restarting Server with AG

  • Hello,
    I have 2 servers(SS2016) which participate in AG in such way that Server1 is a host for primary replica of AG1 and also a host for secondary replica of AG2.
    Server2 is a host for primary replica of AG2 and also a host for secondary replica of AG1.
    Due to the Windows maintenance, I need to schedule an outage for Server2 for a few hours.
    What is the proper (and less intrusive) way  of doing that.
    Thank you

  • Anyone?

  • inHouseDBA - Sunday, October 14, 2018 10:24 PM

    Anyone?

    I thought I had posted something last week, obviously not and my apologies.
    First of all, I'm making the assumption that the log files on Server1 are large enough to contain the backlog of transactions, while Server2 is unavailable. I'm also assuming that you are using a listener, to control the connections
    .
    I would fail-over AG2, so Server1 is now the primary for both AG groups.
    If fail-over is set to automatic, set it to manual, because I suspect it will complain as soon as you do the next step, with automatic fail-over (although I could be wrong, having only ever worked with AG groups that have more than two servers).
    Suspend data movements for all databases within both AG groups. 
    Stop the SQL services on Server2.
    Perform your maintenance on Server2.
    Re-start SQL services on Server2.
    Resume data movements on all databases within both AG groups.
    Once the data movements have caught up, if you originally had automatic fail-over set, then set that again.
    Fail-over AG2 back to Server2.

  • Thank you Steve, It worked out pretty well!

Viewing 4 posts - 1 through 3 (of 3 total)

You must be logged in to reply to this topic. Login to reply