Learning Kubernetes Through the Lens of Metrics
Talkat KubeCon + CloudNativeCon Europe 2025 · on Wednesday, 2025-04-02· at 17:00· Level 1 | Hall Entrance N10 | Room F·
summary
Creates a tiny kind cluster and curl to 6443 to read the API server metric kubernetes_feature_enabled, then installs Prometheus via Helm and queries count(kubernetes_feature_enabled) to find ~94 alpha, 158 beta, 8 deprecated features and 2 hidden metrics; hits kubelet on 10250 to see desired pods (11) and static mirror pods (4), scrapes cAdvisor for container_last_seen and kernel memory, probes kube-proxy (10249) for iptables counts and etcd (2379) for MVCC compaction revision and etcd v3 details.
AI-written summary · about
chapters (10)
- 0:00Introduction
- 1:23Kubernetes metrics insights
- 4:30Enabled features via metrics
- 10:13Pods and containers tracking
- 13:46Kubelet and cAdvisor metrics
- 16:58Kube-proxy and iptables metrics
- 18:45Scheduler queues insights
- 19:28etcd compaction metrics
- 21:59Advanced metrics overview
- 22:43Wrap-up and resources
related talks
36mReveal Your Deepest Kubernetes Metrics - Bob Cotton, Freshtracks.io
19mKeynote: Monitoring Kubernetes Clusters with Prometheus
33mDeep Dive: Kubernetes Metric APIs using Prometheus
35mDeep Dive: Kubernetes (Instrumentation) SIG
35mFrom New Cluster to Insight. Deploying Monitoring and Logging to Kubernetes
32mUnderstanding Observability in Kubernetes
37mWeighing a Cloud: Measuring Your Kubernetes Clusters
24mPractical Kubernetes Monitoring with Prometheus