On Thu 09 Jun 2011 at 19:11:48 +0000, Camaleón wrote: > On Thu, 09 Jun 2011 19:35:48 +0100, Brian wrote: > > > You are talking about *devices* here. > > AFAIK, network devices (NICs) are also managed by udev.
Maybe - but the cable is not a device. With a static address stanza in /e/n/i the interface will happily be upped and configured without a cable attached during booting. Connecting the cable after booting gives network connectivity, but that is nothing to do with udev. > > An ethernet cable is not a device. > > But the card is and the card interacts with it in some way when the > cable is on/ off :-P Yes, but it still has nothing to do with udev. As I've remarked, the kernel notices the cable reconnection and reports 'eth0: link becomes ready'. It does not invoke ifup. > > But I like your turn of phrase. Go on, get your fingers burnt! > > udevadm(8) and unplug and plug the cable back in. See any reaction from > > udev? > > Me? I don't have any problem with "*-hotplug". Neither do I. It's been in my /e/n/i for many years. But I thought you might have wanted to check what part udev played when a network cable was unplugged/plugged in. udevadm says 'none'. > > The question remains - Why would plugging or unplugging the ethernet > > cable be expected to bring the interface up or down? > > It does not have to be an isolated event. > > For example, the computer is hibernated or suspended, the cable was > disonnected, you attach it and then you awake the system that triggers > the restore script for the network service and the interface that was > marked with "allow-hotplug" cannot be up. You're moving the goalposts! -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20110609222939.GQ19914@desktop
