[zeromq-dev] zmq "bug"

Martin Sustrik sustrik at fastmq.com
Mon Dec 22 14:49:34 CET 2008


Comments inlined:

> zmq_server &
> ./q_test.py
> ./q_send.py Q_TEST hello
>  -- q_test prints 'hello'
> killall zmq_server

At this point q_test still runs, however, there is no directory service 
(zmq_server) running that is able to identify actual location of q_test 
on the network.

> ./q_send.py Q_TEST hello

I would expect q_send to fail here as it cannot access zmq_server. Am I 
right?

>  -- q_test IS STILL RUNNING but prints nothing

q_test prints nothing because q_send failed.

>  -- I start zmq_server again
> zmq_server &

>  -- Ctrl-C to q_test.py does nothing.

This is pretty strange. We'll have a look at it.

> I try to terminate it with an
> 'EXIT' message
> ./q_send.py Q_TEST EXIT
> ~/zmq/report|0|$ ./qsend.py Q_TEST EXIT
> python2.6: api_thread.cpp:151: void zmq::api_thread_t::bind(const
> char*, const char*, zmq::i_thread*, zmq::i_thread*): Assertion `false'
> failed.
> Aborted
> 
> THE BUG: q_test.py is now hanging. no new q_test.py can be launched
> b/c the port is in use:
> 
> ~/zmq/report|0|$ ./q_test.py
> Address already in use
> rc == 0 (tcp_listener.cpp:125)
> Aborted

Thanks for reporting the problem!
Martin



More information about the zeromq-dev mailing list