[ale] Examining pre-existing LVM

Daniel Kahn Gillmor dkg at fifthhorseman.net
Fri Mar 21 10:35:35 EDT 2008


On Thu 2008-03-20 19:32:52 -0400, Jeff Hubbs wrote:

> Yeah, I'm thinking now I shouldn't have run vgcreate; the
> information in the dd'ed files made as described in the article now
> reflects my vge and vgf volume groups.  Where I'm stuck now is that
> I don't have any devices /dev/vge and /dev/vgf

aiui, you should never have device nodes for the VGs themselves, since
they aren't block devices.  you'll have device nodes for the PVs (they
must be device nodes before they're PVs, actually), and device nodes
for any active LVs (which are auto-created in
/dev/mapper/$VGNAME-$LVNAME, in my experience, and then linked from
/dev/$VGNAME/$LVNAME as a convenience).

> and vgscan --mknodes prints out those volume groups but makes no new
> nodes.  One other possibility is that the two disks were put in a vg
> together and I suppose I could re-do vgcreate to accomplish that,
> but I'm still concerned that I won't be able to generate a /dev/vgef
> or whatever.

Have you looked at the output of lvs?  before i touch anything to do
with LVM, i always try to get my bearings with the basic reports: 

 pvs && vgs && lvs

and if i need more information, i use one of the more detailed variants:

 pvdisplay
 vgdisplay
 lvdisplay

I know it's a little late, but you should do this kind of exploration
*before* any potentially destructive operations like vgcreate.

hth,

        --dkg
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 826 bytes
Desc: not available
Url : http://mail.ale.org/pipermail/ale/attachments/20080321/b163bc46/attachment.bin 


More information about the Ale mailing list