As per title, is this possible at all?
I was hoping to be able to determine which projects were using the most shared runner resources to help the teams optimised, and to determine when we may need to expand the cluster (I’m using Kubernetes).
None of the GitLab Runner metrics seem to expose project information, so I don’t see how I can determine which project is burning through all the cycles.
Am I simply looking in the wrong place? Or is there some magic I need to do in the Helm values.yaml to make this possible?
Maybe an alternative approach: On GitLab.com, shared runner usage is calculated with compute minutes, and quotas. The same features exist on self-managed instances with instance runners but are disabled by default. Compute minutes administration | GitLab I heard of a customer setup where the platform team uses compute minutes to provide internal budgets for different teams, and also for scaling insights.
Thanks and Happy New Year!
That’s an interesting idea and would certainly seem to meet my needs, but unfortunately “compute minutes” does not seem to get reported as a metric that I can place on to a dashboard.
Or maybe I’m missing it (there’s a lot of metrics!), my hope would be just to track who is using what rather than setting any hard limit.
Edit: I guess what I am looking for is something a bit like telemetry but at a project level and it seems that what I want to do isn’t possible. Bit of a shame, just being able to show “compute minutes” in a chart would have been amazing.
Edit 2: I also don’t see any way to enable the “Pipeline” tab in “Usage Quotas”
I had to ask an AI as I could not find the documentation (only place I can find it is this issue, but it seems “Pipelines” within “Usage Quotas” is only available to Premium or Ultimate license holders.
Unfortunately I am on the free tier.
Still, at least I now know I am not going crazy! 