4 step Network bonding / teaming configuration in Linux

This article explains what is network bonding in Linux. Quick 4 step guide helps you to set up network bonding in your server in minutes.

Network bonding or network teaming is binding two physical NIC (Network Interface Card) together to create once virtual NIC. This virtual NIC serves the purpose of redundancy, fault tolerance, and load balancing.

For an application running on system its a one NIC they are talking to but on bare metal, their requests are being served by two physical cards. Hence in case, one physical card is failed or unplugged, another one still serves beneath virtual NIC, and applications don’t even know about failure. It’s the same as Auto Port Aggregation (APA) in HPUX.

As of now with RHEL7, there are 7 types of NIC bond available :

1. Bond 0: Load balancing (round-robin)
2. Bond 1: Active backup
3. Bond 2: Balance XOR
4. Bond 3: Broadcast
5. Bond 4: 802.3ad
6. Bond 5: Balance TLB
7. Bond 6: Balance ALB

We will see in detail about these types in another post. More commonly used are type 0 and type 1 bond. Let’s see step by step procedure to configure a network bond in Linux.

For this tutorial, we will consider two ethernet cards eth1 and eth2 to configure bond. It is assumed that both are configured/connected to the same network VLAN.

Step 1:

Configure both eths with master bond0 and slave as themselves. For that, open NIC configuration file located in /etc/sysconfig/network-scripts/ifcfg-eth1 & ifcfg-eth2 in vi and edit entries as highlighted below :

DEVICE=eth1
ONBOOT=yes
TYPE=Ethernet
BOOTPROTO=none
USERCTL=no
MASTER=bond0
SLAVE=yes
NM_CONTROLLED=no

For eth2 file, DEVICE name will be eth2.

Step 2:

Create bond0 device file under /etc/sysconfig/network-scripts/ifcfg-bond0. Add the below details in it.

DEVICE=bond0
ONBOOT=yes
IPADDR=10.10.2.5
NETMASK=255.255.255.0
BONDING_OPTS="mode=1 miimon=100"

Under bonding options, we choose mode 1. If you choose to select any other mode out of 7 mentioned above, you need to specify here against mode=

Step 3:

Make sure the bonding module is loaded into the kernel. Add append lines in /etc/modprobe.conf file.

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

Execute module with below command.

# modprobe bonding

Step 4:

That’s it. You are done with configuration. You need to restart networking service and you are good to go. Make sure your network manager service is not running.

# service network restart

Shutting down interface bond0:                             [  OK  ]
Shutting down loopback interface:                          [  OK  ]
Bringing up loopback interface:                            [  OK  ]
Bringing up interface bond0:                               [  OK  ]

You can confirm your bond0 is up with the mentioned IP in ip addr command output. Bonding mode can be verified with below command :

# cat /proc/net/bonding/bond0

Bonding Mode: load balancing (round-robin)
MII Status: up
MII Polling Interval (ms): 100
Up Delay (ms): 100
Down Delay (ms): 100

Slave Interface: eth0
MII Status: up
Link Failure Count: 0
Permanent HW addr: 00:0c:29:b6:be:32

Slave Interface: eth1
MII Status: up
Link Failure Count: 0
Permanent HW addr: 00:0c:29:b6:be:56

Even ifconfig command output will show you bond0 is up with the mentioned IP address.