Messages in this thread | | | Date | Fri, 18 Jun 1999 05:08:05 -0700 (PDT) | From | Alex Belits <> | Subject | Re: Why khttpd is a bad idea (was a pointless argument about devfs) |
| |
On Fri, 18 Jun 1999, Richard Gooch wrote:
> Alan Cox writes: > > > I find khttpd is a reasonable idea. I dont know what youre talking about > > > with "lower performance than a user space httpd", > > > http://www.fenrus.demon.nl/ seems to indicate otherwise -- what > > > performance measurements were you looking at? B) > > For the record, I'm with Alan on this one. I'd rather see a more > general interface rather than just khttpd. > > > phhttpd. > > > > > Now if you plug khttpd in front of apache the advantages become obvious. > > > Apache isnt exactly a speed demon -- but with khttpd it could be.
Existing khttpd code isn't the best implementation on the idea, it's neither optimized nor particularily well designed. Similar by basic operation thing that works as a cache and does clean fallback to the userspace server (khttpd fallback mechanism is primitive and inefficient), will be much more efficient than this.
> > Apache 2.0 should be.
"Should" isn't good enough, it will be nice to know, what kind of design will cause the performance to be increased so dramatically. So far all changes still involve heavy modifications in kernel that are as little portable between systems as in-kernel HTTP caching support.
> > BTW: the big issue with khttpd is a lack of genericness. Its a > > single problem single solution piece of code. There are lots and > > lots of equivalent problems and they all boil own to the same thing.
Any examples? All other protocols except NFS (that is already in kernel) require heavy processing before sending responses, and won't benefit from "generic" solution that really still only benefits one particular implementation of one particular protocol.
> Which is one of my points about devfs: it's a generic solution to a > range of problems.
Properly done in-kernel HTTP support will be more like masquerade and tunnel modules or NFS -- they are specific for a reason. Existing khttpd code isn't such a thing, but IMHO it's a step in the right direction.
-- Alex
---------------------------------------------------------------------- Excellent.. now give users the option to cut your hair you hippie! -- Anonymous Coward
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/
| |