[Beowulf] diskless booting over jumbo frames

Paul Armor parmor at gravity.phys.uwm.edu
Wed Apr 25 09:30:36 PDT 2007


Hi,
we're a redhat/fedora based install, thus use anaconda, and as such...

On Wed, 25 Apr 2007, Mark Hahn wrote:

>>> mount -n -o nolock,rw,udp,nfsvers=3 $NFSROOT /sysroot
>> 
>> Are you using -o nolock when you're mounting?  This caused problems for us, 
>> a few years ago, iirc.
>
> I tend to disable everything I don't know I need ;)
> I guess I assumed that NFS locking would be necessary for things like 
> exclusive access to NFS-mounted mailboxes, but not much else.  we don't do
> mail on clusters at all, and I don't think any of our users are doing 
> flock/lockf-type stuff on their files.  oh, right, and my nfsroot is also
> readonly (as exported from the server), so I figured locking was irrelevant.
> (and forgot about the 'rw' tucked away in the initrd...)
>
> what kind of problems did you run into?

This goes back a few years, so memories are fuzzy, but we found that when 
we went to update from RH9 to FC3 on our cluster, NFS performance during 
anaconda kickstarts was rediculously slow... we found timeout issues 
waiting for locking (iirc, I don't remember if we had so sniff this 
problem, or were able to catch errors on one of the virtual terms)... the 
end result was that for our anaconda installs, we needed to do -o nolock 
mounts.

Cheers,
Paul



More information about the Beowulf mailing list