Unable to exit/quit or even control-c

[javadba]

I have tried all of those and get the following:

2017-08-15 19:27:10 pycos - uncaught exception in _shutdown/140686503097984: Traceback (most recent call last): File "/usr/local/lib/python2.7/dist-packages/pycos/init.py", line 3687, in _schedule retval = task._generator.send(task._value) File "/usr/local/bin/dispynode.py", line 1605, in _shutdown sock = AsyncSocket(socket.socket(addrinfo.family, socket.SOCK_STREAM), AttributeError: 'NoneType' object has no attribute 'family'

This is happening on a client-only node that had not connected succesfully to the scheduler_node

[javadba]

There is a follow-up additional problem: it seems that after I killed the process by kill -9 there is a residual PID file? I am unable to restart the dispy:

2017-08-15 19:29:09 dispynode - dispynode version: 4.8.3, PID: 7836 Traceback (most recent call last): File "/usr/local/bin/dispynode.py", line 1919, in _dispy_node = _DispyNode(**_dispy_config) File "/usr/local/bin/dispynode.py", line 274, in init config.get('pid', None)) Exception: Another dispynode server seems to be running with PID 6285; terminate that process and rerun with "clean" option

[javadba]

Just to be clear: there is no dispy process running: it is a cleanup issue.

$ ps -ef | grep 6285 (nothing ..)

[javadba]

OK - so re-running with --clean takes care of that second issue

[pgiri]

Client won't terminate with Ctrl-C, as thread is blocked waiting for cluster to finish. One way to terminate is to suspend the program (Ctrl-Z) and kill the process.

A node can be shutdown either by giving "quit" command at dispynode command interface (if it is not running in the background), or by calling dispynode_shutdown() in "close" function if the node has been started with --client_shutdown option.