SQL Server Performance Forum – Threads Archive

CHECKDB error not associated with an object

Hi, One of our SQL 2000 Servers went down this morning and we couldn’t start it up again. We called our hardware supplier who suggested a couple of things which helped us start SQL Server again. Once started, the msdb was suspect. This was down to a torn page in the data file. It was so bad that SQL couldn’t attach it, hence it was suspect. After fixing that I thought I’d run a few DBCC CHECKDB commands to check for any other data loss. I noticed that in the top half of one of our databases there were consistency errors with pages not associated with objects. See the following extract. Msg 8939, Sev 16: Table error: Object ID 0, index ID 0, page (1:4268216). Test (m_headerVersion == HEADER_7_0) failed. Values are 0 and 1. [SQLSTATE 42000]
Msg 8939, Sev 16: Table error: Object ID 0, index ID 0, page (1:4268216). Test ((m_type >=DATA_PAGE && m_type <= UNDOFILE_HEADER_PAGE) || (m_type == UNKNOWN_PAGE && level == BASIC_HEADER)) failed. Values are 0 and 101. [SQLSTATE 42000]
Msg 8939, Sev 16: Table error: Object ID 0, index ID 0, page (1:4268216). Test (m_freeData >= PAGEHEADSIZE && m_freeData <= (UINT)PAGESIZE – m_slotCnt * sizeof (Slot)) failed. Values are 0 and 8192. [SQLSTATE 42000]
Msg 8909, Sev 16: Table error: Object ID 0, index ID 0, page ID (1:7637286). The PageId in the page header = (0:0). [SQLSTATE 42000]
Msg 2536, Sev 16: DBCC results for ‘sl_tes’. [SQLSTATE 01000]
Msg 8954, Sev 16: CHECKDB found 0 allocation errors and 4 consistency errors not associated with any single object. [SQLSTATE 01000] Anyone know what this means? all objects in the database seem to be okay, bar a couple – which I can fix, so that’s okay. I just need to know how serious this is. Many thanks.

If you can afford data loss then a simple DBCC CHECKDB (‘DBName’, REPAIR_ALLOW_DATA_LOSS) will do and also check event viewer log for any isssues or information on hardware problem. Satya SKJ
Contributing Editor & Forums Moderator
http://www.SQL-Server-Performance.Com
This posting is provided â€œAS ISâ€ with no rights for the sake of knowledge sharing.

Hi, Is it saying that some of the pages reserved for database growth are screwed? As they aren’t associated with an object…??? I am concerned with running an REPAIR_ALLOW_DATA_LOSS because the errors aren’t really associated with anything. I also know that a hardware error has occured. this was taken from the event log: Predictive Failure reported: Array Disk 1:2 Controller 2, Channel 1 Thoughts?

Ah.. hardware issues due to that it screwed up data.
Check last good known backup and restore, then run DBCC CHECKDB to see any associated errors are present. if possible take a rowcount on all tables. Satya SKJ
Contributing Editor & Forums Moderator
http://www.SQL-Server-Performance.Com
This posting is provided â€œAS ISâ€ with no rights for the sake of knowledge sharing.

Thanks for your advice. But I’d still like to know what the error means with regards to the error "not associated with a single object". I have noticed that the page number is the same as the page number associated with the table in the results. e.g. page, 1:4268216 has errored (above) as has table sl_1211, also on the same page.

It might be associated with TEMPDB, topichttp://www.sqlteam.com/forums/topic.asp?TOPIC_ID=55422 fyi. Satya SKJ
Contributing Editor & Forums Moderator
http://www.SQL-Server-Performance.Com
This posting is provided â€œAS ISâ€ with no rights for the sake of knowledge sharing.

]]>