I would say the overhead is non measurable due to other external, dominant factors. There is definitely an additional overhead from a protocol with a handshake on top of TCP compared to simple ICMP. That result just means there is something else which is the dominant factor.
On Thu, 2018-05-03 at 16:13 +0300, Ernest Zed wrote: > If I get ping ~50us and zmq performance reports ~50us it means that > zmq has > virtually zero overhead, right? > > On Thu, May 3, 2018 at 4:05 PM, Luca Boccassi <[email protected] > m> > wrote: > > > On Thu, 2018-05-03 at 15:44 +0300, Ernest Zed wrote: > > > Hi All, > > > I'm running performance tests in accordance with this document > > > http://zeromq.org/results:perf-howto > > > > > > It is stated that the latency measured just one-way. I guess this > > > document > > > is very old and I wanted to ask it is still holds true. > > > > The measuring code is very simple, I suggest to just look at it. > > The latency is measured for a roundtrip. > > > > > Second, what is the > > > expected overhead of zmq? For example, if I ping one machine from > > > the > > > second one and get, say, 50us, what is the latency I could expect > > > from zmq > > > when running local/remote_lat test? 25, 50, 100, 200us? > > > > > > Sincerely, > > > Ernest > > > > The environment can vary the performances, so you should simply > > measure > > it and see what's it like with your hardware and network. > > > > -- > > Kind regards, > > Luca Boccassi > > _______________________________________________ > > zeromq-dev mailing list > > [email protected] > > https://lists.zeromq.org/mailman/listinfo/zeromq-dev > > > > > > _______________________________________________ > zeromq-dev mailing list > [email protected] > https://lists.zeromq.org/mailman/listinfo/zeromq-dev -- Kind regards, Luca Boccassi
signature.asc
Description: This is a digitally signed message part
_______________________________________________ zeromq-dev mailing list [email protected] https://lists.zeromq.org/mailman/listinfo/zeromq-dev
