Examples for Using Process Monitoring

The following contains examples for the use of Process Monitoring:

Status Check/Error Determination

Administrators must check the status of their processes regularly. Here you must make sure that all process instances have run completely. If an instance has not run completely, the reason for this must be analyzed.

To give you an initial general overview, the statistical data of the Process Overview (Process Monitoring initial screen) tells you the total number of instances as well as the number of instances with a particular status (such as with errors, or completed). You also see the average duration, as well as the minimum and maximum duration of the process instances. The minimum and maximum durations are predominantly of significance for Service Level Agreements. The process flow graphic also allows you to see immediately the point at which problems have occurred.

If you require more information about processes that have a particular status, such as processes that have errors, you can click the number of processes with errors in the statistical data to display a list of all the instances that have errors. If you select an instance from this list, you are taken to the corresponding Instance View.

In the instance view, you can see the status for the individual instance steps. If you select a step that has errors, detailed information about the step is displayed. This allows you to view information for a single IDoc, for example.

Locating Performance Bottlenecks

Slow performance is not as obvious as for example the status of processes or process steps, as it does not always trigger an alert. Slow performance is only automatically displayed as an error in the monitoring interface if the processing time exceeds a predefined threshold value.

You can find information about process performance in general in the process overview. There you can find the maximum and minimum processing duration under Statistical Data. If you notice that the performance is not satisfactory, you must look at one of the instances in greater detail. To do this, select a process instance that has a long processing duration, or even one that has the maximum processing duration. The instance view for the instance you have just selected is displayed. Here you can see the sequence of process steps of the instance. The start time is available for each step. In this way you can identify which process step took too long. If you then choose to view the Attribute Display for this step, you are given information required for the detailed performance analysis. 

You may want to view additional information about a process step, such as memory consumption. You can find this information in the functional trace (transaction STATTRACE), which saved the statistical data record for a process step. The statistical data record contains detailed information about the response time, database time, memory consumption, amount of data transferred and so on. You can use the transaction ID in the attribute display to find the corresponding statistical data record in the functional trace.

Process Documentation

By using the PMI user interface, new colleagues in your IT organization can get a quick overview of the structure of your processes. For example, these colleagues have to know which technical steps the process consists of, and at which points processing the steps occurs asynchronously. This allows the new colleagues to get used to the processes quickly, which saves on training costs.

 Process Monitoring Display and Administration