May 15, 2014 at 12:11 pm
Hello,
Dealing with Server 2012 and SQL Server 2012 fail-over cluster in an Active/Passive set up. I am trying to find out how to fail node 1 over to node 2. I don't have a lot of experience with clustering, what little I have was in 2008. If I remembered correctly in 2008, you would right click the active node select move, and select the node you want to move it to.
I am not seeing this in SQL 2012. If I right click my active node, go to more actions (as I don't see a move option) and my two options are Stop Cluster Service or Evict. Neither one of those sound like viable options.
Googling this subject, I see stuff about right clicking the active node, select pause, and select Drain Roles which does not sound right either. And stuff about resource groups.
If anyone can advise me on how to perform a manual fail-over, I would greatly appreciate it!!
Thanks
May 15, 2014 at 12:14 pm
Also, I looked into the option that an Availability group was used, but within SSMS, I attempt to expand the AlwaysOn High Availability section and I immediately get an error telling me I need to enable it before I use it.
Thanks Again!
May 15, 2014 at 12:30 pm
If you want to move a cluster group manually from C drive, use below:
Please note SQL3 would be the destination node, SQL1 (cluster group) is your source.
c:\>cluster group
Listing status for all available resource groups:
Group Node Status
-------------------- --------------- ------
Cluster Group SQL1 Online
Available Storage SQL1 Online
SQL Server (TEST) TESTSQL2 Online
SQL Server (TEST2) TESTSQL2 Online
SQLWINDTC SISSQL2 Online
c:\>cluster group "cluster group" /move:SQL3
Moving resource group 'cluster group'...
Group Node Status
-------------------- --------------- ------
cluster group SQL3 Online
I hope this helps.
Lava
May 15, 2014 at 12:34 pm
Thanks Lava, I will look into this. Unfortunately I am new to this job and there is no test or dev environment for me to test these commands out. Is there a way to do this via the Fail Over Cluster Manager UI?
Thanks Again for your reply
May 15, 2014 at 12:41 pm
I don't have a Always ON failover, but from what I have I can expand services. Right click on the service I need to failover to, and it gives a node name to where I need to failover to it. Have you checked Youtube to see if there are clips out there demonstrating the failover?
May 15, 2014 at 12:57 pm
If you screw something up when dealing with HA/DR the company can lose a tremendous amount of revenue (and even wind up going out of business). I have a client that measures some of their systems' downtime at over $1M PER HOUR. You should consider letting your company know that managing this system is not your strong point and that you need training and perhaps mentoring before being able to safely take over the reigns.
Best,
Kevin G. Boles
SQL Server Consultant
SQL MVP 2007-2012
TheSQLGuru on googles mail service
May 15, 2014 at 1:19 pm
Thanks for the reply Kevin...that is the plan here with my current company. In the meantime, can you help point me in the right direction?
May 15, 2014 at 1:50 pm
Post up a screen shot of the tool you are using that you described in the first post.
As for Always On, see here for DMVs you can use to get information about that subsystem: http://technet.microsoft.com/en-us/library/ff877943.aspx
Best,
Kevin G. Boles
SQL Server Consultant
SQL MVP 2007-2012
TheSQLGuru on googles mail service
May 15, 2014 at 2:15 pm
What I am looking at is attached...
I am thinking that while I have the "Roles" section highlighted on the left, in my center pane I can right click my role, select move and move the role and it's associated resources to my 2nd node. That maybe how you perform a manual failover in SQL Server 2012 via the Failover Cluster Manager when Availability groups are not enabled??? If anyone can confirm or deny that, it would be greatly appreciated.
May 15, 2014 at 7:36 pm
Are you sure you have an "Active/Passive" failover cluster service or application in place (i.e. SQL Server Failover Cluster Instance et al) since you say you DON'T have Always On in place? Seems to me you may just have a Windows Server Failover Cluster, which to my knowledge doesn't have a notion of one of the servers being active and the other passive. They are just "tied together" in a fashion. I note that I have had 2 beers tonight, which is 1.9 beers over my average, so my recollection could be faulty. 😎
Best,
Kevin G. Boles
SQL Server Consultant
SQL MVP 2007-2012
TheSQLGuru on googles mail service
May 16, 2014 at 4:58 am
GBeezy (5/15/2014)
If I remembered correctly in 2008, you would right click the active node select move, and select the node you want to move it to.I am not seeing this in SQL 2012. If I right click my active node, go to more actions (as I don't see a move option) and my two options are Stop Cluster Service or Evict. Neither one of those sound like viable options.
No, this is where you are getting the problems, this is not done at the node level.
Is this a Windows 2008 cluster or a Windows 2012 cluster?
In Failover Cluster Manager you must right click the role you want to move and then from the pop up menu select "Move". On the sub menu select either "Best possible" or "Select Node".
Or you could use the following Powershell commands (assuming you're logged onto a cluster node)
List the non system cluster groups
Get-ClusterGroup | ?{$_.name -NotIn "Available Storage","Cluster Group"} | ft name
Move your chosen group
Move-ClusterGroup -Name yourgroupname -Node nodetomoveto -Verbose
Do not evict the node or you'll likely make yourself very unpopular 😉
-----------------------------------------------------------------------------------------------------------
"Ya can't make an omelette without breaking just a few eggs" 😉
May 16, 2014 at 6:30 am
Thank You Perry!!! Evicting and Draining within Fail-over Cluster Manager definitely didn't sound right, I am sure they have their purpose though.
May 16, 2014 at 6:34 am
GBeezy (5/16/2014)
Evicting and Draining within Fail-over Cluster Manager definitely didn't sound right
That one will have you typing your resume quicker than you can select "Node" -> "Evict Node" 😉
You should be just fine with the Powershell commands above
-----------------------------------------------------------------------------------------------------------
"Ya can't make an omelette without breaking just a few eggs" 😉
May 16, 2014 at 7:20 am
Two things: As I mentioned, first go to C drive and type c:\>cluster group to see all the nodes connected to your cluster. This command only LIST groups, it does not do anything else. And 2) use the move command if you want to move a group. Here is a good link, http://blogs.technet.com/b/askcore/archive/2011/08/12/how-to-failover-the-cluster-group-and-available-storage-cluster-groups.aspx
May 16, 2014 at 7:38 am
Hold the phone guys!
GBeezy,
Fail over is not that hard at all. First let's find out what clustering you are using. When SQL 2012 was setup did you\they setup AlwaysOn? Do you have a Availability Group group in SSMS?
If this is not AlwaysOn and it is traditional clustering like SQL08 is setup then go into failover clustering admin under administration tools, under services and application, choose and right click node1 and move to passive node2.
Do not drain, evict, or anything else in the failover cluster admin console. Only move to Node2.
Also another way to know if this is not always on, are your disks shared? or are they independent?
MCSE SQL Server 2012\2014\2016
Viewing 15 posts - 1 through 15 (of 17 total)
You must be logged in to reply to this topic. Login to reply