A. State of the Art and Practice
Performance testing of enterprise applications is manual,
laborious, costly, and not particulary effective. Currently, a
prevalent method for performance testing is intuitive testing,
which is a method for testers to exercise the AUT based
on their intuition and experience, surmising probable errors
[13]. Intuitive testing was first introduced in 1970s as an approach
to use experience of test engineers to focus on errorprone
and relevant system functions without writing timeconsuming
test specifications thus lowering pre-investment
and procedural overhead costs [13].When running many different
test cases and observing application’s behavior, testers
intuitively sense that there are certain properties of test cases
that are likely to reveal performance bugs. Distilling these
properties automatically into rules that describe how these
properties affect performance of the application is a subgoal
of our approach.