System Monitor

The System Monitor dialog shows live CPU, load, memory, and network throughput for the SkyView host. It is available from the Server settings panel and can be toggled on or off in configuration.

System Monitor

What you see

  • CPU usage – Instantaneous CPU utilization across all cores, plus the core count.
  • Server load (1 / 5 / 15) – Standard Unix load averages, useful for spotting sustained CPU pressure.
  • Memory usage – Used vs total memory with percentage, helpful for identifying pressure or leaks.
  • Network in/out – Current throughput (Mbit/s) and cumulative bytes since boot, useful for diagnosing bandwidth saturation.

Each metric also appears as a short, textual summary above the charts for quick at-a-glance checks.

Live vs historical ranges

  • Latest (live) – Shows recent in-memory samples (a few minutes) with 5-second refresh. This works even when Analytics is disabled.
  • Historical ranges (5m, 15m, 1h, 3h, 12h, 24h, All) – Available when the Analytics option is enabled. System monitor samples are written to TimescaleDB; the backend returns raw data for short windows and aggregated buckets for longer windows. The charts display the full selected window, filling missing buckets with zeros so gaps or downtime remain visible.

Aggregation details (Analytics enabled)

  • Up to 30 minutes: raw samples.
  • 30 minutes to 6 hours: 5-second buckets.
  • Beyond 6 hours: 1-minute buckets.

Tips

  • Use Server load alongside CPU usage to tell transient spikes from sustained pressure.
  • Network in/out quickly reveals throttling or unexpected spikes in bandwidth.
  • If you select a long range shortly after enabling Analytics, early windows may be sparse until more samples accumulate. Zeros will appear where no data was recorded, keeping the timeline aligned with your selected range.