Global Performance Analysis

Software, system landscapes and software development processes are becoming ever more complex in environments in which business processes run across several systems. Manual performance tests of individual programs or transactions, with standard test and analysis tools, such as Performance Trace (ST05), ABAP Runtime Analysis (SE30) or Business Transaction Analysis (STAD), are therefore no longer sufficient, because performance will usually depend upon the interaction of components in different systems.

Global Performance Analysis (transaction ST30, component BC-TWB-TST-P-GPA)avoids time and resource-consuming manual performance tests, with the following functions:

• Cross-system performance tests for ABAP and non-ABAP components
• Get performance statistics for all application and system components of a business process
• Automatic performance of performance analyses, e.g. for regression or scalability tests
• Record performance degradation during development cycle
  Caution

  ST30 is not suitable for stress tests. Benchmark tools (benchmark = simulation of system load with a lot of users) are used for this purpose.

The performance tests performed in the tested systems by ST30 are saved in a central database. You can access it at any time to perform analyses, such as long-term comparisons or statistical analyses (scalability or regression tests, etc.)

ST30 calls the performance statistics created in the tests, by using:

• the statistics records (see the transactions STAD, ST03N, ST03G, etc.) and data generated by the performance analysis (transaction ST05)
• the code inspector function (transaction SCI)

You can access the performance statistics in the central database, from transaction ST30 or ST33.

Main ST30 functions:

• Automatic Performance Tests
• Display Performance Test Results

You can also perform manual performance tests in ST30, but it is mainly intended for automatic tests.

Automatic performance tests in ST30 are automatic and repeated runs of business processes or scenarios, using all relevant system landscape components. The tool collects all performance-relevant data from all systems involved, and saves it centrally in a database.

The performance tests are bases on eCATT test configurations (extended Compter-Aided Test Tool, transaction (SECATT)). eCATT test configurations are also essential for automatic tests with ST30, although they were developed mainly for functional tests. They are based on test scripts which contain all steps which were performed in the business process or scenario to be tested. Test scripts can be recorded or written manually, and edited.

The most important attributes of ST30 are:

• The performance tests performed with ST30 run in a central test system. They are dynamic (they are performed in runtime), as opposed to static checks (not performed in runtime), as, for example, the Code Inspector in the program source text.
• The performance statistics display may not be consistent, because the system components in a test can have different releases. ST30 always displays the statistics data in one unit of measurement (UOM), regardless of which UOM is used in the tested system component. For example memory usage is always in MB.
• 64-bit support for a system is automatically recognized and taken into account.
• Differences in time zones between systems (e.g. USA and Europe), are taken into account automatically. The values do not need to be converted.

An automatic performance test comprises the following steps (assuming that the scenario to be tested is already in an eCATT test script):

1. ST30 starts the eCATT test script in the central test system.
2. The eCATT test script runs the scenario using all system landscape components (programs, transactions).
3. ST30 gets the performance statistics from all system components involved.
4. ST30 saves the collected performance statistics in the central database.

More Information:

Automatic Performance Tests

Manual Performance Tests

Display Performance Test Results