vcctl only reads the ~/.kube/config and neglects $KUBECONFIG

[jiangkaihua]

Is this a BUG REPORT or FEATURE REQUEST?:

/kind feature

What happened:
When I created a local k8s cluster using ~/kubernetes/hack/local-up-cluster.sh, the guideline suggested me to export KUBECONFIG=/var/run/kubernetes/admin.kubeconfig instead of creating a new config in ~/.kube/. Then the vcctl command line reported Failed to list job: stat /root/.kube/config: no such file or directory and failed to execute.
After I copyed the config file referred to $KUBECONFIG to ~/.kube/ , vcctl run correctly. So I doubt that the vcctl command line is set to read ~/.kube/config only and neglects the enviroment variable $KUBECONFIG.

What you expected to happen:
When vcctl cannot find ~/.kube/config, it shall get $KUBECONFIG as an alternative.

How to reproduce it (as minimally and precisely as possible):

Anything else we need to know?:

Environment:

• Volcano Version: release-0.2
• Kubernetes version (use kubectl version): v1.15.3
• Cloud provider or hardware configuration:
• OS (e.g. from /etc/os-release): Ubuntu 18.04
• Kernel (e.g. uname -a):
• Install tools:
• Others:

[hzxuzhonghu]

duplicate with #353

[Rui-Tang]

I would like to work on it.

[Rui-Tang]

So it just need to look for KUBECONFIG if default config file is missed. Is that correct?

[hzxuzhonghu]

/assign @tall49er

correct