Comments on: Configure Network Bonding on CentOS 6 / RHEL 6

By: Pradeep Kumar

Pradeep Kumar — Thu, 29 Dec 2016 03:28:26 +0000

In reply to kiran.

Hi Kiran,

Are you testing bonding on Virtual Machines hosted on Virtual Box …?

If you want to achieve fault tolerance and load balancing then either go for mode 5 or mode 6. As the name suggests in round robin mode packets will flow in round robin way.

By: kiran

kiran — Wed, 28 Dec 2016 19:24:21 +0000

Hi,
For round-robin if have some problem. if I down the the nic1 from the console there is no interruption in my putty session, but if I down nic2 then my putty session becoming in active until I restart the network service. Do we really need to restart the network service once the nic goes bad?

By: Chris Enslow

Chris Enslow — Mon, 14 Sep 2015 16:14:37 +0000

I’m noticing in RHEL documentation that they are recommending *NOT* to use modeprobe.conf with later RHEL releases – they recommend using “BONDING_OPTS=” in the ifcfg-bond[#] file instead so that it’s all in one place. It is supposed to work in later kernel releases.

By: Pradeep Kumar

Pradeep Kumar — Tue, 08 Sep 2015 02:54:51 +0000

Hi Navneet ,

Please check whether second interface like eth1 is detected and its link is up using the command “ethtool eth1”.

By: Navneet

Navneet — Mon, 07 Sep 2015 13:56:48 +0000

When I did one of the nic(eth0) down, bonding stopped working but /proc/net/bonding/bond0 was showing everything fine and bond0 was showing IP too. As soon as I enabled eth0, server again started pinging. Please help