Resource Management Overview

Resource Management

Kubernetes typically abstracts the organization of workload details away from the user. However, it sometimes becomes necessary to guarantee resources like CPU, memory, and device locality optimizations needed for the nodes to operate at peak efficiency.

Situations that require high-efficiency processing like data analytics, machine learning, scientific computing, financial services, and telecommunications can be adjusted by policies that better regulate the actions governing these elements. The two management policies that specifically address these demands are the Topology Manager and the CPU Manager. These management tools are part of the Kubelet agent that coordinates the components responsible for these optimizations.