Whenever a replica has to restore the primary’s complete database backup set (as part of replica enrolment, immediately after failback to the primary has completed, or whenever the primary detects that the replica has experienced a serious error) several things may happen:
• The primary may be unable to push additional updates to the replica until the restore is complete
• The replica’s web UI may be unavailable
• The primary’s cluster status page may be unable to query the replica’s status
The reason for this is that the replica is busy restoring the database backup set and while this is occurring the databases are generally unavailable. When the database is unavailable, the web UI is also generally unavailable.
The process to un-force a failover and synchronize the data on the replica with the primary takes some time. The amount of time that this takes is variable. It depends on how big the database backup set is, both in terms of the size of the files (full DB backup, plus the size of the individual log backups, a function of DB activity over time) and the number of files (a function of the amount of time since the last full backup). However, once the operation is complete, the web UI is available.
Before attempting any HA related activities please insure a recent backup has been taken on the primary appliance (preferably that same day).