Monitoring: Follow the Data from the Darkest Corners
summary
Shows exactly where observability data comes from and how it flows: kubelet’s metrics-server summary API and cAdvisor provide only current resource samples, while kube-state-metrics exposes object status; Prometheus uses a pull model with node_exporter and linear-prediction alerts (example: predict disk full within 24 hours) to catch outages early. Stackdriver demonstrates a push model with a daemonset metadata agent and fluentd logs. Custom Metrics API is pluggable so vendors can implement autoscaling-ready metrics.
AI-written summary · about
chapters (6)
- 0:00Introduction
- 2:45Monitoring in Kubernetes
- 8:21Prometheus
- 14:00Kubernetes
- 21:25Real Data
- 27:32Conclusion
related talks
19mKeynote: Monitoring Kubernetes Clusters with Prometheus
24mPractical Kubernetes Monitoring with Prometheus
32mDemocratizing Application Monitoring with Prometheus
32mUnderstanding Observability in Kubernetes
37mMonitoring Kubernetes with Prometheus
14mFrom Metrics to Mastery: Monitoring vital signs of Kubernetes
48mMetrics All the Way: Data Driven Operations
31mEvolution of Metric Monitoring and Alerting: Upgrade Your Prometheus Today