Re: 2.6.0 NFS-server low to 0 performance

[Muli Ben-Yehuda <mulix@mulix.org>]

[-- Attachment #1: Type: text/plain, Size: 1357 bytes --]

On Mon, Jan 12, 2004 at 10:22:31AM -0500, Trond Myklebust wrote:
> P? m? , 12/01/2004 klokka 09:40, skreiv James Pearson:
> > If mount defaults to trying TCP first then UDP if the TCP mount fails,
> > should there be separate options for [rw]size depending on what type of
> > mount actually takes place? e.g. 'ursize' and 'uwsize' for UDP and
> > 'trsize' and 'twsize' for TCP ?
> 
> No. The number of "mount" options is complex enough as it is. I don't
> see the above as being useful.
> If you need the above tweak, you should be able to get round the above
> problem by first attempting to force the TCP protocol yourself, and then
> retrying using UDP if it fails.

I have a patch, sent to the util-linux maintainer, that adds a couple
of new mount options to nfsmount. Those allow you to force either of
tcp, udp, tcp then udp, and udp then tcp, using the existing proto=xxx
syntax. It's available at 
http://www.mulix.org/code/patches/util-linux/tcp-udp-mount-ordering-A3.diff

It also cleans up nfsmount() somewhat, although it could certainly do
with further rewrite^Wcleanups. I'm waiting to hear from the
util-linux maintainer before embarking on that, though. 

Cheers, 
Muli 
-- 
Muli Ben-Yehuda
http://www.mulix.org | http://mulix.livejournal.com/

"the nucleus of linux oscillates my world" - gccbot@#offtopic


[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 189 bytes --]