Results 1 to 8 of 8
  1. #1
    Join Date
    Aug 2011
    Posts
    33

    Unanswered: Hang on Database

    We have one Java Virtual machine(appserver) which connects to the database and perform work.This JVM spawns 20 threads to the database to do its work.Now,the real problem is all the 20 threads are being used up and JVM is unable to spawn a new thread and the application is struck and eventually not responding and our website is down.In the database ,when this incident occurs,I am seeing only one or two applications in 'UOW Executing' status.Any idea,what's going on here?

  2. #2
    Join Date
    Jan 2003
    Posts
    4,292
    Provided Answers: 5
    The problem is at the JVM side. Either it is not sharing the connections properly or the users of the connections are not behaving properly.

    Andy

  3. #3
    Join Date
    Aug 2011
    Posts
    33
    Any tests I can perform to rule out DB2 database as culprit?

  4. #4
    Join Date
    Jan 2003
    Posts
    4,292
    Provided Answers: 5
    On the DB2 server, try to connect to the database. If it works then DB2 is not the problem.

    Andy

  5. #5
    Join Date
    Aug 2011
    Posts
    33
    Thankyou...but the issue doesn't occur everytime...it connects to the db well and continues its work for 10 to 12 hrs and then suddenly the hang occurs..I am able to connect ,run sql's take snapshots etc when this issue happens..so you say its not db2.Any idea if db2 server(lpar) not the database may be cuasing this?

  6. #6
    Join Date
    Jan 2003
    Posts
    4,292
    Provided Answers: 5
    If you are able to use the database outside the application while the application is hanged, then it is not the database.

    Andy

  7. #7
    Join Date
    Aug 2011
    Posts
    33
    thankyou..I think that answers my question..appreciate it

  8. #8
    Join Date
    Jan 2007
    Location
    Jena, Germany
    Posts
    2,721
    My guess would be that you probably have a deadlock in your application. It takes some time until the conditions are meat, i.e. the right code is running concurrently in multiple threads to run into the deadlock. That would at least explain the 10-12h.
    Knut Stolze
    IBM DB2 Analytics Accelerator
    IBM Germany Research & Development

Posting Permissions

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