ROX-34942: increase collect-service-logs timeout to 30m#21045
Conversation
Reverts the Gradle 8.13 downgrade (PR #20384) and fixes the root cause: 1. junit-platform-launcher runtimeOnly dependency — Gradle 9 removed auto-provisioning. Without this, test tasks fail with "Failed to load JUnit Platform." 2. testClassesDirs + classpath wiring in configureEach — Gradle 9's register<Test> tasks don't inherit the test source set. Without this, custom test tasks (testBAT, testSMOKE, etc.) report NO-SOURCE even after compilation succeeds. Root cause verified locally and on CI: - compileTestGroovy produces class files - testSMOKE/testBAT find and execute tests - 78+ BAT tests pass on KinD Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
…radle 9 The protobuf plugin 0.8.19 uses the deprecated 'convention' API which was removed in Gradle 9. Upgrading to protobuf plugin 0.10+ requires API migration in both build.gradle.kts (sourceSet.java → sourceSet.extensions) and protobuf.gradle (DSL changes). Keep the JUnit Platform and testClassesDirs fixes — they're needed for both Gradle 8 and 9. The Gradle 9 upgrade needs to be paired with the protobuf plugin migration as a separate effort. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
EKS nightly e2e jobs fail with "Post failed: exit 1" when collect-service-logs.sh exceeds 900s on namespaces with many resources. On the last passing run (May 31) post-test took ~40 minutes; on June 7 all tests passed but the job failed due to this timeout. Double the per-namespace COLLECT_TIMEOUT to 30 minutes. Partially generated by AI. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
|
Skipping CI for Draft Pull Request. |
|
Important Review skippedAuto reviews are disabled on this repository. Please check the settings in the CodeRabbit UI or the ⚙️ Run configurationConfiguration used: Organization UI Review profile: CHILL Plan: Enterprise Run ID: You can disable this status message by setting the Use the checkbox below for a quick retry:
✨ Finishing Touches🧪 Generate unit tests (beta)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
🚀 Build Images ReadyImages are ready for commit c447aa9. To use with deploy scripts: export MAIN_IMAGE_TAG=4.12.x-138-gc447aa95cf |
|
/test eks-qa-e2e-tests |
|
@davdhacs: No presubmit jobs available for stackrox/stackrox@davdhacs/gradle9-fix DetailsIn response to this:
Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes-sigs/prow repository. |
|
/test eks-qa-e2e-tests |
|
/test eks-qa-e2e-tests |
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
|
/test eks-qa-e2e-tests |
davdhacs
added
the
auto-retest
msugakov
changed the title
|
/retest |
3 participants
Description
EKS nightly e2e jobs (ROX-34942) fail with
Post failed: exit 1whencollect-service-logs.shexceeds the 900s COLLECT_TIMEOUT on namespaces with many resources. On the last passing run (May 31) post-test log collection took ~40 minutes per part; on June 7 all 850+ tests passed but the job failed because thestackroxnamespace log collection timed out at 15 minutes.Doubles
COLLECT_TIMEOUTfrom 15 to 30 minutes.Stacked on #20429 (Gradle 9 test discovery fix).
Testing and quality
How I validated my change
Analyzed EKS nightly CI runs from May 28 to June 8:
collect-service-logs.sh stackroxtimed out after 900s during Part II post-testThe 900s timeout is too tight for the
stackroxnamespace on EKS. 30 minutes provides headroom without being so long that a genuinely stuck collection blocks the job indefinitely.