9.6.2015, 9:24, Antti Karanta kirjoitti: > The error is reproducible at least by killing the process to which the > messages are addressed to before it has received all of them. It seems > to occur randomly at other times as well.
I turned off ZMQ_ROUTER_MANDATORY and the problem seems to disappear (at least for the case where the receiving dealer is crashed). According to http://api.zeromq.org/4-0:zmq-setsockopt#toc24 zmq should return EHOSTUNREACH when a message can not be routed - why do I get unknown error instead? ::Antti:: _______________________________________________ zeromq-dev mailing list [email protected] http://lists.zeromq.org/mailman/listinfo/zeromq-dev
