Re: [OSR-users] New Preview RPMs for next Release Candidate of comoonics-bootimage

[Dan M. <dan...@or...>]

If you can point out the place(s) in the initrd scripts
that does the actual "mount" of the ocfs2 fs, I will
try adding the "-o heartbeat=local" option to see
if the problem goes away.

Dan

> -----Original Message-----
> From: Dan Magenheimer 
> Sent: Wednesday, April 29, 2009 1:26 PM
> To: Dan Magenheimer; Marc Grimme;
> ope...@li...
> Subject: RE: [OSR-users] New Preview RPMs for next Release 
> Candidate of
> comoonics-bootimage
> 
> 
> > > > I *think* an ocfs2 kernel patch is required, or possibly
> > > > a "-o heartbeat=local" option might need to be specified.
> > > This I didn't really understand. Cause for me it worked 
> > > without something like this.
> > 
> > I think you are probably working with a xen-linux guest with
> > kernel version 2.6.18?  I think the patch that causes
> > this problem came after linux-2.6.19 in this Linux commit:
> > 
> > http://www.kernel.org/hg/index.cgi/linux-2.6/rev/710f6c6bd06c
> 
> OK, here's what the ocfs2 team told me:
> 
> The patch in Linux is a safety check for handling the case
> where the heartbeat is on the same disk as the filesystem
> data.  If this is the case (and I think this is always true
> for OSR), you must either use the mount.ocfs2 command
> to do the mount (which does the right thing and notifies
> the kernel about it) or must specify "-o heartbeat=local" if
> you use the mount command.
> 
> So since I am using an upstream Linux (2.6.29), I had to
> workaround this by hacking an ocfs2 kernel source file.
> 
> Dan
>