Marc Herbert wrote, On 05/27/2009 11:13 AM:
I think one issue here is that, despite all your efforts, NM still indirectly relies on the device name, since "NM_CONTROLLED=no" is read in a file unfortunately named "ifcfg-eth1".
But again: This file _is_ about eth1 and doesn't necessarily say which MAC address eth1 has. So I think the name is OK for what it is. I appreciate that NM does it in another way which in all ways is better. But when interfacing to ifcfg then we have to play by ifcfgs rules.
So if instead you make NM read the "NM_CONTROLLED=no" flag in some other "/etc/NetworkManager/HWADDR" configuration item (NOT referencing the device name in anyway), then the duplicate mapping problem goes away. This would make NetworkManager really ethX-agnostic, isolated from this major ifcfg design flaw (duplicating udev).
I agree that that could be useful for some purposes. But AFAIK NM already has functionality for "blacklisting" devices?
But having NM_CONTROLLED=no specified in ifcfg is a very good choice for co-existence with the old ifcfg system. If the MAC address should be listed somewhere else then it would be more error-prone.
(Btw: Why isn't NM_CONTROLLED=no the default for ifcfg files? If someone has created an ifcfg file then it probably was in order to use it the old way, and any NM interference with that would probably not be appreciated.)
/Mads
Attachment:
smime.p7s
Description: S/MIME Cryptographic Signature