Are there significant performance differences between epoll and kqueue?

My development machine is a MacBook (which of course has kqueue). However, in production we use Linux (which, of course, uses epoll). Obviously, to know the performance characteristics of my code, I need to run it with epoll. However, is the performance that I see under kqueue a worthy approximation of what I will see with epoll? Or are there situations where productivity can be significantly different? For the most part, it seems that kqueue and epoll are pretty much similar in terms of performance, but I really didn’t check very carefully.

If that matters, I use tornadoes in Python.

+4
source share
2 answers

kqueue is superior to epoll according to Berkeley University, mainly because epoll does not support multiple percent updates in a single system call, while kqueue can do this with kevent ().

There is also a white paper on the differences between comparison results 2 and performance.

http://www.eecs.berkeley.edu/~sangjin/2012/12/21/epoll-vs-kqueue.html

+5
source

All Articles