-
Bug
-
Resolution: Unresolved
-
Normal
-
None
-
4.17, 4.18
-
Moderate
-
None
-
NE Sprint 260, NE Sprint 261, NE Sprint 262
-
3
-
Rejected
-
False
-
-
-
Bug Fix
-
In Progress
Description of problem:
Reviewing some 4.17 cluster-ingress-operator logs, I found many (215) of these even when the GatewayAPI feature was disabled: 024-09-03T08:20:03.726Z INFO operator.gatewayapi_controller controller/controller.go:114 reconciling {"request": {"name":"cluster"}} This makes it look like the feature was enabled when it was not. Also check for same in the other gatewayapi controllers in the gateway-service-dns and gatewayclass folders. A search for r.config.GatewayAPIEnabled should show where we are checking whether the feature is enabled.
Version-Release number of selected component (if applicable):
4.17, 4.18 should have the fix
How reproducible:
This was showing up in the CI logs for the e2e-vsphere-ovn-upi test: https://gcsweb-ci.apps.ci.l2s4.p1.openshiftapps.com/gcs/test-platform-results/logs/periodic-ci-openshift-release-master-nightly-4.17-e2e-vsphere-ovn-upi/1830872105227390976/artifacts/e2e-vsphere-ovn-upi/gather-extra/artifacts/pods/openshift-ingress-operator_ingress-operator-64df7b9cd4-hqkmh_ingress-operator.log It is probably showing up in all logs to varying degrees.
Steps to Reproduce:
1. Deploy 4.17 2. Review cluster-ingress-operator logs
Actual results:
Seeing a log that makes it look like the GatewayAPI feature is enabled even when it is not.
Expected results:
Only see the log when the GatewayAPI feature is enabled.
Additional info:
The GatewayAPI feature is enabled in the e2e-aws-gatewayapi PR test and any techpreview PR test, and can be manually enabled on a test cluster by running: oc patch featuregates/cluster --type=merge --patch='{"spec":{"featureSet":"CustomNoUpgrade","customNoUpgrade":{"enabled":["GatewayAPI"]}}}'
- is caused by
-
NE-1273 Add a watch to the ingress operator so it will recreate the gwapi crds
- Closed
- links to