Lightweight Observability with Performance Co-Pilot and Grafana
summary
Shows a working observability pipeline that uses Performance Co-Pilot (pmcd + pmlogger) to read cgroup metrics (CPU, memory, disk I/O, network) and Grafana (PCP plugin, datasource like PCP Vector, URL port 4432) to visualize them, all running on a single-host laptop with no agents inside containers. Demo monitors two demo containers, stores history in PCP archives, and keeps overhead low (claim: ~500 MB footprint) while warning it does not scale to large multi-node clusters.
AI-written summary · about
chapters (16)
- 0:00Introduction
- 1:05Observability for containers
- 2:54What is lightweight observability
- 6:47Container resource sharing
- 8:11Performance Co-Pilot basics
- 10:38Performance Co-Pilot architecture explained
- 13:08Grafana for lightweight setups
- 14:53Container resource sharing vs virtual machines
- 15:35Demo setup and PCP install
- 16:40Enabling container metrics
- 17:44Grafana PCP plugin config
- 18:26Adding PCP data source
- 19:06Grafana dashboards for containers
- 20:14Demo key takeaways
- 21:59Trade-offs and alternatives
- 22:27Q&A
related talks
17mFrom noise to signal: piping your way to clear observability
31mProgramming the Performance Co-Pilot toolkit
36mCilium & Grafana LGTM! - Raymond de Jong
23mHow to Monitor Application Performance in a Container-Based World
27mFrom Zero Visibility to Full Observability with Grafana
36mObservability at Scale With Monitoring as Code...
26mObserving Node.js: Using Metrics to Improve your Application Performance
29mParanoid Observability –Success Stories and Misfortunes observing an observability stack