Thanks for the response, JettR.
You have correctly described the problem I'm having. Let me clarify a couple points. As it turns out, it doesn't seem to matter whether I enter and exit the synchronized section using the batch sync steps or if I set the properties of the individual steps to run in "one-thread-only" mode. It also doesn't seem to matter whether I've assigned names to the synchronized sections or not. One other thing - it doesn't matter if an error occurred - the problem is an execution terminating... I typically terminate an execution (continue to clean-up) if an error occurs, but this can also be reproduced by manually terminating the execution. Also, the problem is more than just the terminated execution running the synchronized section before the other executions are ready - the section is actually executed twice. (Once by the terminated execution, and once by the next execution in line.)
In short, I seems like one potential work-around would be to add some additional code to jump past the cleanup sync section if 1) the execution is terminating and 2) it isn't the only execution running. But it seems like there should be a better way than that.
I've attached a simple sequence that can be used to reproduce the problem. If you run it without terminating any threads, everything will behave as expected. But if you terminate an execution while in the delay in the Main step group, you should see the one-thread-only synchronized step in the cleanup group (a TestStand dialog) get displayed twice.
Let me know if you have any questions. Thanks.
