problem description revised [was: Re: udev-110 ->111 changed behaviour for naming net-devices?]
drifter at bcsoft.de
Fri Jun 8 06:01:14 PDT 2007
Jens Stroebel wrote:
> On Mon, May 21, 2007 at 06:22:07PM -0400, Bryan Kadzban wrote:
>> But if you remove all your rules and reboot, udev will generate new
>> rules for all your NICs, so Alexander's suggestion of removing them all
>> is a good idea. Just beware that what you want isn't possible.
>> (Why do you want it, by the way?)
> That's because we install a laptop with a pcmcia-plugged network card.
> Should someone at a later time change this card (e.g. it is defect),
> the MAC would change, too.
> This would lead to the pcmcia network card not getting the name the
> vanished card had, in spite of being in the same place and having to
> fulfill the same tasks (which it won't, because it'd have another name
> and the network configs wouldn't match anymore).
First of all:
The observed behavior is (unfortunately) not dependent on udev-111 <->
udev-110. We have seen this behavior more often, lately, still trying to
live without persistent, auto-generated network rules.
Maybe we'll have to refrain from that, but still udev's naming behavior
puzzles me a bit:
We have card a -> eth0 (custom rule via PCI-id)
card b -> eth1 (should become that automatically,
In the normal boot process (over 50% of the time... *sigh* ..), card a
is found first and becomes eth0 straight away-> no problem; card b gets
found second and becomes eth1.
Problems arise when card b is found first, because:
card b becomes eth0
our custom rule wants to assign card a -> eth0
card b gets moved out of the way; because card a is still eth1, card b
card a becomes eth0; but card b stays eth1_rename
Shouldn't the final step (eth1_rename -> eth1) be done somehow by
something? Is this indeed expected behavior?
drifter at bcsoft.de
By caffeine alone I set my mind in motion, By the beans of Java
do thoughts acquire speed, hands acquire shaking, the shaking
becomes a warning, By caffeine alone do I set my mind in motion
More information about the lfs-dev