[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: upgraded systems won't boot from UUID volumes

On Mon, 2013-04-08 at 15:46 +0200, Daniel Pocock wrote:
> On 08/04/13 13:53, Wouter Verhelst wrote:
> > On 08-04-13 08:53, Daniel Pocock wrote:
> >> I'm not suggesting that squeeze systems were installed that way by
> >> default, although people who have migrated an FS from a raw partition
> >> to an LV may have this in fstab.
> > And that fact alone makes it a non-RC bug -- if it's even a bug at all.
> >
> > Changing the way the root filesystem is mounted without performing a
> > reinstallation is something that's fairly advanced. I'm not saying we
> > shouldn't support people who wish to do something like that, but if they
> > do it, they should make sure that whatever configuration they're using
> > afterwards is still a valid configuration.
> >
> > Having a root filesystem on a logical volume, specified by UUID, is not
> > strictly a valid configuration. It may work if you're not using
> > snapshots, but it might have unforeseen consequences. So Don't Do That
> > Then(TM).
> >
> > If Debian exhibits "wrong" behaviour upon encountering a "strange"
> > configuration that, while valid, is not possible to generate using any
> > of Debian's tools, then that is probably a bug; but I fail to see why it
> > should be release-critical.
> >
> The coverage of UUID on the Debian wiki makes it seem like it is a good
> idea to use it and makes no warning about the LVM snapshot issue:
>   http://wiki.debian.org/fstab#UUIDs
>   http://wiki.debian.org/Part-UUID
> so maybe it would be good if somebody who knows this issue in more depth
> than myself was to update that.


Ben Hutchings
Life would be so much easier if we could look at the source code.

Attachment: signature.asc
Description: This is a digitally signed message part

Reply to: