In Part 3, I look at capacity utilization, bin-packing, cgroups, Kubernetes scheduling, and how Descheduler and Cluster Autoscaler can work together to make better use of the infrastructure we already have.
In Part 2, I look at rightsizing, instance flexibility, scheduling, OOC handling, graceful degradation, and how applications and platforms can work together to make more of the available capacity actually usable.
Cloud capacity isn't "infinite" anymore. Let's look at what's happening to compute capacity in 2026, why classical capacity planning is back, and why forecasts, CSP cooperation and some old assumptions need a serious revisit.