6.1.4 Exadata Alerts

Alerts draw attention to potential and actual problems and other events of interest to an administrator.

There are three types of alerts; informational alerts, warning alerts, and critical alerts.

Metrics observations are automatically compared with stored thresholds. When a metric crosses a specified threshold, an alert is automatically generated. Metrics can be associated with warning and critical thresholds. When a metric value crosses a warning threshold, then a warning alert is generated. When a metric crosses a critical threshold, then a critical alert is generated.

Exadata administrators can specify the warning and critical threshold for most metrics. There are also some preset thresholds. For example, there are built-in thresholds for ambient temperature. If the temperature is too low or too high, an alert is automatically generated.

Some system conditions and state changes can generate informational alerts; for example, a server chassis is opened. More serious system errors and state changes can also generate warning or critical alerts; for example, a hard disk device is reporting disk errors.

This system of thresholds and alerts enables administrators to focus on the most important events, while maintaining awareness other interesting events.

You can use CellCLI commands to manage and display Exadata alerts and thresholds. You can also configure alert notification using email and SNMP. Alert information is also included in the AWR report.

Like Exadata metrics, alerts are managed separately on each server.