venkateshwaracholan opened a new pull request, #1040:
URL: https://github.com/apache/yunikorn-k8shim/pull/1040
## What is this PR for?
Stabilize the predicates e2e cleanup path for YUNIKORN-3298.
The failures occur in `AfterEach` after namespace teardown:
```text
Name: Negative resources
Message: Nodes with negative resources: ["yk8s-worker"]
```
We don't yet know whether this is a transient reconciliation delay or an
underlying scheduler issue. This change makes the cleanup path wait for
scheduler health to recover instead of assuming it is immediately stable.
## What does this PR do?
* Wait for helper pods to fully terminate after deletion.
* Replace the immediate post-cleanup health check with polling that waits
for:
* non-negative node resources, and
* scheduler health checks to pass.
* Continue to fail if health does not recover within 60 seconds.
## How should this be tested?
```bash
go test ./test/e2e/framework/helpers/yunikorn/... -count=10 -v
```
```bash
make kind-e2e E2E_TEST="--label-filter=TestPredicates"
```
Added tests:
* TestHasNegativeNodeResources
* TestWaitForSchedulerHealth
* TestWaitForSchedulerHealthNegativeNodes
* TestWaitForSchedulerHealthTimeout
Helper tests: Passed (10/10)
## What type of PR is this?
* Test Stabilization
## Jira
https://issues.apache.org/jira/browse/YUNIKORN-3298
### Questions:
* [ ] - The licenses files need update.
* [ ] - There is breaking changes for older versions.
* [ ] - It needs documentation.
--
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]
