1.2.7.1.1.1 Alert Properties

The following table lists the Alert Properties options.

Table 1-60 Alert Properties Options

Options	Descriptions
Name	The name of the alert. It must be unique among the alerts globally. The name value is limited to 255 characters and must be unique within the project. The initial character must not be "."
Type	The type of the alert; it is based on monitoring policy. The drop-down list includes the following items: Call Path Service When you select a service alert, the Metrics Independent Checking check box appears. It allows Oracle TSAM Plus to check service timeouts independent of metrics collection. The alert can be generated while the service is running. If checked, “Drop Message” is removed from the “Action” option drop-down menu. Service Statistics You can define one or more service statistics alerts for a Tuxedo domain or a domain group. When you select a service statistics alert, the configuration panel becomes different than other alert types. You can configure TSAM Plus to compare the instant metric data with a static threshold or historical data to evaluate and trigger the service statistics alerts. For more information, see Adaptive Settings (For Service Statistics Alerts Only). XA Transaction GWTDOMAIN BRIDGE GWWS
Severity	The severity drop-down list contains the following severity levels: Fatal Critical Warn Information
Interval	Accepts an integer value in seconds. The current alert is evaluated only once during this interval. The purpose is to throttle alert volume in similar scenario. If the interval value is 0, any alert is reported.
Action	Specifies what kind of action is executed when an alert definition is set to "true". The following actions are supported: Publish to Tuxedo Event Broker An event is posted to the Oracle Tuxedo event broker. The event name is the alert name by default; you can also specify it manually. The buffer is an FML32 buffer containing the metric snapshot. Drop Message If this alert evaluation is true, the request message is dropped and the call fails (TPESYSTEM). This only applies to Call Path and Service type alerts. The evaluation points come before the request sent to the IPC queue (call path) and after it is retrieved from the IPC queue (call path, service). This alert can be used to avoid unneeded service processing for stagnant requests. For example, if the request message "Wait Time" in the IPC queue exceeds a particular threshold, the client may have already timed out. In this case, the action can drop the stagnant request and the application server will not process it. Caution: If the drop message action is executed in a embedded Tuxedo call, the whole call fails.