-
Story
-
Resolution: Done
-
Undefined
-
None
-
None
-
Quality / Stability / Reliability
-
False
-
-
False
-
None
-
None
-
None
-
None
-
None
HIVE-2416 introduces some extra status updates in the machinepool controller, increasing the probability of 409 Conflict errors when managing a MachinePool, particularly programmatically. To wit, e2e for that card's PR had one such failure:
=== RUN TestScaleMachinePool
time="2025-03-26T17:11:35Z" level=info msg="expected Machine name prefix: hiveci-67e42826-a44-54t2j-worker-" pool=hiveci-67e42826-a44-worker test=TestScaleMachinePool
time="2025-03-26T17:11:35Z" level=info msg="scaling pool to 1 replicas" pool=hiveci-67e42826-a44-worker test=TestScaleMachinePool
infra_test.go:67:
Error Trace: /go/src/github.com/openshift/hive/test/e2e/postinstall/machinesets/infra_test.go:67
Error: Received unexpected error:
Operation cannot be fulfilled on machinepools.hive.openshift.io "hiveci-67e42826-a44-worker": the object has been modified; please apply your changes to the latest version and try again
Test: TestScaleMachinePool
Messages: cannot update worker machine pool to reduce replicas
--- FAIL: TestScaleMachinePool (0.19s)
Refactor that suite to a) better avoid, and b) retry upon encountering 409s.
