Post reply

AG setup issues

  • February 19, 2021 at 7:10 am

    #3847135

    I'm pretty much a veteran at AG... I'm trying to set up a 3 node cluster right now on SQL Server 2019 and Windows 2019.  The primary and a secondary are at our local data center, and another secondary is on Azure on a standalone box.  There are 2 issues I've never seen really before (at least not during setup), and I see no cases for this when searching around.

    1. There is no option for Automatic for Failover mode, only Manual.  I'm not using any External failover mechanisms, so None for Cluster Type, plus I checked off Database Level Heath Detection.  When I get to the part of specifying my Replicas, there's no option for Automatic, only Manual.  This despite the fact that I am selecting Synchronous Synchronization State.  Actually, I don't think it matters, the option does not appear at all on the drop down.  When I try to manually run the Transact-SQL command the error I get is:  Msg 47101, Level 16, State 0, Line 7

      The cluster type of availability group 'xxxxx' only supports MANUAL failover mode. Verify that the specified availability group cluster type is correct, then retry the operation.

    2. When I try to do a manual failover, I get a failover readiness error for both replicas, with a Data Loss Warning - This replica has one or more databases that are not synchronized, blah blah blah.  This is really weird, as I have literally a) added the replica database server  2) chose automated seeding and created the replica 3) immediately tried to failover.  Both servers shows the database (I have a test AG with one test DB with one table with 4 rows in it) as Synchronized, and I verified that the one table in each of the replicas has 4 rows.  So what could be the problem here?  I've only seen this problem in the past when things are not synchronized or in the middle of synchronizing.

  • February 19, 2021 at 6:32 pm

    #3847443

    IT and I solved the issue.  I chose a IP for the cluster that I thought wasn't in use, but apparently it was, so the cluster would go down after a short while (not so obvious actually but I guess I didn't look carefully enough).  Without the WSFC cluster, the AG assumes clusterless, so that's why only MANUAL was available.  I also had machines running across different subnets, which also caused some ordering problems.  Anyhow, with the cluster, the data loss problems seem to have disappeared magically as well.

  • February 27, 2021 at 4:35 am

    This was removed by the editor as SPAM

Viewing 3 posts - 1 through 2 (of 2 total)

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