Fleet Timeline
All servers plotted on a shared time axis — spot cascading failures instantly.
Reading the chart
Each coloured line represents one server. The Y axis is percentage (0–100%). Switch between CPU, RAM, and Disk using the metric selector.
Time range
Choose 15 min, 1 hr, 6 hr, or 24 hr. Historical data comes from the metrics database; the rightmost portion is live from WebSocket.
Isolating servers
Click a host pill below the controls to hide it. Click again to show it. Reset with "All hosts" or "Busiest N".
Tips
- A simultaneous spike across all servers usually indicates a shared upstream issue — network, power, or a shared dependency.