Tune It Up! Enabling Low Latency in Kubernetes Clusters
CNCF [Cloud Native Computing Foundation] via YouTube
Overview
Explore how to enable low latency in Kubernetes clusters for Cloud-native Network Functions (CNF) and 5G deployments in this 25-minute conference talk from KubeCon + CloudNativeCon Europe 2021. Gain insights into fine-tuning cluster settings for optimal low-latency behavior, including kubelet configurations, kernel settings, and workload guarantees. Discover the interplay between topology and CPU managers, tuned kernel settings, and upcoming enhancements like the kubelet memory manager. Learn about Operators, Real Time Kernel, Node Tuning Operator, CPU manager, Device manager, Topology manager, and topology-aware resource allocation. Understand how to address NUMA-aligned resource challenges and leverage the Performance Addon Operator as an orchestrator for demanding workloads in Kubernetes environments.
Syllabus
Intro
Feature use case
Operators
Real Time Kernel
Micro Demo
Node Tuning Operator
CPU manager
Device manager
Topology manager
Topology-aware resource allocation
Example: Not-NUMA-Aligned resources
Performance Addon Operator as orchestrator
Taught by
CNCF [Cloud Native Computing Foundation]