Glances

Glances is a cross-platform, open-source system monitoring tool that provides a real-time, unified view of key host metrics. It consolidates CPU, memory, disk I/O, network, processes, sensors, and more into a compact terminal UI, with an optional web interface and APIs.

It is designed for operators who need quick, low-overhead visibility: sysadmins, DevOps engineers, developers, and hobbyists. Glances works well for ad-hoc troubleshooting, incident triage, lightweight remote monitoring, and as a metrics source for other tools.

Use Cases

  • Rapid incident triage on a single host: identify CPU, memory, I/O bottlenecks and top processes without switching tools.
  • Remote monitoring across multiple machines using client/server mode or the web UI.
  • Monitoring container hosts: view Docker/LXC container metrics alongside host resources.
  • Low-overhead monitoring on resource-constrained nodes (home labs, edge devices, small VMs).
  • Ad-hoc performance profiling before/after deploys or configuration changes.
  • Exporting short-term metrics snapshots (CSV/JSON) or forwarding to time-series databases via exporters for dashboards elsewhere.

Strengths

  • Real-time unified dashboard: compact curses UI (and web UI) shows CPU, memory, disk, network, processes, temperatures, and more in one place.
  • Adaptive layout: maximizes useful information across small terminals and large screens.
  • Flexible access modes: local terminal, web interface, and client/server remote monitoring.
  • Lightweight and easy to install: Python-based, available via package managers, pip, and containers.
  • Process monitoring and filtering: quickly pinpoint resource-hungry processes.
  • Configurable alerts and thresholds: color-coded warnings highlight anomalies early.
  • Plugin and export architecture: extend with custom checks and send metrics to files or external stores.
  • Container and sensor awareness: displays container metrics and hardware sensors (where available).
  • Cross-platform: runs on Linux, FreeBSD, macOS, and Windows (with some feature differences).
  • Open source and actively maintained: community contributions and transparent development.

Limitations

  • Not a long-term metrics store: lacks native historical retention and rich time-series visualization. Pair with a TSDB and dashboard stack for trends.
  • Text-focused visualization: the curses UI is efficient but not a substitute for interactive charts.
  • Platform differences: hardware sensors and certain integrations work best on Linux; expect variability on Windows/macOS.
  • Plugin maturity varies: community plugins/exporters may be inconsistently maintained; validate before production use.
  • Not an enterprise alerting/aggregation system: no multi-tenant dashboards, RBAC, or advanced alert routing.

Final Thoughts

Glances is a practical, low-friction choice for real-time system insight. Use it for quick diagnostics and continuous lightweight monitoring, and integrate its exporters if you need historical dashboards elsewhere.

Practical starting points: pip install glances then run glances locally; launch a web view with glances -w; enable remote mode with glances -s (server) and connect via glances -c <host> (client). Configure thresholds and plugins via the config file, and export to CSV/JSON or a time-series backend as needed.

References