Multiple NICs, partial success, partial weirdness

[Carsten Aulbert]

Sorry, rewriting my email again (again wrong from address, sorry Thomas,
it's now sitting in the moderator queue)

Thomas Lange schrieb:
>>>>>> On Tue, 03 Mar 2009 12:34:31 +0100, Carsten Aulbert <carsten at welcomes-you.com> said:
> 
>     > Well, we do. One for the subnet on eth0 (10.0.0.0/8, data network, fast
>     > large MTU) and one for eth3 (172.16.0.0/12, mgmt. network, slow) and
>     > both MAC addresses are known, thus deny unknown-client won't help. So in
>     > that respect the set-u pis correct and consistant, right?
> Yep. You should disable PXE booting from your nic that is connected to
> the data network so the computer will only PXE boot from the mgt network.
> 

PXE booting is not the problem, we only PXE boot via the data network
anyway...

> There may be still a problem, that boot.log network settings for both
> network cards, but you have to check that.

The problem lies with the initramfs which is probing all found NICs
simultaneously with DHCP (when ip=dhcp is given) and then the fastest or
slowest reply survives in boot.log.

To solve this one could
o disable all but one MAC for DHCP, but that's not really a good
solution when doing a full cluster reinstall, since a lot of DHCP server
tables need to be modified and servers restarted
o create a special hook/script taking care of this issue
o start playing with ip=::::... again

Probably we will take the latter route *sigh* and I was so happy to get
away from the custom chboot setup we currently have :(

Or is there any other way to achieve this?

Cheers

Carsten