strimzi-kafka-operator
strimzi-kafka-operator copied to clipboard
strimzi
fix: Update TopicOperatorTest to handle async flow correctly
Type of change
Select the type of your PR
- Bugfix
Description
Update the TopicOperatorTest to fix problems with asynchronous tests being completed early before all assertions are run. This fixes issue #7072.
Main changes are:
- Update the resourceAdded and similar functions so they return a future, rather than creating and flagging a checkpoint
- Change checkpoints to context.completeNow() where possible
- Remove the use of CountdownLatch in favour of vertx future composition
Checklist
Please go through this checklist and make sure all applicable tasks have been done
- [ ] Write tests
- [ ] Make sure all tests pass
- [ ] Update documentation
- [ ] Check RBAC rights for Kubernetes / OpenShift roles
- [ ] Try your changes from Pod inside your Kubernetes and OpenShift cluster, not just locally
- [ ] Reference relevant issue(s) and close them after merging
- [ ] Update CHANGELOG.md
- [ ] Supply screenshots for visual changes, such as Grafana dashboards
This change is complete, but as you can see three of the tests fail. These are:
- testReconcile_withResource_withKafka_noPrivate_matching
- testReconcileMetricsWithPausedTopic
- testReconcileMetricsDeletedTopic
The first one the error is that the expected metrics for the topic cannot be found. These assertions weren't being run previously due to the checkpoint being flagged before the end of the test.
The second two are both failing within the addResource or deleteResource functions and are complaining that the topic isn't in the expected state. Previously these functions ran in parallel with the reconcile call so it is unclear what order they are supposed to be in to pass correctly. If I comment out the specific assertions that are failing, the others assertions in those tests do pass.
