Appropriately scale cloud

Created by: daxmc99

We scaled cloud heavily for the dot-com launch. We are now in a stable state and can start to ensure we have not overprovisioned our resources.

TODOs to be done in-order:

• Ensure QoS Guaranteed for all stateful services PR
• Ensure all services have resource limits PR
• Enable K8s autoscaling (allow scale down) Vertical node autoscaling enabled Horizontal node autoscaling enabled Horizontal Pod Autoscaling (HPA) only enabled for sourcegraph-docsite
• Setup default logging containers for all services PR
• Setup requested resources/limits based on the usage (cost savings) PR