Note: This is not part of the Strimzi CNCF project!

Strimzi Shutdown is a simple utility to temporarily stop or restart your Strimzi-based Apache Kafka cluster.

You can download one of the release binaries from one of the GitHub releases and use it. Alternatively, you can also use the provided container image to run it from a Kubernetes Pod or locally as a container.

Strimzi Shutdown supports several command line options:

You can stop your Strimzi-based Apache Kafka cluster using the stop command. The stop command will execute the following steps:

1. Pause the reconciliation of the Kafka cluster and wait for the Strimzi Cluster Operator to confirm it
2. Stop all auxiliary tools (Cruise Control, Kafka Exporter, and Entity Operator)
3. Stop all Kafka nodes with the broker role only
4. Stop all Kafka nodes with the controller role

The following snippet shows an example of the stop command:

$ ./strimzi-shutdown stop --namespace myproject --name my-cluster
2025/07/22 11:43:08 Using kubeconfig /Users/scholzj/.kube/config
2025/07/22 11:43:08 Pausing reconciliation of Kafka cluster my-cluster in namespace myproject
2025/07/22 11:43:08 Waiting for Kafka cluster my-cluster in namespace myproject reconciliation to be paused
2025/07/22 11:43:08 Reconciliation of Kafka cluster my-cluster in namespace myproject is paused
2025/07/22 11:43:08 Deleting Deployment my-cluster-entity-operator in namespace myproject
2025/07/22 11:43:08 Deleting Deployment my-cluster-cruise-control in namespace myproject
2025/07/22 11:43:08 Deleting Deployment my-cluster-kafka-exporter in namespace myproject
2025/07/22 11:43:10 Deployment my-cluster-kafka-exporter in namespace myproject has been deleted
2025/07/22 11:43:10 Deployment my-cluster-entity-operator in namespace myproject has been deleted
2025/07/22 11:43:10 Stopping all Kafka nodes with broker role only
2025/07/22 11:43:10 Deleting StrimziPodSet my-cluster-witton for KafkaNodePool witton
2025/07/22 11:43:10 Deleting StrimziPodSet my-cluster-aston for KafkaNodePool aston
2025/07/22 11:43:10 Deleting StrimziPodSet my-cluster-bodymoor for KafkaNodePool bodymoor
2025/07/22 11:43:13 Stopping all remaining Kafka nodes
2025/07/22 11:43:13 Deleting StrimziPodSet my-cluster-controllers for KafkaNodePool controllers
2025/07/22 11:43:20 Kafka cluster my-cluster in namespace myproject has been stopped

You can restart your stopped Strimzi-based Apache Kafka cluster using the continue command. The continue command will unpause the reconciliation of the Apache Kafka cluster, and the Strimzi Cluster Operator will restore it. The continue command will wait until the Kafka cluster is ready before exiting.

The following snippet shows an example of the continue command:

$ ./strimzi-shutdown continue --namespace myproject --name my-cluster
2025/07/22 11:44:32 Using kubeconfig /Users/scholzj/.kube/config
2025/07/22 11:44:32 Reconciliation of Kafka cluster my-cluster in namespace myproject will be unpaused
2025/07/22 11:44:32 Unpausing reconciliation of Kafka cluster my-cluster in namespace myproject
2025/07/22 11:44:32 Waiting for Kafka cluster my-cluster in namespace myproject to get ready.
2025/07/22 11:47:02 Kafka cluster my-cluster in namespace myproject has been restarted and should be ready

You can use Kubernetes CronJob with Strimzi Shutdown to stop or restart your Kafka cluster at a specific time. To do so, you can use the following CronJob:

apiVersion: v1
kind: ServiceAccount
metadata:
  name: strimzi-shutdown

---
apiVersion: rbac.authorization.k8s.io/v1
kind: Role
metadata:
  name: strimzi-shutdown
rules:
  - apiGroups: [""]
    resources: ["pods"]
    verbs: ["get", "list"]
  - apiGroups: ["apps"]
    resources: ["deployments"]
    verbs: ["get", "watch", "list", "delete"]
  - apiGroups: ["kafka.strimzi.io"]
    resources: ["kafkas", "kafkanodepools"]
    verbs: ["get", "watch", "list", "update"]
  - apiGroups: ["core.strimzi.io"]
    resources: ["strimzipodsets"]
    verbs: ["get", "list", "delete"]

---
apiVersion: rbac.authorization.k8s.io/v1
kind: RoleBinding
metadata:
  name: strimzi-shutdown
subjects:
  - kind: ServiceAccount
    name: strimzi-shutdown
roleRef:
  kind: Role
  name: strimzi-shutdown
  apiGroup: rbac.authorization.k8s.io

---
apiVersion: batch/v1
kind: CronJob
metadata:
  name: shutdown-my-cluster
spec:
  schedule: "00 20 * * 1-5"
  jobTemplate:
    spec:
      template:
        spec:
          serviceAccountName: strimzi-shutdown
          containers:
            - name: strimzi-shutdown
              image: ghcr.io/scholzj/strimzi-shutdown:0.1.0
              command:
                - /strimzi-shutdown
                - stop
                - --name=my-cluster
          restartPolicy: OnFailure

---
apiVersion: batch/v1
kind: CronJob
metadata:
  name: restart-my-cluster
spec:
  schedule: "00 8 * * 1-5"
  jobTemplate:
    spec:
      template:
        spec:
          serviceAccountName: strimzi-shutdown
          containers:
            - name: strimzi-shutdown
              image: ghcr.io/scholzj/strimzi-shutdown:0.1.0
              command:
                - /strimzi-shutdown
                - continue
                - --name=my-cluster
          restartPolicy: OnFailure

This example will stop the Kafka cluster named my-cluster every Monday to Friday at 8 pm. And it will start the same Kafka cluster again every Monday to Friday at 8 am.

Note: If you want to deploy the CronJob into a different namespace from where the Kafka cluster is running, you would need to adjust the RBAC resources accordingly and use the --namespace option.

No, Strimzi shutdown currently supports only KRaft-based Apache Kafka clusters.

No, the other Strimzi operands, such as Kafka Connect, Mirror Maker, or Bridge are stateless and do not need any special process to be stopped. If you want to stop them, you can scale them to 0 by editing the resource and changing the .spec.replicas field or using kubectl scale.