Predictable Network Traffic in Kubernetes
summary
Describes giving Kubernetes applications predictable network performance by tagging high-priority apps and dedicating NIC resources to them. Covers application-specific queuing and steering that assigns hardware NIC queues to application threads to reduce jitter and provide deterministic packet processing in large clusters.
AI-written summary · about
related talks
34mNetwork QoS Support for Kubernetes Applications
34mRunning prioritised workload in Kubernetes
26mEnhancing K8s Networking with SmartNICs
32mThe Easy--Don't Drive Yourself Crazy--Way to Kubernetes Networking
20mKubernetes Networking 101
42mFiltering Network Flows from Kubernetes Clusters: Why and How?
36mNetwork-aware Scheduling in Kubernetes
40mThe Network in Kubernetes: Can We See It More Clearly?