July 2, 2009 at 1:58 pm
We have been experiencing this error message when a particular procedure is executed...
The full error message is as follows
DESCRIPTION:Error: 625, Severity: 20, State: 3
Could not retrieve row from logical page (0:165) by RID because the entry in the offset table (2) for that RID (2) is less than or equal to 0.
This a SQL 7.0 server SP4.
Can someone please give some recommendations as what steps to follow to troubleshoot?
The procedure is a lengthy one that touches several different databases outside of the calling database.
I have run DBCC CheckDB on the calling database with a clean bill of health...
CHECKDB found 0 allocation errors and 0 consistency errors in database 'CurwWorkObj'.
DBCC execution completed. If DBCC printed error messages, contact your system administrator.
Any recommendations are appreciated.
Thanks,
Lee
July 2, 2009 at 2:03 pm
Can you run CheckDB on all databases that procedure touches?
Gail Shaw
Microsoft Certified Master: SQL Server, MVP, M.Sc (Comp Sci)
SQL In The Wild: Discussions on DB performance with occasional diversions into recoverability
July 2, 2009 at 2:15 pm
GilaMonster (7/2/2009)
Can you run CheckDB on all databases that procedure touches?
I am holding off for the moment...
The calling database was very small (300 mb's) so the DBCC command went pretty quick as far as impacting any users...
The other two databases are a little larger (8 gb's and 6 gb's) in comparison and I was asked to wait until end of day to run any DBCC checkDB commands against those but I do plan on it...after hours in a little bit (it is month end and also the server is slow by today's standards)
QUESTION
In the event I run the DBCC CheckDB on all the db's involved and they also come back with no issues where would you recommend looking next?
I had found a undocumented command DBCC Page that allows you to review information about the specific page...however I am not sure that the FileID the error message is reporting is even valid?
Can you have a Fileid of 0?
I tried the following command in the procedure calling db...and I got a 'corrupt table' error message...but I am guessing that is because table ID 0 is invalid cause?
DBCC TRACEON(3604)
GO
DBCC PAGE (CurwWorkObj, 0, 165, 0)
Server: Msg 8968, Level 16, State 1, Line 1
Table Corrupt: DBCC PAGE page (0:165) (object ID 0, index ID 0) is out of the range of this database.
PAGE:
DBCC execution completed. If DBCC printed error messages, contact your system administrator.
Thanks for the quick reply
Lee
July 2, 2009 at 3:25 pm
Lee Hart (7/2/2009)
QUESTIONIn the event I run the DBCC CheckDB on all the db's involved and they also come back with no issues where would you recommend looking next?
Not sure. A bridge to cross if it arrives
Can you have a Fileid of 0?
No. FileIDs start at 1.
I have seen CheckDB give file no of 0 or tableid of 0 when it can't work out from the page what the value should be. Sometimes it's just a case of not enough info, sometimes it's due to damage on the page.
Gail Shaw
Microsoft Certified Master: SQL Server, MVP, M.Sc (Comp Sci)
SQL In The Wild: Discussions on DB performance with occasional diversions into recoverability
July 2, 2009 at 10:03 pm
I spoke with my boss regarding this issue and he said it happens as a result of the combo of SQL 7 and the manner in which this process is ran...
When it is month end occurs and if other variables are involved (like they were today) a large number of updates can be generated by this process which is what was happening.
I wasn't given low level details on the specific 'why' but the gist was in regards to how SQL 7 handles memory the large number of updates this process generates...
Once the activity died down for the day, the process caught up...no more errors...and all is well...
The server and process are due for sun setting in about a year so no work is planned and is something that has to be tolerated when it happens.
Thanks Gail for assisting me!
July 3, 2009 at 12:02 am
I would still suggest running a checkDB on all the databases involved. Are you running regular integrity checks?
Gail Shaw
Microsoft Certified Master: SQL Server, MVP, M.Sc (Comp Sci)
SQL In The Wild: Discussions on DB performance with occasional diversions into recoverability
July 4, 2009 at 4:27 pm
Hi Gail,
There are regular scheduled maintenance checks performed nightly on this server and all have been clean.
Viewing 7 posts - 1 through 6 (of 6 total)
You must be logged in to reply to this topic. Login to reply