Bonding ifup-script for LFS 6.2
dbn.lists at gmail.com
Tue Oct 10 11:30:44 PDT 2006
On 10/10/06, Philox / Jonathan Salomon <joni at philox.eu> wrote:
> Basically the script ignores the slave bonding devices until the master
> is set. Then the script checks using ethtool if a device is a bonding
> device and if so it will find the config files that reference to that
> bonding device as their master and enslave them to the bonding master.
> Therefore the bonding slaves should have a config similar as follows:
Sounds nice. I think there are a handful of fixes that should go into
our if* scripts. Could you do me a favor and repost with a diff to the
original ifup? `diff -u <orig> <new>'.
Also, instead of using ethtool and ifconfig for everything, I'd
suggest looking in sysfs to see if these values exist there. It would
be easier than parsing output. Since we don't install ethtool (that I
know about) in BLFS, it would be hard to put it in the network
Look in /sys/class/net/<interface>/ at the different files there.
Also, try looking in /proc/net.
More information about the lfs-dev