Skip to content

Random build failures of the GlassFish project

Summary

For some 2 weeks we experience instability of GlassFish builds.

Note: The build first compiles and installs all artifacts without a test execution. Tests are executed later.

Steps to reproduce

Create a PR or rerun some build.

What is the current bug behavior?

  • Build starts - sometimes passes, sometimes fails. Every time in different part(s) (some are more probable, some less)
  • Once it crashed while it was cloning a github repo
  • Several times in Ant tests
  • Several times in Maven tests
  • Several times while compiling sources or generating javadoc jar file.
  • Builds are now also slower, and I believe it is not caused by changes in the project as we don't see it in local builds.

What is the expected correct behavior?

Reproducible builds.

Relevant logs and/or screenshots

Priority

  • Urgent
  • High
  • Medium
  • Low

Severity

  • Blocker
  • Major
  • Normal
  • Low

Impact

  • We have to rerun the build several times for nearly every PR
  • Builds are not reproducible - every time crashes different part.

Related Questions

  • Last year OmniFish joined Eclipse Foundation and we assigned our resource packs to GlassFish.
  • How many resource packs is currently assigned to GlassFish?
  • Where can I find some records about it or even be notified about changes and possible consequences?
  • Did something change in January 2023?
  • What else? Is there something we could do?
Edited by David Matějček