Restart the cluster

Restart all nodes in a cluster in a rolling fashion to provide zero downtime to applications. Restart nodes in a DataStax Enterprise (DSE), Hyper-Converged Database (HCD), or Apache Cassandra® cluster after adding or terminating one or more nodes.

Create the CassandraTask or K8ssandraTask custom resource (CR) that defines a restart operation in the data plane Kubernetes cluster where the target cassandraDatacenter or MissionControlCluster resource and the datacenter are deployed.

Mission Control detects the restart task custom resource definition (CRD), iterates one rack at a time, and triggers and monitors restart operations one pod at a time. Starting with version 1.18.0, you can configure parallel restarts to restart all nodes in a rack simultaneously for faster operations. Mission Control reports task progress and status.

Performance impact

Running a restart task has no significant implications on the performance or stability of the cluster as long as the replication factor (RF) is set to 3 or higher. This RF setting tolerates one node being down while maintaining quorum read and write operations.

Prerequisites

Restart the cluster

Decide to restart one or multiple datacenters, or all datacenters in the cluster. You have granular control over choosing target datacenters and their assigned rack.

You can use the Mission Control UI or CLI to configure backup storage:

Use the UI to restart one or more DCs

  1. Access Mission Control’s UI.

  2. In the Home Clusters dialog, click the target cluster namespace.

  3. In the Nodes section of the Overview tab:

    1. Click the row checkbox for your target datacenter.

    2. Click more_vert More Options on your target datacenter.

    3. Click Restart.

    4. To restart more than one datacenter, repeat substeps a, b, and c for each target datacenter.

      The restart activity starts immediately.

      To view notifications from the restart operation, see Monitor restart activity status.

Use the UI to restart all DCs

  1. Access Mission Control’s UI.

  2. In the Home Clusters dialog, click the target cluster namespace.

  3. In the Nodes section of the Overview tab, click the Name checkbox. This selects all datacenter row checkboxes.

  4. Click Bulk Actions.

  5. In the Bulk Actions dialog:

    1. Make sure that the Action type is Restart.

    2. Make sure that All is your target datacenter selection.

      In the Racks field, All is the default and only selection.

    3. Click Run.

      The restart activity starts immediately.

      To view notifications from the restart operation, see Monitor restart activity status.

Use the CLI to restart clusters

  1. Access Mission Control’s CLI.

  2. Create the restart task manifest using either CassandraTask or K8ssandraTask. This is a YAML configuration file that defines the task and the command in the kind section.

    For CassandraTask: You must create a separate restart task per datacenter.

    For K8ssandraTask: You can specify multiple datacenters in a single task using the spec.datacenters list.
    CassandraTask example

    The filename for this example is restart-task.cassandratask.yaml.

    apiVersion: control.k8ssandra.io/v1alpha1
kind: CassandraTask
metadata:
  name: restart-dc
spec:
  datacenter:
    name: dc1
    namespace: demo
  jobs:
    - name: restart-dc1
      command: restart

    Key options:

    • metadata.name: A unique identifier within the Kubernetes namespace where the task is submitted. While the name can be any value, consider including the cluster and datacenter names to prevent collision with other actions.

    • spec.datacenter: A unique namespace and name combination used to determine which datacenter to target with this operation.

    • spec.jobs[0].command: MUST be restart for this operation.

      Submit the restart CassandraTask CRD with this kubectl command:

      kubectl apply -f restart-task.cassandratask.yaml
    K8ssandraTask example

    The filename for this example is restart-task.k8ssandratask.yaml.

    apiVersion: control.k8ssandra.io/v1alpha1
kind: K8ssandraTask
metadata:
  name: restart-dc
spec:
  cluster:
    name: CLUSTER_NAME
    namespace: CLUSTER_NAMESPACE
  datacenters:
    - dc1
  jobs:
    - name: restart-dc1
      command: restart

    Key options:

    • metadata.name: A unique identifier within the Kubernetes namespace where the task is submitted. While the name can be any value, consider including the cluster and datacenter names to prevent collision with other actions.

    • spec.cluster: A unique namespace and name combination used to determine which cluster to target with this operation.

    • spec.datacenters: A list of datacenter names to restart within the specified cluster.

    • spec.jobs[0].command: MUST be restart for this operation.

    • Optional: spec.maxConcurrentPods: Controls the number of pods that can restart concurrently. Set to a positive integer to limit concurrent restarts. Omit for sequential restarts (default).

      Submit the restart K8ssandraTask CRD with this kubectl command:

      kubectl apply -f restart-task.k8ssandratask.yaml
      Submit the restart task object to the Kubernetes cluster where the specified datacenter is deployed.

      To view notifications from the restart operation, see Monitor restart activity status.

Configure parallel rack restarts

By default, Mission Control restarts nodes sequentially within each rack to maintain cluster stability. For faster restart operations, you can configure Mission Control to restart all nodes in a rack in parallel in Mission Control version 1.18.0 and later.

Parallel rack restarts are useful when:

  • You need to minimize downtime during maintenance windows.

  • Your cluster has sufficient replication factor (RF ≥ 3) to tolerate multiple nodes being down simultaneously.

  • You want to expedite configuration changes that require a restart.

Parallel rack restarts temporarily reduce cluster availability within the affected rack. Ensure your replication factor is set to 3 or higher before enabling parallel restarts.

To configure parallel rack restarts, use a K8ssandraTask manifest with the maxConcurrentPods parameter:

apiVersion: control.k8ssandra.io/v1alpha1
kind: K8ssandraTask
metadata:
  name: restart-dc
spec:
  cluster:
    name: CLUSTER_NAME
    namespace: CLUSTER_NAMESPACE
  datacenters:
    - dc1
  maxConcurrentPods: 0
  jobs:
    - name: restart-dc1
      command: restart

Set maxConcurrentPods to a positive integer to limit the number of pods that can restart concurrently.

Monitor restart activity status

You can use the Mission Control UI or CLI to monitor the status of the restart operation:

Use the UI to monitor status

  1. In the Mission Control UI main navigation, click Activities.

  2. Review Status notifications for the progress of the restart activity.

    A status of SUCCESS indicates the restart operation completed without issue. Timestamps are issued for the Start and End of the restart activity.

    The Activities pane refreshes often and automatically.

Use the CLI to monitor status

  1. Access Mission Control’s CLI.

  2. Review the restart status with one of these commands:

    For CassandraTask:

    kubectl get cassandratask restart-dc | yq .status

    For K8ssandraTask:

    kubectl get k8ssandratask restart-dc | yq .status

    Upon completion of the restart task, Mission Control updates or adds the following fields in the task YAML file to indicate the status:

    • either a status.succeeded or status.failed field with a count of the pods that have been restarted and whether each restart was successful or unsuccessful

    • status.completionTime field: timestamp

    • status.condition.status: True

    • status.condition.type: Complete

      Result
      kind: CassandraTask
metadata:
  annotations:
    kubectl.kubernetes.io/last-applied-configuration: |
      {"apiVersion":"control.k8ssandra.io/v1alpha1","kind":"CassandraTask","metadata":{"annotations":{},"name":"restart-dc","namespace":"mission-control"},"spec":{"datacenter":{"name":"dc1","namespace":"mission-control"},"jobs":[{"command":"restart","name":"restart-dc1"}]}}
  creationTimestamp: "2022-10-15T21:24:35Z"
  generation: 2
  labels:
    app.kubernetes.io/created-by: cass-operator
    app.kubernetes.io/instance: cassandra-demo
    app.kubernetes.io/managed-by: cass-operator
    app.kubernetes.io/name: cassandra
    app.kubernetes.io/version: 6.8.26
    cassandra.datastax.com/cluster: demo
    cassandra.datastax.com/datacenter: dc1
    control.k8ssandra.io/status: completed
  name: restart-dc
  namespace: mission-control
  ownerReferences:
  - apiVersion: cassandra.datastax.com/v1beta1
    blockOwnerDeletion: true
    controller: true
    kind: CassandraDatacenter
    name: dc1
    uid: 51561d2f-2d20-4a16-b90c-fe9b0655f1ba
  resourceVersion: "780428"
  uid: 2202880f-16f0-4b83-a5ce-ab3057cd6d70
spec:
  concurrencyPolicy: Forbid
  datacenter:
    name: dc1
    namespace: mission-control
  jobs:
  - args: {}
    command: restart
    name: restart-dc1
  restartPolicy: Never
status:
  completionTime: "2022-10-15T21:36:54Z"
  conditions:
  - lastTransitionTime: "2022-10-15T21:36:54Z"
    status: "True"
    type: Complete
  startTime: "2022-10-15T21:24:35Z"

Was this helpful?

Give Feedback

How can we improve the documentation?

© Copyright IBM Corporation 2026 | Privacy policy | Terms of use |  Manage Privacy Choices

Apache, Apache Cassandra, Cassandra, Apache Tomcat, Tomcat, Apache Lucene, Apache Solr, Apache Hadoop, Hadoop, Apache Pulsar, Pulsar, Apache Spark, Spark, Apache TinkerPop, TinkerPop, Apache Kafka and Kafka are either registered trademarks or trademarks of the Apache Software Foundation or its subsidiaries in Canada, the United States and/or other countries. Kubernetes is the registered trademark of the Linux Foundation.

General Inquiries: Contact IBM