I've created the following set-up:
a two-node 2012 R2 failover cluster, file-share quorum.
Installed SQL 2016 Basic on each node, enabled High Availability (Basic used due to budget concerns)
Set up several databases, created a basic HA group for each database. Automatic fail-over is enabled.
My problem is as follows:
Manual failover of all databases works. Shutting down, rebooting, or disconnecting the NIC on either server triggers the databases to fail over quickly & reliably.
If I simulate the SQL server service having a problem (kill sqlservr.exe or gracefully stop it), then fail-over is inconsistent. Sometimes all databases remain in a resolving state, sometimes some fail over and some are stuck resolving. I need to resolve
this before I can trust the system as live.
Thanks in advance for any advice!