Github (Enterprise) Action that launches a runner on Kubernetes (HP Ezmeral) cluster

Hi I’m trying to achieve the launch-runner step from here on our kubernetes cluster. After adding steps to install terraform, kubectl and a kubectl-plugin in the action I run the following

cml runner launch --cloud=kubernetes --log=debug --labels-cml

and get a pretty long log but I think this is the most relevant part:

{"level":"info","message":"2022-11-29T13:40:16.566+0100 [INFO]  provider.terraform-provider-iterative: 2022/11/29 13:40:16 [DEBUG] Received error: &url.Error{Op:\"Get\", URL:\"[https://redacted_url.com:port/apis/batch/v1/namespaces/redacted/jobs/iterative-e0kvt3bw1hba\](https://redacted_url.com:port/apis/batch/v1/namespaces/redacted/jobs/iterative-e0kvt3bw1hba/)", Err:(*errors.errorString)(0xc000990270)}: timestamp=2022-11-29T13:40:16.565+0100"}
{"level":"info","message":"iterative_cml_runner.runner: Creation errored after 4s"}
{"level":"error","message":"terraform error: Error: Failed disposing the machine: Get \"https://redacted_url.com:port/apis/batch/v1/namespaces/redacted/jobs/iterative-e0kvt3bw1hba\": getting credentials: *** executable kubectl failed with exit code 1"}
{"level":"error","message":"terraform error: Error: Failed creating the machine: Failed to create Job! API error: Post \"https://redacted_url.com:port/apis/batch/v1/namespaces/redacted/jobs\": getting credentials: *** executable kubectl failed with exit code 1"}
{"level":"info","message":"2022-11-29T13:40:16.575+0100 [DEBUG] provider: plugin process exited: path=.terraform/providers/registry.terraform.io/iterative/iterative/0.11.13/linux_amd[64](https://redacted_url.com/cml-test/cml-tutorial-1/runs/382?check_suite_focus=true#step:10:65)/terraform-provider-iterative pid=95314"}
{"level":"info","message":"2022-11-29T13:40:16.575+0100 [DEBUG] provider: plugin exited"}

Any help is appreciated!

[rick-van-veen]

Yes. It is fixed. It was an authentication issue. Had to run kubectl hpecp authenticate and give it my password. We can move forward. Thank you!

[0x2b3bfa0]

I’m not familiar with hpecp authentication, but it seems like this issue was caused by kubectl trying to read a password from the standard input and failing to do so. How are you generating the kubconfig file? Running kubectl hpecp refresh locally?