[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

bonding failure in Debian 3.1 Kernel 2.6.8-2-smp


I am running three machines with 1 Xeon (2 Cores) and
a machine with 2 Pentium III with 1,26 Ghz. In all machines I have 2 Intel Network Cards.

The Xeon machine (2Cores) is running the 2.6.8-2-smp kernel without bonding problems.
The PIII Machine with the Kernel 2.6.8-2 runs in identical konfiguration as the system above
When I use the 2.6.8-2-smp kernel on the PIII Machine with the same konfiguration 
the bonding does not work korrekt. (One Network card is disconnected from Network
so the failure let the machine work. With both cards connected I get a loss of 50 Percent 
of the network packages)

here the Messages:

e100: eth0: e100_watchdog: link up , 100 Mbps, full-duplex
bonding: bond0: enslaving eth0 as an active interface with an up link.
bonding: bond0: enslaving eth1 as an active interface with an up link.

As you can read there is no master defined and both Bonding adapters should have
an up link and are active. The network cable of eth1 is disconnected.

Why could this happen.

The Bond0 startparameters are in the networking script under case "$1" in .. behind ifup -a
modprobe bonding mode=1 miimon=100
ifup bond0
ifenslave bond0 eth0 eth1 

under modules.conf  alias bond0 bonding

under interfaces auto bond0, iface bond0 inet static, address ........ only normal parameters. 

Could You tell me why the P III processor with the 2.6.8-2-smp kernel handels the bonding
in another way then under the 2.6.8-2 kernel. (With one ore two processors it is the same).

Holger Harling
Thann-Matzbacher-Strasse 4
84435 Lengdorf

Reply to: