(in place) Vertical pod auto scaling

See Kubernetes 1.27: In-place Resource Resize for Kubernetes Pods (alpha) | Kubernetes

So I have an application hosted on AWS via k8s. Most of the time the application is pretty quiet, serving occasional requests. Every evening however it processes a large request that consumes multiple cores. I came across the link above that allows the pod to autoscale without restarting. So most of the time it would run on a single core, but during busy times, it would scale to multiple cores.

How would you configure the Erlang VM to handle this?

I assume you just configure the number of schedulers to a value the same as the maximum number of cores?

Would the schedulers handle cores appearing out of nowhere.

3 Likes