Please start any new threads on our new
site at https://forums.sqlteam.com. We've got lots of great SQL Server
experts to answer whatever question you can come up with.
Author |
Topic |
sandlu
Starting Member
17 Posts |
Posted - 2010-03-24 : 14:06:34
|
In one sql 2005 with sp2 X64, integrity check job failed with the following error:Process ID 58 attempted to unlock a resource it does not own: OBJECT: 2:127727253:0 . Retry the transaction, because this error may be caused by a timing conditionThe job failed in the last few weeks (it is a weekly job). It successed sometimes if I reran it, but It failed a few times last week when I reran the job. |
|
tkizer
Almighty SQL Goddess
38200 Posts |
|
sandlu
Starting Member
17 Posts |
Posted - 2010-03-24 : 14:31:39
|
Thanks, Tara, checkdb did not show any error |
 |
|
tkizer
Almighty SQL Goddess
38200 Posts |
Posted - 2010-03-24 : 14:35:18
|
If this timing/blocking issue is a frequent occurence, I would suggest running the DBCC CHECKDB job on a copy of production. DBCC CHECKDB is a very I/O intensive task, so it is not wise to rerun it especially during normal business hours. You could greatly hurt the performance of production. We can't even run it at night as we don't have a quiet period, so we grab the full backup daily and restore it to a test machine. On the test machine is where we run DBCC CHECKDB.Tara KizerMicrosoft MVP for Windows Server System - SQL Serverhttp://weblogs.sqlteam.com/tarad/Subscribe to my blog |
 |
|
sandlu
Starting Member
17 Posts |
Posted - 2010-03-24 : 15:11:24
|
Thanks, Tera. The integrity check is running in Sunday afternoon, which is pretty quite for us. Our database is too big to create a backup to do the integrity check.Thanks |
 |
|
russell
Pyro-ma-ni-yak
5072 Posts |
Posted - 2010-03-25 : 13:31:35
|
I hope that doesn't mean you aren't taking backups.There is no such thing as a database being too big to do backups. I would argue that the bigger it is, the more you need them. |
 |
|
Kristen
Test
22859 Posts |
Posted - 2010-03-25 : 13:54:57
|
Might mean that they have no machine with enough space to do a restore onto (for testing etc).That would worry me as much as not being able to do backups. It is important to do trial restores (to another "temporary" database, on another machine if you like) otherwise you have no way of knowing if backups are restore-able or not!Not to mention that the backup copy can then be checked for corruption using DBCC CHECKDBAnd being able to test out thing like Service Pack upgrades and so on.If you ain't got a spare machine that you can restore the DB onto then go buy one!! It doesn't have to have fancy disks - but having a machine identical to the Production machine has obvious benefits for fail-over etc. |
 |
|
|
|
|
|
|