In message <20031213230941.GA24329@rek.tjls.com>Thor Lancelot Simon writes:
>On Sat, Dec 13, 2003 at 02:58:58PM -0800, Jonathan Stone wrote:

>I can think of a bunch of ways to attack this problem and the exact
>problem you describe at the same time, but I think they both highlight
>why the current way is Bad.

Sure. I'm  just micro-focused right now.

NFS really needs enough outstanding requests to fill our old friend,
delay*bandwidth.  For NFS, we dont come anywhere close to that for
contemporary networks and moderately-loaded servers.

If we were getting only 4 Mbyte/sec read from a local 10,000 rpm SCSI
disk, then I'd say, sure, fix it everywhere. Right now I'd sooner stop
NFS from haemmorraging to death now, whilst we think about how to
address the global problem later.