Search code examples
pythonzeromqdistributed-computingpyzmqlow-latency

pyzmq: socket breaks with "ZMQError: Operation cannot be accomplished in current state"


I am very new to ZeroMQ and trying to construct a very basic messaging system. The code is very much based on examples from here with some twists

For some reason following a last message arriving to a frontend socket (hbm ,tx) the code throws errors that I'm not sure of it's origin

Sorry for the long code/output I felt it is important to stress what is the issue and help in debug


Below is my code + output :

s_frontend = "ipc://frontend.ipc"
s_backend  = "ipc://backend.ipc"

import time

def hbm():
    """Worker task, using a REQ socket to do load-balancing."""
    socket = zmq.Context().socket(zmq.REQ)
    socket.identity = u"hbm".encode("ascii")
    socket.connect(s_frontend)
    # Tell broker we're ready for work
    socket.send(b"READY")
    while True:
        msgs = socket.recv_multipart()
        print("hbm got something", msgs)
        if "BE READY" in msgs:
            print("hbm:BE is ready for some work")
        time.sleep(3)

def tx():
    """Worker task, using a REQ socket to do load-balancing."""
    socket = zmq.Context().socket(zmq.REQ)
    socket.identity = u"txtx".encode("ascii")
    socket.connect(s_frontend)
    # Tell broker we're ready for work
    socket.send(b"READY")
    while True:
        msgs = socket.recv_multipart()
        print("tx got something ", msgs)
        if "BE READY" in msgs:
            print("tx:BE is ready for some work")
        time.sleep(3)


def dev():
    """Worker task, using a REQ socket to do load-balancing."""
    socket = zmq.Context().socket(zmq.REQ)
    socket.identity = u"dev".encode("ascii")
    socket.connect(s_backend)
    # Tell broker we're ready for work
    socket.send(b"READY")


def main():
    # Prepare context and sockets
    context = zmq.Context.instance()
    frontend = context.socket(zmq.ROUTER)
    frontend.bind( s_frontend )                 # ( "ipc://frontend.ipc" )
    backend = context.socket(zmq.ROUTER)
    backend.bind( s_backend )                   # ( "ipc://backend.ipc" )

    def start(task, *args):
        process = multiprocessing.Process(target=task, args=args)
        process.daemon = True
        process.start()

    start(hbm)
    start(tx)


    time.sleep(1)
    start(dev)

    clients = []
    poller = zmq.Poller()
    poller.register(backend, zmq.POLLIN)
    #poller.register(frontend, zmq.POLLIN)
    all_is_ready = False
    while True:
        sockets = dict(poller.poll(timeout=1))
        #print(sockets)
        soc = None
        if backend in sockets:
            print("got something from backend")
            msg = backend.recv_multipart()
            print(msg)
            print("adding frontend to poller")
            poller.register(frontend, zmq.POLLIN)
            print("backend is ready, notify frontend")

        elif frontend in sockets:
            print("got something from frontend")
            msg = frontend.recv_multipart()
            print(msg)
            clients.append(bytes(msg[0]))

        elif len(clients) == 2 and all_is_ready is False:
            all_is_ready = True
            for c in clients:
                print("sending response to", c)
                time.sleep(0.1) # just to prevent print overlap
                frontend.send_multipart([c, b"", b"BE READY"])
        else:
            print("so much work, no rest, sleeping for 3")
            time.sleep(3)


    # Clean up
    backend.close()
    frontend.close()
    context.term()


if __name__ == "__main__":
    main()

Running this code results in a following output :

so much work, no rest, sleeping for 3
got something from backend
['dev', '', 'READY']
adding frontend to poller
backend is ready, notify frontend
got something from frontend
['hbm', '', 'READY']
got something from frontend
['txtx', '', 'READY']
sending response to hbm
sending response to txtx
hbm got something ['BE READY']
hbm:BE is ready for some work
tx got something  ['BE READY']
tx:BE is ready for some work
so much work, no rest, sleeping for 3
Process Process-1:
Traceback (most recent call last):
  File "/usr/lib/python2.7/multiprocessing/process.py", line 258, in _bootstrap
    self.run()
  File "/usr/lib/python2.7/multiprocessing/process.py", line 114, in run
    self._target(*self._args, **self._kwargs)
  File "/home/vm/repo/pyzmq_examples/hal_example/example1.py", line 104, in hbm
    msgs = socket.recv_multipart()
  File "/usr/local/lib/python2.7/dist-packages/zmq/sugar/socket.py", line 475, in recv_multipart
    parts = [self.recv(flags, copy=copy, track=track)]
  File "zmq/backend/cython/socket.pyx", line 791, in zmq.backend.cython.socket.Socket.recv
  File "zmq/backend/cython/socket.pyx", line 827, in zmq.backend.cython.socket.Socket.recv
  File "zmq/backend/cython/socket.pyx", line 191, in zmq.backend.cython.socket._recv_copy
  File "zmq/backend/cython/socket.pyx", line 186, in zmq.backend.cython.socket._recv_copy
  File "zmq/backend/cython/checkrc.pxd", line 25, in zmq.backend.cython.checkrc._check_rc
    raise ZMQError(errno)
ZMQError: Operation cannot be accomplished in current state
so much work, no rest, sleeping for 3
Process Process-2:
Traceback (most recent call last):
  File "/usr/lib/python2.7/multiprocessing/process.py", line 258, in _bootstrap
    self.run()
  File "/usr/lib/python2.7/multiprocessing/process.py", line 114, in run
    self._target(*self._args, **self._kwargs)
  File "/home/vm/repo/pyzmq_examples/hal_example/example1.py", line 130, in tx
    msgs = socket.recv_multipart()
  File "/usr/local/lib/python2.7/dist-packages/zmq/sugar/socket.py", line 475, in recv_multipart
    parts = [self.recv(flags, copy=copy, track=track)]
  File "zmq/backend/cython/socket.pyx", line 791, in zmq.backend.cython.socket.Socket.recv
  File "zmq/backend/cython/socket.pyx", line 827, in zmq.backend.cython.socket.Socket.recv
  File "zmq/backend/cython/socket.pyx", line 191, in zmq.backend.cython.socket._recv_copy
  File "zmq/backend/cython/socket.pyx", line 186, in zmq.backend.cython.socket._recv_copy
  File "zmq/backend/cython/checkrc.pxd", line 25, in zmq.backend.cython.checkrc._check_rc
    raise ZMQError(errno)
ZMQError: Operation cannot be accomplished in current state
so much work, no rest, sleeping for 3
Traceback (most recent call last):
  File "/home/vm/repo/pyzmq_examples/hal_example/example1.py", line 244, in <module>
    main()
  File "/home/vm/repo/pyzmq_examples/hal_example/example1.py", line 233, in main
    time.sleep(3)
KeyboardInterrupt

Process finished with exit code 1


Solution

  • Q : "... throws error ... I'm not sure of it's origin "

    ZMQError: Operation cannot be accomplished in current state
    

    The origin is that in each use-case, where REQ-Archetype was used, the code was not meeting the native API conditions :

    ...
    socket = zmq.Context().socket( zmq.REQ ) #--------------------- REQ socket Archetype
    socket.identity = u"hbm".encode( "ascii" )
    socket.connect( s_frontend )
    socket.send( b"READY" ) #-------------------------------------- REQ.send()-s
    while True:             # ..................................... REQ next can .recv()
        #      socket-FSA( of a type of REQ ) can execute a .recv() iff.send() preceded.....?
        msgs = socket.recv_multipart() #- - - - - - - - - - - - - - REQ.recv_multipart()-s
        #.......................................................... REQ next can .send()
        continue                       #-?-?-?-?-?-?: DID IT TRY TO REQ.send()? NO, NEVER...!
    

    Do [you] mean by that that REQ must do iterative recv send and cannot do any other way ? (send recv recv) – LordTitiKaka

    REQ-Archetype has a hardwired expectation of a mandatory sequencing of being kept :
    .send() - .recv() - .send() - .recv() - .send() - .recv() - ... ad infinitum

    Any violation of this flip/flop-changing its internal Finite-State-Automaton (FSA) internal-state causes the said ZMQError: Operation cannot be accomplished in current state