Upgrading by re-initializing HADR deployments - Failure scenarios

(Primary instance) The instance upgrade fails during the upgrade job

You can identify this issue by reviewing the logs in the Db2 or Db2 Warehouse engine pod.

1. Exec into the Db2 engine pod as the db2inst1 user and run the following command to re-run the instance upgrade.

   db2_update_upgrade

2. If the re-run operation fails:
   1. Failover to the standby instance by redirecting all your applications.
   2. Open a support case with IBM to review the primary database instance.
3. If the re-run operation fails and high availability is not tolerable, restore to the version before the upgrade.
   1. Deploy a new Db2 instance for the version prior to your upgrade. For more information, see Deploying Db2 or Db2 Warehouse.
   2. Restore your database from a backup image. For more information, see Backing up and restoring Db2 or Db2 Warehouse.
   3. On the standby database instances, run the following command from the Db2 engine pod to stop HADR.

      manage_hadr -stop

   4. Configure and start HADR. For more information see configuring and starting HADR.

(Primary instance) The db2_update_upgrade --databases command fails to upgrade databases

You can identify this issue by reviewing the logs in the Db2 or Db2 Warehouse engine pod.

1. Exec into the Db2 engine pod as the db2inst1 user and run the following command to re-run the database upgrade.

   db2_update_upgrade --databases

2. If the re-run operation fails:
   1. Failover to the standby instance by redirecting all your applications.
   2. Open a support case with IBM to review the primary database instance.
3. If the re-run operation fails and high availability is not tolerable, restore to the version before the upgrade.
   1. Deploy a new Db2 instance for the version prior to your upgrade. For more information, see Deploying Db2 or Db2 Warehouse.
   2. Restore your database from a backup image. For more information, see Backing up and restoring Db2 or Db2 Warehouse.
   3. On the standby database instances, run the following command from the Db2 engine pod to stop HADR.

      manage_hadr -stop

   4. Configure and start HADR. For more information see configuring and starting HADR.

(Primary instance) The db2_update_upgrade --post-upgrade post upgrade operations fail

You can identify this issue by reviewing the logs in the Db2 or Db2 Warehouse engine pod. This failure scenario occurs when the instance and databases are upgraded successfully, but the post upgrade operations fails.

1. Exec into the Db2 engine pod as the db2inst1 user and run the following command to re-run the instance upgrade.

   db2_update_upgrade --post-upgrade

2. If the re-run operation fails:
   1. Failover to the standby instance by redirecting all your applications.
   2. Open a support case with IBM to review the primary database instance.

(Standby instance) The instance upgrade fails during the db2ckupgrade upgrade job