While trying to troubleshoot the previous issue, I set clustermq.ssh.timeout (to 10, 100, 1000), and my jobs started to, and then kept crashing with the following error:
Error in private$zmq$poll(sid, timeout) : 1 peer(s) lost
Removing the timeout flag and reloading the package into a new R session doesn't seem to help. Does the ssh server persist? Do I need to bring it down manually somehow?
While trying to troubleshoot the previous issue, I set clustermq.ssh.timeout (to 10, 100, 1000), and my jobs started to, and then kept crashing with the following error: Error in private$zmq$poll(sid, timeout) : 1 peer(s) lost
Removing the timeout flag and reloading the package into a new R session doesn't seem to help. Does the ssh server persist? Do I need to bring it down manually somehow?