Red Hat Bugzilla – Bug 142697
Network link status changes provoke no further action
Last modified: 2014-03-16 22:51:13 EDT
From Bugzilla Helper: User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.7) Gecko/20040616 Description of problem: When plugging in a network cable, the kernel detects the new link, but no further action is taken Version-Release number of selected component (if applicable): hotplug-2004_04_01-8 How reproducible: Always Steps to Reproduce: 1. Boot machine with network cable unplugged 2. Once booted, plug in network cable 3. Actual Results: The kernel logs the link state change: Dec 13 00:47:04 localhost kernel: b44: eth0: Link is up at 100 Mbps, full duplex. Dec 13 00:47:04 localhost kernel: b44: eth0: Flow control is on for TX and on for RX. But that's all. Expected Results: It should trigger that interface to be configured (e.g., an ifup eth0). On most machines, that would cause dhclient to be invoked to try and get an IP address. As it stands, if you boot without the network cable connected, you're screwed unless you know enough to do a "service network restart" from the command line after reconnecting the cable. Additional info: Logging this against hotplug as it's the closest to an appropriate place I can think of.
You may want to try the NetworkManager package; it handles things like this.
Not really suitable. From the README: >It is intended only for the desktop use-case, and is not intended >for usage on servers. Since I'm wanting to do this on a headless box, it doesn't look like NetworkManager is a solution to the problem.
There is also netplugd, part of net-tools.
That looks much better (even if it launches a DoS on my filesystem by filling up /var/log/messages -- but that's a separate bug :-) I'd still have expected this to be hotplug, rather than requiring a separate daemon, but it works, which is all I'm bothered about right now. Thanks.