Backup and restore requirements for a two data center deployment
Points to consider when you configure backup and restore settings on a two data center disaster recovery (2DCDR) deployment.
The 2DCDR backup requirements are different for the Portal and Management subsystems.
Ensure that you have read and understand the concepts of 2DCDR. See Two data center deployment strategy on Kubernetes and OpenShift and Key concepts of 2DCDR and failure scenarios
Management subsystem 2DCDR backup and restore requirements
- Different remote backup locations must be specified for each data center.
- Both active and warm-standby Management subsystems run the scheduled backup job.
- To restore from a management backup, follow these steps:
- Make each management data center a stand-alone API Connect deployment:
Remove the
multiSiteHA
section from thespec
section of the management CRs on both data centers. Keep a copy of themultiSiteHA
sections that you remove, they are added back at the end of the restore process. - You now have two stand-alone API Connect deployments. Decide which data center is to be the active, and restore the backup from that data center. Do not attempt to restore the backup from one data center, to the other data center.
- Add the
multiSiteHA
section back to the management CR in the restored data center. Ensure that themultiSiteHA
section defines the data center as the active. - Add the
multiSiteHA
section back to the management CR in the other data center. Ensure that themultiSiteHA
section defines the data center as the warm-standby (passive). - The management database takes some time to resync across the data centers. To confirm that replication is working, see: Verifying replication between data centers.
- Make each management data center a stand-alone API Connect deployment:
Remove the
Portal subsystem 2DCDR backup requirements
- The backup settings for both the active and warm-standby portal must be identical.
- Only the active data center portal takes backups. On failover, the backup job of the new warm-standby is disabled, and the backup job of the new active is enabled.
- Portal restores are only done on the active data center.
Configuring backups for 2DCDR deployments
The steps for configuring backups for both subsystems are the same as for a single API Connect deployment, but
keep the additional 2DCDR
requirements in mind when following them: