You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
This repository has been archived by the owner on Jul 22, 2022. It is now read-only.
Next, I upload the kubeconfig of usercluster as a configmap in the metacluster.
Then, I attach the configmap as a volume to the statefulset of metacontroller and also update the metacontroller container command to add --client-config-path and pass the kubeconfig path to it in metacluster.
Next, I create a CRD "hello-world" in usercluster
I create a compositecontroller for this CRD in metacluster and deploy the webhook in metacluster
With this setup ready, I now try to create a k8s object of type "hello-world" in usercluster
Expectation: Child resources of "hello-world" CRD should spin up in usercluster Observation: No child resources spin up. I see the below errors in metacontoller pod in metacluster : metacontroller.app/client/generated/informer/externalversions/factory.go:117: Failed to list *v1alpha1.ControllerRevision: the server could not find the requested resource (get controllerrevisions.metacontroller.k8s.io) metacontroller.app/client/generated/informer/externalversions/factory.go:117: Failed to list *v1alpha1.DecoratorController: the server could not find the requested resource (get decoratorcontrollers.metacontroller.k8s.io) metacontroller.app/client/generated/informer/externalversions/factory.go:117: Failed to list *v1alpha1.CompositeController: the server could not find the requested resource (get compositecontrollers.metacontroller.k8s.io)
I can see from the logs that the metacontroller pod in metacluster is looking for controllerrevision, decoratorcontroller and compositecontroller in the usercluster instead of the metacluster. From my understanding, it should be looking for it in the metacluster instead.
Can anyone explain how to get this scenario to work? Is my understanding of the change correct?
The text was updated successfully, but these errors were encountered:
I am trying out the changes introduced in db0ecdf for the issue #111
Suppose you have 2 clusters - "metacluster" and "usercluster". Below are the steps I am following :
Expectation: Child resources of "hello-world" CRD should spin up in usercluster
Observation: No child resources spin up. I see the below errors in metacontoller pod in metacluster :
metacontroller.app/client/generated/informer/externalversions/factory.go:117: Failed to list *v1alpha1.ControllerRevision: the server could not find the requested resource (get controllerrevisions.metacontroller.k8s.io) metacontroller.app/client/generated/informer/externalversions/factory.go:117: Failed to list *v1alpha1.DecoratorController: the server could not find the requested resource (get decoratorcontrollers.metacontroller.k8s.io) metacontroller.app/client/generated/informer/externalversions/factory.go:117: Failed to list *v1alpha1.CompositeController: the server could not find the requested resource (get compositecontrollers.metacontroller.k8s.io)
I can see from the logs that the metacontroller pod in metacluster is looking for controllerrevision, decoratorcontroller and compositecontroller in the usercluster instead of the metacluster. From my understanding, it should be looking for it in the metacluster instead.
Can anyone explain how to get this scenario to work? Is my understanding of the change correct?
The text was updated successfully, but these errors were encountered: