Currently, the fsm tests capture errors and log them after joining all the threads. This makes it challenging to corelate the timing of when the error ocurred to the rest of the events that happened during the test.