You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
In later versions of k8s, a request to //version is invalid, whilst a request to /version responds as expected. This appears to be an upstream bug, but it can be handled gracefully enough by simply "squashing" the URL path. ie.
Thanks @roycaihw, I will test this tomorrow. It certainly looks like it should work around the issue (FWIW, I'd already removed the trailing / from my kubeconfig to get around this initially.
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
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
Rotten issues close after 30d of inactivity.
Reopen the issue with /reopen.
Mark the issue as fresh with /remove-lifecycle rotten.
Send feedback to sig-testing, kubernetes/test-infra and/or fejta.
/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.
Activity
roycaihw commentedon Nov 12, 2018
@glbyers Could you take a look at kubernetes-client/python-base#45 to see if it solves your problem?
Besides the host url containing trailing slash in kubeconfig file, do you have any other specific cases that cause invalid url in this client package?
glbyers commentedon Nov 13, 2018
Thanks @roycaihw, I will test this tomorrow. It certainly looks like it should work around the issue (FWIW, I'd already removed the trailing / from my kubeconfig to get around this initially.
fejta-bot commentedon Apr 26, 2019
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 commentedon May 26, 2019
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
fejta-bot commentedon Jun 26, 2019
Rotten issues close after 30d of inactivity.
Reopen the issue with
/reopen.Mark the issue as fresh with
/remove-lifecycle rotten.Send feedback to sig-testing, kubernetes/test-infra and/or fejta.
/close
k8s-ci-robot commentedon Jun 26, 2019
@fejta-bot: Closing this issue.
Details
In response to this:
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.