adoroszlai opened a new pull request, #7483:
URL: https://github.com/apache/ozone/pull/7483
## What changes were proposed in this pull request?
Background:
- CI checks collect problems that caused the failure into a report file
(usually `summary.txt`). This file is shown in the jobs' `Summary of failures`
step.
- Checks are marked as failed if the report file is not empty.
- There may be failures that affect the check, but are not specific to its
validation. E.g. findbugs may fail due to compile error. This is not
reflected in the report file, so the check may be passing despite compile
errors.
On current `master`, Kubernetes check runs into error:
```
RuntimeException: Failed to set directory permissions for /data: /data:
Operation not permitted
at
org.apache.hadoop.hdds.server.ServerUtils.getDirectoryFromConfig(ServerUtils.java:188)
at
org.apache.hadoop.hdds.utils.db.DBDefinition.getDBLocation(DBDefinition.java:58)
at
org.apache.hadoop.hdds.utils.db.DBStoreBuilder.getDBDirPath(DBStoreBuilder.java:169)
at
org.apache.hadoop.ozone.container.metadata.WitnessedContainerMetadataStoreImpl.get(WitnessedContainerMetadataStoreImpl.java:45)
at
org.apache.hadoop.ozone.container.ozoneimpl.OzoneContainer.<init>(OzoneContainer.java:192)
at
org.apache.hadoop.ozone.container.common.statemachine.DatanodeStateMachine.<init>(DatanodeStateMachine.java:183)
at
org.apache.hadoop.ozone.HddsDatanodeService.start(HddsDatanodeService.java:297)
```
which was introduced in HDDS-11650, but went unnoticed due to the false
"passing" status.
This change ensures that check status reflects errors:
1. Add common post-processing step in check scripts to make sure
`summary.txt` is not left empty if any failures are encountered.
2. Remove `continue-on-error` from main step of checks. Previously this was
needed to make GitHub UI auto-open the `Summary of failures` step instead of
`Execute tests`.
It also fixes the permission problem in `kubernetes` check
(e09ba8f228838f15f857b5fcf0e667abfc91d34b).
https://issues.apache.org/jira/browse/HDDS-11718
## How was this patch tested?
Verified that `kubernetes` check is failing before applying the fix for the
permission problem:
https://github.com/adoroszlai/ozone/actions/runs/12012294888/job/33484088312
Temporarily introduced compile error, verified `findbugs` (as well as
`build` and `checkstyle`) fails:
https://github.com/adoroszlai/ozone/actions/runs/12012941631/job/33485335513
Regular CI:
https://github.com/adoroszlai/ozone/actions/runs/12012816442
Kubernetes is passing (without failures):
```
2024-11-25 15:19:38 INFO SCMSafeModeManager:248 - SCM exiting safe mode.
...
Basic Freon smoketest | PASS
|
...
```
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
