SIG Instrumentation Intro
summary
SIG Instrumentation describes concrete pieces: core resource metrics API in beta, custom metrics API in beta, and a metric-server (incubator) that polls nodes every minute and metrics-server stores one minute snapshots in memory; it scales to ~5,000 nodes with ~30–50 pods per node in benchmarks. kube-state-metrics is GA for object-level metadata, Prometheus-style metrics are exposed on /metrics, events live in etcd (risking load at high scale), and log rotation defaults to 10MB with limited retention.
AI-written summary · about
chapters (20)
- 0:00Intro
- 1:00CoreOS Mission
- 3:33CoreOS Matrix Pipeline
- 4:24Logging Pipeline
- 5:17Monitoring Extensions
- 6:49Monitoring Integrations
- 7:54Leads
- 8:17Biweekly meetings
- 8:37Core Matrix Pipeline
- 9:14Core Matrix API
- 10:54Metadata
- 13:07Resource Metrics API
- 15:28Resource Metrics Migration
- 18:29Events
- 20:05Prometheus
- 21:43Prometheus Metric
- 22:09Prometheus Registry
- 23:50Gogo Metrics
- 25:09Metrics Repository
- 29:19Kubernetes Monitoring
related talks
28mSIG Instrumentation Introduction and Deep Dive
28mKubernetes SIG Instrumentation - Intro
24mSIG Instrumentation Introduction and Deep Dive
40mSIG Instrumentation Introduction And Deep Dive
26mIntro and Deep Dive: Kubernetes SIG Instrumentation
35mSIG Instrumentation Introduction and Deep Dive
28mIntro and Deep Dive: Kubernetes SIG Instrumentation
30mIntro: Kubernetes (Instrumentation) SIG