[Lustre-discuss] root on lustre and timeouts

[Andreas Dilger]

On May 01, 2009  16:33 +0100, Daire Byrne wrote:
> FYI we have been testing FS-Cache from Redhat (landed in 2.6.30) -
> http://people.redhat.com/~dhowells/fscache. A common read-only NFS root
> is cached by the clients to the local disk. Not quite a "diskless" system
> but a good hybrid between booting entirely from network and booting from
> the local drive. You can make the NFS lookups a little more lazy with
> the mount options "actimeo=7200,nocto" (say). It has been working well so
> far but more testing is required before we go into production. Combined
> with UnionFS for writes we can network boot desktop machines without
> users noticing the difference between a local HD install.

There was an external project to add fscache support to Lustre (was for
use over a WAN, but would be equally valuable for many diskless clients).
It would be nice to hear if that project was still underway and what
the status is.  Is anyone involved reading this?

> ----- "Brian J. Murrell" <Brian.Murrell at Sun.COM> wrote:
> 
> > On Thu, 2009-04-30 at 11:48 -0400, Robin Humble wrote:
> > > BTW, as was pointed out in one talk of this years LUG, Lustre 1.8's
> > > OSS read cache should help things like root-on-Lustre because small
> > > commonly used files will likely be cached in the OSS's and won't
> > > result in disk accesses.
> > 
> > Yes, imagine what the ROSS cache can do for 150 clients all booting
> > (and executing the same scripts/binaries) at the same time.  Imagine
> > what the OSS disk did/does before the cache.  :-)
> > 
> > Certainly, I am not without bias, but the feature set of 1.8 looks
> > compelling enough to make me want to upgrade my own little "dogfood"
> > cluster here to 1.8.  :-)

Cheers, Andreas
--
Andreas Dilger
Sr. Staff Engineer, Lustre Group
Sun Microsystems of Canada, Inc.