bringing up interface eth0 rtnetlink

Linux and Unix-Like Linux
1

Hi all,

OS: fedora release 10

when i try to restart network service, ending up with following errors. There are two interfaces eth0 and eth1

bringing up interface eth0 rtnetlink answers file exists
error adding <IP Address here> to eth0

bringing up interface eth0 rtnetlink answers invalid argument

below are ifcfg-eth0 and eth1 files under /etc/sysconfig/network-scripts/ directory

# cat /etc/sysconfig/network-scripts/ifcfg-eth0
DEVICE=eth0
BOOTPROTO=none
BROADCAST=19.16.12.255
HWADDR=<MAC ADDRESS HERE>
IPADDR=19.16.12.125
IPV6_AUTOCONF=yes
NETMASK=255.255.255.0
NETWORK=19.16.12.0
ONBOOT=yes
TYPE=Ethernet
USERCTL=no
IPV6INIT=no
PEERDNS=yes
NM_CONTROLLED=yes

# cat /etc/sysconfig/network-scripts/ifcfg-eth1
DEVICE=eth1
ONBOOT=yes
BOOTPROTO=none
HWADDR=<MAC ADDRESS HERE>
TYPE=Ethernet
USERCTL=no
IPV6INIT=no
NETMASK=255.255.255.0
IPADDR=<IP ADDRESS HERE>
GATEWAY=<GATEWAY HERE>
NM_CONTROLLED=yes
PEERDNS=No
DNS1=127.0.0.1
DNS2=<DNS IP HERE>


# service network status
Configured devices:
lo eth0 eth1
Currently active devices:
lo eth0 eth1


# service NetworkManager status
NetworkManager (pid  4236) is running...

There is another similar server, and ifcfg-eth0 and ifcfg-eth1 are similar(ofcourse different IP etc).
The other server is working fine.

difference is

on WORKING server

# route
Kernel IP routing table
Destination     Gateway         Genmask         Flags Metric Ref    Use Iface
IP ADDRESS HERE *               255.255.255.0   U     1      0        0 eth1
19.16.12.1      *               255.255.255.0   U     1      0        0 eth0
default         GATEWAY HERE    0.0.0.0         UG    0      0        0 eth1


on NON-WORKING server
Kernel IP routing table
Destination     Gateway         Genmask         Flags Metric Ref    Use Iface
IP ADDRESS HERE  *               255.255.255.0   U     1      0        0 eth1
19.16.12.0      *               255.255.255.0   U     1      0        0 eth0
link-local      *               255.255.0.0   U     1002      0        0 eth0
link-local      *               255.255.0.0   U     1003      0        0 eth1
default         GATEWAY HERE    0.0.0.0         UG    0      0        0 eth1

and WORKING server is able to resolve IP but NON-WORKING server is not able resolve IP and ending up with an error when it try to use nslookup command to resolve some other server

on WORKING server
# more /etc/resolv.conf
# Generated by NetworkManager
nameserver 127.0.0.1
nameserver <DNS IP HERE>

on NON-WORKING server
# more /etc/resolv.conf
# Generated by NetworkManager
nameserver 127.0.0.1
nameserver <SAME DNS IP HERE AS ABOVE FOR WORKING SERVER>

Please help to solve this issue.
Thanks in advance!

2

BootProto=Static ?

3

It's saying "file exists" because the interface is already defined at the time the script tries to define it again. I'm pretty sure (not certain) this is because you're mixing old-style ifcfg files with NetworkManager. Since these are two different ways of accomplishing the same end, you're going to have to pick one or the other. It sounds like you're using ifcfg on the other server as well.

Personally, I prefer ifcfg, so I would stop and chkconfig off the NetworkManager service, change NM_CONTROLLED to be no in each ifcfg and reboot the system (sometimes you can get away with just restarting the network service but if it's already off the network you might as well use this as a chance to clear out anything NM has done in the background before people start depending on this server again).