lkml.org 
[lkml]   [1998]   [Jan]   [12]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: disk naming proposal & devfs
On Mon, Jan 12, 1998 at 08:52:59AM -0800, Randolph Bentson wrote:
> The one which triggered my note was a proposal to have the kernel
> take some kind of automatic action based on data stored in a special
> place on the disk. It sort of begs the question "how do you address
> a disk which hasn't been labeled?"

You have a potted naming scheme. If a partition isn't labelled, you use its
UUID (which will have been added if fsck has been run once on it) and if
you can't do that then you encode it in some other way. By device/partition
as has been discussed in other threads should work.

> part of the startup sequence--but only as user level programs.

True, it could be done in mount. mount would translate device 'names' in
fstab or on the command line into real device names based on volume label.
That could work fine. But having /dev/usr has some very nice properties.
It also has some ugly ones - for instance, I _like_ knowing the layout of
my disks. SCO annoys me 'cos you can't find out, besides exmaing
major/minor numbers and looking things up in the kernel config, which
filesystems come from which physical disks. df doesn't tell you. /dev/usr
could be on any disk. But if the system supports both or craetes them as
virtaul symlinks, that's all fine.

--
[======================================================================]
[ Kevin Lentin Email: K.Lentin@cs.monash.edu.au ]
[ finger kevinl@fangorn.cs.monash.edu.au for PGP public key block. ]
[ KeyId: 06808EED FingerPrint: 6024308DE1F84314 811B511DBA6FD596 ]
[======================================================================]

\
 
 \ /
  Last update: 2005-03-22 13:41    [W:0.034 / U:0.148 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site