On 03/27/2008 01:03 PM Roger Heflin wrote:
David Chambers wrote:
On 03/27/2008 10:32 AM Heitor Moraes wrote:
Servers running Fedora 8?
yes. All of them.
Network interfaces are e1000.
I had 2 of 12 machines lose their config. One had eth0-eth3 and the
other had eth0 and eth1. It sounds suspiciously like this might be
related to having multiple interfaces configured, as none of the
single interface machines failed, nor did the ones where only eth0
was activated at boot time.
- D
The e1000 driver recently split into 2 different parts (I could wrong
on the exact details of what happened-it looks like it split in
2.6.23) one part called e1000 (pci/pci-x) and one part called e1000e
(pcie), and there appears to be a lot of changes being done on the
e1000 interface after that change, which could probably change the
order of the interfaces and which would confuse the new hardware
configuration tool and/or the current configuration and make it do bad
things because of the order changing around in an unexpected way.
Roger
That would fit with the problem occurring after the recent kernel update
(to 2.6.24.3-50.fc8). Presumably if there was a change in e1000, that
made kudzu think the hardware was different, so it created default
config files.
I chmod -w'd my ifcfg-eth* files in any case :-)
- david
--
fedora-list mailing list
fedora-list@xxxxxxxxxx
To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list
