When a vRealize Automation appliance in a multiple-node, high availability (HA) configuration has failed, you might need to replace the faulty node.

Caution: Before proceeding, VMware recommends that you contact technical support to troubleshoot the HA issue and verify that the problem is isolated to one node.

If technical support determines that you need to replace the node, take the following steps.

  1. In vCenter, take backup snapshots of every appliance in the HA configuration.

    In the backup snapshots, don't include virtual machine memory.

  2. Shut down the faulty node.
  3. Make note of the faulty node vRealize Automation software build number, and network settings.

    Note the FQDN, IP address, gateway, DNS servers, and especially MAC address. Later, you assign the same values to the replacement node.

  4. Check the status of the primary database node. From a root command line on any healthy node, run the following:
    > kubectl get pod `vracli status | jq -r '.databaseNodes[] | select(.["Role"] == "primary") | .["Node name"]' | cut -d '.' -f 1` -n prelude -o wide --no-headers=true
    
    primary-db-node-name 1/1 Running 0 39h 12.123.2.14 vc-vm-224-84.company.com <none> <none>
    Important: The primary database node must be one of the healthy nodes.

    If the primary database node is faulty, contact technical support instead of proceeding.

  5. From the root command line of the healthy node, remove the faulty node.

    vracli cluster remove faulty-node-FQDN

  6. Use vCenter to deploy a new, replacement vRealize Automation node.

    Deploy the same vRealize Automation software build number, and apply the network settings from the faulty node. Include the FQDN, IP address, gateway, DNS servers, and especially MAC address that you noted earlier.

  7. Power on the replacement node.
  8. Log in as root to the command line of the replacement node.
  9. Verify that the initial boot sequence has finished by running the following command.

    vracli status first-boot

    Look for a First boot complete message.

  10. From the replacement node, join the vRealize Automation cluster.

    vracli cluster join primary-DB-node-FQDN

  11. Log in as root to the command line of the primary database node.
  12. Deploy the repaired cluster by running the following script.

    /opt/scripts/deploy.sh