Bug 1868384

Description of problem:

It is not possible to log into multiple OCP clusters with the same username using the OCP 4 CLI without having to re-enter credentials every time.

Version-Release number of selected component (if applicable):

4.x


How reproducible:

Always


Steps to Reproduce:

$ oc login -u user https://openshift.cluster.one:443
Authentication required for https://openshift.cluster.one:443 (openshift)
Username: user
Password: 
Login successful.

$ oc login -u user https://openshift.cluster.one:443
Logged into https://openshift.cluster.one:443" as "user" using existing credentials

$ oc login -u user https://openshift.cluster.two:443
Authentication required for https://openshift.cluster.two:443 (openshift)
Username: user
Password: 
Login successful.

$ oc login -u user https://openshift.cluster.one:443
Authentication required for https://openshift.cluster.one:443 (openshift)
Username: user
Password: 
Login successful.

Actual results:

Users have to re-enter their credentials every time they try to switch between clusters.

Expected results:

Users can log in to clusters where they have already entered credentials for. This was the case in OCP 3.


Additional info:

The OCP 4 CLI does not distinguish between usernames on different OCP clusters. As a result, the user token is overwritten with each log in. Example:

- The contents of ~/.kube/config generated by the OCP 4 CLI after logging in to multiple OCP clusters with the same username:

```
...
 kind: Config
 preferences: {}
 users:
 - name: user
   user:
-    token: jzd-HQB_bY6cIfJhgJhJY53zVzoOw0qnU21kBUTL7lY
+    token: JAWCHMbOG6TuqCfqe9h5mKqMNqyXcJWwaByeEqHjSsM
...
```

The same file after the same interaction with the OCP 4 CLI:

```
...
 kind: Config
 preferences: {}
 users:
 - name: user/openshift.cluster.one:443
   user:
     token: 4lkgcqM7oHP6PDSl_945oMymylib24ZXWNQvnXAsufY
 - name: user/openshift.cluster.two:443
   user:
     token: aJp4cNXVZNZAn3HdZLV7JcjrJHUXoQcDw5--G3hdS8I
...
```

Can we restore the "old" behaviour?