Repository navigation
MINOR: [CI] Upload JVM crash files when the Binaries job fails - #1428
Merged
Merged
Conversation
A crashed test JVM leaves hs_err_pid<pid>.log in the module directory and its crash banner in surefire's .dumpstream files, and the job uploaded neither, so a fork crash could not be diagnosed from CI.
andygrove
requested review from
jbonofre,
kou,
laurentgo,
lidavidm and
wgtmac
as code owners
November 1, 2026 20:34
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
What's Changed
When a test JVM crashes in the RC workflow's Binaries job, the log says only
Error occurred in starting fork, check output in log,Corrupted channel by directly writing to native stream in forked JVM NandAborted (core dumped). The details stay on the runner. HotSpot writeshs_err_pid<pid>.logto the module directory, and surefire saves the crash banner the JVM printed to stdout intarget/surefire-reports/*.dumpstream. The job uploads neither.For example, the Binaries job on #1419 failed twice this way (attempt 1, attempt 2). All 46 arrow-dataset tests passed, then the JVM that had run 35 of them aborted, and nothing in the log shows where it crashed.
This adds a step that uploads those files as a
binaries-jvm-crash-filesartifact when the job fails. It uses the same pinnedactions/upload-artifactas the job's other upload steps.if-no-files-found: ignorekeeps the step quiet when a failure isn't a crash.I checked the patterns against a local arrow-dataset crash, which left
dataset/hs_err_pid<pid>.loganddataset/target/surefire-reports/<timestamp>-jvmRun<N>.dumpstream. actionlint reports nothing new inrc.yml. The step runs only when the job fails, so this PR's own CI run won't exercise it.