More than just a server load test, we WANT the serverkit to crash. By running the serverkit through the debugger, we can see each line of code being executed as it happens, or rather, when the kit crashes, the debugger will be showing us the last line of code that was being executed.
Instead of trying to find a needle in a haystack, this will tell us precisely what is causing the kit to crash, and therefore, how to fix it precisely. It may only take 10 minutes, for all we know. It is also quite possible, depending upon the nature of the bugs found, that we will fix, recompile and restart the test on the fly, attempting to locate as many bugs as we can, thus in this fashion, we can get quite a lot of work done and make huge leaps in finally getting the serverkit to work bug free!!