Click here to monitor SSC
SQLServerCentral is supported by Red Gate Software Ltd.
 
Log in  ::  Register  ::  Not logged in
 
 
 
        
Home       Members    Calendar    Who's On


Add to briefcase 12»»

SQL SERVER 2008 Cluster Failover message Expand / Collapse
Author
Message
Posted Thursday, June 20, 2013 12:06 PM
Grasshopper

GrasshopperGrasshopperGrasshopperGrasshopperGrasshopperGrasshopperGrasshopperGrasshopper

Group: General Forum Members
Last Login: Wednesday, February 5, 2014 2:28 PM
Points: 11, Visits: 41
Hi,
I have just installed a 2 node one instance SQL Server Cluster. When I fail over from one node to the other everything fails over fine but I am seeing this in the event log:

"LogWriter: Operating system error 21(The device is not ready.) encountered.
Write error during log flush.
The log for database 'temdb' is not available.
Error: 17053, Severity: 16, State:1."

Our configuration has all drives on SAN disk which we have for several other clusters as well and they work fine. Has anyone encountered this before? I'm kind of stumped given every thing seems to be working fine and fail over is successful. I get this failing either cluster over. I appreciate an input.
Post #1465878
Posted Thursday, June 20, 2013 12:34 PM
SSCertifiable

SSCertifiableSSCertifiableSSCertifiableSSCertifiableSSCertifiableSSCertifiableSSCertifiableSSCertifiableSSCertifiable

Group: General Forum Members
Last Login: Thursday, August 21, 2014 3:52 PM
Points: 5,988, Visits: 12,923
it sounds like SQL is starting before that drive is accessible, are your dependencies correct?

---------------------------------------------------------------------

Post #1465894
Posted Thursday, June 20, 2013 1:40 PM
Grasshopper

GrasshopperGrasshopperGrasshopperGrasshopperGrasshopperGrasshopperGrasshopperGrasshopper

Group: General Forum Members
Last Login: Wednesday, February 5, 2014 2:28 PM
Points: 11, Visits: 41
Hi,
The SQL server is running on both nodes but is "passive" on the failover node. If I am on Node 1 and failover to Node 2 I see this error message on Node 1.

Thanks,
Kevin
Post #1465913
Posted Friday, June 21, 2013 12:17 AM


SSCertifiable

SSCertifiableSSCertifiableSSCertifiableSSCertifiableSSCertifiableSSCertifiableSSCertifiableSSCertifiableSSCertifiable

Group: General Forum Members
Last Login: Today @ 7:41 AM
Points: 6,300, Visits: 13,586
As George said have you checked the resource dependencies in the cluster group\application?

-----------------------------------------------------------------------------------------------------------

"Ya can't make an omelette without breaking just a few eggs"
Post #1466056
Posted Friday, June 21, 2013 8:29 AM
Grasshopper

GrasshopperGrasshopperGrasshopperGrasshopperGrasshopperGrasshopperGrasshopperGrasshopper

Group: General Forum Members
Last Login: Wednesday, February 5, 2014 2:28 PM
Points: 11, Visits: 41
George/ Perry,
First George I apologize to you ..I was thick as a brick when I read your response. Perry thanks for reinforcing. What I had to do to fix the issue was there:

There was an existing dependency for SQL Server that the mount point be online. What I had to do was place a dependency on the mount point that the disk arrays that hold "data, backup and tempdb" are online before the mount point shows online.

I didn't have to set this dependency on previous clusters I've built but this is a different SAN so I've asked that team to take a look if there is an issue.

Thanks for the input!!

Kevin
Post #1466213
Posted Friday, June 21, 2013 8:35 AM
SSCertifiable

SSCertifiableSSCertifiableSSCertifiableSSCertifiableSSCertifiableSSCertifiableSSCertifiableSSCertifiableSSCertifiable

Group: General Forum Members
Last Login: Thursday, August 21, 2014 3:52 PM
Points: 5,988, Visits: 12,923
thanks for feeding back, so many people just don't bother.

perhaps this cluster is the first with mount points?


---------------------------------------------------------------------

Post #1466218
Posted Friday, June 21, 2013 9:19 AM
Grasshopper

GrasshopperGrasshopperGrasshopperGrasshopperGrasshopperGrasshopperGrasshopperGrasshopper

Group: General Forum Members
Last Login: Wednesday, February 5, 2014 2:28 PM
Points: 11, Visits: 41
Actually we have 5 clusters and they all use mount points from SAN devices. This is the only one that I'm aware of that we have on this particular SAN so I'm not sure if they have some sort of latency issue ... hopefully it is not a sign of problems to come since this is an instance that will support adfs single sign on

Any thoughts on from a "best practice" standpoint I should go back and update my other clusters with a similar dependency?

Thanks,
Kevin
Post #1466248
Posted Friday, June 21, 2013 9:48 AM


SSCertifiable

SSCertifiableSSCertifiableSSCertifiableSSCertifiableSSCertifiableSSCertifiableSSCertifiableSSCertifiableSSCertifiable

Group: General Forum Members
Last Login: Today @ 7:41 AM
Points: 6,300, Visits: 13,586
See my article here and also the relevant Microsoft KB

-----------------------------------------------------------------------------------------------------------

"Ya can't make an omelette without breaking just a few eggs"
Post #1466269
Posted Friday, June 21, 2013 9:52 AM


SSCertifiable

SSCertifiableSSCertifiableSSCertifiableSSCertifiableSSCertifiableSSCertifiableSSCertifiableSSCertifiableSSCertifiable

Group: General Forum Members
Last Login: Today @ 7:41 AM
Points: 6,300, Visits: 13,586
kjtyson05 (6/21/2013)
What I had to do was place a dependency on the mount point that the disk arrays that hold "data, backup and tempdb" are online before the mount point shows online.

Correct, the root drive must be online before the mounted volume comes online.


-----------------------------------------------------------------------------------------------------------

"Ya can't make an omelette without breaking just a few eggs"
Post #1466271
Posted Friday, June 21, 2013 10:17 AM
SSCertifiable

SSCertifiableSSCertifiableSSCertifiableSSCertifiableSSCertifiableSSCertifiableSSCertifiableSSCertifiableSSCertifiable

Group: General Forum Members
Last Login: Thursday, August 21, 2014 3:52 PM
Points: 5,988, Visits: 12,923
thanks Perry I thought you would have relevant links to hand.

but i still get my point


---------------------------------------------------------------------

Post #1466296
« Prev Topic | Next Topic »

Add to briefcase 12»»

Permissions Expand / Collapse