[zeromq-dev] Increasing transparently ZMQ performance up to 2400%

Dave Lambley dave.lambley at gmail.com
Thu Feb 27 16:18:26 CET 2014


On 27 February 2014 14:56, Lindley French <lindleyf at gmail.com> wrote:

> One question: Don't you have to get to the kernel eventually if you're
> really going out on the network? How does UFS help in this case? It may
> bypass the kernel in the loopback case, but you don't need to be using
> TCP/IP in that case anyway; you can use ipc or inproc.
>

I have not come across UFS before, but the idea of moving TCP handling out
of the kernel has been floating around for a couple of years.

http://highscalability.com/blog/2013/5/13/the-secret-to-10-million-concurrent-connections-the-kernel-i.html

Cheers,
Dave
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.zeromq.org/pipermail/zeromq-dev/attachments/20140227/fbcf5943/attachment.htm>


More information about the zeromq-dev mailing list