Prometheus metrics endpoint
bufferoverflow opened this issue · comments
Would be cool the have Prometheus metrics served via /metrics to provide insights, e.g. number of webhooks processed etc.
@bufferoverflow out of curiosity, what does your ideal deployment look like? I went this route early on, but when deploying into an environment with varying #'s of transient buz instances (gcp cloud run, lambda, etc) it provided limited value (since it was a large # of small instances).
Right now we plan to use buz as webhook receiver for GitLab System Hooks, see also https://gitlab.com/gitlab-org/gitlab/-/merge_requests/118616 to make it work smooth in combination with buz. To monitor /health is helpful but as we use Prometheus all over it would be cool to also get insights on buz running on the machine.
Sweet, makes perfect sense! Thanks for the insight.