We are running cockroach DB on Kubernetes via a stateful set. We were running into some performance issues so we began to review https://www.cockroachlabs.com/docs/stable/kubernetes-performance.html
Upon working through the doc we discovered that each pod was only requesting 100m of CPU.
Following the instructions, I attempted to increase the CPU request. While the yaml saves properly it does not take. IE, when the yaml is reopened the cpu request, has not saved.
I have attempted to add
resources: requests: cpu: "3500m"
via the cmd line and GKE’s website.
We are using the free tier of Cockroackdb.
Any feedback would be very helpful!