Collecting MustGather data for Out of Memory issues

Prepare the following MustGather information before you call HCL Support to help with Out of Memory issues.

Before you begin

Ensure that you have enabled the Health Center for the containers that you want to monitor. See Setting up IBM Health Center for your production environment.

You can collect core dump and heap dump data by killing the running server process using the kill -3 command.

Procedure

  1. Enable verbose garbage collection (verboseGC) in WebSphere Application Server.
  2. Enable verbose garbage collection (verboseGC) in WebSphere Liberty.
  3. Reproduce the problem.

    Record the specific steps that you used to reproduce the problem and any details that might be relevant.

  4. Collect the following files.
    • In the Transaction server Docker container:
      • /opt/WebSphere/AppServer/profiles/default/core*.dmp
      • /opt/WebSphere/AppServer/profiles/default/heapdump*.phd
      • /opt/WebSphere/AppServer/profiles/default/javacore*.txt
      • /opt/WebSphere/AppServer/profiles/default/Snap*.trc
      • /opt/WebSphere/AppServer/profiles/default/logs/server1/*.log
    • In the Search server Docker container, Store server Docker container, Customization server Docker container:
      • /opt/WebSphere/Liberty/usr/servers/default/core*.dmp
      • /opt/WebSphere/Liberty/usr/servers/default/heapdump*.phd
      • /opt/WebSphere/Liberty/usr/servers/default/javacore*.txt
      • /opt/WebSphere/Liberty/usr/servers/default/Snap*.trc
      • /opt/WebSphere/Liberty/usr/servers/default/logs/messages.log
      • /opt/WebSphere/Liberty/usr/servers/default/logs/console.log