Re: Old SUN NFS performance papers.

From: Kris Kennaway <kris_at_obsecurity.org>
Date: Tue, 27 Jan 2004 13:04:45 -0800
On Tue, Jan 27, 2004 at 08:17:58PM +0100, Bj?rn Gr?nvall wrote:
> On Sat, 24 Jan 2004 21:14:51 -0500 (EST)
> Robert Watson <rwatson_at_freebsd.org> wrote:
> 
> > I haven't done much benchmarking on NFS lately, but something worth
> > remembering is that people have spent a lot of time researching and
> > optimizing TCP for a variety of connection types, whereas the NFS code
> > basically has a static implementation of RPC backoff and flow control that
> > hasn't evolved much.
> 
> One reason that FreeBSD users experience poor NFSv3/TCP performance is
> that the defaults for rsize and wsize is unusually small, only
> 8k. Solaris and HP-UX defaults to 32k for a good reason. I guess TCP
> simply needs a little bit more data to chew on to be efficient.
> 
> I tested this on 5.2-CURRENT and found that large file read
> performance went up from 56Mbit/s to 80Mbit/s, an improvement by 43%.
> 
> I have written a patch that makes FreeBSD use the same defaults as
> Solaris and HP-UX. Note that with NFSv3 there is no risk associated
> with specifying to large values for [rw]size. The server automatically
> limits these values in the fsinfo rpc. Patch is attached.

If no-one picks this up in the next few days, can you please send-pr
it so it does not get lost?

Kris

Received on Tue Jan 27 2004 - 12:07:07 UTC

This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:37:40 UTC