[fedora] RTNETLINK answers: no such process

[Gatt]

OK this has me stumped

 

Recently updated my fedora 10 box with latest updates via Yum as well as installing a 1GB Switch

 

it has 2 NICs - eth0 (Staff) and eth1 (Pupils)

 

When it brings up eth0 I get the following message

 

RTNETLINK answers: no such process

 

eth0 appears to start with [OK]

 

however, I cannot ping the server or ping out using eth0 - Destination Host unreachable

 

I have checked the configs in /etc/sysconfig/network-scripts/ifcfg-staff_lan and it is identical to the config for the eth1 port (save for the IP addreses)

 

Eth0 config:

DEVICE=eth0
BOOTPROTO=none
BROADCAST=10.4.25.255
HWADDR=00:18:71:78:68:B6
IPADDR=10.4.24.236
NETMASK=255.255.254.0
NETWORK=10.4.24.0
GATEWAY=10.4.24.255
ONBOOT=yes
DNS1=10.4.24.200
SEARCH="staff.internal"
NM_CONTROLLED=no
TYPE=Ethernet
USERCTL=no
IPV6INIT=no

 

Eth1 Config:

DEVICE=eth1
BOOTPROTO=none
BROADCAST=192.168.255.255
HWADDR=00:18:71:78:68:B4
IPADDR=192.168.0.100
NETMASK=255.255.0.0
NETWORK=192.168.0.0
ONBOOT=yes
DNS1=192.168.0.200
SEARCH="pupils.moorsdie.internal"
NM_CONTROLLED=no
TYPE=Ethernet
USERCTL=no
IPV6INIT=no

 

Dunno why Eth1 is working but eth0 is not

[AdamK]

Im abit confussed as to why you named it ifcfg-staff_lan, I would of just kept it as eth0

 

But a few things to try:

 

Rename it to ifcfg-eth0

then do:

 

ifup eth0

 

Or just try with your current settings...

 

ifup staff_lan

 

 

Comparing it with my config files it looks right, so the only thing that is jumping to mind is the file name...

 

Adam

[mac_shinobi]

DEVICE=eth0

BOOTPROTO=none

BROADCAST=10.4.25.255

HWADDR=00:18:71:78:68:B6

IPADDR=10.4.24.236

NETMASK=255.255.254.0

NETWORK=10.4.24.0

GATEWAY=10.4.24.255

ONBOOT=yes

DNS1=10.4.24.200

SEARCH="staff.internal"

NM_CONTROLLED=no

TYPE=Ethernet

USERCTL=no

IPV6INIT=no

 

shouldnt the broadcast address be 10.4.255.255

 

?????

 

Not sure if that makes any difference but was just something I noticed

[AdamK]

Well spotted!. I missed that one too

[Gatt]

@mac_shenob - no its a subnet for 10.4.24.1 thru 10.4.25.255

 

Will try it though..

[mac_shinobi]

Not sure if its the same as obviously each ethernet adaptor ( network card ) is on a different subnet and noticed the student ethernet one the 192.168.x.x was on a 255.255 broadcast range where as the other one was missing a 5 ie 25.255 instead of 255.255.

 

As I said above not sure if it would help or not.

[Geoff]

Look at the syslog when you restart networking to see what is causing the RTNETLINK error

[Gatt]

@geoff - nothing in any of the log files

 

looks like its sorted now.. didnt tocuh anything - or change anything so no idea what the porblem was, what caused it,

 

I hate those kinda faults..

[mac_shinobi]

those kinda of faults at least in my windows experience means that the services were not up and running correctly

 

Not sure why linux would do that *shrugs*

 

Am guessing a reboot or using the correct commands to restart the services would of done the trick possibly.

[Geoff]

@geoff - nothing in any of the log files

 

looks like its sorted now.. didnt tocuh anything - or change anything so no idea what the porblem was, what caused it,

 

I hate those kinda faults..

 

Ok then the other reason you'd get that error is when the kernel is trying to setup a route to somewhere that doesn't exist or effectively duplicates an existing route. Check your routing table(s) with 'route -n'. A common example of this problem is on servers with dual nics. You setup both nics with sensible defaults and you give them both the same gateway address (because they are plugged into the same LAN). You will get the above error when the kernel tries to setup the default route to the default gateway upon initializing the second Ethernet device.

Edited April 8, 2009 by Geoff