Kops: Add --kubeconfig flag to `kops export kubecfg`, to set the location of the kubeconfig file to create

Created on 18 Apr 2017 · 18Comments · Source: kubernetes/kops

Without setting environment variable KUBECONFIG,
kops export kubecfg does not allow to export to a different file than ~/.kube/config

Please add a parameter --kubeconfig to set the location of the kubeconfig file to create.

--kubeconfig should take priority over the environment variable KUBECONFIG if both are set.
And if none are set, the default location should be ~/.kube/config.

Example usage

$ kops export kubecfg --name k8s.test.eu-west-1.aws.redacted.net --state s3://k8s/kops/  --kubeconfig $HOME/test-kubeconfig.yaml
# ---> creates a file kubeconfig at the location $HOME/test-kubeconfig.yaml

Current workaround

When setting the environment variable KUBECONFIG, it is possible to set the location of the config file to be created.

$ export KUBECONFIG=$HOME/test-kubeconfig.yaml
$ kops export kubecfg --name k8s.test.eu-west-1.aws.redacted.net --state s3://k8s/kops/
$ cat $HOME/test-kubeconfig.yaml
apiVersion: v1
clusters:
- cluster:
...

Feature Request lifecyclrotten

Source

kenden

👍17

Most helpful comment

Surprisingly not fixed yet (nor the documentation about --config).

Still have to use the export KUBECONFIG workaround. Pretty bad when you need to share a kubeconfig with someone.

awbacker on 4 May 2018

👍7

All 18 comments

As far as I understand this is as designed and models kubectl. Would more documentation help, or do we have a flag that is not working??

chrislovecnm on 23 Apr 2017

Oh ... --config is actually not for kubeconfig. It is for the viper config file config... we need to fix the help section on that. The viper config is part of the cli tooling that we use. Think if it as a kops base config file. I actually never have used it. Not sure how and if it works ;). I am sure it does, but you should be able to set stuff like state store location. Not certain.

chrislovecnm on 23 Apr 2017

👍1

@chrislovecnm Thanks!
kubectl has the parameters --kubeconfig to indicate where to find the kubeconfig file.
More documentation about the kops export --config parameter would be good, yes! (there is a ticket for that here).

Is there a param to set the location of the kubeconfig file then?
If not, this ticket is a request to add one.

kenden on 23 Apr 2017

Nope no --kubeconfig flag here https://github.com/kubernetes/kops/blob/master/cmd/kops/export_kubecfg.go

Great idea... I am going to update the title of the issue if you do not mind. This is a great starter level issue for someone to address!

chrislovecnm on 23 Apr 2017

Please edit the title if I am not understanding the feature request clearly.

chrislovecnm on 23 Apr 2017

@chrislovecnm that's great thanks! I also updated the description

kenden on 23 Apr 2017

🎉2

if no one is working on it yet, can i take this. as a first time contributor, it really do look like 'good-starter-issue'

rajatjindal on 31 Oct 2017

👍2

We can use one liner like this

KUBECONFIG=$HOME/test-kubeconfig.yaml kops export kubecfg --name k8s.test.eu-west-1.aws.redacted.net --state s3://k8s/kops/

pandeybk on 14 Feb 2018

👍1

Is is take ? @rajatchopra do you handle this?

jsenon on 21 Mar 2018

Not sure if this is fixed already. Proposed a fix in #4746
PTAL. Thanks.

rajatchopra on 21 Mar 2018

Surprisingly not fixed yet (nor the documentation about --config).

Still have to use the export KUBECONFIG workaround. Pretty bad when you need to share a kubeconfig with someone.

awbacker on 4 May 2018

👍7

Is this a good place for a bad programmer with the go book to start?

ms4720 on 11 Jul 2018

😄2

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 on 9 Oct 2018

/remove-lifecycle stale

ms4720 on 9 Oct 2018

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 on 8 Nov 2018

So we have a PR for this now - #5955. I guess I'm wondering why we don't just use the KUBECONFIG env var?

We use kubeconfig in a few places:

delete (when removing the configuration after a cluster delete)
update/export (when writing the configuration)
generally, to get the "current cluster", which makes specify the cluster optional in most places.

I think we'd have to add the kubeconfig flag essentially to all our commands.

justinsb on 19 Nov 2018

👍1

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

fejta-bot on 19 Dec 2018

@fejta-bot: Closing this issue.

In response to this:

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.