cmd-path not respected

[tcolgate]

The python client does not respect cmd-path in the gcp provider.

We abuse this feature to retrieve credentials for our non-gke clusters (hosted in AWS). This works well with k8s.io/client-go, but does not seem to be supported by the clients in the kubernetes-client org, with the exception of the nodejs client.

[fejta-bot]

Issues go stale after 90d of inactivity.
Mark the issue as fresh with /remove-lifecycle stale.
Stale issues rot after an additional 30d of inactivity and eventually close.

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

Send feedback to sig-testing, kubernetes/test-infra and/or fejta.
/lifecycle stale

[fejta-bot]

Stale issues rot after 30d of inactivity.
Mark the issue as fresh with /remove-lifecycle rotten.
Rotten issues close after an additional 30d of inactivity.

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

Send feedback to sig-testing, kubernetes/test-infra and/or fejta.
/lifecycle rotten

[jfrabaute]

We are having the same issue with a standard GKE cluster.
The kube config has "cmd-path" and the client is failing to refresh the access token.
Exception raised is:

google.auth.exceptions.DefaultCredentialsError: Could not automatically determine credentials. Please set GOOGLE_APPLICATION_CREDENTIALS or explicitly create credentials and re-run the application. For more information, please see https://cloud.google.com/docs/authentication/getting-started

Is there a plan to make this python client with with "cmd-path"?

[jfrabaute]

/remove-lifecycle rotten

[fejta-bot]

Issues go stale after 90d of inactivity.
Mark the issue as fresh with /remove-lifecycle stale.
Stale issues rot after an additional 30d of inactivity and eventually close.

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

Send feedback to sig-testing, kubernetes/test-infra and/or fejta.
/lifecycle stale

[jfrabaute]

We're having the same problem with all our gke clusters which are using cmd-path.
I've just send a pull request to try to fix this: kubernetes-client/python-base#168

[fejta-bot]

Stale issues rot after 30d of inactivity.
Mark the issue as fresh with /remove-lifecycle rotten.
Rotten issues close after an additional 30d of inactivity.

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

Send feedback to sig-testing, kubernetes/test-infra and/or fejta.
/lifecycle rotten