Alerts

Introduction

An alert is triggered when a device metric (e.g., ping, CPU usage) crosses its configured threshold for a specified duration (tolerance). A recovery alert is sent when the metric returns to normal.

Alerts are based on Alert Settings defined for each metric, each setting includes:

  • Operator: defines the condition to evaluate the metric value:

    • Greater than: triggers when the value exceeds the threshold.

    • Less than: triggers when the value is below the threshold.

  • Threshold: the metric value that triggers the alert.

  • Tolerance: the duration (in minutes) the threshold must be continuously breached before triggering an alert. Helps reduce noise from flapping metrics.

Note

You can override the default alert settings globally using the OPENWISP_MONITORING_METRICS setting, or on a per-device basis as explained in the Device Checks & Alert Settings section.

The built-in alerts are explained below.

Ping

Triggers when the device becomes unreachable via ping.

Alert enabled by default? Yes.

Collected via: Ping Check.

Charts: Ping Chart.

Default Alert Settings:

Operator

< (less than)

Threshold

1

Tolerance

30 minutes

Config Applied

Triggers when the device fails to apply configuration changes within the specified time.

Alert enabled by default? Yes.

Collected via: Config Applied Check.

Default Alert Settings:

Operator

< (less than)

Threshold

1

Tolerance

10 minutes

Data Collected

Triggers when no metric data has been collected from the device.

Alert enabled by default? Yes.

Collected via: Config Applied Check.

Default Alert Settings:

Operator

< (less than)

Threshold

1

Tolerance

30 minutes

Memory Usage

Triggers when memory usage exceeds the threshold.

Alert enabled by default? Yes.

Collected via: OpenWrt Monitoring Agent.

Charts: Memory Usage Chart.

Default Alert Settings:

Operator

> (greater than)

Threshold

95 (percent)

Tolerance

30 minutes

CPU Load Average

Triggers when CPU usage exceeds the threshold.

Alert enabled by default? Yes.

Collected via: OpenWrt Monitoring Agent.

Charts: CPU Load Chart.

Default Alert Settings:

Operator

> (greater than)

Threshold

90 (percent)

Tolerance

30 minutes

Disk Usage

Triggers when disk usage exceeds the threshold.

Alert enabled by default? Yes.

Collected via: OpenWrt Monitoring Agent.

Charts: Disk Usage Chart.

Default Alert Settings:

Operator

> (greater than)

Threshold

90 (percent)

Tolerance

0 minutes

WiFi Clients (Max)

Triggers when the number of connected WiFi clients exceeds the threshold.

Alert enabled by default? No (see WiFi Clients Check for details on how to enable it).

Collected via: the WiFi clients information is collected through the OpenWrt Monitoring Agent, but the alert is triggered by the WiFi Clients Check.

Default Alert Settings:

Operator

> (greater than)

Threshold

50

Tolerance

120 minutes

WiFi Clients (Min)

Triggers when the number of connected WiFi clients falls below the threshold.

Alert enabled by default? No (see WiFi Clients Check for details on how to enable it).

Collected via: the WiFi clients information is collected through the OpenWrt Monitoring Agent, but the alert is triggered by the WiFi Clients Check.

Default Alert Settings:

Operator

< (less than)

Threshold

1

Tolerance

0 minutes