Re: KNFSD question + others

Olaf Kirch (okir@monad.swb.de)
Wed, 11 Nov 1998 13:37:00 +0100


On Fri, Nov 06, 1998 at 11:16:43AM +0100, Lenart Gabor wrote:
> I've got a question on knfsd. I don't know too much about Linux kernel
> internals (maybe in the future :) so can you suggest a document/URL
> which can explain why a kernel based NFS implementation gives better
> performance than one running in user space.

There's a somewhat dated paper about the Linux NFS implementation, from
a talk I gave at the 1997 Linux Expo. You'll find it, as well as the
slides, on
ftp://linux.mathematik.tu-darmstadt.de/pub/linux/people/okir/expo97*

> The other : how will knfsd
> share CPU time ? I mean if there is a server which is mounted by many
> clients through NFS, kernel will stay more and more time in kernel mode
> beacuse of the kernel based NFS implementation. This will decrease
> the usable CPU resources by user processes. But what can I do if (for
> example) interactive performance is much important than NFS' on a busy
> server ? Is there any document on new kernel features ? I'd like to get
> closer with kernel hacking but I've got major problems like this one.

knfsd is just another process, so it's subject to scheduling just as
any other process.

Olaf

-- 
Olaf Kirch         |  --- o --- Nous sommes du soleil we love when we play
okir@monad.swb.de  |    / | \   sol.dhoop.naytheet.ah kin.ir.samse.qurax
okir@caldera.de    +-------------------- Why Not?! -----------------------
         UNIX, n.: Spanish manufacturer of fire extinguishers.            

- 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/