September 14, 2006 at 10:39 am
I have a very strange thing occuring. I have an Active/Active SQL 2005 64Bit Enterprise Edition cluster with 2 instances: default and named instance called NODE02.
Both instances work fine and installed without any problem. Issue is on NODE02 I installed the system databases onto the incorrect disk. I believe I followed all MS KB recommendations for moving them to the correct disk, however I just noted that I have (2) sets of data/log files for msdb. A .mdf/.ldf on T:\... and one on U:\...
Stranger yet is that the set on U is the same size as my tempdb that resides on that volume - a whopping 10GB/2GB. I only grew the tempdb to this size. Stranger yet is that when I either query sysdatabases or look at the properties of msdb I see only the files on T that are the ones I expect to use and that are 200MB/100MB.
There are no references I see to the set of files on U yet I can not delete/rename/move/etc because I get a "File in use" msg. If I shut down/take offline SQL I can rename the set on U, but then when I bring SQL back online it fails. This instance obviously thinks it needs both sets of files (timestamps continue to match up between T and U whenever I perform actions that would update data.) I need to break the dependancy and get rid of the files on U however.
Any help is appreciated!
- Tim Ford, MCSD Spectrum-Health
"Support SQL PASS!"
September 14, 2006 at 11:01 am
Hello Tim,
Do you mean 2 SQL instances on each of the cluster nodes?
If so, can you check for the physical files of each instance "msdb" database on the second node NODE02.
Hope this might give a clue for you to proceed further.
Thanks
Lucky
September 14, 2006 at 11:14 am
1 instance per virtual node. Default on virtual node X and named (NODE02) instance on virtual node Y. Neither instance on either node of this cluster references the files I am attempting to remove.
Viewing 3 posts - 1 through 2 (of 2 total)
You must be logged in to reply to this topic. Login to reply