oc project not works with cluster proxy.

[wanghaoran1988]

We are using a proxy as the apiserver that can reverse proxy to multiple clusters, so we has config like:

apiVersion: v1
clusters:
- cluster:
    server: http://localhost:8001/backplane/cluster/1h2cu64jqeufe3dnk7gvpl7j3ui3msut/
  name: 1h2cu64jqeufe3dnk7gvpl7j3ui3msut
contexts:
- context:
    cluster: 1h2cu64jqeufe3dnk7gvpl7j3ui3msut
    namespace: openshift-logging
    user: haoran.openshift
  name: default/1h2cu64jqeufe3dnk7gvpl7j3ui3msut/haoran.openshift
current-context: default/1h2cu64jqeufe3dnk7gvpl7j3ui3msut/haoran.openshift
kind: Config
preferences: {}
users:
- name: haoran.openshift
  user:
    exec:
      apiVersion: client.authentication.k8s.io/v1beta1
      args:
      - /Users/haowang/.kube/ocm-token
      command: bash
      env: null

When we perform oc project
It will create a new context&server&user as follows:

apiVersion: v1
clusters:
- cluster:
    server: http://localhost:8001/backplane/cluster/1h2cu64jqeufe3dnk7gvpl7j3ui3msut/
  name: 1h2cu64jqeufe3dnk7gvpl7j3ui3msut
- cluster:
    server: http://localhost:8001/backplane/cluster/1h2cu64jqeufe3dnk7gvpl7j3ui3msut/
  name: localhost:8001
contexts:
- context:
    cluster: 1h2cu64jqeufe3dnk7gvpl7j3ui3msut
    namespace: openshift-monitoring
    user: haoran.openshift
  name: default/1h2cu64jqeufe3dnk7gvpl7j3ui3msut/haoran.openshift
- context:
    cluster: localhost:8001
    namespace: default
    user: system:serviceaccount:openshift-backplane-srep:89bb76f8aa82917707763cfb8c4a01a5/localhost:8001
  name: default/localhost:8001/system:serviceaccount:openshift-backplane-srep:89bb76f8aa82917707763cfb8c4a01a5
current-context: default/localhost:8001/system:serviceaccount:openshift-backplane-srep:89bb76f8aa82917707763cfb8c4a01a5
kind: Config
preferences: {}
users:
- name: haoran.openshift
  user:
    exec:
      apiVersion: client.authentication.k8s.io/v1beta1
      args:
      - /Users/haowang/.kube/ocm-token
      command: bash
      env: null
- name: system:serviceaccount:openshift-backplane-srep:89bb76f8aa82917707763cfb8c4a01a5/localhost:8001
  user: {}

oc project should respect the current context, and only change the namespace in the current-context

[feichashao]

The current oc project <project-name> will always generate a new cluster&context based on the RESTConfig, and fetch the username from whoami. Link.

if len(userNameInUse) == 0 {  // <--- Always true here.
	user, err := project.WhoAmI(o.RESTConfig)
	if err != nil {
		return fmt.Errorf("unable to default to a user name: %v", err)
	}
	userNameInUse = user.Name
}

kubeconfig, err := cliconfig.CreateConfig(projectName, userNameInUse, o.RESTConfig)

CreateConfig will generate the cluster name based on host:port with getClusterNicknameFromConfig, which doesn't respect the cluster name set in kubeconfig. Also, CreateConfig only copy the BearerToken from RESTConfig, and drops other info like ExecProvider. Link.

This works fine if the kubeconfig was generated by oc login. But if the kubeconfig was pre-generated by other tools or was made manually by user, information could be lost.

To make it works for both cases, we can just re-use the info from current context, and change the namespace to current project name. If there's no current context | no cluster info | no authinfo (how can it happen?) , we can then generate a new one from CreateConfig.

Not sure if it would break anything.

Update:
Users may pass --context --cluster --user options to oc, so the source of truth should be RESTConfig instead currentContext.

[wanghaoran1988]

@soltysh Hi, could you take a look at this issue?

[openshift-bot]

Issues go stale after 90d of inactivity.

Mark the issue as fresh by commenting /remove-lifecycle stale.
Stale issues rot after an additional 30d of inactivity and eventually close.
Exclude this issue from closing by commenting /lifecycle frozen.

If this issue is safe to close now please do so with /close.

/lifecycle stale

[wanghaoran1988]

/remove-lifecycle stale

[openshift-bot]

Issues go stale after 90d of inactivity.

Mark the issue as fresh by commenting /remove-lifecycle stale.
Stale issues rot after an additional 30d of inactivity and eventually close.
Exclude this issue from closing by commenting /lifecycle frozen.

If this issue is safe to close now please do so with /close.

/lifecycle stale

[openshift-bot]

Stale issues rot after 30d of inactivity.

Mark the issue as fresh by commenting /remove-lifecycle rotten.
Rotten issues close after an additional 30d of inactivity.
Exclude this issue from closing by commenting /lifecycle frozen.

If this issue is safe to close now please do so with /close.

/lifecycle rotten
/remove-lifecycle stale

[clcollins]

/remove-lifecycle rotten

[wanghaoran1988]

This should already fixed in #692

[soltysh]

/close

[openshift-ci]

@soltysh: Closing this issue.

In response to this:

/close

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository.