Description of problem: No version for openshift-controller-manager-operator when get clusteroperator Version-Release number of selected component (if applicable): ./openshift-install v0.10.0 payload: quay.io/openshift-release-dev/ocp-release@sha256:66cee7428ba0d3cb983bd2a437e576b2289e7fd5abafa70256200a5408b26644 version: 4.0.0-0.1 How reproducible: always Steps to Reproduce: 1.$ oc get clusteroperator openshift-cluster-samples-operator 4.0.0-alpha1-701a10fea True True True 20h openshift-controller-manager-operator True False False 8m openshift-dns-operator 0.0.1 True False False 22h 2. 3. Actual results: No version for openshift-controller-manager-operator Expected results: Should have version for openshift-controller-manager-operator Additional info:
openshift-apiserver kube-apiserver have no version too
The version field as currently described is flawed. It doesn't allow us to support upgrades. We need https://github.com/openshift/api/pull/176 to allow something like https://docs.google.com/document/d/1YV_rJ6qR46_DV1s6RwobTYHX0CCE4CIBnt_VRPtN8Nw/edit#heading=h.j35cxjc8vibg to allow upgrades and payloads to mix and match. This won't be changed until that pull merges and even then the actual readable version won't be present until the CVO gets updated.
The master team pieces 1. kube-apiserver 2. kube-controller-manager 3. openshift-apiserver report versions. The openshift-controller-manager is actually owned by bparees. I suspect the Build component is the best fit.
@ben I saw master team had version of clusteroperator like[1],but openshift-controller-manager clusteroperator doesn't have versions[2], could you help to check it？ [1] $ oc describe clusteroperator kube-apiserver Name: kube-apiserver Namespace: Labels: <none> Annotations: <none> API Version: config.openshift.io/v1 Kind: ClusterOperator Metadata: Creation Timestamp: 2019-02-22T02:29:36Z Generation: 1 Resource Version: 16235 Self Link: /apis/config.openshift.io/v1/clusteroperators/kube-apiserver UID: b265d684-3649-11e9-993e-02be42712cee Spec: Status: Conditions: Last Transition Time: 2019-02-22T02:45:45Z Status: False Type: Failing Last Transition Time: 2019-02-22T02:45:54Z Status: False Type: Progressing Last Transition Time: 2019-02-22T02:30:18Z Status: True Type: Available Last Transition Time: 2019-02-22T02:29:36Z Status: True Type: Upgradeable Extension: <nil> Related Objects: Group: operator.openshift.io Name: cluster Resource: kubeapiservers Group: Name: openshift-config Resource: namespaces Group: Name: openshift-config-managed Resource: namespaces Group: Name: openshift-kube-apiserver-operator Resource: namespaces Group: Name: openshift-kube-apiserver Resource: namespaces Versions: ####had versions Name: operator Version: 0.0.0_version_cluster-kube-apiserver-operator Name: kube-apiserver Version: 0.0.0_version_hypershift Events: <none> [2]$ oc describe clusteroperator openshift-controller-manager Name: openshift-controller-manager Namespace: Labels: <none> Annotations: <none> API Version: config.openshift.io/v1 Kind: ClusterOperator Metadata: Creation Timestamp: 2019-02-22T02:39:05Z Generation: 1 Resource Version: 12148 Self Link: /apis/config.openshift.io/v1/clusteroperators/openshift-controller-manager UID: 0575c150-364b-11e9-94e8-0615dd2afa9e Spec: Status: Conditions: Last Transition Time: 2019-02-22T02:39:05Z Status: False Type: Failing Last Transition Time: 2019-02-22T02:39:38Z Status: True Type: Available Last Transition Time: 2019-02-22T02:41:25Z Status: False Type: Progressing Extension: <nil> Related Objects: Versions: <nil> ### no vesions Events: <none>
that's correct. to my knowledge we are still waiting for a finalization of the versioning plan (last i heard there was going to be some revisiting of what david implemented for the control plane operators, and the payload bits of that implementation weren't in place anyway). While the openshift apiserver has a version value, it's not a valid/meaningful one (0.0.0_version_cluster-kube-apiserver-operator) that said, since the operators are owned by different teams, it would make sense to clone this bug, one for david and one for me, to ensure that we get real versions into both operators before we ship.
Thanks ben quick response, I reported one(https://bugzilla.redhat.com/show_bug.cgi?id=1679871) for you, and let xxia report another one for david which is for master team. *** This bug has been marked as a duplicate of bug 1679871 ***