Outages of one data center (DC) that last up to 72 hours don't require this disaster recovery procedure. When the inaccessible DC becomes available again, Premium High Availability (PHA) automatically repairs the affected DC and restores Managed Cluster operations. For details, go to Multi-data center failover.
To avoid data inconsistency while a DC is unavailable, shut down the server service on all nodes in the affected DC. Start services only when network connectivity is stable again.
After 72 hours, Mission Control (MC) marks the Managed Cluster as not repaired after failover. Such a Managed Cluster isn't reliable. As a result, you must restore or recreate the failed DC from either an operational DC or a backup.
To recover from another data center, complete the following steps:
For the detailed procedure, go to Rebuild data center.