[CentOS] network interface up but no data

Nat N

phenisha at gmail.com
Thu Oct 21 10:52:52 UTC 2010


Hi List,

I am running a x86_64 Centos server, after a reboot the network has decided to
not work.. this particular machine has 8 eth ports eth0-7. eth0 and eth3 were
bonded (mode 1 ) after reboot the bond did not work. It came up OK but no
data was passing through it. all other interfaces on the machine were working.

I decided to remove the bond, so after removing the bond i was left with eth0
using the same IP as the bond used. however it still does the same thing,
it starts up with no errors but i cannot even ping the default gateway,
tcpdump shows nothing on the interface.

I have made sure of the following
* no iptables installed
* kernel modules for the NIC are loaded with no errors
* the bond kernel module is not installed / all bond configs have been removed

Below is some info on the interface, it should be noted that all others are OK
the only 2 that were having the issue are eth0 and eth3 (the ex-bonded ones)

server1:/etc/sysconfig/network-scripts# ip route  | grep 172.18
172.18.16.0/24 dev eth0  proto kernel  scope link  src 172.18.16.1
default via 172.18.16.254 dev eth0


# ifconfig eth0
eth0      Link encap:Ethernet  HWaddr 00:24:49:C8:66:98
          inet addr:172.18.16.1  Bcast:172.18.16.255  Mask:255.255.255.0
          UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
          RX packets:0 errors:0 dropped:0 overruns:0 frame:0
          TX packets:7223 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:1000
          RX bytes:0 (0.0 b)  TX bytes:463918 (453.0 KiB)
          Interrupt:226 Memory:dc000000-dc012800


If you look closely TX  and RX are massively different, but I am not
seeing anything on tcpdump even there is a large number of TX packets.

Any ideas?

Nathan



More information about the CentOS mailing list