Database corruption

  • We have a server that has 4 x 3.0 GB CPU and 20 GB RAM.  It also has 3 drive chassis with 14 drives in each (having primarily RAID 5 and RAID 1 configurations) The box is a pretty hefty machine.  The problem we seem to be having is frequent database corruption errors.  We have checked with HP and verified all the drivers are up to date on the server and controllers.  We've disabled the caching on all controllers.  Our next steps are to try and lower the memory that SQL server is using.  Currently, we have 16 GB allocated to SQL.

    Does anyone have a similar configuration and experienced these problems?

    Marco A. Ramirez

    DBA, Data Management & Administration

    mramirez@landstar.com

     

     

     

  • Can you detail what corruption you had experienced? Did you configure OS with /3GB or /PAE and SQL Server with AWE enabled?

  • The most common error we get is Error:605, Severity 21, State:1 Attempt to fetch logical page in database 'DataWarehouse' belongs to object '180', not to object 'SAMPLE_FACT'

    We run DBCC CHECKDB commands on the database and it tells us we have corruption.  We then execute DBCC CHECKDB with REPAIR option to fix the errors.  The database has multiple filegroups however, all the errors don't happen on same drive controller.  It seems to happen randomly without a pattern.

    The /3 GB flag has been disabled and SQLServer with AWE is disabled.

     

     

  • You have read the explanations in BOL on 605? I would guess hardware causing this error, but a call to MS PSS might be a good choice here anyway.

    --
    Frank Kalis
    Microsoft SQL Server MVP
    Webmaster: http://www.insidesql.org/blogs
    My blog: http://www.insidesql.org/blogs/frankkalis/[/url]

  • Try this link :

    http://support.microsoft.com/default.aspx?scid=kb;en-us;826433

    It details corruption errors without physical hardware errors. Load SP4 change your disk cache settings and give it a go.

     

    PS. Got the same problem on 2 HP Servers. HP swears its not the hardware, which in a sense it isn't, but the problem is caused by caching and Lost Writes and Stale Reads.

    Good luck

Viewing 5 posts - 1 through 4 (of 4 total)

You must be logged in to reply to this topic. Login to reply