February 6, 2010 at 9:07 pm
You guys are not going to believe this, but I was on a support call with Microsoft the other day (yes, back to more issues with my Win2008 cluster and SQL2005), and here's what the support engineer said was the root cause of the problem with DBMail, maintenance plans, etc -
"Known Issue" when using lower case server node names instead of upper case server node names. I almost fell out of my chair.
Take a look at your SQL errorlog. Note that the resourse database is still in 1399 build level? This is a result of the the lower case server node name "known issue".
So here's what I need to do now...
a) with all resources running on node-a, remove node-b from the cluster
b) rename node-b to NODE-B
c) attach NODE-B to the cluster
d) failover all resources to NODE-B
e) remove node-a from the cluster
f) rename node-a to NODE-A
g) attach NODE-A to the cluster
h) failover resources a few times to make sure it all works
i) re-apply my most recent patch (in my case 9.00.4266) and check if the resource db build level has been updated.
There will probably not be any hotfixes to correct this. The "simple" solution is to build your nodes with UPPER-CASE node names.
Yeah, I thought it was pretty dern funny too.
February 7, 2010 at 12:58 pm
Thank you very much for this valuable feedback.
So this is a windows 2008 clustering issue ... who would have come up with that ??
I think I've been lucky because my cluster administrator always used upper case namings :Whistling:
If you can, provide a KB ref. so I can forward it to my peers.
Johan
Johan
Learn to play, play to learn !
Dont drive faster than your guardian angel can fly ...
but keeping both feet on the ground wont get you anywhere :w00t:
- How to post Performance Problems
- How to post data/code to get the best help[/url]
- How to prevent a sore throat after hours of presenting ppt
press F1 for solution, press shift+F1 for urgent solution 😀
Need a bit of Powershell? How about this
Who am I ? Sometimes this is me but most of the time this is me
February 8, 2010 at 1:08 pm
Please see this MSDN blog...
We are still trying to get this issue resolved. I'm currently working with MS. They spend 3 hours on our new clustered servers this morning, but still haven't fixed it.
February 9, 2010 at 12:18 pm
Dale Hughes-356583 (2/6/2010)
You guys are not going to believe this, but I was on a support call with Microsoft the other day (yes, back to more issues with my Win2008 cluster and SQL2005), and here's what the support engineer said was the root cause of the problem with DBMail, maintenance plans, etc -"Known Issue" when using lower case server node names instead of upper case server node names. I almost fell out of my chair.
Take a look at your SQL errorlog. Note that the resourse database is still in 1399 build level? This is a result of the the lower case server node name "known issue".
So here's what I need to do now...
a) with all resources running on node-a, remove node-b from the cluster
b) rename node-b to NODE-B
c) attach NODE-B to the cluster
d) failover all resources to NODE-B
e) remove node-a from the cluster
f) rename node-a to NODE-A
g) attach NODE-A to the cluster
h) failover resources a few times to make sure it all works
i) re-apply my most recent patch (in my case 9.00.4266) and check if the resource db build level has been updated.
There will probably not be any hotfixes to correct this. The "simple" solution is to build your nodes with UPPER-CASE node names.
Yeah, I thought it was pretty dern funny too.
Microsoft was unable to get the above to fix our new cluster, so we've decided to rebuilt from the OS up.
There are some posts of people putting production systems onto new clusters and then latter discovering this problem. They are now stuck at the base installation level, 1399, for SQL Server 2005 - they're unable to apply any service packs or CUs without rebuilding from scratch.
We, too, thought we had fixed this when we ran the sqldbupg.exe mentioned earlier. But that's a false resolution. Running sqldbupg.exe got mail working and fixed the maintenance plan error, but your service pack installation may still be incomplete, even if it appears to install successfully.
As I said, we thought we were okay after running the sqldbupg.exe, but then we encountered a problem when trying to setup replication. We were getting errors stating:" Procedure or function sp_MSadd_logreader_history has too many arguments specified" That's when we finally found the blog stating that there is a known issue when node names are lower case characters (see my post above).
You can verify that your SP upgrade complete and successful by run this script:
SELECT @@VERSION
SELECT SERVERPROPERTY('ResourceVersion')
The returned values must match, or your service pack installation is incomplete.
What burns me is that this was a known issue back in March, 2009. Windows 2008-R2 wasn't released until October of 2009, and it's not mentioned, or fixed, in the R2 release.
June 23, 2010 at 3:34 pm
AWU,
I am having the same problem with sql 2005 sp3 - cluster on Wiindows 2008 server with DB mail. Did Dale's suggestion fix your problem?
gw
July 20, 2010 at 11:26 pm
This issue has been fixed in Cumulative Update 9 for SQL Server 2005 SP3. Please refer to the KB below for more information:
http://support.microsoft.com/kb/978308
This relieves you from the hassle of evicting your nodes, changing the case and adding it back to cluster.
Viewing 6 posts - 16 through 20 (of 20 total)
You must be logged in to reply to this topic. Login to reply