Skip to main content
Warning This document has not been updated for a while now. It may be out of date.
Last updated: 5 Apr 2022

Troubleshoot CI Jenkins Performance

There are many reasons Jenkins could perform poorly, this document attempts to outline some common approaches to diagnosing/monitoring and outlines steps to resolve common issues.

A useful link is /monitoring which shows a variety of diagnostic information. This can be used to identify which aspects of Jenkins performance/usage have changed over periods of time.

There are resources Jenkins was not able to dispose automatically

This is an error that can be seen on the /manage page in Jenkins and indicates that the jenkins user does not have sufficient permissions to delete items it is trying to. When lots of these build up the load can increase enormously and make Jenkins near unusuable.

A common culprit in this problem is builds that make use of Docker. As Docker runs as a root user any files Docker creates on volumes that are mounted have a root owner. Jenkins then cannot delete these files as part of a workspace cleanup.

To remove these manually you can log into each of the agent machines and perform the following commands:

sudo find /var/lib/jenkins/workspace/*ws-cleanup*/ -user root will show a list of files that will be deleted, you should scan these to check there is nothing surprising.

Then sudo find /var/lib/jenkins/workspace/*ws-cleanup*/ -user root -delete can be run to delete everything that was found.

You should then look to ensure that the build which is leaving these files is not doing it in the course of a regular build. However even if a job does clean up after itself some of these may still exist if a job is aborted prematurely.