1.2.4 Alerting the user about an operational problem in the virtual environment
Besides being used to display the performance of the virtual environment in the form of a report, performance data is used to alert the user when an operational problem or error occurs.
For example, when physical CPU usage rate exceeds 90%, an e-mail notification can be sent to the user. To send this notification, you set usage of the physical CPU exceeds 90% as the threshold for an abnormal condition, and you specify that an e-mail notification be sent to the user when this threshold is reached. The action that is taken when the threshold is reached is called an action. The following types of actions can be taken:
An entity that defines a threshold and an action is called an alarm. A table that lists more than one alarm is called an alarm table. After you define an alarm table, you link it to PFM - RM for Virtual Machine. The act of linking an alarm table to PFM - RM for Virtual Machine is called binding. Once an alarm table is bound to PFM - RM for Virtual Machine, the user is notified when the performance data being collected by PFM - RM for Virtual Machine reaches the threshold defined by an alarm.
By defining alarms and actions in this way, you can discover operational problems in the virtual environment early, and take the necessary corrective actions.
For details about how to set alarms and actions, see the chapter that explains operation monitoring using alarms in the JP1/Performance Management User's Guide.