June 14, 2012 at 6:45 am
I have installed sql server 2012 on two vms called sql2012-1 and sql2012-2. Both machines are running on windows 2008 r2 core mode. I created cluster called sql2012 between both nodes. I configured always on availability group for a test database, created a listener.
I am able to connect to the database via listener.
I am also able to failover the availability group from one node to other via gui/t-sql.
Problem - when i shut down the vm or stop sql services on the active node the group in not automatically failing over. Any help would be helpful.
Also do we need to add any shared disk for the cluster? My setup does not have any shared disk.
June 14, 2012 at 7:49 am
If you take a look at the dashboard, are all your replicas synchronized or synchronizing? I've run into issues when the automatic failover was not synchronized. I just ran a test on my virtuals, shutting down the primary, one of the secondaries that was defined as the automatic failover point picked it right up.
"The credit belongs to the man who is actually in the arena, whose face is marred by dust and sweat and blood"
- Theodore Roosevelt
Author of:
SQL Server Execution Plans
SQL Server Query Performance Tuning
June 14, 2012 at 7:50 am
And no, you shouldn't have to share storage in any way between the servers. That's the whole point of this new mechanism for AlwaysOn.
"The credit belongs to the man who is actually in the arena, whose face is marred by dust and sweat and blood"
- Theodore Roosevelt
Author of:
SQL Server Execution Plans
SQL Server Query Performance Tuning
June 14, 2012 at 8:05 am
Thanks Grant. Yes the db is synchronized. Its a test db with no transaction. I was thinking about failover timeout n other aspects. I had setup a plain cluster using cluster.exe and didnt change any defaults parameters of the cluster.. I was thinking about changing failover timeout but not sure how to do that..
June 14, 2012 at 8:24 am
Hmmm.... Not sure. Are you getting any errors? What does the cluster look like in the Dashboard with the primary offline? You can look at the health event file? There should be some indications of errors in there. Have you done a manual failover? Did that work?
It has to be something.
"The credit belongs to the man who is actually in the arena, whose face is marred by dust and sweat and blood"
- Theodore Roosevelt
Author of:
SQL Server Execution Plans
SQL Server Query Performance Tuning
June 14, 2012 at 10:39 am
Yes, the manual failover works but with the following warning.
The replica is on a WSFC node that does not have a quorum vote.
I manually failed over the availability group AG1 from sql2012-1 to sql2012-2 node
Then I shut down the VM sql2012-2 hoping for sql2012-1 to take over. but the Availability group shows as AG1(Resolving)
The warning gave me this link http://msdn.microsoft.com/en-us/library/hh270280.aspx%5B/url%5D
June 14, 2012 at 11:10 am
Ah well, there you go. Something up with your quorum settings. Makes sense. Change that & you should be good to go.
"The credit belongs to the man who is actually in the arena, whose face is marred by dust and sweat and blood"
- Theodore Roosevelt
Author of:
SQL Server Execution Plans
SQL Server Query Performance Tuning
June 14, 2012 at 11:42 am
My setup has only 2 nodes so "Node Majority" quorum will not work (which is currently setup). The other type of quorum (node and disk) needs a shared disk...
Looks like I need to add a quorum disk..
June 14, 2012 at 11:59 am
Hmmm.... Take a look at this article. It might help. They only use two servers. I'm using five, plus a domain controller.
"The credit belongs to the man who is actually in the arena, whose face is marred by dust and sweat and blood"
- Theodore Roosevelt
Author of:
SQL Server Execution Plans
SQL Server Query Performance Tuning
June 15, 2012 at 3:41 am
You do need a quorum disk as the cluster witness.
This could be just a network share hosted in a separate file server (network and file share majority, see 1st link below).
Of course it will need to be accessible to the service account of your cluster so all nodes can "see" it at all times.
See:
http://www.brentozar.com/archive/2012/06/everything-know-about-clustering-wrong/ (section "Cluster Quorums Manage Themselves")
http://technet.microsoft.com/en-us/library/hh212952.aspx
__________________________________________________________________________________
SQL Server 2016 Columnstore Index Enhancements - System Views for Disk-Based Tables[/url]
Persisting SQL Server Index-Usage Statistics with MERGE[/url]
Turbocharge Your Database Maintenance With Service Broker: Part 2[/url]
June 15, 2012 at 3:51 am
Thanks. I have asked my windows guy to give me a small quorum disk from the SAN or a third VM.. I will try the node+file share majority as well.
Viewing 11 posts - 1 through 10 (of 10 total)
You must be logged in to reply to this topic. Login to reply