The alert monitor uses thresholds and rules to generate alerts if anything abnormal occurs in your R/3 System or its environment. Alerts direct your attention to critical situations so that you do not have to discover these for yourself. With auto-reaction methods, the alert monitor can notify you of a problem or take action even if you are not currently working in the alert monitor.

For the system administrator, an alert is a "trouble ticket" or a "service request." You should respond to every alert. Use the alert browser and the analysis methods provided by the alert monitor to investigate the alert. When you have resolved the problem or know that you can safely ignore it, you can set the alert to complete. It is then deleted from the display and saved in the R/3 database.

The alert monitor reports alerts up through the monitoring tree. This means that the color-coding for a monitoring tree element (MTE) always represents the highest alert in all MTEs in its branch. For example, if a host system is color-coded red, then one or more of the components in the host system monitoring tree has a red alert.

An alert is uniquely assigned to a node in the monitoring tree (a monitor tree element, MTE), and created when values are sent by a collection method to a monitor tree element. The alert monitor evaluates this incoming information against alert thresholds. If the incoming data exceeds or violates a threshold, then the alert monitor triggers an alert. The alert monitor generates an alert for every abnormal event and keeps track of every alert until it is set to status "complete".

Any MTE that produces an alert is automatically visible in the alert monitor. This is true even if the offending MTE is intended for display only at a higher display level than is currently active (overview, analysis, expert analysis).