thread-(dead)lock

[Thomas Thiele]

I have a server (derived from mutithreading SocketServer.py) and a
client running on the same host, connectet via AF_INET-Socket.

The server creates for every request a new thread. So I have three
threads now. The main thread, a second thread that executes a
loop-function, in which my problematic function is called, and a third
thread created by a request from my client.

class Requesthandler:
    def __init__(self):
        .....
        self.lock = thread.allocate_lock()

    def Loop(self):                             #runs in one thread
        time.sleep(self.updatetime)
        self.ProblemFunction()

    def ProblemFunction(self):
            print "before aquire()"
         self.lock.aquire()
            print "locked"
        ....do something....
        self.lock.release()
            print "unlocked"

    def HandelRequestFromClient(self): #runs in another thread
        self.ProblemFunction()

    .........


The client causes the server periodically to call its ProblemFunction()
too.

class Client:
    def __init__(self):
        .....
        signal.signal(signal.SIGALRM, self.LoopSignalHandler)
        signal.alarm(self.updatetime)

    def LoopSignalHandler(self):
        Cause_Server_to_execute_ProblemFunction()


    ........

My problem is, that it works for 20 minutes an then I have a deadlock.
"before aquire()" is printed and the function is blocked in both
threads.
The problem is not the "do something"-stuff!
Any ideas why?

Thanks Thomas