[Linux-cluster] gfs2, kvm setup

[J. Bruce Fields]

On Fri, Jun 27, 2008 at 12:18:45PM -0500, David Teigland wrote:
> On Thu, Jun 26, 2008 at 05:10:52PM -0400, J. Bruce Fields wrote:
> > > > So, the first mount (on "piglet1") succeeds.  The second (on "piglet2")
> > > > returns immediately without mounting, and leaves this in the logs:
> > > > 
> > > > 	gfs_controld[3035]: segfault at 0 ip 08051361 sp bfd88ae0 error 4 in gfs_controld[8048000+1d000]
> > > 
> > > Looking at the object file, that appears to be in purge_plocks().
> > > 
> > > After rerunning configure with --debug and rebuilding, the second mount
> > > hangs instead of returning immediately without mounting.
> > 
> > By the way, as I've said, I'm using the STABLE2 branch with the one
> > patch from you to bring it uptodate with the latest kernel interface.  I
> > might actually prefer to try the master branch instead (with the stuff
> > labeled "cluster3" in the wiki), but didn't know whether it was actually
> > working at this point.  Is it?
> > 
> > My goal is just to be able to experiment with your latest posix locking
> > code and with nfsd exports.
> 
> The master branch is probably too unstable to attempt right now.
> I'd really like to try this out, but none of the 2.6.26-rc kernels
> will work on my test machines!  (mpt fusion driver doesn't work)

Oog.  Well, I see that's been reported a few times.  Maybe someone is
taking a look....

> Not yet sure what to try next...

I'll see if I can get any more information out of my setup.

--b.