Parallel Python (pp) and subprocess : task never ends on the client

Question

I'm trying to use the module "Parallel Python" (pp) to distribute the computation of a "big" tool (gdal2tiles, if you've heard of it). I was running a few simple tests to get familiar with pp, but I ran into a problem that I can't explain.

Here is the code :

import pp

def fun():
    import subprocess
    p = subprocess.Popen("dir", shell=True)
    p.wait()

ppservers=("*",)

job_server=pp.Server(ncpus=0,ppservers=ppservers)

if __name__=='__main__':

    FORMAT = u'%(asctime)-15s %(message)s'
    import logging
    logging.basicConfig( level=logging.DEBUG, format=FORMAT )
    f = job_server.submit(fun)
    r=f()
    job_server.print_stats()

The "dir" in subprocess.Popen is just here to pass a valid shell command.

On my ppserver, I run "ppserver.py -a -d". On the client's console, it appears that the task never ends, whereas it does end on the server.

Before going any further in my tests, I would like to understand what's happening here, because my final program is way more complicated.

Thanks!