Ben Greear wrote:
> There was another reference count bug in the arp state machine
> that was fixed in 2.6.14 (or maybe 2.6.15, I don't remember precisely)
> If you can reproduce this bug in 2.6.15 or later, it might be
> something new.
> Otherwise, you probably just need to upgrade the kernel.
Now I am with 2.6.16.19.
The thing what is happen is, that vlans are showed as up interfaces,
but they are not.
# rcnetwork restart
vlan135
vlan135 IP address: 192.168.23.250/30 done
vlan905
vlan905 IP address: 192.168.151.130/30 done
Setting up service network . . . . . . . . . . .
. . . done
***********************
# ip addr ls
1: lo: <LOOPBACK,UP> mtu 16436 qdisc noqueue
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
inet 127.0.0.1/8 scope host lo
2: eth0: <BROADCAST,MULTICAST,UP> mtu 1504 qdisc pfifo_fast qlen 1000
link/ether 00:50:45:bb:d6:d4 brd ff:ff:ff:ff:ff:ff
inet 169.254.100.105/24 brd 169.254.100.255 scope global eth0
3: eth1: <BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast qlen 1000
link/ether 00:50:45:bb:d6:d5 brd ff:ff:ff:ff:ff:ff
inet 192.168.21.2/24 brd 192.168.21.255 scope global eth1
# dmesg
vlan135: add 01:00:5e:00:00:01 mcast address to master interface
register_vlan_device: ALREADY had VLAN registered
vlan135: del 01:00:5e:00:00:01 mcast address from master interface
vlan135: add 01:00:5e:00:00:01 mcast address to master interface
register_vlan_device: ALREADY had VLAN registered
register_vlan_device: ALREADY had VLAN registered
vlan905: add 01:00:5e:00:00:01 mcast address to master interface
register_vlan_device: ALREADY had VLAN registered
vlan135: del 01:00:5e:00:00:01 mcast address from master interface
vlan905: del 01:00:5e:00:00:01 mcast address from master interface
e1000: eth0: e1000_watchdog_task: NIC Link is Up 1000 Mbps Full Duplex
e1000: eth1: e1000_watchdog_task: NIC Link is Up 1000 Mbps Full Duplex
*******************
This is the full dmesg after restarting network service...
>
> The Intel driver and hardware is some of the best, so I doubt that that
> is the problem.
I am with the same opinion for the Intel network hardware.
>
> Ben
>
Best regards
