I am testing my application using webload for stress testing. During the same i came across issue's which were not there during few runs like API taking too much time or queries taking too much. So now am planning to check these things using profiling tool like jprofiler. But jprofiler crashed when started recording CPU data with JDBC and method statistics. This happened multiple times. Am using weblogic app server. After so many tries i thought profiling tool cannot be attached during stress testing.
Try using App Dynamics, they claim that the overhead is very low (2-3 %) http://www.appdynamics.com/products-free-download.php.
In case you are using the Jrockit JVM, you can also use a tool called Jrockit Mission Control, which also (as claimed by Oracle) has a low overhead.
Another option is that you could connect to JVM using Visual VM to see how threads are executing under stress, is there contention, which can be figured out by seeing if thread are waiting, (will show in red color) or is executing fine (all green). In case you are seeing lot of reds taking a thread dump might be useful.