I intended to file a bug because of the shutdown problem but found out there is this ticket already.

In our application we sometimes run into this problem that you can't exit because of a hanging TCP connection used inside a ThreadPoolExecutor task. 

executor.shutdown(wait=False) does not help - the process still hangs.

Another problem was inside our tests were a bug caused a hanging task inside a thread pool executor and the test runner (pytest) would not terminate because of this.

I actually went so far to drop the thread reference inside the concurrent.futures.thread module. :-(

A documented way to drop a hanging executor thread would be very much appreciated.