Analyzing Trace Events in the Aggregated Call Tree

Remember, the Call Sequence tool provides you with a time sequence of trace events as a hierarchical tree. By contrast, the Aggregated Call Tree shows the trace events in an aggregated form. In the latter case, the trace events are aggregated By Call Stack and displayed in a tree hierarchy.

Aggregation By Call Stack means that the runtime data of the multiple executions that occurred via the same call path (the same caller) is added up to a single trace event entry. In the results display of the Aggregated Call Tree, the Executions column displays how many trace events were aggregated in a single trace event entry.

Imagine a class with methods M1, M2, and M3. Now imagine a program PROG that calls M1 twice from the same call position. The first call of M1 implies also the call of method M2. Finally, M2 is called through the method M3, as shown in the figure below.

In the Aggregated Call Tree, the ABAP Profiler creates a single trace entry for M1. The Executions column has the value 2. Due to the fact that both calls to M1 occurred via the same call path, the ABAP Profiler aggregates the two calls to M1. The number of Executions indicates that two executions of method M1 have been added up.

Figure 1: Trace events in the Call Sequence and in the Aggregated Call Tree

Use the Aggregated Call Tree to perform tasks such as the following:

• Analyze all usages of a procedure (such as a method) - how many executions of a method occur via a particular call path?
• Analyze how many executions of specific events occur with the same call stack.
• Analyze which specific statements are executed and how often within a procedure, and which call stacks occurred for outgoing calls of the procedure.