gluster-kubernetes
gluster-kubernetes copied to clipboard
gk_deploy: add --show-all switch to kubectl call
Inside gk_deploy script, in check() function definition, when calling kubectl to check if one of the jobs has finished, the call is missing "--show-all" so it also shows jobs that are in Completed
status, otherwise if the job has finished it is not listed and the script keeps waiting, eventually timing out.
Sample output:
$ kubectl get pods --no-headers -l job-name=heketi-storage-copy-job
No resources found, use --show-all to see completed objects.
$ kubectl get pods --show-all --no-headers -l job-name=heketi-storage-copy-job
heketi-storage-copy-job-8nfz7 0/1 Completed 0 19m
Can one of the admins verify this patch?
ok to test
This undoes half of PR #502 ...
@SaravanaStorageNetwork @jarrpa what to do here?
Just want to see whether the @centos-ci had not run when triggered by @nixpanic or if it just does not show up in the list of checks:
ok to test
Please note that --show-all
shouldn't be used anymore, since it is deprecated. The default behaviour of kubectl
in current version is equivalent to the previous show all. See: https://github.com/kubernetes/kubernetes/issues/67895