Indeed, AWS even operates VMs in an overcommitted state for certain VM series, such as the T-series. Consequently, the actual CPU and memory utilization of a VM is managed by the VM orchestrator. Additionally, I acknowledge that the title of this post is not entirely accurate, and to prevent confusion, I will begin by introducing a few assumptions.

The primary focus of this post is to delve into potential issues that may arise when migrating from a VM to Kubernetes. It's worth noting that overcommitment is not within the scope of this discussion, as a similar overcommitment strategy can be applied to containers.