Network Interface Card (NIC) bonding (also known as teaming) enables you to join two NICs as a single physical device so that they appear as one interface. This is usually performed to provide failover capabilities or load balancing. Bonding can be configured on the BMC Atrium Discovery hardware appliance. It is not something that you should consider on a virtual appliance, because failover and load balancing should be provided by the virtualization platform.

Network configuration When changing network configuration on your appliance, you should always be able to access the system console in case the new configuration does not work correctly.

Netadmin does not work with bonded NICs If you configure your appliance to use bonded NICs, you can no longer use the netadmin user to perform any networking configuration.

The BMC Atrium Discovery appliance relies on eth0 having an IP address. When the appliance starts, it requires an IP address on eth0 for configuration and start-up services. In a default bonded NIC configuration, the IP address would be assigned to the bond point bond0, and the BMC Atrium Discovery services would be unable to start without additional configuration.

The ifconfig output for bonded NIC cards is shown in the following example. Note that bond0 has an IP address but eth0 and eth1 do not.

bond0   Link encap:Ethernet HWaddr 00:11:43:FD:9A:B1
 inet addr:192.168.0.25 Bcast:192.168.0.255 Mask:255.255.255.0
 UP BROADCAST RUNNING MASTER MULTICAST MTU:1500 Metric:1
 RX packets:58071 errors:0 dropped:0 overruns:0 frame:0
 TX packets:1465 errors:0 dropped:0 overruns:0 carrier:0
 collisions:0 txqueuelen:0
 RX bytes:4315472 (4.1 Mb) TX bytes:120360 (117.5 Kb)

eth0   Link encap:Ethernet HWaddr 00:11:43:FD:9A:B1
 UP BROADCAST RUNNING SLAVE MULTICAST MTU:1500 Metric:1
 RX packets:26447 errors:0 dropped:0 overruns:0 frame:0
 TX packets:1262 errors:0 dropped:0 overruns:0 carrier:0
 collisions:0 txqueuelen:1000
 RX bytes:1992430 (1.9 Mb) TX bytes:95078 (92.8 Kb)
 Interrupt:16

eth1   Link encap:Ethernet HWaddr 00:11:43:FD:9A:B1
 UP BROADCAST RUNNING SLAVE MULTICAST MTU:1500 Metric:1
 RX packets:31624 errors:0 dropped:0 overruns:0 frame:0
 TX packets:203 errors:0 dropped:0 overruns:0 carrier:0
 collisions:0 txqueuelen:1000
 RX bytes:2323042 (2.2 Mb) TX bytes:25282 (24.6 Kb)
 Interrupt:17 

Creating a bond point called eth0

To create a bond point called eth0 you need to:

1. Configure the network interface scripts in /etc/sysconfig/network-scripts to force the hardware to bond to the names eth1 and eth2 using their MAC addresses. Note the "SLAVE=yes" and "MASTER=eth0" options.
2. Create the eth0 interface which is actually assigned the IP configuration options.
3. Configure the boot process to include the bonding module and to use eth0 as the bond point, with the various options associated with that bond.

To do this, log in as the tideway user and perform the following steps:

1. Stop the tideway, omniNames and appliance services. Enter:

   [tideway@localhost] $ sudo /sbin/service tideway stop
   [tideway@localhost] $ sudo /sbin/service omniNames stop
   [tideway@localhost] $ sudo /sbin/service appliance stop

2. Change to the /etc/sysconfig/network-scripts directory and determine the MAC addresses (HWaddr of the network cards that you want to bond. Enter:

   [tideway@localhost] $ cd /etc/sysconfig/network-scripts
   [tideway@localhost network-scripts] $ /sbin/ifconfig -a | grep HWaddr
   eth0      Link encap:Ethernet  HWaddr 00:11:43:FD:9A:B1
   eth1      Link encap:Ethernet  HWaddr 00:11:43:FD:9A:B2
   [tideway@localhost network-scripts] $ 

   
   In this example the MAC Addresses are as follows:

   • eth0 – 00:11:43:FD:9A:B1
   • eth1 – 00:11:43:FD:9A:B2
3. Change to the root user and using a text editor, edit, or create and edit ifcfg-eth1 to contain the following:
   

   DEVICE=eth1
   BOOTPROTO=none
   ONBOOT=yes
   HWADDR=00:11:43:FD:9A:B1 # The MAC address of eth1 as determined in the previous step.
   TYPE=Ethernet
   MASTER=eth0
   SLAVE=yes 

4. Using a text editor, edit or create and edit ifcfg-eth2 to contain the following:
   

   DEVICE=eth2
   BOOTPROTO=none
   ONBOOT=yes
   HWADDR=00:11:43:FD:9A:B2 # The MAC address of eth2 as determined in the previous step.
   TYPE=Ethernet
   MASTER=eth0
   SLAVE=yes 

5. Using a text editor, edit or create and edit ifcfg-eth0 to contain the following. Complete the IPADDR, NETMASK, and GATEWAY information
   

   DEVICE=eth0
   BOOTPROTO=none
   ONBOOT=yes
   IPADDR=<required IP address>
   NETMASK=<required netmask>
   GATEWAY=<required gateway> 

   Note that options such as BROADCAST and NETWORK must be defined in ifcfg-eth0 if they are required.

6. Modify /etc/modprobe.conf. The lines that must be changed contain eth0 and possibly eth1. These need to be modified to specify that eth0 is now a bonded interface. For example, if the file looks like this:
   

   alias eth0 e1000
   alias eth1 e1000
   alias scsi_hostadapter ata_piix
   

   Depending on the options required, edit it to look like this:

   alias eth0 bonding
   options eth0 miimon=100 mode=balance-rr
   alias scsi_hostadapter ata_piix
   

   Note The entries after options eth0 are the bonding options. They might not be required, or might be device specific, but the ones shown are good defaults. miimon=100 means check the link state every 100 milliseconds, and mode=balance-rr provides load balancing and tolerance. The balance-rr mode uses sequential order transmissions. There are other mode options available, some of which require changes to other items.

7. Reboot the system.
8. Log in as the tideway user and run the following command:
   

   [tideway@localhost ] $ /sbin/ifconfig -a

   The output should look like the following, note the bond0 interface:

    bond0  Link encap:Ethernet HWaddr 00:00:00:00:00:00
       BROADCAST MASTER MULTICAST  MTU:1500  Metric:1
       RX packets:0 errors:0 dropped:0 overruns:0 frame:0
       TX packets:0 errors:0 dropped:0 overruns:0 carrier:0
       collisions:0 txqueuelen:0
       RX bytes:0 (0.0 b)  TX bytes:0 (0.0 b)
   
   eth0  Link encap:Ethernet HWaddr 00:11:43:FD:9A:B1
     inet addr:192.168.0.5  Bcast:192.168.0.255  Mask:255.255.255.0
     inet6 addr: fe80::250:56ff:fea7:362/64 Scope:Link
     UP BROADCAST RUNNING MASTER MULTICAST  MTU:1500  Metric:1
     RX packets:21026 errors:0 dropped:0 overruns:0 frame:0
     TX packets:175 errors:0 dropped:0 overruns:0 carrier:0
     collisions:0 txqueuelen:0
     RX bytes:1405048 (1.3 MiB)  TX bytes:61541 (60.0 KiB)
   
   eth1  Link encap:Ethernet HWaddr 00:11:43:FD:9A:B1
   UP BROADCAST RUNNING SLAVE MULTICAST  MTU:1500  Metric:1
   RX packets:10529 errors:0 dropped:0 overruns:0 frame:0
   TX packets:87 errors:0 dropped:0 overruns:0 carrier:0
   collisions:0 txqueuelen:1000
   RX bytes:703475 (686.9 KiB)  TX bytes:30280 (29.5 KiB)
   
   eth2  Link encap:Ethernet HWaddr 00:11:43:FD:9A:B1
   UP BROADCAST RUNNING SLAVE MULTICAST  MTU:1500  Metric:1
   RX packets:10498 errors:0 dropped:0 overruns:0 frame:0
   TX packets:90 errors:0 dropped:0 overruns:0 carrier:0
   collisions:0 txqueuelen:1000
   RX bytes:701633 (685.1 KiB)  TX bytes:31793 (31.0 KiB)

9. Check /var/log/messages to check for a successful bonding:

   Nov 18 15:47:18 localhost kernel: Ethernet Channel Bonding Driver: v3.4.0-1 (October 7, 2008)
   Nov 18 15:47:18 localhost kernel: bonding: MII link monitoring set to 100 ms
   Nov 18 15:47:18 localhost kernel: bonding: eth0 is being created...
   Nov 18 15:47:18 localhost kernel: ADDRCONF(NETDEV_UP): eth0: link is not ready
   Nov 18 15:47:18 localhost kernel: bonding: eth0: Adding slave eth1.
   Nov 18 15:47:18 localhost kernel: bonding: eth0: enslaving eth1 as an active interface with a down link.
   Nov 18 15:47:18 localhost kernel: e1000: eth1 NIC Link is Up 1000 Mbps Full Duplex, Flow Control: None
   Nov 18 15:47:18 localhost kernel: bonding: eth0: link status definitely up for interface eth1.
   Nov 18 15:47:18 localhost kernel: ADDRCONF(NETDEV_CHANGE): eth0: link becomes ready
   Nov 18 15:47:18 localhost kernel: bonding: eth0: Adding slave eth2.
   Nov 18 15:47:18 localhost kernel: e1000: eth2 NIC Link is Up 1000 Mbps Full Duplex, Flow Control: None
   Nov 18 15:47:18 localhost kernel: bonding: eth0: enslaving eth2 as an active interface with an up link.

   The list of running kernel modules should include the bonding module.

10. Run the following command:
    

    [tideway@localhost ] $ /sbin/lsmod | sort
    ac                     38729  0
    acpi_memhotplug        40517  0
    asus_acpi              50917  0
    ata_piix               57541  0
    autofs4                63049  3
    backlight              39873  1 video
    battery                43849  0
    *bonding               142537  0*
    button                 40545  0
    cdrom                  68713  1 ide_cd
    crypto_api             42945  1 xfrm_nalgo
    ...