
To grant Kafka permissions to user accounts or service accounts, you first create an OAuth client specification in Keycloak for the Kafka cluster.
You then specify Keycloak Authorization Services rules on the client.
Typically, the client ID of the OAuth client that represents the Kafka cluster is kafka
.
The example configuration files provided with Strimzi use kafka
as the OAuth client id.
Note
|
If you have multiple Kafka clusters, you can use a single OAuth client ( |
The kafka
client definition must have the Authorization Enabled option enabled in the Keycloak Admin Console.
For information on enabling authorization services, see the guide to Keycloak Authorization Services.
All permissions exist within the scope of the kafka
client.
If you have different Kafka clusters configured with different OAuth client IDs, they each need a separate set of permissions even though they’re part of the same Keycloak realm.
When the Kafka client uses OAUTHBEARER authentication, the Keycloak authorizer (KeycloakAuthorizer
) uses the access token of the current session to retrieve a list of grants from the Keycloak server.
To grant permissions, the authorizer evaluates the grants list (received and cached) from Keycloak Authorization Services based on the access token owner’s policies and permissions.
An initial Keycloak configuration usually involves uploading authorization scopes to create a list of all possible actions that can be performed on each Kafka resource type. This step is performed once only, before defining any permissions. You can add authorization scopes manually instead of uploading them.
Authorization scopes should contain the following Kafka permissions regardless of the resource type:
-
Create
-
Write
-
Read
-
Delete
-
Describe
-
Alter
-
DescribeConfigs
-
AlterConfigs
-
ClusterAction
-
IdempotentWrite
If you’re certain you won’t need a permission (for example, IdempotentWrite
), you can omit it from the list of authorization scopes.
However, that permission won’t be available to target on Kafka resources.
Note
|
The |
Resource patterns are used for pattern matching against the targeted resources when performing permission checks.
The general pattern format is <resource_type>:<pattern_name>
.
The resource types mirror the Kafka authorization model. The pattern allows for two matching options:
-
Exact matching (when the pattern does not end with
*
) -
Prefix matching (when the pattern ends with
*
)
Topic:my-topic
Topic:orders-*
Group:orders-*
Cluster:*
Additionally, the general pattern format can be prefixed by kafka-cluster:<cluster_name>
followed by a comma, where <cluster_name>
refers to the metadata.name
in the Kafka custom resource.
kafka-cluster:my-cluster,Topic:*
kafka-cluster:*,Group:b_*
When the kafka-cluster
prefix is missing, it is assumed to be kafka-cluster:*
.
When defining a resource, you can associate it with a list of possible authorization scopes which are relevant to the resource. Set whatever actions make sense for the targeted resource type.
Though you may add any authorization scope to any resource, only the scopes supported by the resource type are considered for access control.
Policies are used to target permissions to one or more user accounts or service accounts. Targeting can refer to:
-
Specific user or service accounts
-
Realm roles or client roles
-
User groups
A policy is given a unique name and can be reused to target multiple permissions to multiple resources.
Use fine-grained permissions to pull together the policies, resources, and authorization scopes that grant access to users.
The name of each permission should clearly define which permissions it grants to which users.
For example, Dev Team B can read from topics starting with x
.
16.4.3. Permissions for common Kafka operations
The following examples demonstrate the user permissions required for performing common operations on Kafka.
To create a topic, the Create
permission is required for the specific topic, or for Cluster:kafka-cluster
.
bin/kafka-topics.sh --create --topic my-topic \
--bootstrap-server my-cluster-kafka-bootstrap:9092 --command-config=/tmp/config.properties
If a user has the Describe
permission on a specified topic, the topic is listed.
bin/kafka-topics.sh --list \
--bootstrap-server my-cluster-kafka-bootstrap:9092 --command-config=/tmp/config.properties
To display a topic’s details, Describe
and DescribeConfigs
permissions are required on the topic.
bin/kafka-topics.sh --describe --topic my-topic \
--bootstrap-server my-cluster-kafka-bootstrap:9092 --command-config=/tmp/config.properties
To produce messages to a topic, Describe
and Write
permissions are required on the topic.
If the topic hasn’t been created yet, and topic auto-creation is enabled, the permissions to create a topic are required.
bin/kafka-console-producer.sh --topic my-topic \
--bootstrap-server my-cluster-kafka-bootstrap:9092 --producer.config=/tmp/config.properties
To consume messages from a topic, Describe
and Read
permissions are required on the topic.
Consuming from the topic normally relies on storing the consumer offsets in a consumer group, which requires additional Describe
and Read
permissions on the consumer group.
Two resources
are needed for matching. For example:
Topic:my-topic
Group:my-group-*
bin/kafka-console-consumer.sh --topic my-topic --group my-group-1 --from-beginning \
--bootstrap-server my-cluster-kafka-bootstrap:9092 --consumer.config /tmp/config.properties
As well as the permissions for producing to a topic, an additional IdempotentWrite
permission is required on the
Cluster:kafka-cluster
resource.
Two resources
are needed for matching. For example:
Topic:my-topic
Cluster:kafka-cluster
bin/kafka-console-producer.sh --topic my-topic \
--bootstrap-server my-cluster-kafka-bootstrap:9092 --producer.config=/tmp/config.properties --producer-property enable.idempotence=true --request-required-acks -1
When listing consumer groups, only the groups on which the user has the Describe
permissions are returned.
Alternatively, if the user has the Describe
permission on the Cluster:kafka-cluster
, all the consumer groups are returned.
bin/kafka-consumer-groups.sh --list \
--bootstrap-server my-cluster-kafka-bootstrap:9092 --command-config=/tmp/config.properties
To display a consumer group’s details, the Describe
permission is required on the group and the topics associated with the group.
bin/kafka-consumer-groups.sh --describe --group my-group-1 \
--bootstrap-server my-cluster-kafka-bootstrap:9092 --command-config=/tmp/config.properties
To change a topic’s configuration, the Describe
and Alter
permissions are required on the topic.
bin/kafka-topics.sh --alter --topic my-topic --partitions 2 \
--bootstrap-server my-cluster-kafka-bootstrap:9092 --command-config=/tmp/config.properties
In order to use kafka-configs.sh
to get a broker’s configuration, the DescribeConfigs
permission is required on the
Cluster:kafka-cluster
.
bin/kafka-configs.sh --entity-type brokers --entity-name 0 --describe --all \
--bootstrap-server my-cluster-kafka-bootstrap:9092 --command-config=/tmp/config.properties
To change a Kafka broker’s configuration, DescribeConfigs
and AlterConfigs
permissions are required on Cluster:kafka-cluster
.
bin/kafka-configs --entity-type brokers --entity-name 0 --alter --add-config log.cleaner.threads=2 \
--bootstrap-server my-cluster-kafka-bootstrap:9092 --command-config=/tmp/config.properties
To delete a topic, the Describe
and Delete
permissions are required on the topic.
bin/kafka-topics.sh --delete --topic my-topic \
--bootstrap-server my-cluster-kafka-bootstrap:9092 --command-config=/tmp/config.properties
To run leader selection for topic partitions, the Alter
permission is required on the Cluster:kafka-cluster
.
bin/kafka-leader-election.sh --topic my-topic --partition 0 --election-type PREFERRED /
--bootstrap-server my-cluster-kafka-bootstrap:9092 --admin.config /tmp/config.properties
To generate a partition reassignment file, Describe
permissions are required on the topics involved.
bin/kafka-reassign-partitions.sh --topics-to-move-json-file /tmp/topics-to-move.json --broker-list "0,1" --generate \
--bootstrap-server my-cluster-kafka-bootstrap:9092 --command-config /tmp/config.properties > /tmp/partition-reassignment.json
To execute the partition reassignment, Describe
and Alter
permissions are required on Cluster:kafka-cluster
. Also,
Describe
permissions are required on the topics involved.
bin/kafka-reassign-partitions.sh --reassignment-json-file /tmp/partition-reassignment.json --execute \
--bootstrap-server my-cluster-kafka-bootstrap:9092 --command-config /tmp/config.properties
To verify partition reassignment, Describe
, and AlterConfigs
permissions are required on Cluster:kafka-cluster
, and on each
of the topics involved.
bin/kafka-reassign-partitions.sh --reassignment-json-file /tmp/partition-reassignment.json --verify \
--bootstrap-server my-cluster-kafka-bootstrap:9092 --command-config /tmp/config.properties
16.4.4. Example: Setting up Keycloak Authorization Services
If you are using OAuth 2.0 with Keycloak for token-based authentication,
you can also use Keycloak to configure authorization rules to constrain client access to Kafka brokers.
This example explains how to use Keycloak Authorization Services with keycloak
authorization.
Set up Keycloak Authorization Services to enforce access restrictions on Kafka clients.
Keycloak Authorization Services use authorization scopes, policies and permissions to define and apply access control to resources.
Keycloak Authorization Services REST endpoints provide a list of granted permissions on resources for authenticated users. The list of grants (permissions) is fetched from the Keycloak server as the first action after an authenticated session is established by the Kafka client. The list is refreshed in the background so that changes to the grants are detected. Grants are cached and enforced locally on the Kafka broker for each user session to provide fast authorization decisions.
Strimzi provides example configuration files. These include the following example files for setting up Keycloak:
kafka-ephemeral-oauth-single-keycloak-authz.yaml
-
An example
Kafka
custom resource configured for OAuth 2.0 token-based authorization using Keycloak. You can use the custom resource to deploy a Kafka cluster that useskeycloak
authorization and token-basedoauth
authentication. kafka-authz-realm.json
-
An example Keycloak realm configured with sample groups, users, roles and clients. You can import the realm into a Keycloak instance to set up fine-grained permissions to access Kafka.
If you want to try the example with Keycloak, use these files to perform the tasks outlined in this section in the order shown.
When you configure token-based oauth
authentication, you specify a jwksEndpointUri
as the URI for local JWT validation.
When you configure keycloak
authorization, you specify a tokenEndpointUri
as the URI of the Keycloak token endpoint.
The hostname for both URIs must be the same.
In Keycloak, confidential clients with service accounts enabled can authenticate to the server in their own name using a client ID and a secret.
This is convenient for microservices that typically act in their own name, and not as agents of a particular user (like a web site).
Service accounts can have roles assigned like regular users.
They cannot, however, have groups assigned.
As a consequence, if you want to target permissions to microservices using service accounts, you cannot use group policies, and should instead use role policies.
Conversely, if you want to limit certain permissions only to regular user accounts where authentication with a username and password is required, you can achieve that as a side effect of using the group policies rather than the role policies.
This is what is used in this example for permissions that start with ClusterManager
.
Performing cluster management is usually done interactively using CLI tools.
It makes sense to require the user to log in before using the resulting access token to authenticate to the Kafka broker.
In this case, the access token represents the specific user, rather than the client application.
Setting up permissions in Keycloak
Set up Keycloak, then connect to its Admin Console and add the preconfigured realm.
Use the example kafka-authz-realm.json
file to import the realm.
You can check the authorization rules defined for the realm in the Admin Console.
The rules grant access to the resources on the Kafka cluster configured to use the example Keycloak realm.
-
A running Kubernetes cluster.
-
The Strimzi
examples/security/keycloak-authorization/kafka-authz-realm.json
file that contains the preconfigured realm.
-
Install the Keycloak server using the Keycloak Operator as described in Installing the Keycloak Operator in the Keycloak documentation.
-
Wait until the Keycloak instance is running.
-
Get the external hostname to be able to access the Admin Console.
NS=sso kubectl get ingress keycloak -n $NS
In this example, we assume the Keycloak server is running in the
sso
namespace. -
Get the password for the
admin
user.kubectl get -n $NS pod keycloak-0 -o yaml | less
The password is stored as a secret, so get the configuration YAML file for the Keycloak instance to identify the name of the secret (
secretKeyRef.name
). -
Use the name of the secret to obtain the clear text password.
SECRET_NAME=credential-keycloak kubectl get -n $NS secret $SECRET_NAME -o yaml | grep PASSWORD | awk '{print $2}' | base64 -D
In this example, we assume the name of the secret is
credential-keycloak
. -
Log in to the Admin Console with the username
admin
and the password you obtained.Use
https://HOSTNAME
to access the KubernetesIngress
.You can now upload the example realm to Keycloak using the Admin Console.
-
Click Add Realm to import the example realm.
-
Add the
examples/security/keycloak-authorization/kafka-authz-realm.json
file, and then click Create.You now have
kafka-authz
as your current realm in the Admin Console.The default view displays the Master realm.
-
In the Keycloak Admin Console, go to Clients > kafka > Authorization > Settings and check that Decision Strategy is set to Affirmative.
An affirmative policy means that at least one policy must be satisfied for a client to access the Kafka cluster.
-
In the Keycloak Admin Console, go to Groups, Users, Roles and Clients to view the realm configuration.
- Groups
-
Groups
organize users and set permissions. Groups can be linked to an LDAP identity provider and used to compartmentalize users, such as by region or department. Users can be added to groups through a custom LDAP interface to manage permissions for Kafka resources. For more information on using KeyCloak’s LDAP provider, see the guide to Keycloak Server Administration. - Users
-
Users
define individual users. In this example,alice
(in theClusterManager
group) andbob
(inClusterManager-my-cluster
) are created. Users can also be stored in an LDAP identity provider. - Roles
-
Roles
assign specific permissions to users or clients. Roles function like tags to give users certain rights. While roles cannot be stored in LDAP, you can add Keycloak roles to groups to combine both roles and permissions. - Clients
-
Clients
define configurations for Kafka interactions.-
The
kafka
client handles OAuth 2.0 token validation for brokers and contains authorization policies (which require authorization to be enabled). -
The
kafka-cli
client is used by command line tools to obtain access or refresh tokens. -
team-a-client
andteam-b-client
represent services with partial access to specific Kafka topics.
-
-
In the Keycloak Admin Console, go to Authorization > Permissions to see the granted permissions that use the resources and policies defined for the realm.
For example, the
kafka
client has the following permissions:Dev Team A can write to topics that start with x_ on any cluster Dev Team B can read from topics that start with x_ on any cluster Dev Team B can update consumer group offsets that start with x_ on any cluster ClusterManager of my-cluster Group has full access to cluster config on my-cluster ClusterManager of my-cluster Group has full access to consumer groups on my-cluster ClusterManager of my-cluster Group has full access to topics on my-cluster
- Dev Team A
-
The Dev Team A realm role can write to topics that start with
x_
on any cluster. This combines a resource calledTopic:x_*
,Describe
andWrite
scopes, and theDev Team A
policy. TheDev Team A
policy matches all users that have a realm role calledDev Team A
. - Dev Team B
-
The Dev Team B realm role can read from topics that start with
x_
on any cluster. This combinesTopic:x_*
,Group:x_*
resources,Describe
andRead
scopes, and theDev Team B
policy. TheDev Team B
policy matches all users that have a realm role calledDev Team B
. Matching users and clients have the ability to read from topics, and update the consumed offsets for topics and consumer groups that have names starting withx_
.
Deploying a Kafka cluster with Keycloak authorization
Deploy a Kafka cluster configured to connect to the Keycloak server.
Use the example kafka-ephemeral-oauth-single-keycloak-authz.yaml
file to deploy the Kafka cluster as a Kafka
custom resource.
The example deploys a single-node Kafka cluster with keycloak
authorization and oauth
authentication.
-
The Keycloak authorization server is deployed to your Kubernetes cluster and loaded with the example realm.
-
The Cluster Operator is deployed to your Kubernetes cluster.
-
The Strimzi
examples/security/keycloak-authorization/kafka-ephemeral-oauth-single-keycloak-authz.yaml
custom resource.
-
Use the hostname of the Keycloak instance you deployed to prepare a truststore certificate for Kafka brokers to communicate with the Keycloak server.
SSO_HOST=SSO-HOSTNAME SSO_HOST_PORT=$SSO_HOST:443 STOREPASS=storepass echo "Q" | openssl s_client -showcerts -connect $SSO_HOST_PORT 2>/dev/null | awk ' /BEGIN CERTIFICATE/,/END CERTIFICATE/ { print $0 } ' > /tmp/sso.pem
The certificate is required as Kubernetes
Ingress
is used to make a secure (HTTPS) connection.Usually there is not one single certificate, but a certificate chain. You only have to provide the top-most issuer CA, which is listed last in the
/tmp/sso.pem
file. You can extract it manually or using the following commands:Example command to extract the top CA certificate in a certificate chainsplit -p "-----BEGIN CERTIFICATE-----" sso.pem sso- for f in $(ls sso-*); do mv $f $f.pem; done cp $(ls sso-* | sort -r | head -n 1) sso-ca.crt
NoteA trusted CA certificate is normally obtained from a trusted source, and not by using the openssl
command. -
Deploy the certificate to Kubernetes as a secret.
kubectl create secret generic oauth-server-cert --from-file=/tmp/sso-ca.crt -n $NS
-
Set the hostname as an environment variable
SSO_HOST=SSO-HOSTNAME
-
Create and deploy the example Kafka cluster.
cat examples/security/keycloak-authorization/kafka-ephemeral-oauth-single-keycloak-authz.yaml | sed -E 's#\${SSO_HOST}'"#$SSO_HOST#" | kubectl create -n $NS -f -
Preparing TLS connectivity for a CLI Kafka client session
Create a new pod for an interactive CLI session. Set up a truststore with a Keycloak certificate for TLS connectivity. The truststore is to connect to Keycloak and the Kafka broker.
-
The Keycloak authorization server is deployed to your Kubernetes cluster and loaded with the example realm.
In the Keycloak Admin Console, check the roles assigned to the clients are displayed in Clients > Service Account Roles.
-
The Kafka cluster configured to connect with Keycloak is deployed to your Kubernetes cluster.
-
Run a new interactive pod container using the Kafka image to connect to a running Kafka broker.
NS=sso kubectl run -ti --restart=Never --image=quay.io/strimzi/kafka:latest-kafka-3.9.0 kafka-cli -n $NS -- /bin/sh
NoteIf kubectl
times out waiting on the image download, subsequent attempts may result in an AlreadyExists error. -
Attach to the pod container.
kubectl attach -ti kafka-cli -n $NS
-
Use the hostname of the Keycloak instance to prepare a certificate for client connection using TLS.
SSO_HOST=SSO-HOSTNAME SSO_HOST_PORT=$SSO_HOST:443 STOREPASS=storepass echo "Q" | openssl s_client -showcerts -connect $SSO_HOST_PORT 2>/dev/null | awk ' /BEGIN CERTIFICATE/,/END CERTIFICATE/ { print $0 } ' > /tmp/sso.pem
Usually there is not one single certificate, but a certificate chain. You only have to provide the top-most issuer CA, which is listed last in the
/tmp/sso.pem
file. You can extract it manually or using the following command:Example command to extract the top CA certificate in a certificate chainsplit -p "-----BEGIN CERTIFICATE-----" sso.pem sso- for f in $(ls sso-*); do mv $f $f.pem; done cp $(ls sso-* | sort -r | head -n 1) sso-ca.crt
NoteA trusted CA certificate is normally obtained from a trusted source, and not by using the openssl
command. -
Create a truststore for TLS connection to the Kafka brokers.
keytool -keystore /tmp/truststore.p12 -storetype pkcs12 -alias sso -storepass $STOREPASS -import -file /tmp/sso-ca.crt -noprompt
-
Use the Kafka bootstrap address as the hostname of the Kafka broker and the
tls
listener port (9093) to prepare a certificate for the Kafka broker.KAFKA_HOST_PORT=my-cluster-kafka-bootstrap:9093 STOREPASS=storepass echo "Q" | openssl s_client -showcerts -connect $KAFKA_HOST_PORT 2>/dev/null | awk ' /BEGIN CERTIFICATE/,/END CERTIFICATE/ { print $0 } ' > /tmp/my-cluster-kafka.pem
The obtained
.pem
file is usually not one single certificate, but a certificate chain. You only have to provide the top-most issuer CA, which is listed last in the/tmp/my-cluster-kafka.pem
file. You can extract it manually or using the following command:Example command to extract the top CA certificate in a certificate chainsplit -p "-----BEGIN CERTIFICATE-----" /tmp/my-cluster-kafka.pem kafka- for f in $(ls kafka-*); do mv $f $f.pem; done cp $(ls kafka-* | sort -r | head -n 1) my-cluster-kafka-ca.crt
NoteA trusted CA certificate is normally obtained from a trusted source, and not by using the openssl
command. For this example we assume the client is running in a pod in the same namespace where the Kafka cluster was deployed. If the client is accessing the Kafka cluster from outside the Kubernetes cluster, you would have to first determine the bootstrap address. In that case you can also get the cluster certificate directly from the Kubernetes secret, and there is no need foropenssl
. For more information, see Setting up client access to a Kafka cluster. -
Add the certificate for the Kafka broker to the truststore.
keytool -keystore /tmp/truststore.p12 -storetype pkcs12 -alias my-cluster-kafka -storepass $STOREPASS -import -file /tmp/my-cluster-kafka-ca.crt -noprompt
Keep the session open to check authorized access.
Checking authorized access to Kafka using a CLI Kafka client session
Check the authorization rules applied through the Keycloak realm using an interactive CLI session. Apply the checks using Kafka’s example producer and consumer clients to create topics with user and service accounts that have different levels of access.
Use the team-a-client
and team-b-client
clients to check the authorization rules.
Use the alice
admin user to perform additional administrative tasks on Kafka.
The Kafka image used in this example contains Kafka producer and consumer binaries.
-
ZooKeeper and Kafka are running in the Kubernetes cluster to be able to send and receive messages.
-
The interactive CLI Kafka client session is started.
-
Prepare a Kafka configuration file with authentication properties for the
team-a-client
client.SSO_HOST=SSO-HOSTNAME cat > /tmp/team-a-client.properties << EOF security.protocol=SASL_SSL ssl.truststore.location=/tmp/truststore.p12 ssl.truststore.password=$STOREPASS ssl.truststore.type=PKCS12 sasl.mechanism=OAUTHBEARER sasl.jaas.config=org.apache.kafka.common.security.oauthbearer.OAuthBearerLoginModule required \ oauth.client.id="team-a-client" \ oauth.client.secret="team-a-client-secret" \ oauth.ssl.truststore.location="/tmp/truststore.p12" \ oauth.ssl.truststore.password="$STOREPASS" \ oauth.ssl.truststore.type="PKCS12" \ oauth.token.endpoint.uri="https://$SSO_HOST/realms/kafka-authz/protocol/openid-connect/token" ; sasl.login.callback.handler.class=io.strimzi.kafka.oauth.client.JaasClientOauthLoginCallbackHandler EOF
The SASL
OAUTHBEARER
mechanism is used. This mechanism requires a client ID and client secret, which means the client first connects to the Keycloak server to obtain an access token. The client then connects to the Kafka broker and uses the access token to authenticate. -
Prepare a Kafka configuration file with authentication properties for the
team-b-client
client.cat > /tmp/team-b-client.properties << EOF security.protocol=SASL_SSL ssl.truststore.location=/tmp/truststore.p12 ssl.truststore.password=$STOREPASS ssl.truststore.type=PKCS12 sasl.mechanism=OAUTHBEARER sasl.jaas.config=org.apache.kafka.common.security.oauthbearer.OAuthBearerLoginModule required \ oauth.client.id="team-b-client" \ oauth.client.secret="team-b-client-secret" \ oauth.ssl.truststore.location="/tmp/truststore.p12" \ oauth.ssl.truststore.password="$STOREPASS" \ oauth.ssl.truststore.type="PKCS12" \ oauth.token.endpoint.uri="https://$SSO_HOST/realms/kafka-authz/protocol/openid-connect/token" ; sasl.login.callback.handler.class=io.strimzi.kafka.oauth.client.JaasClientOauthLoginCallbackHandler EOF
-
Authenticate admin user
alice
by usingcurl
and performing a password grant authentication to obtain a refresh token.USERNAME=alice PASSWORD=alice-password GRANT_RESPONSE=$(curl -X POST "https://$SSO_HOST/realms/kafka-authz/protocol/openid-connect/token" -H 'Content-Type: application/x-www-form-urlencoded' -d "grant_type=password&username=$USERNAME&password=$PASSWORD&client_id=kafka-cli&scope=offline_access" -s -k) REFRESH_TOKEN=$(echo $GRANT_RESPONSE | awk -F "refresh_token\":\"" '{printf $2}' | awk -F "\"" '{printf $1}')
The refresh token is an offline token that is long-lived and does not expire.
-
Prepare a Kafka configuration file with authentication properties for the admin user
alice
.cat > /tmp/alice.properties << EOF security.protocol=SASL_SSL ssl.truststore.location=/tmp/truststore.p12 ssl.truststore.password=$STOREPASS ssl.truststore.type=PKCS12 sasl.mechanism=OAUTHBEARER sasl.jaas.config=org.apache.kafka.common.security.oauthbearer.OAuthBearerLoginModule required \ oauth.refresh.token="$REFRESH_TOKEN"