-
Bug
-
Resolution: Done
-
Blocker
-
Pipelines 1.16.0
-
None
-
False
-
None
-
False
-
-
Description of problem:
Event listener created with 1.15.1 is crashing after operator is upgraded to 1.16.0
Workaround
Prerequisites (if any, like setup, operators/versions):
OpenShift 4.15-4.18
OpenShift Pipelines 1.15.1 -> 1.16.0
Steps to Reproduce
# Install Pipelines operator 1.15.1
- gauge run --log-level=debug --verbose --tags pre-upgrade specs/operator/pre-upgrade.spec
- update operator to 1.16.0
- gauge run --log-level=debug --verbose --tags post-upgrade specs/operator/post-upgrade.spec:60
Actual results:
Test fails with the following message
Specification: specs/operator/post-upgrade.spec:61 Error Message: sink did not return 2xx response. Got status code: 503
which is actually caused by crashing event listener
$ oc get pod -n releasetest-upgrade-tls NAME READY STATUS RESTARTS AGE el-listener-embed-binding-56fd69b487-95pkt 0/1 CrashLoopBackOff 7 (2m51s ago) 10m el-listener-embed-binding-597497f998-5ltpk 0/1 CrashLoopBackOff 64 (4m56s ago) 3h14m
pod logs:
W0925 15:01:18.175142 1 reflector.go:539] knative.dev/pkg/controller/controller.go:732: failed to list *v1alpha1.ClusterInterceptor: clusterinterceptors.triggers.tekton.dev is forbidden: User "system:serviceaccount:releasetest-upgrade-tls:pipeline" cannot list resource "clusterinterceptors" in API group "triggers.tekton.dev" at the cluster scope E0925 15:01:18.175174 1 reflector.go:147] knative.dev/pkg/controller/controller.go:732: Failed to watch *v1alpha1.ClusterInterceptor: failed to list *v1alpha1.ClusterInterceptor: clusterinterceptors.triggers.tekton.dev is forbidden: User "system:serviceaccount:releasetest-upgrade-tls:pipeline" cannot list resource "clusterinterceptors" in API group "triggers.tekton.dev" at the cluster scope
Expected results:
test pass
Reproducibility (Always/Intermittent/Only Once):
always
Acceptance criteria:
Definition of Done:
Build Details:
Additional info (Such as Logs, Screenshots, etc):