• Icon: Task Task
    • Resolution: Fixed
    • Icon: Major Major
    • core, monitoring-plugin
    • None
    • Windows 2008

      Not sure is this a bug or not.
      In this chart http://localhost/monitoring?part=graph&graph=httpSessions, I see the HTTP session count is increased from 0 to 100,110,120... after restart. But in this page http://localhost/monitoring?part=sessions I only see less than 10 sessions.
      I think the data of the 2 charts should be aligned, is it correct?

      Btw, for the HTTP session is always increased, I think it's an issue and just post in community to get help http://jenkins-ci.361315.n4.nabble.com/HTTP-session-increases-and-out-of-memory-td4680122.html

          [JENKINS-20532] HTTP session count

          I see the same issue in our 1.539 instance.
          What's worse is that I see a significant increase in sessions before the service becomes unresponsive and has to be restarted.

          Currently we have ~200 sessions in the counter, but only ~12 sessions listed.

          Klaus Schniedergers added a comment - I see the same issue in our 1.539 instance. What's worse is that I see a significant increase in sessions before the service becomes unresponsive and has to be restarted. Currently we have ~200 sessions in the counter, but only ~12 sessions listed.

          evernat added a comment -

          The issue is probably related to the change from Winstone to Jetty8, in Jenkins v1.535:
          http://jenkins-ci.org/changelog#v1.535

          I have reproduced the issue locally after a load test (the sessions chart shows a high number of sessions, but there are only one session in the list of sessions).
          In the report http://localhost/monitoring?part=heaphisto or in a heap dump, I can also see that there is exactly the same high number of instances of org.eclipse.jetty.server.session.HashedSession existing in memory, as in the chart.

          The cause can be related to the following jetty bug (fixed in Jetty 9.1, which is later than Jetty 8 used by Jenkins):
          https://bugs.eclipse.org/bugs/show_bug.cgi?id=413019

          So the invalidated http sessions are not garbage collected, there is a small memory leak in the monitoring plugin, and the sessions chart is "not right".
          I will fix that, with a workaround for the jetty bug.

          evernat added a comment - The issue is probably related to the change from Winstone to Jetty8, in Jenkins v1.535: http://jenkins-ci.org/changelog#v1.535 I have reproduced the issue locally after a load test (the sessions chart shows a high number of sessions, but there are only one session in the list of sessions). In the report http://localhost/monitoring?part=heaphisto or in a heap dump, I can also see that there is exactly the same high number of instances of org.eclipse.jetty.server.session.HashedSession existing in memory, as in the chart. The cause can be related to the following jetty bug (fixed in Jetty 9.1, which is later than Jetty 8 used by Jenkins): https://bugs.eclipse.org/bugs/show_bug.cgi?id=413019 So the invalidated http sessions are not garbage collected, there is a small memory leak in the monitoring plugin, and the sessions chart is "not right". I will fix that, with a workaround for the jetty bug.

          Dennys Hsieh added a comment -

          Since this is a Jetty's bug, does it mean that Jenkins has memory leak issue?

          Dennys Hsieh added a comment - Since this is a Jetty's bug, does it mean that Jenkins has memory leak issue?

          evernat added a comment -

          @Dennys
          No, just the monitoring plugin

          evernat added a comment - @Dennys No, just the monitoring plugin

          evernat added a comment - - edited

          attached chart from related issue JENKINS-20640

          evernat added a comment - - edited attached chart from related issue JENKINS-20640

          evernat added a comment -

          This is now fixed in the monitoring plugin v1.48, which was just published.
          The monitoring plugin v1.48 should be available in the update center of your Jenkins, or from http://mirrors.jenkins-ci.org/plugins/monitoring/, in a few hours (24h max?).

          evernat added a comment - This is now fixed in the monitoring plugin v1.48, which was just published. The monitoring plugin v1.48 should be available in the update center of your Jenkins, or from http://mirrors.jenkins-ci.org/plugins/monitoring/ , in a few hours (24h max?).

          Dennys Hsieh added a comment -

          After upgrading to 1.48, it's fixed, thanks.

          Dennys Hsieh added a comment - After upgrading to 1.48, it's fixed, thanks.

            evernat evernat
            dennys Dennys Hsieh
            Votes:
            0 Vote for this issue
            Watchers:
            3 Start watching this issue

              Created:
              Updated:
              Resolved: