On Mon, 2009-09-21 at 15:40 +0200, Guy Deleeuw wrote: > Hello > > I have a problem with a blade server Siemens PrimeCenter with 2 switchs > Gbe IBP (Gigabits Intelligent Blade panel) . > This server has 2x3 nic's, but I can only setting up the 2 first > interfaces, one by IBP. > > Example > sFw01:~# ifup eth2 > In my syslog : > Sep 18 13:27:30 sFw01 kernel: [90584.043038] ADDRCONF(NETDEV_UP): eth2: > link is not ready > and the result of eth2 ip link say NO CARRIER. > 4: eth2: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast > state DOWN qlen 1000 > link/ether 00:1b:24:5e:d1:cc brd ff:ff:ff:ff:ff:ff > > Our provider replace the blade but the problem persist, and he say that > the problem come from Debian5. > > I check the network cable :-) it is ok. [...] Are you sure that the cable was connected to the port named eth2? The interfaces might not be named in the order you expect. If the ports have visible LEDs, then 'ethtool -p eth2' should make eth2's LED blink. Otherwise, bring up all the interfaces with ifconfig and then check their link state. Linux does not monitor or report the link state for interfaces that have not been brought up. If some of the ports really don't work under Debian 5.0, please file a bug using reportbug. Ask your provider which distributions are tested and known to be working with this hardware, so we can compare the driver versions. Ben. -- Ben Hutchings I haven't lost my mind; it's backed up on tape somewhere.
Attachment:
signature.asc
Description: This is a digitally signed message part