February 17, 2004 at 5:44 am
Recently we upgraded the RAM and OS to WIN 2003 , on our SQL 2000 Active/Active cluster and after that we have been experiencing problems of High Disk Queue Length. On investigation it was found that it is read disk queue length which is high.
No Changes has been made to the SQL Server in the upgrade process except for RAM , we have enabled awe memory and have specified the MAX size of memory.
Is this a known issue/bug? Any ideas?
February 20, 2004 at 8:00 am
This was removed by the editor as SPAM
February 20, 2004 at 10:02 am
Did you do both upgrades at once?
Haven't done much on W2K3, getting ready to do a few here, no clusters, but with monitoing I should be able to see if performance overall has changed. Don't monitor the read Q unless I think there's an issue, however.
Nothing else changed? No SAN changes? No disk failures, code changes, etc.?
February 21, 2004 at 9:05 pm
If you check SQL Server current activity, do you see any processes with some data under Wait Resource? I would guess you should see some processes with Wait Resources and likely should spell the resource they are waiting for. You may want to check what processes these are and what they are waiting for (use sp_locks, dbcc page etc,.). Usually in upgrades like these there may have been "some old application code" that crept back in.. It all takes few processes to ruin the performance.. Let us know what you find..
Viewing 4 posts - 1 through 3 (of 3 total)
You must be logged in to reply to this topic. Login to reply