Standby Bare Metal Node Failure or Maintenance

This section describes the procedures involved in replacing a Standby node which has either failed (unplanned) or removed for maintenance (planned). To replace a Standby node, use the following:

  1. Shutdown the Standby node.

    Note

    Upgrade or synchronize the remote clusters using the Standby node.

    For more information on upgrading the SMI Cluster Manager in HA mode, see Upgrading SMI Cluster Manager in HA section

    in UCC SMI Cluster Manager Deployment Guide.

  2. Add the Standby node to the same HA configuration when the Standby node is up and running.

  3. Run the cluster synchronization.

    clusters cluster_name actions sync run debug true 

    Example:

    SMI Cluster Deployer# clusters trysjc-ha actions sync run debug true
    This will run sync.  Are you sure? [no,yes] yes
    message accepted
  4. Verify the status of the cluster and pods when the Active and Standby nodes are up and running.

    clusters cluster_name actions k8s cluster-status 
    kubectl get pods -n namespace 
  5. Log in to the Ops Center service and verify whether the cluster configuration and CEE data are retained.

NOTES:

  • clusters cluster_name - Specifies the K8s cluster.

  • actions sync run debug true - Synchronizes the cluster configuration.

  • actions k8s cluster-status - Displays the status of the cluster.