[Linux-cluster] gfs2 convert hosed all VMs

[Steven Whitehouse]

Hi,

On Mon, 2008-11-10 at 09:00 -0800, Dave Costakos wrote:
> I'm using the kernel from RHEL 5.2 release.
> 
> $ uname -rm
> 2.6.18-92.el5xen x86_64
> 
Well that is quite old, but even so I'm surprised that you had a problem
like that. What kind of messages did fsck spit out? I can't really debug
this without any information on what is going wrong,

Steve.

> I got no syslog messages, but when I ran gfs_fsck, it complained quite
> a bit.  I presume it destroyed my files. Sadly, I don't have too much
> time to reproduce the error since I need to restore 30 virtual
> machines from backup.  Luckily, it's just our lab environment.
> 
> On Mon, Nov 10, 2008 at 1:24 AM, Steven Whitehouse
> <swhiteho at redhat.com> wrote:
>         Hi,
>         
>         You don't say what kernel version you are using. I'd suspect
>         that maybe
>         its too old. Do you get any messages in syslog at all?
>         
>         Steve.
>         
>         
>         On Fri, 2008-11-07 at 19:53 -0800, Dave Costakos wrote:
>         >
>         > I just converted a shared file-backed Xen VM GFS filesystem
>         to a GFS2
>         > filesystem.  The conversion was successfully and all my
>         files appear
>         > intact.  I followed the GFS instructions by unmounting the
>         filesystem
>         > on all machines, running gfs_fsck, and gfs2_covert.
>         >
>         > Since I converted the filesystem, all my file-backed Xen VMs
>         can no
>         > longer boot.  pygrub reports errors that the boot loader
>         isn't
>         > returning any data.  If I edit the Xen config to boot of a
>         kernel on
>         > the DomU, VMs still can't start up because LVM cannot
>         identify any
>         > volume groups.
>         >
>         > If I try to access the VM device files locally via losetup
>         and kpartx,
>         > I get "read errors".
>         >
>         > So what's the deal  I know GFS2 is a preview, but I have to
>         assume
>         > I've missed some crucial step here.
>         >
>         > --
>         > Dave Costakos
>         > mailto:david.costakos at gmail.com
>         
>         > --
>         > Linux-cluster mailing list
>         > Linux-cluster at redhat.com
>         > https://www.redhat.com/mailman/listinfo/linux-cluster
>         
>         --
>         Linux-cluster mailing list
>         Linux-cluster at redhat.com
>         https://www.redhat.com/mailman/listinfo/linux-cluster
> 
> 
> 
> -- 
> Dave Costakos
> mailto:david.costakos at gmail.com
> --
> Linux-cluster mailing list
> Linux-cluster at redhat.com
> https://www.redhat.com/mailman/listinfo/linux-cluster