Re: Fedora 8 not recognizing eth0 after imaging

["Paulo Cavalcanti" <promac@xxxxxxxxx>]

On Sun, Mar 9, 2008 at 7:29 AM, Michael D. Setzer II <mikes@xxxxxxxxxxxxxxxx> wrote:

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Since Redhat 9, I've been images of my lab systems to all the other
machines. I uncheck the bind to mac, and the systems would boot and then
use there same nic as eth0. But after doing the upgrades to Fedora 8, the
original machine would have eth0, but all the other machines would come up
with eth1 and eth0 deactivated. With the latest updates from yesterday, it
has gotten worse. The machines boot up after an image, but gives a
message during boot that eth0 not found, but doesn't activate eth1, so the
machines are all off-line.

The solution that I have found, is use netowk option to delete the inactive
eth0, and then create a new devices. In hardware it shows eth0 and eth1,
and selecting eth0 shows messaage that hardware not found, but selecting
eth1 and then activating it then works fine, This takes only a couple minutes,
but doing it on 19 other machines is a problem.

After the imaging eth0 is renamed eth0.bak and the active interface is eth1, but using dhcp.
What I do is delete eth0.bak using the graphical network tool and make eth1 use a fix IP again,
restart the network service, and reboot.

After rebooting, if I want the identifier eth0 back, I move

/etc/sysconfig/network-scripts/ifcfg-eth1 to

/etc/sysconfig/network-scripts/ifcfg-eth0

and edit ifcfg-eth0 to change
DEVICE=eth1 for DEVICE=eth0

and then reboot again. If I do anything different, I end up with eth2, eth3, etc ...
and it is really messy.

-- 
Paulo Roma Cavalcanti
LCG - UFRJ