Results 1 to 9 of 9
  1. #1
    Join Date
    Jun 2009
    Posts
    272

    Unanswered: Instance coming down instantaneously

    Environment: DB2 9.7 FP3a
    DPF environment on linux
    17 partitions spanned across 1 physical node:

    Whenever I issue db2start, db2 is coming up and with in seconds it is coming down.

    The only error message I see in db2diag.log is

    Reason code for abrupt termination of system controller: 6

    Any ideas ??


    See below, I issued db2start, it came up and after 2 seconds it came down.


    N:/db2home/arcdwq4> db2start
    08/18/2011 23:43:32 0 0 SQL1063N DB2START processing was successful.
    08/18/2011 23:43:32 1 0 SQL1063N DB2START processing was successful.
    08/18/2011 23:43:32 2 0 SQL1063N DB2START processing was successful.
    08/18/2011 23:43:32 3 0 SQL1063N DB2START processing was successful.
    08/18/2011 23:43:32 4 0 SQL1063N DB2START processing was successful.
    08/18/2011 23:43:32 5 0 SQL1063N DB2START processing was successful.
    08/18/2011 23:43:32 7 0 SQL1063N DB2START processing was successful.
    08/18/2011 23:43:32 8 0 SQL1063N DB2START processing was successful.
    08/18/2011 23:43:32 11 0 SQL1063N DB2START processing was successful.
    08/18/2011 23:43:32 12 0 SQL1063N DB2START processing was successful.
    08/18/2011 23:43:32 14 0 SQL1063N DB2START processing was successful.
    08/18/2011 23:43:33 10 0 SQL1063N DB2START processing was successful.
    08/18/2011 23:43:33 9 0 SQL1063N DB2START processing was successful.
    08/18/2011 23:43:33 6 0 SQL1063N DB2START processing was successful.
    08/18/2011 23:43:33 15 0 SQL1063N DB2START processing was successful.
    08/18/2011 23:43:33 16 0 SQL1063N DB2START processing was successful.
    08/18/2011 23:43:33 13 0 SQL1063N DB2START processing was successful.
    SQL1063N DB2START processing was successful.
    N:/db2home/arcdwq4> ps -ef | grep -i db2sysc
    arcdwq4 1322 1 4 23:43 ? 00:00:00 db2sysc 0
    arcdwq4 1615 1 3 23:43 ? 00:00:00 db2sysc 1
    arcdwq4 1882 1 3 23:43 ? 00:00:00 db2sysc 2
    arcdwq4 2243 1 3 23:43 ? 00:00:00 db2sysc 3
    arcdwq4 2653 1 3 23:43 ? 00:00:00 db2sysc 4
    arcdwq4 3029 1 3 23:43 ? 00:00:00 db2sysc 5
    arcdwq4 3259 1 3 23:43 ? 00:00:00 db2sysc 6
    arcdwq4 3588 1 3 23:43 ? 00:00:00 db2sysc 7
    arcdwq4 4071 1 4 23:43 ? 00:00:00 db2sysc 8
    arcdwq4 4510 1 4 23:43 ? 00:00:00 db2sysc 9
    arcdwq4 4571 1 4 23:43 ? 00:00:00 db2sysc 10
    arcdwq4 5102 1 4 23:43 ? 00:00:00 db2sysc 11
    arcdwq4 5425 1 4 23:43 ? 00:00:00 db2sysc 12
    arcdwq4 5611 1 5 23:43 ? 00:00:00 db2sysc 13
    arcdwq4 5959 1 5 23:43 ? 00:00:00 db2sysc 14
    arcdwq4 6279 1 5 23:43 ? 00:00:00 db2sysc 15
    arcdwq4 6538 1 5 23:43 ? 00:00:00 db2sysc 16
    arcdwq4 6601 19375 0 23:43 pts/12 00:00:00 grep -i db2sysc
    arcdwq4 6603 19375 0 23:43 pts/12 00:00:00 grep -i db2sysc

    N:/db2home/arcdwq4> ps -ef | grep -i db2sysc
    arcdwq4 6624 19375 0 23:43 pts/12 00:00:00 grep -i db2sysc

  2. #2
    Join Date
    Nov 2010
    Posts
    14
    blazer789,

    DB2 log (db2diag.log) may contain additional useful information the event. Have you checked db2diag. There could be an EDU crash logged in the db2diag.log.

    Event logs level of "Severe" and "Error" are the potential events to investigate with.

    Code:
     db2diag -time "2011-08-18-23.43" -level "Severe, Error" | more
    Thanks,
    SK Chandra

  3. #3
    Join Date
    Jun 2009
    Posts
    272
    No specific error messages other than the one I mentioned earlier.

    Reason code for abrupt termination of system controller: 6

  4. #4
    Join Date
    Jun 2009
    Posts
    272
    I even searched all the archived and current diaglogs for any EDU crash error messages but couldn't find any.

    N:/db2home/arcdwq4/sqllib/db2dump> cat db2diag* | grep -i crashed
    N:/db2home/arcdwq4/sqllib/db2dump>

  5. #5
    Join Date
    Nov 2010
    Posts
    14
    Can you give the excerpts of the db2diag.log?

    Code:
    db2diag -time "2011-08-18-23.43.30:2011-08-18-23.43.59"
    Thanks,
    SK Chandra

  6. #6
    Join Date
    May 2003
    Location
    USA
    Posts
    5,737
    With 17 partitions on one physical node, you would need an extremely large amount of memory on that server. For example, when you define bufferpools, each partition (node) tries to allocate that amount, so it would need 17 times the memory defined in syscat.bufferpools (and 17 times the memory for the instance and other db allocations also).
    M. A. Feldman
    IBM Certified DBA on DB2 for Linux, UNIX, and Windows
    IBM Certified DBA on DB2 for z/OS and OS/390

  7. #7
    Join Date
    Jun 2009
    Posts
    272
    Box has enough memory to accomodate all the 17 logical nodes. Memory is not an issue. Even I tried by restarting the instance by setting the registry variable DB2_OVERRIDE_BPF to a lower value. It didn't help. Instance goes down in the same way after it starts up. Very interesting !!! The only errors I see in diaglog are with “Starting FMP failed” and “watch dog died after forking db2sysc”.

  8. #8
    Join Date
    Dec 2013
    Posts
    1
    I had the same problem. Please check your mount point for your db2 instance (eg: /home/db2inst1) at /etc/fstab and verify if it has the suid option

  9. #9
    Join Date
    Jun 2009
    Posts
    272
    Yes, I remember: The issue was with the mount point itself and we had to change the SUID option to get the issue resolved.

Posting Permissions

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