On Tue, Oct 24, 2006 at 08:45:50PM +0930, Brett Lymn wrote:
> On Mon, Oct 23, 2006 at 08:20:00PM +0100, David Laight wrote:
> > 
> > And it is persistent across machine reboots......
> > 
> 
> and tmpfs is? ;)

There are two problems:

1) The client keeps on retrying - not much can be done about that.
2) The same FH gets allocated to a different file (maybe even in a
   different FS) after a server reboot.  [1]

	David

[1] I've seen the effects of this with other an exported layered FS
    on a different OS.  Having data written to the wrong files was horrid!

-- 
David Laight: david@l8s.co.uk