Closed jawnsy closed 11 months ago
I know we have limits now, but just wanted to address:
In our cluster, the operator is extremely lightweight
This is intentional, and is part of the reason that https://github.com/authzed/controller-idioms exists. It's too easy to write controllers with ballooning caches and lots of churn if you're not careful, we wanted to make it easier to write efficient operators.
Summary
Add CPU and memory request/limits for SpiceDB Operator deployment
Details
The SpiceDB Operator manifest is missing resource specs, which means that the resulting pod will have
BestEffort
quality of service.This is typically a low-priority issue because the pod is mainly idle, is not very resource-intensive, and most clusters have spare capacity available for
BestEffort
pods. For completeness, adding a request would be useful to force the cluster autoscaler to make room for the operator (guaranteeing that some forward progress will be made) and limits would be useful to make issues like leaks more apparent.Proposal
Add a small request/limit for the pod based on current usage. In our cluster, the operator is extremely lightweight:
So something like this seems suitable: