On 30/01/15 11:48, Martin Lucina wrote: >> Are there any observable advantages over the previous httpd? > > Well, Nginx is the #3 server used in production on the web. So we're > getting pretty close to an equivalent of a real-world production setup.
We can speculate that it is a highly desirable selling point, and that it may indeed be invaluable. It's not observable per se, though, at least not currently. > Further, Nginx FastCGI implementation does keep-alive so we don't run into > port re-use problems with many TCP connections being opened in succession. Ok, that is an observable advantage. > I also researched PostgreSQL, this will be problematic as their > architecture is the classic UNIX "fork a process for each connection" > model. It was pointed out to me last night that we might be able to support fork() if we copy memory at switchtime -- the rump kernel rfork supports forks of everything else already. The approach would require tracking user memory allocations... which we should do anyway so that exit() works correctly. I'll probably do some experiments when I get back from FOSDEM, unless I get distracted with something else. Of course, it will be less efficient than threads, but at least it'll bring almost all server programs into the realm of potential support. ------------------------------------------------------------------------------ Dive into the World of Parallel Programming. The Go Parallel Website, sponsored by Intel and developed in partnership with Slashdot Media, is your hub for all things parallel software development, from weekly thought leadership blogs to news, videos, case studies, tutorials and more. Take a look and join the conversation now. http://goparallel.sourceforge.net/ _______________________________________________ rumpkernel-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/rumpkernel-users
