Disaster recovery of a 2018 deployment

You can optionally return to a 2018 deployment, if necessary, after completion of a successful upgrade from 2018 to v10.

About this task

You can return to the 2018 system by performing a disaster recovery restore on each subsystem.

Note:
  • You must have upgraded the subsystems in the order specified in Upgrade requirements and limitations.
  • To return to a 2018 system, you must remove all version 10 subsystems. To uninstall API Connect entirely, see Removing API Connect. You can optionally follow the links below to remove individual subsystems, but you must remove all subsystems to successfully restore a 2018 system.

Procedure

  1. Management subsystem

    Perform disaster recovery of the management subsystem if the 2018 Management subsystem was successfully upgraded and all 2018 Management resources were deleted but you now need to revert to version 2018. You also must perform disaster recovery of the management subsystem if the 2018 Portal subsystem was successfully upgraded and all 2018 Portal resources were deleted, but you now need to revert to version 2018.

    1. To uninstall the v10 management subsystem, see Uninstalling API Connect subsystems.
    2. Follow the 2018 instructions in Restoring the management database in a Kubernetes environment.

      Note:

      • You will need to restore both the Management subsystem and the Portal subsystem. Be sure to choose backups from the same time to perform a restore.
      • You must perform disaster recovery for Portal, as described in Step 2, immediately after the Management restore. Carefully follow all the prerequisites of 2018 Portal disaster recovery.
  2. Developer Portal subsystem

    Perform disaster recovery of the 2018 Portal if the 2018 Portal subsystem was successfully upgraded and all 2018 portal resources were deleted but you now need to revert to version 2018

    1. You must first complete disaster recovery for the Management subsystem, as described in Step 1, and then perform Portal disaster recovery.
    2. Uninstall the version 10 Portal:
    3. Follow the instructions in Backing up and restoring the Developer Portal in a Kubernetes environment, being sure to follow all the prerequisites.

      When restoring, be sure to choose backups for both the Management subsystem and the Portal subsystem from the same time.

  3. Gateway subsystem

    You will want to perform disaster recovery of the 2018 Gateway subsystem if the 2018 Gateway subsystem was successfully upgraded and all the 2018 gateway resources were deleted, but you now need to revert to 2018.

    Note that although the upgrade from 2018 to 10.0.1.x-eus Gateway on Kubernetes is designed for zero downtime (when using a 3 node setup), reverting to the 2018 gateway will incur downtime.

    You must perform Management disaster recovery first and then perform Gateway disaster recovery.

    1. Uninstall the v10 Gateway:
    2. Install the 2018 Gateway subsystem. See Installing the Gateway subsystem into a Kubernetes environment.
  4. Analytics subsystem

    You will want to perform disaster recovery of the 2018 Analytics subsystem if your 2018 Analytics subsystem was successfully upgraded and all the 2018 Analytics resources were deleted, but you now need to revert to 2018.

    If Management and Portal subsystems are already upgraded to 10.0.1.x-eus, you must complete disaster recovery for Management subsystem, as described in Step 1, and then disaster recovery for the Portal subsystem, as described in Step 2.

    1. Uninstall the v10 Analytics subsystem:
    2. Install 2018 Analytics. See Installing the Analytics subsystem into a Kubernetes environment.
    3. Restore the 2018 backup that you took before upgrading. See Restoring the Analytics database.
    4. Re-associate the Analytics service with your Gateways.