On Mon, Sep 18, 2006 at 06:19:52PM +1000, Sarton O'Brien wrote:
> On Monday 18 September 2006 18:07, Ivan Vari wrote:
> > One thing I noticed though and it's I/O related. Extracting pkgsrc under
> > /usr stopped several times. It looked like crashed but eventually it
> > didn't. I use FFSv2 and softdep is enabled. I have also exported a 30GB
> > LVM managed block device into my VM for squid cache. It's xbd1e after
> > labelling. The newfs -i 2 -O 1 /dev/xbd1e has been running for 5 hours
> > and I have no idea why it's taking so long...? I'm using real chunks of
> > my real storage not a file backend. Is it normal?
> 
> 
> Is it running continuously or is it stalling? I had possibly the same problem. 
> Starting and stopping xentop made it move faster for some reason. 
> 
> I'm pretty sure one of the "noapic acpi=off nosmp" options fixed it.
> 
> Manuel, is it possible the domu requires these features to be disabled under a 
> linux dom0?

It should not, but there may be race conditions in the frontend drivers that
only show up (or are more likely to show up) in SMP configurations.
I admit Xen3 has not been tested as extensively as Xen2 in this area yet.

-- 
Manuel Bouyer, LIP6, Universite Paris VI.           Manuel.Bouyer@lip6.fr
     NetBSD: 26 ans d'experience feront toujours la difference
--