We had a SQL 2014 (12.0.5000.0 version) windows 2012 R2 shared disk sql server cluster with two nodes in primary and one node in muti-site DR. We have an AG listener from primary to a multi-site secondary. We had an issue where we had to failover the cluster to node 2 from node 1. The AG listener stayed on node 1 and was up. The databases stayed in "not synchronizing/recovery pending" until we started and stopped the AG listener so it failed over to node 2 with the rest of the cluster. We did not have DB corruption as the cause of the recovery pending and all of the DBs were fine after AG listener start and stop. Can someone please explain why the AG listener isn't failing over as it should with the cluster and why it could not recover the DBs until it was on node 1 with the cluster resources?
Thanks!