Public SourceMage Discussion List

[Jeremy Blosser <jblosser-smgl AT firinn.org>]

On Jan 07, David Brown [dmlb2000 AT gmail.com] wrote:
> I've been kinda curious as to the status of lvm (specifically lvm
> root).  I know the iso has yet to support lvm root creation. There was
> a wiki hack that referenced a bug that has since been reopened and
> there are a couple of open lvm bugs and a couple of reopened lvm bugs.
...
> The only unknown I have before I start on this venture is where the
> state of the initscripts are. If someone with past experience and
> knowledge of lvm could help me out with what works and what doesn't
> (cause either it hasn't been done or it did and got dropped) that
> would be much appreciated.

The open bugs remain accurate.  9177 is the main one.  The current problem
(documented in that bug) is the order things are done in does not work with
an LVM root/initrd.  http://bugs.sourcemage.org/show_bug.cgi?id=9177#c25
lists the steps and the conflicting needs of the LVM root / non-LVM root
setups.

What probably makes the most sense is to move all the LVM init stuff out to
its own init script, make that part of the LVM spell, and give it logic to
know when in the boot to run.

As for the other new stuff you're talking about, supporting that is all
fine and well but we wouldn't want to just support one particular way of
doing things.  Not that I claim to follow all of the bleeding edge kernel
stuff or how it affects existing stuff.

Attachment: pgpmh7i7KRWDC.pgp
Description: PGP signature