Lightweight Observability with Performance Co-Pilot and Grafana

Learn to build a fast, resource-efficient performance monitoring setup using Performance Co-Pilot (PCP) and Grafana in this 28-minute conference talk from DevConf.in 2026. Discover how to monitor containers without deploying heavyweight observability stacks, making it perfect for local development, testing, or lightweight production environments. Explore the installation and configuration of PCP to collect essential system metrics including CPU usage, memory pressure, disk I/O, and network activity, while mastering the streaming and storage of performance data locally for both real-time and historical analysis. Master the creation of intuitive Grafana dashboards that answer critical questions about container performance, system load, and bottleneck identification. Understand how to observe container workloads directly from the host without requiring Kubernetes or Prometheus, including accessing per-container resource usage from a system-level perspective using standard Fedora/RHEL tools. Gain practical knowledge of observability fundamentals including sampling, storing, and visualizing performance metrics using tools built into the Linux ecosystem. Walk away with a complete, working observability stack that's easy to install, understand, and powerful enough for real-world container debugging and performance monitoring, requiring no prior experience with PCP or Grafana.