Results 1 to 3 of 3
  1. #1
    Join Date
    Jan 2004
    Location
    Munich/Germany
    Posts
    12

    Angry Unanswered: update statistics during DB dump? anyone?

    We have 8 production servers, each has one customer database. on Weekly basis make a dump of all DBs and load them on a separate "Report Server". Three weeks ago, we start running update statistics on 3 large tables durning the dump process on the production servers. No errors reported on the production sides, but while loading these dumps on the "Report Server" we are getting the following error

    [dtloadcmd 8149] DT-03016W: [SYBMSG 1108] Cannot deallocate extent 131941448, database 8. Object id 1314103722, index id 4, status 0 in extent does not match object id 1314103722, index id 3, status 0 in object being deallocated. Run DBCC CHECKALLOC.

    I suspect it is the update statistics process that is doing this.

    Any ideas?

    Thank you in advance guys

  2. #2
    Join Date
    Feb 2002
    Location
    Willy is on vacation
    Posts
    1,208
    Update stats wouldn't have anything to do with this error. It simply updates 2 system tables.

    If you notice from your error, the problem is reported on a user table object id = 1314103722. Run TABLEALLOC on this table, the problem seems to be on non-clustered indexes.

    You should do this on the source database. Fix the corruptions and dump it again.

  3. #3
    Join Date
    Jan 2004
    Location
    Munich/Germany
    Posts
    12
    I appreciate your reply, the strange thing is this error a)occurs on all databases copied from different severs - same object, b)no errors reported on the produciton sides, c)started occuring exactly the same weekend we started running update stat, d)we run update stat at that time only on 3 large tables, the faulty one here is one of these tables - again the othere two showed no problems!!!

    I thought Update stat reads index pages into cache first to accumlate statistics. Is it not possible that something is going wrong there, while dump process reading certain blocks on disk at the same time they're being flushed back from cache. This is just a theory, I find it unlikely that all databases (on different servers) have faulty indexe on the same object. We tried to browse through the application error log (Arbor BP), but nothing was mentioned there.

    Thanks for your time



    Originally posted by willy_and_the_ci
    Update stats wouldn't have anything to do with this error. It simply updates 2 system tables.

    If you notice from your error, the problem is reported on a user table object id = 1314103722. Run TABLEALLOC on this table, the problem seems to be on non-clustered indexes.

    You should do this on the source database. Fix the corruptions and dump it again.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •