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

Re: RFC: Possible (partial) solution for device persistence issue

Op 15-10-2006 om 19:06 schreef Frans Pop:
> I have written a very simple (or at least small) script intended to be run 
> during finish-install that converts devices listed in /target/etc/fstab 
> from regular device names to uuids (as far as possible).
> Although this is not a perfect solution, I feel it is an acceptable 
> solution for Etch.
> By doing this at the very end of the installation, there is no chance that 
> we break anything in the installer itself, while still making the fstab 
> less vulnerable to changes in the device order.
> The script relies on the fact that the kernel currently also generates 
> "normal" device names for disks.
> It is only a partial solution as the set up of the boot loader (especially 
> the definition of the root partition) may still be wrong.
> I've tested this successfully on a LVM setup. The attached fstab is from 
> that install.
> Comments welcome.
> Cheers,

> # /etc/fstab: static file system information.
> #
> # <file system> <mount point>   <type>  <options>       <dump>  <pass>
> proc            /proc           proc    defaults        0       0
> # /dev/mapper/Debian-root
> UUID=54eb0e28-f41b-486b-9b47-2e8d484e9417 /               ext3    defaults,errors=remount-ro 0       1
> # /dev/sda1
> UUID=297e4922-7842-4dd2-9f56-09581968bc64 /boot           ext3    defaults        0       2
> /dev/mapper/Debian-swap_1 none            swap    sw              0       0
> /dev/hdc        /media/cdrom0   udf,iso9660 user,noauto     0       0
> /dev/fd0        /media/floppy0  auto    rw,user,noauto  0       0

I'm not happy with /etc/fstab like that.
The comments like '/dev/mapper/Debian-root' and '/dev/sda1' are fine,
but I'm really unhappy with the entries that start with 'UUID='.

I also have the unpleasent feeling that "fixing" device persistence by
forcing persistence by UUID usage is solving a wrong described problem.

The thing I would like to see is that the _difference_ in device naming
between d-i kernel plus fellows and installed kernel plus fellows is

Infact I don't understand why device naming does differ. AFAIK are the
d-i kernel and the installed kernel from the same build. That d-i uses
different tools (busybox, libc?) then installed kernel, will have it
good reason. But it is no reason that devices naming differs ...

Geert Stappers

Reply to: