Alarms and gauges

Each PingDirectory server includes a set of gauges for tracking various metrics over time.

Each gauge is associated with a monitor attribute and can be configured with warning, minor, major, and critical thresholds. If the monitor value crosses one of those threshold values, then the server raises an alarm condition that might indicate a problem such as low disk space or unavailability of an external server, and can optionally trigger administrative alerts.

Like alerts, alarms have a name, severity, and message. They also have a condition that is indicated by the alarm, like “Disk Usage,” and they can be associated with a specific resource, such as which filesystem is running low on disk space. If alarm conditions are published using SNMP, then they can also include probable cause and alarm type properties.

Alarms are exposed in the cn=alarms backend. Their values can change over time, and they can be cleared whenever the monitored value returns to normal. Active alarms can be viewed using the status tool.