Preparing Workflow Process Service for disaster recovery

Prepare Workflow Process Service for disaster recovery. If the primary environment cannot provide service, the secondary environment must be able to.

Note: The disaster recovery procedure might vary in different environment and deployment configuration.

These steps are for an Workflow Process Service instance that is deployed successfully on a primary environment. Both the primary site and secondary site connect to the same PostgreSQL cluster.

Procedure

  1. For the PVC prefixed with datasave, use the Kuberenetes command to get its definition and backup the necessary part in the definition.
    1. Apply the PVC definition for datasave on the secondary site.
    2. Backup the files under the folder folder_for_datasave_PV/messaging.
  2. On the secondary site, restore the messaging folder for datasave PV, keeping the same folder structure.
  3. On the secondary site, check the uid-range in the used namespace. For example, the output for execute oc describe project wfpsproject is:
    Name:                   wfpsproject
          Created:                3 hours ago
          ...
          Annotations:            ... ...
                                  openshift.io/sa.scc.supplemental-groups=1000750000/10000
                                  openshift.io/sa.scc.uid-range=1000750000/10000
  4. Modify the copied messaging folders to the allowed uid. For example, chown -R 1000750000:root messaging.
  5. Update the CR definition to include the uid, securityContext.runAsUser: 1000750000.
      node:
        ... 
        securityContext:
          runAsUser: 1000750000
       ... 
  6. Apply the CR definition to update the Workflow Process Service deployment.