End-to-End Exception Analysis

The E2E Exception Analysis follows the same approach than the other previously described tools but displays all kinds of exception in the managed systems. This starts with dumps or syslog messages in an ABAP system but also includes for example Out of memory situations in a J2EE engine. As the information about the exceptions is persisted on the Solution Manager, it is prevented that the exceptions are overwritten by a round robin log mechanism or cleanup procedures in the managed system. Of course, not all information is persisted on the Solution Manager system (only relevant header data which is needed for error statistics as type of the exception, description, severity, time and date, user that caused the exception,…)

Based on the available statistical information on exceptions, one of the main use cases is that you want to perform an exception trend analysis to find out if after implementing a special path or update more or less exceptions are occurring. Another use case is to ensure that with a specific patch or a SAP note a problem is fixed and the exception does not occur anymore.

Figure: Overview of exceptions

In the overview (see above figure) you can identify in the history diagram the distribution of exceptions over the selected timeframe,

The drill down allows you to identify the exact exception which is only summarized in this diagram. Depending on the exception type more data is stored in BW to allow an efficient first analysis, but the complete exception data is always available in the managed system itself and can be reached with a jump-in from the Solution Manager system. In the following example you can see that many dumps have been triggered in the selected timeframe. Using the green jump-on, you could now navigate to the managed system and analyze the dump in transaction ST22.

Figure: List of dumps

Similar possibilities exist for all kind of supported exception, like system and application errors in a J2EE engine with a jump-in to the Log Viewer.