Uploaded image for project: 'Puppet Server'
  1. Puppet Server
  2. SERVER-1368

Graceful OutOfMemoryError handling please?

    XMLWordPrintable

    Details

    • Type: Bug
    • Status: Closed
    • Priority: Major
    • Resolution: Won't Fix
    • Affects Version/s: SERVER 2.4.0
    • Fix Version/s: None
    • Component/s: Puppet Server
    • Labels:
      None
    • Environment:

      RHEL6, Java 8, PuppetDB, PostgreSQL 9.5

    • Template:

      Description

      With the -XX:OnOutOfMemoryError="kill -9 %p" it seems fairly trivial to cause a denial of service... by removing the -XX:OnOutOfMemoryError depending on the cause of the OutOfMemoryError, Java is actually able to recover intelligently as once the request has completed (by reporting the OutOfMemoryError to the puppet-agent) Puppet Server's gc pressure drops back to reasonable levels.

      We are running into an issue with a specific type of nodes that we have, they all cause an OutOfMemoryError during catalog creation, with the -XX:OnOutOfMemoryError="kill -9 %p" the process was just dying with no details on location/thread of where the error occurs.

      I would think the better approach would be to wrap the JRuby pool with OutOfMemoryError catching to recycle the worker automatically instead of just killing the entire JVM.

        Attachments

          Activity

            People

            Assignee:
            Unassigned
            Reporter:
            ruckc Curtis Ruck
            Votes:
            0 Vote for this issue
            Watchers:
            3 Start watching this issue

              Dates

              Created:
              Updated:
              Resolved:

                Zendesk Support