I've set up a readscale AG with 1 primary and 3 secondary read-only replicas (all async). No AD, no cluster.
Everything seems to work--the read-only replicas update properly as expected and are only slightly behind the primary, as expected.
When I force a failover with data loss to a secondary replica, it knocks the other replica databases out of replication and I am unable to get them to recognize the new primary replica. After a forced failover the secondaries still have the original primary name listed in the dashboard.
Is there a way to force the secondaries to recognize a new primary replica?