NIC Bonding in Fedora 13 DNS issue?

[Evil-I]

Hi all,

Have set up bonding on my Fedora 13 server as per the tutorial :http://www.cyberciti.biz/tips/linux-...interface.html

Now the interfaces are up and the bond seems to be working on its static IP address of 192.168.0.100, I just no longer seem to have internet access on the server and can't work out why? I think I may have an issue at this stage:

Step # 3: Load bond driver/module

Make sure bonding module is loaded when the channel-bonding interface (bond0) is brought up. You need to modify kernel modules configuration file:# vi /etc/modprobe.confAppend following two lines:alias bond0 bonding
options bond0 mode=balance-alb miimon=100Save file and exit to shell prompt. You can learn more about all bounding options in kernel source documentation file (click here to read file online).

The file contains no information at all, so i did add these 2 lines of code as requested,

alias bond0 bonding
options bond0 mode=balance-alb miimon=100

but it did seem odd that it was blank before I added that code in, I'm wondering if Fedora 13 works differently to what this is suggesting?

I've got: GATEWAY=192.168.0.1 as part of the bond config, which is my default gateway and I've even tried manually adding the nameservers in the bond config file as either NAMESERVER=XXX.XXX.XXX.XXX and NAMESERVERS=XXX.XXX.XXX.XXX and that didn't appear to help. It almost seems like a DNS issue as I can ssh and VNC to the server remotely so it is accesible via the internet it just doesn't appear to be resolving addresses?

I've no idea whats wrong, as far as I can tell in every other way the bond is working fine, it's just the internet access side thats not right.

Any suggestions?

Thanks in advance,

E-I
__________________

 

[Evil-I]

Hi All,

Kinda got this sorted, I was using nameservers: in my bond0 config rather than dns1: dns2:

Not sure where I got led astray, but found out what I was doing wrong in the end.

I'm using Mode6 adaptive load balancing (balance-alb) and it seems to work okay, hard to tell on a home network, file access from a single computer uses a single nic on the server, but as soon as you have multiple machines accessing the server it suddenly makes a real difference with load being spread over both nic's, althoug worth noting that a single machine accessing the server only gets benefit from one nic. I'm using mode 6 as its non switch specific, so I don't need to replace my cheap and cheerful 8 port netgear gigabit switch with some nutter bas**rd cisco switch to make it all work.

My /etc/sysconfig/network-scripts/ifcfg-bond0 reads:

DEVICE=bond0
IPADDR=xxx.xxx.xxx.xxx
GATEWAY=xxx.xxx.xxx.xxx
NETMASK=xxx.xxx.xxx.xxx
NETWORK=xxx.xxx.xxx.xxx
DNS1=xxx.xxx.xxx.xxx
DNS2=xxx.xxx.xxx.xxx
USERCTL=no
BOOTPROTO=none
ONBOOT=yes

My /etc/sysconfig/network-scripts/ifcfg-eth1 reads:

# Intel Corporation 82546EB Gigabit Ethernet Controller (Copper)
DEVICE=eth1
USERCTL=no
ONBOOT=yes
MASTER=bond0
SLAVE=yes
BOOTPROTO=none

eth2 is as above but with the expected changes.

Not sure I actually need bonding, but then I don't reall need a multi-cpu hardware raid base server either. It's something to tinker with in my spare time and by god does it make a good media server

Anyway, thanks for the replies,

Best,

E-I