Flaky fuzzing failures in CI #11810
Description
Sometimes I see a report that the fuzzing CI job failed, e.g.: https://github.com/prometheus/prometheus/actions/runs/3840368422/jobs/6539337120
It says:
2023-01-04 18:04:26,855 - root - INFO - Fuzzer: fuzzParseMetricSelector. Detected bug.
2023-01-04 18:04:26,856 - root - INFO - Trying to reproduce crash using: /tmp/tmpdw5piina/tmplxscns1i.
2023-01-04 18:04:29,[52](https://github.com/prometheus/prometheus/actions/runs/3840368422/jobs/6539337120#step:5:53)9 - root - INFO - Reproduce command returned: 1. Reproducible on /github/workspace/build-out/fuzzParseMetricSelector.
2023-01-04 18:04:29,[53](https://github.com/prometheus/prometheus/actions/runs/3840368422/jobs/6539337120#step:5:54)0 - root - INFO - Crash is reproducible.
However it does not say what the crash was. This does not seem very useful.
It also says:
Container for artifact "artifacts" successfully created. Starting upload of file(s)
Total size of all the files uploaded is 195 bytes
I cannot see where this file went.
The same job for the same PR passed on retry, hence I describe it as flaky.