kubewharf / katalyst-core

Katalyst aims to provide a universal solution to help improve resource utilization and optimize the overall costs in the cloud. This is the core components in Katalyst system, including multiple agents and centralized components

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

[Umbrella] Support topology-aware scheduling

caohe opened this issue · comments

commented

This is an umbrella tracking the items that support topology-aware scheduling.

Why is this needed?

Currently, Katalyst supports numa-binding and numa-exclusive enhancements for the dedicated_cores QoS class in the colocation scenario.

In situations outside of colocation, it is also necessary to be aware of the NUMA and device topology when scheduling and allocating resources, so as to improve the performance of containers. Furthermore, in this particular scenario, Katalyst's resource allocation strategy needs to be compatible with Kubernetes' native allocation strategy.

What would you like to be added?