-
-
Notifications
You must be signed in to change notification settings - Fork 32
Open
Description
Our build is currently plagued by tests that occasionally fail when they run via github actions, but don't seem to fail when running locally. My strong suspicion is that the difference lies in the availability of processing cores and stalled threads are not reviving before timeouts in the tests. This is often resolvable by simply re-running the workflow, but that's annoying and leaves us looking like there are problems. This issue will collect and count the various failures. I have an idea for a solution that will first show up in the test I am writing now for SendToOpenSearchProcessor but for now I'm keeping track here in case the pattern sheds additional light.
This table will track counts, comments will document occurances
| Test | Count |
|---|---|
| org.jesterj.ingest.indexing.JdbcScannerImplFTITest > testScanWithMemory FAILED | 6 |
| org.jesterj.ingest.indexing.HeuristicFatalFTITest > test3errorsMarkedDead | 3 |
| org.jesterj.ingest.scanners.SimpleFileScannerImplTest > testFilter FAILED | 1 |
Reactions are currently unavailable
Metadata
Metadata
Assignees
Labels
No labels