Server monitoring is either complex (Prometheus/Grafana) or expensive (Datadog). Small teams need a lightweight, self-hosted dashboard with metrics and log analysis in one place.
A FastAPI dashboard collecting CPU/memory/disk/network metrics every 5 seconds via psutil with live Chart.js graphs, Nginx access + error log parsing, a downloadable remote agent, and PDF report export.
Nginx log parsing across different formats required careful regex — OpenCode helped me write and test the patterns. The self-collect + forward architecture needed careful async loop management; OpenCode diagnosed a deadlock I'd introduced. WeasyPrint PDF rendering had font issues that OpenCode helped me resolve by finding the right CSS workarounds.
OpenCode helped me build the full client-server telemetry protocol with Pydantic validation in one session. It taught me how to design a real-time monitoring system, parse logs at scale, and structure a FastAPI project for production. The built-in agent download endpoint was an OpenCode suggestion I wouldn't have thought of.
Add alerting thresholds, multi-server topology view, Prometheus export — all architectural patterns OpenCode already helped establish.
Great job delivering a functional, self‑hosted monitoring dashboard within a short timeframe—your real‑time charts, log parsing, and PDF reports work well. To strengthen the project, publish the code (e.g., on GitHub) with clear setup instructions and add basic tests and authentication for the agent endpoint. Consider containerizing the app and expanding log format support to make deployment easier for other teams.