Post reply

Problem with database incosnsitency or HP disk array

  • June 10, 2008 at 6:37 am

    #381049

    Hi,

    My SQL Server 2000 sp4 is using HP StorageWorks MSA 1510i disk array. After installation of this storage, from time to time I have getting error messages about database inconsistency:

    Error: 7987

    A possible database consistency problem has been detected on database 'master'. DBCC CHECKDB and DBCC CHECKCATALOG should be run on database 'master'..

    Error: 5180

    Could not open FCB for invalid file ID 0 in database 'XYZ'..

    Any CAUSE described by Miccorsoft Support hadn't fit to my environment. DBCC CHECKDB had been run with no error messages.

    I waited for next error messages and yesterday I got it after executing simple select statement with where clause. Today this select statement is running fine, no errors.

    Have you ever encountered these set of errors? Could it be storage problem?

    Thanks in advance for reply.

    Bartek

  • June 10, 2008 at 7:38 am

    #826639

    Most consistency errors are raced back to hardware problems. Potentially fabric or controller errors could be present.

    I'd be sure that someone carefully runs diagnostics on your storage.

  • June 10, 2008 at 8:03 am

    #826663

    Hardware or driver errors often will have these events in the system event log:

    The Driver was unable to obtain the HBA slot number for ScsiPort 3. No extended information in Dump Data.

    The Driver has detected a path failure to Subsystem ID 50001FE15005C4C0. Dump Data 0 contains the Phys Path Info . Dump Data 1 contains the Driver Status. Dump Data 2 contains the HBA Slot Number (ffffffff if unavailable). Dump Data 3 contains extended Driver Status.

    These messages indicate that database corruption or backup failures are likely to have occured. Typical backup failure events due to this root cause are:

    18210 :

    BackupIoRequest::WaitForIoCompletion: read failure on backup device 'D:\mssql\data\....mdf'. Operating system error 1117(The request could not be performed because of an I/O device error.).

    SQL = Scarcely Qualifies as a Language

  • June 10, 2008 at 12:02 pm

    #826868

    Have you run CHECKDB on master?

    Looks like a transient problem in the IO subsystem if the issue goes away and comes back - e.g. a stale read from a disk controller.

    Paul Randal
    CEO, SQLskills.com: Check out SQLskills online training!
    Blog:www.SQLskills.com/blogs/paul Twitter: @PaulRandal
    SQL MVP, Microsoft RD, Contributing Editor of TechNet Magazine
    Author of DBCC CHECKDB/repair (and other Storage Engine) code of SQL Server 2005

  • June 11, 2008 at 5:08 am

    #827150

    Thank you for all your posts.

    Of course, I had done CHECKDB against master database and no error messages.

    I also checked system log, none of those messages appeared.

  • June 11, 2008 at 6:03 am

    #827178

    So definitely sounds like a transient IO problem. Run IO diagnostics and the SQLIOSim tool to help flush out any IO issues you may have.

    Thanks

    Paul Randal
    CEO, SQLskills.com: Check out SQLskills online training!
    Blog:www.SQLskills.com/blogs/paul Twitter: @PaulRandal
    SQL MVP, Microsoft RD, Contributing Editor of TechNet Magazine
    Author of DBCC CHECKDB/repair (and other Storage Engine) code of SQL Server 2005

  • June 12, 2008 at 12:50 am

    #827664

    Definietly I have to run this tool which you've suggested.

    Thanks for help.

    Bartek

  • June 13, 2008 at 1:12 pm

    #828753

    Make sure that every single driver and firmware component between the CPUs and the physical disks are patched to the latest available level. That is likely the first thing that Microsoft (or anyone else) would require if you called them with this problem.

    Best,
    Kevin G. Boles
    SQL Server Consultant
    SQL MVP 2007-2012
    TheSQLGuru on googles mail service

Viewing 8 posts - 1 through 7 (of 7 total)

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