Normally, when ordering a new Nutanix cluster, or additional nodes, the systems are configured and sized for your environment. Now and then, modifications are made after the systems are ordered. Lately, a customer decided to order additional NICs to eliminate the network interface card as a single point of failure. This short article shows the symptoms and the resolution to comnfigure the nic appropriately.

What do you notice?

• The command “manage_ovs show_interfaces” on a CVM will show all NICs, but the new interfaces are down, even when cabled well.
• There are no configuration files for the new interfaces present in the folder /etc/sysconfig/network-scripts on the AHV host.
• Changing the virtual switch configuration using the Prism UI won’t be possible, as the new interfaces are not shown.

What to do with the new interfaces?

The resolution

Well, after some searches on the Nutanix portal, the resolution was quite obvious. We can just consider this new NIC as a replacement. Nutanix documented this really well.

When making changes to the network settings of a host, I always prefer to place the host in maintenance mode. The procedure on the Nutanix support site ask us to use the IPMI to connect to the console. Because it is a secondary NIC, a ssh session will do the trick as well.

• When logged on to the AHV host, you simply run the command “/usr/local/bin/nic_add_or_replace”.
  This command will identify the new interfaces and rename them. Also the configuration file is created in the /etc/sysconfig/network-scripts folder.

After performing these steps on all hosts in the cluster, you will be able to change the vs0 using the Prism UI. The new interfaces are shown now.

For Nutanix customers, I used this link.