Doron, we need to fix the man page then. Can you read the RFC http://rfc.zeromq.org/spec:41 and tell me which pieces exactly are incorrect? I'll do the rest.
On Mon, Apr 4, 2016 at 10:04 AM, Doron Somech <[email protected]> wrote: > You have to specify ZMQ_DONT_WAIT for the send method to return eagain > instead of blocking. > > On Mon, Apr 4, 2016 at 5:31 AM, Robert Holterhoff <[email protected]> > wrote: >> >> I'm using the "unstable" 4.2.0 release and have a question regarding >> ZMQ_SERVER sockets. >> >> The documentation suggests that ZMQ_SERVER sockets will not block but >> return EAGAIN when the HWM is reached. I am finding that the send() >> blocks instead. >> >> Actually, this is exactly the behaviour I want so I'm not complaining. My >> app works beautifully, running at full rate, quickly filling up various >> clients queues and then automatically rate-limiting to the slowest client >> because of the blocking behaviour. >> >> However, I made the observation that the "back-off" part of the code which >> I had written expecting EGAIN to be returned is never exercised. >> >> I just want to know if I can rely on this behaviour in the future so I can >> remove my hacky arbitrary "sleep-for-a-bit back-off" code, or if this is a >> bug because SERVER/CLIENT sockets are still in their infancy. >> >> Thanks, >> >> Rob Holterhoff >> >> >> _______________________________________________ >> zeromq-dev mailing list >> [email protected] >> http://lists.zeromq.org/mailman/listinfo/zeromq-dev > > > > _______________________________________________ > zeromq-dev mailing list > [email protected] > http://lists.zeromq.org/mailman/listinfo/zeromq-dev > _______________________________________________ zeromq-dev mailing list [email protected] http://lists.zeromq.org/mailman/listinfo/zeromq-dev
