You have many GKE microservices with incorrect CPU and memory limits. How do you ensure each microservice has right-sized limits?

Vertical Pod Autoscaler (VPA) automatically adjusts CPU and memory requests and limits for containers based on their historical usage. This directly addresses the problem of incorrect resource limits for individual microservices, ensuring they are "right-sized." Changing the node pool machine type or enabling cluster autoscaling would provide more resources at the node or cluster level, but wouldn't optimize individual microservice resource allocations. Horizontal Pod Autoscaler (HPA) scales the number of pod replicas based on metrics like CPU utilization, which helps with demand fluctuations but doesn't right-size the resources within each pod.