Can somebody explain me why the user-deployments have a liveness/startup probes but not a readiness probe? https://github.com/dagster-io/dagster/blob/c10f6d445d9ecbc95c7f30fc7976a3bbfb7833bd/helm/dagster/values.yaml#L313-L336

It’s something we’ve been considering changing for all the servers. We have them out of fear of processes getting wedged, but I don’t think we’ve observed that in any recent version.

I think it would be a good idea. Happy to submit a PR if needed