Today’s compute workloads and workflows are growing, varied and with many customer choices. Underlying networks and fabrics can either curtail or unlock growth. In some ways, things used to revolve more around the somewhat simpler ‘speeds and feeds,’ but how much are elements like complexity and security being factored into modern data center cost-benefit analysis?
If you care about performance of your application stack you’d be much better off using the concept of swimlanes – parallel application stacks deployed in different regions or availability zones. A performance problem in a single availability zone would hit only a single swimlane, and you can easily take the whole swimlane out of operation by removing the first-layer instance (web server) from the top load balancer pool.