Case Study: Jenkins speeds up CI failure debugging with Undo Live Recorder

A Undo Case Study

Preview of the Jenkins Case Study

Jenkins - Customer Case Study

Jenkins, the open-source continuous integration system, needed a faster way to diagnose failing tests, especially intermittent and hard-to-reproduce issues that often left developers with only a terse error report. Undo’s Live Recorder was used to capture failing test runs so engineers could inspect the exact execution leading up to the failure.

Undo added a small change to the test runner so any failed test would be re-run and recorded to an UndoDB trace file. With Undo’s recording and reverse-debugging, developers could replay the crash, step backward through execution, and identify the root cause quickly: a missing `_PyRandom_Init()` call before command-line parsing in cpython. The result was a much faster path to resolution for CI failures, with the case study highlighting reduced debugging effort and improved handling of both reproducible and intermittent bugs.


Open case study document...

Jenkins

Kohsuke Kawaguchi

Founder


Undo

5 Case Studies