Uploaded image for project: 'ActiveMQ Classic'
  1. ActiveMQ Classic
  2. AMQ-481

BrokerService out of memory error - possible memory leak

Details

    • Bug
    • Status: Resolved
    • Major
    • Resolution: Cannot Reproduce
    • 3.2.1, 4.0
    • 4.0 M4
    • None
    • None

    Attachments

      Activity

        shwang shawn hwang added a comment -

        Files necessary for reproducing the bug have been attached. Files depend on ActiveMQ M3 and log4j to build and run. TestProducer and TestConsumer are the classes with main() methods to run. Run the broker service "out of the box" with a journaled Derby DB instance. Let both the producer and consumer run, see memory usage creep up. Eventually, broker service with quit with an "out of Java heap space" error.

        shwang shawn hwang added a comment - Files necessary for reproducing the bug have been attached. Files depend on ActiveMQ M3 and log4j to build and run. TestProducer and TestConsumer are the classes with main() methods to run. Run the broker service "out of the box" with a journaled Derby DB instance. Let both the producer and consumer run, see memory usage creep up. Eventually, broker service with quit with an "out of Java heap space" error.
        shwang shawn hwang added a comment -

        Actually, I'm only getting the out of memory error on 4.0 M3. 3.2.1 seems to work fine when I self-throttle the number of messages produced.

        shwang shawn hwang added a comment - Actually, I'm only getting the out of memory error on 4.0 M3. 3.2.1 seems to work fine when I self-throttle the number of messages produced.
        jlim Jonas B. Lim added a comment -

        Unable to replicate the issue using the latest snapshot from svn. I ran the test for an hour and the memory usage was normal. I'm running it on a Pentium M 1.73G, 1 GB machine.

        There was a few bugs that were fixed after M3 release that might have already fixed this issue (ie. http://forums.logicblaze.com/posts/list/69.page).

        jlim Jonas B. Lim added a comment - Unable to replicate the issue using the latest snapshot from svn. I ran the test for an hour and the memory usage was normal. I'm running it on a Pentium M 1.73G, 1 GB machine. There was a few bugs that were fixed after M3 release that might have already fixed this issue (ie. http://forums.logicblaze.com/posts/list/69.page ).

        People

          jlim Jonas B. Lim
          jlim Jonas B. Lim
          Votes:
          0 Vote for this issue
          Watchers:
          2 Start watching this issue

          Dates

            Created:
            Updated:
            Resolved: