Streamline Your Cluster with Automated Node Labels,Taints, and Annotations
summary
Node Feature Discovery (NFD) v0.15 runs as master + worker (+ topology-updater) and auto-detects hardware/software to create NodeFeature CRD labels, annotations and extended resources that real schedulers consume. It demos on minikube: workers scan in <30 ms, default refresh is 1 minute default refresh, scaling nodes triggers NodeFeatureRule-based taints to reserve new nodes, and vendor sidecars emit GPU-specific CRDs used by NVIDIA/Intel in production. CRD-driven rules filter which attributes get created and can auto-add taints or extended resources.
AI-written summary · about
chapters (12)
- 0:00Introduction
- 2:59Heterogeneous Kubernetes clusters
- 4:46Manual labeling challenges
- 6:31Node Feature Discovery
- 8:01NFD components
- 9:25How NFD uses Custom Resource Definitions
- 11:10Automated node labels
- 12:57Automated taints and control
- 15:47Live demo of NFD
- 23:05Q&A
- 31:29Advanced NFD use cases
- 34:00Closing
related talks
33mNFD and My Adventure in the Cloud Native Project Jungle
30mTaming Unbounded Resources with Node Feature Discovery
33mNFD and My Adventure in the Cloud Native Project Jungle
25mK8s Labels Everywhere! Decluttering With Node Profile Discovery.
26mResource Orchestration of HPC on Kubernetes: Where We Are Now and the Journey Ahead!
34mWorking your Cluster: Smarter Scheduling Decisions for Your Work...
33mCNCF TAG-Runtime: Cloud Native Open...
22mAnatomy of a production grade Kubernetes cluster