Results 1 to 7 of 7
  1. #1
    Join Date
    Aug 2001
    Location
    UK
    Posts
    4,650

    Unanswered: Reorg & Backup - Reason for -911

    DB2 8.2 on AIX
    Online database backup fails.(SQLCODE -911, Reason Code 68) if an INPLACE REORG is in progress on a table in the database.
    Curious to know what the conflicting lock is here, that causes a timeout? Any idea?

    If it makes a difference, the table is in SMS tablespace.

    Thanks
    Visit the new-look IDUG Website , register to gain access to the excellent content.

  2. #2
    Join Date
    Jan 2003
    Posts
    4,292
    Provided Answers: 5
    I have seen the same thing on V8.2 for linux. Maybe if you examined which locks the reorg is holding, you might be able to figure it out.

    Let us know.

    Andy

  3. #3
    Join Date
    Aug 2001
    Location
    UK
    Posts
    4,650
    Yep, I did try lock snapshots before posting here. I could not see anything useful.
    I'll post some snapshots here tomorrow

    Thanks
    Visit the new-look IDUG Website , register to gain access to the excellent content.

  4. #4
    Join Date
    Aug 2001
    Location
    UK
    Posts
    4,650
    Here's the diaglog. I have also attached the lock snapshot, taken about 5-6 seconds before the timeout.

    The locktimeout value on this database is 15 secs.

    I'm not sure which ones are actually causing the timeout. any clues? thanks.


    2007-11-30-14.28.51.120707+060 I267738C429 LEVEL: Error
    PID : 1257700 TID : 1 PROC : db2bm.1417402.3 0
    INSTANCE: pqrs_t NODE : 000
    APPHDL : 0-221
    FUNCTION: DB2 UDB, database utilities, SqlubTableLock::request, probe:81
    DATA #1 : String, 45 bytes
    Error getting lock.
    PoolId: 2, ObjectId: 16
    DATA #2 : Hexdump, 4 bytes
    0x2FF1EEE0 : 8010 0044 ...D


    2007-11-30-14.28.51.121746+060 I268168C370 LEVEL: Severe
    PID : 1257700 TID : 1 PROC : db2bm.1417402.3 0
    INSTANCE: pqrs_t NODE : 000
    APPHDL : 0-221
    FUNCTION: DB2 UDB, database utilities, sqlubreadsms, probe:30
    MESSAGE : DiagData
    DATA #1 : Hexdump, 8 bytes
    0x2005A400 : 8010 0044 0000 0044 ...D...D


    2007-11-30-14.28.51.121889+060 I268539C350 LEVEL: Error
    PID : 1257700 TID : 1 PROC : db2bm.1417402.3 0
    INSTANCE: pqrs_t NODE : 000
    APPHDL : 0-221
    FUNCTION: DB2 UDB, database utilities, sqlubreadsms, probe:30
    DATA #1 : Hexdump, 8 bytes
    0x2005A400 : 8010 0044 0000 0044 ...D...D


    2007-11-30-14.28.52.541576+060 I268890C358 LEVEL: Severe
    PID : 1417402 TID : 1 PROC : db2agent (pqrs_T) 0
    INSTANCE: pqrs_t NODE : 000 DB : pqrs_T
    APPHDL : 0-221 APPID: *LOCAL.pqrs_t.071130132818
    FUNCTION: DB2 UDB, database utilities, sqlubcka, probe:140
    MESSAGE : Backup Terminated.
    Attached Files Attached Files
    Visit the new-look IDUG Website , register to gain access to the excellent content.

  5. #5
    Join Date
    Jan 2003
    Posts
    4,292
    Provided Answers: 5
    I am not sure what is causing the problem either. It looks like the reorg has a bunch of Intent-None (IN) row locks. The only thing that this would prevent is a super-exclusive (Z) lock, and I do not see how the online backup would try that.

    The only thing I can suggest is to monitor the lock chain that occurs during the lock wait. This should give you a better insight as to what is causing the problem. Other than that, you can open a PMR with IBM to get an answer.

    Andy

  6. #6
    Join Date
    Aug 2001
    Location
    UK
    Posts
    4,650
    Thanks Andy. The issue here is that I do not see the backup in lock-wait state at all :-)
    I guess I'll ask the client staff to raise a PMR
    Thanks
    Visit the new-look IDUG Website , register to gain access to the excellent content.

  7. #7
    Join Date
    Jan 2003
    Posts
    4,292
    Provided Answers: 5
    The problem of addressing a lockwait timeout issue is catching the lockwait. You have to constantly keep snapshot monitoring the lock situation with an interval between snapshots less than the lockwait timeout value. So, if as in your condition, the lockwait timeout value is 15 seconds, you have to get the snapshots every 10 seconds or so. This can be very tedious and rather boring if done manually. If you automate it, you will have tons of data to sift through.

    Andy

Posting Permissions

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