Here's the scenario:
- Primary Site A - SQL Node 1 - Windows 2012 R2/SQL 2012 SP2
- Recovery Site B - SQL Node 2 - Windows 2012 R2/SQL 2012 SP2
- There is no option to put a third node in site A.
- Reasonably reliable low latency link between the two geographically separate sites.
- Availability group with asynchronous replication created for a database that needs to be highly available, manual failover is OK.
What is the correct quorum configuration for the cluster here?
Do I have this correct...
I setup dynamic file share witness at Site A. If Node 1 goes down, then vote from witness is dynamically removed but Node 2 comes won't come online automatically because replication is asynchronous. Have to manually bring node 2 online.
If link between sites goes down temporarily, Node 2 can't see either FSW or Node 1 so doesn't do anything. All OK.
If site A...