For the new GitLab 16.xx-versions, I was wondering, if it is a problem, to use the same runnerToken for multiple runner-instances in a K8s-Cluster.
As the values.yaml supports HPA(Horizontal Pod Autoscaler), there can be created multiple gitlab-runner-pods, which are using the same runnerToken (which has been created via GitLab UI) .
Is this intended?
For big clusters with runners for many projects, it would be good to have multiple replicas in order to provide high availability. Sure, I could use n Helm-releases with different runnerTokens, but the hpa seems like the cleaner way of doing it.
I have tried it locally (not in prod) but didn’t notice any troubles.