It’s an internal cluster on bare metal/vm, but it is a new cluster. I deployed it through Rancher, and have been following the config described here: https://rancher.com/blog/2018/2018-08-07-cicd-pipeline-k8s-autodevops-rancher-and-gitlab/
I have the latest Gitlab version (11.5.3 omnibus), and have no firewall between Gitlab and the Kubernetes cluster. I have established that network connectivity is not the issue by testing with ping and curl to the target IP.
On the link you cited, someone says “you’ll need to disable rbac to add deploy apps to your cluster.” But there is an RBAC check box, which I have checked, as we are using RBAC. I’d rather not disable.
The ServiceAccount was set up by applying this gist (from the above blog): http://x.co/rm082018
I believe it has the permissions needed.
Questions:
- I can find nothing in Gitlab logs on the attempted connection. Which log (might) it be logged in, and if the logging needs to be enhanced/turned on, which logs should I enhance/enable? I’m also getting debug-level logging from the cluster, with no mention of helm or gitlab (or its IP address) showing up.
- Is there a way to test the connection from the Gitlab server shell via command line using the same pem/token/target format? If so I’m not finding how to do it on internet searches.