Checkmate
DocumentationInfrastructure

Setting hardware thresholds

Alert when CPU, memory, disk or temperature cross a limit.

Hardware monitors fire alerts when the Capture agent reports a metric above its threshold. Thresholds live on the monitor itself and are configured when you create or edit a hardware monitor.

Hardware alert thresholds

Configure thresholds

  1. Open Infrastructure in the sidebar.
  2. Create a new hardware monitor or open an existing one and click Configure.
  3. Under Alert thresholds, set values for:
    • cpuAlertThreshold: CPU usage (percent)
    • memoryAlertThreshold: memory usage (percent)
    • diskAlertThreshold: disk usage (percent)
    • tempAlertThreshold: CPU/disk temperature (degrees Celsius)
  4. Save.

Each threshold is a required percentage between 0 and 100 and defaults to 100. To stop alerting on a metric, leave it at 100 rather than clearing the field.

How thresholds trigger

Checkmate compares each Capture reading against the saved threshold. A single breaching reading is not enough: a metric must breach on five consecutive checks before the monitor's status flips to breached and notifications fire. Any passing reading resets that metric's counter through the channels attached to that monitor. When the metric drops back below the threshold, the monitor returns to up.

Suggested starting points

ResourceTypical alert level
CPU85%
Memory90%
Disk85%
Temperature75 °C

Tune from there based on what your workloads run at.

See also