Bringing a failed, previously primary/active DAG member back online RRS feed

  • Question

  • This could just be my paranoia after going on 2 hours of sleep and dealing with a failed DAG, but I would just like someone to confirm something here before I go trying to power this server back on.

    On one DAG I have two sites running DACP. Site 1 has a primary and a standby, then site 2 has a standby. Last night I had both servers at site 1 go down on me, and the primary active server didn't come back online, VMware problem. Long story short the databases were all kinds of jacked up and I spent hours getting them finally mounted on site 1's standby server. So site 1's secondary server and site 2's standby server are healthy at the moment.

    Now, I believe I have the VMware issue worked out, and want to try to boot up site 1's primary server. My concern is that something will go wrong, it will boot up and not realize the secondary server is now active, mount of the databases on itself and I'll have more issues.

    Can someone either calm my fears here and explain to me briefly why this won't be an issue, the server will realize the other server is running the databases, and all will be good. Or is there something I should do as a precaution before booting the server back up? Currently the mailbox server in site 2 is the active manager and the alternate file server witness box in site 2 is active as well (assuming still like this from the datacenter failover even though I left site 1 secondary server online and brought mailboxes back online on it). The primary server in site 1 will have full connectivity to all these servers once it's brought back online.

    Monday, June 11, 2018 7:54 PM


All replies