ZeroHTTPd

Very insightful article. Thanks for putting it together. Looking forward to an update with io_uring.

Great article, explains things better than most O'Reilly books.

Heh, you beat me to publishing my own simple HTTP server. Though the one I'm working on is even more dumbed down, only supporting GET to serve prepared HTTP messages from a hash…

Had a quick glance and while the content seems worthwhile I am a little saddened by the code. There are numerous calls to strcpy() which is a buffer overflow waiting to happen. I…

strcpy() is actually fine when used with literal/constant strings and the buffer is large enough. The compiler can even optimize it and issue warnings in case of overflow.…

So there is no keep alive in these tests including lb to application? Makes sense why the qps is so low on a single core for all versions tested if nginx is having to reopening a…

Testing "Linux server architectures" without considering keepalive is worse than useless.

> I guess it is 1.1 Negative ghost rider, it's hard-coded to 1.0: https://github.com/shuveb/zerohttpd/blob/master/01_iterative/main.c#L468 (in reality it's barely even that e.g.…

Need a uring based one! http://kernel.dk/io_uring.pdf

Needs async IO variant as well...

Notably lacks part VIII, kqueue-based server. This is because Linux doesn't support kqueue. Got to make do with their pointless NiH syndrome fueled epoll().

I've never really done any serious work on anything other than Linux. But I always wanted to try out kqueue(). Even if I don't add a separate section on kqueue(), I think it…

It would be nice to also compare memory use, which will vary between implementations.

Awesome comparison, super curious how this would scale out on the current generation of big CPU machines. E.g. Epyc with 64 cores, would threads still perform that well?

OpenBSD doesn't have sendfile as far as I know so maybe s/Open// there too

That manpage is an import from freebsd, not part of base -- As far as I can see on my system it's not there: OpenBSD: ``` $ man 2 sendfile man: No entry for sendfile in section 2…

Yes, IIRC, even Nginx uses the general structure you're describing.

nginx and others use worker processes instead of worker threads, with epoll in each worker. This is for robustness and garbage collection, AFAIK.…

Another project in the performance oriented Linux HTTP servers is https://lwan.ws/ (using pervasive zero copy), but as far as I know it does not cover a lot of features yet.

So does Apache with mpm_event, which has been around for close to 15 years now. The fact that you can run Apache in a process pool model doesn't mean you should. That mode was…