On Sat, Jan 8, 2022 at 10:40 AM Reinis Rozitis <r@roze.lv> wrote:

> >>
> https://www.nginx.com/blog/testing-the-performance-of-nginx-and-nginx-plus-web-servers/
> >
> > I don't view the test described as valid because the test is between one
> client and one server. I'm interested in testing with one server and many
> clients.
>
> wrk (used in the tests) [1] is a multithread/connection benchmark tool.
> You can simulate how many clients you want (basically you are just limited
> by ephemeral port tuple count ~65k .. but then again can work around it by
> assigning multiple IPs to the server/client).
>
>
>
Yes. I've used wrk in my own tests. It's a very good tool. The problem is I
don't think wrk really can simulate communication from many clients. There
is something different between simulating many clients with one real client
and actually testing with many real different clients. I can't put my
finger on what the difference is. But there must be a difference. Otherwise
why is my application running into such performance limits as mentioned in
this question on stackoverflow
https://stackoverflow.com/questions/70584121/why-doesnt-my-epoll-based-program-improve-performance-by-increasing-the-number
?


> [1] https://github.com/wg/wrk
>
> rr
>
> _______________________________________________
> nginx mailing list
> nginx@nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx
>
_______________________________________________
nginx mailing list
nginx@nginx.org
http://mailman.nginx.org/mailman/listinfo/nginx