[PATCH] user_friendly_names risks

[Ritesh Raj Sarraf]

Hi Guido,

On Thursday 09 Apr 2009 12:57:43 you wrote:
> I fully agree that there are possible problems but without leaving a
> choise in d-i (since it doesn't support labeled mounts or mounts by uuid
> yet) we shouldn't highly discourage the default installation method.
>
> The bindings file is being updated automatically by the initramfs-hook
> which leaves us mostly with the locking problem. RedHat worked around
> this by pushing the bindings file timeout to 30 seconds:
>
>  https://bugzilla.redhat.com/show_bug.cgi?id=445268
>
> We could do the same if it really hits people.
>
> Maybe putting an explanation how to switch to labeled mounts from
> user_friendly_names=yes would help?

Yeah!! Just checked. Dave's increased it to 30 secs.

Like d-i, anaconda also has the same problem. The default installation of 
anaconda uses friendly_names. So the user is stuck, at least for the root LUN.

I'll reword and send the patch because I think this needs be documented. And 
also because referring the LUNs by their 128bit UID is the most hassle free 
and persistent way. And that covers raw LUNs also.

Ritesh
-- 
Ritesh Raj Sarraf
RESEARCHUT - http://www.researchut.com
"Necessity is the mother of invention."

-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 197 bytes
Desc: This is a digitally signed message part.
URL: <http://lists.alioth.debian.org/pipermail/pkg-lvm-maintainers/attachments/20090410/62b63f7d/attachment.pgp>