Create Workload Cluster fails to add context about tkg-lab HOT 7 CLOSED

@agregory999 do you think we should just add tkg get cluster.. followed by kubectl config use-context ... after the tkg create cluster

from tkg-lab.

I think this probably needs to be checked & corrected in more places.
On TKG 1.0.0 I experienced the TKG CLI was creating the new context and setting it as the active one after a tkg create cluster. That's no longer the case with TKG 1.1

Now every time we create a cluster we need to manually run tkg get credentials and then set the active context (via kubectl)

from tkg-lab.

Agreed - I don't recall facing any issues with the lab until I changed to 1.1. I'm wondering if each lab should do a use-context and fail if that context isn't there. That way we avoid running any workload in the wrong cluster. Should we make a generic script?

from tkg-lab.

Sounds good to me.
@doddatpivotal ?

from tkg-lab.

I went to add the kubectl config use-context ... commands to the deploy-workload-cluster.sh script and it looks like it is already in place. If this is the case, can this issue be closed? @jaimegag @agregory999

from tkg-lab.

Yeah deploy-workload-cluster.sh seems to be all set. And I realize now that it is used for both shared-services and workload-clusters. So we are good to go.

What I can't remember is:
What happens after tkg init now. I know we get the credentials in kubeconfig, but I'm not 100% if the context is set.
Anyhow, the problem reported in this issue is solved

from tkg-lab.

tkg init does not set the context to the newly created management cluster. However, the 03-post-deploy-mgmt-cluster.sh does set the context.

from tkg-lab.