-
Notifications
You must be signed in to change notification settings - Fork 1.4k
Description
I just noticed that we have a huge number of error logs in our e2e tests that look like this
E1115 09:40:33.696629 1 controller.go:474] "Reconciler error" err="failed to discover variables for ClusterClass in-memory-1: failed to call DiscoverVariables for patch test-patch: failed to call extension handler "discover-variables.scale": failed to get extension handler "discover-variables.scale" from registry: handler with name "discover-variables.scale" has not been registered" controller="clusterclass" controllerGroup="cluster.x-k8s.io" controllerKind="ClusterClass" ClusterClass="scale/in-memory-1" namespace="scale" name="in-memory-1" reconcileID="5ab76484-5d97-4d73-a2d2-7a86507f5d18"
https://storage.googleapis.com/kubernetes-ci-logs/logs/periodic-cluster-api-e2e-main/1989612616221200384/artifacts/clusters/bootstrap/logs/capi-system/capi-controller-manager/capi-controller-manager-5cdf4c75bc-8zr27/manager.log
I think the root cause is that the scale e2e test () is not cleaning up the test namespace and its ClusterClasses at the end of the test. We should probably do that
(xref:
Lines 519 to 528 in d897aa3
| AfterEach(func() { | |
| if !input.SkipCleanup { | |
| if input.ExtensionServiceNamespace != "" && input.ExtensionServiceName != "" { | |
| Eventually(func() error { | |
| return input.BootstrapClusterProxy.GetClient().Delete(ctx, &runtimev1.ExtensionConfig{ObjectMeta: metav1.ObjectMeta{Name: input.ExtensionConfigName}}) | |
| }, 10*time.Second, 1*time.Second).Should(Succeed(), "Deleting ExtensionConfig failed") | |
| } | |
| } | |
| cancelWatches() | |
| }) |