Before talking about the problems I experienced lately with Nutanix Foundation to install a few new clusters, I want to emphasize I am quite happy with the Foundation tool. But when things are following the manual, there is no need to write a blog about it. Right now I want to spend a few words about my latest experience with the Foundation and the excellent support delivered by Nutanix.

For a customer I had to implement 5 clusters for branch offices. No problem at all at first sight. How much time do you need for an installation of a couple of NX-1365 clusters? Not too much according the books, but unfortunately, this day was one of those days.

The installation of all five clusters started just fine… Using the Foundation applet, all nodes were detected quite fine. Each time it was possible to launch the Foundation page of one of the nodes in the cluster. The manual was still valid, because I was able to enter all the necessary information onto the first pages. So, name servers, subnet masks, gateways were all filled in. I was even succesfull in entering all the hostnames, and the IP addresses for the hypervisor, the CVM and the IPMI interface.

But then… Unfortunately, the problems started… Anyone familiar with the Foundation knows, the network configuration is validated.And this step was not completed successfully… At all tries the validation was stuck at 0%. To troubleshoot this, I tried a lot… I changed network cables, tried different switch ports, nodes were reseated in the enclosure. Despite all effort, nothing worked. But luckily for us, there is always something called Nutanix support.

After some troubleshooting, it turned out to be a misconfiguration of node A. During the manufacturing process, not all IP addresses were removed from this node… Due to the presence of this address, Foundation was not able to set the provided IP address in the CVM config.

Here is the procedure, I had to follow to enter the valid IP configuration.

• Attach the affected node, in this situation node A, to a monitor and keyboard.
• Log in using the root account.
• Start a ssh session to the controller VM using “ssh nutanix@192.168.5.2“
• start editing the config of eth0 using “sudo vi /etc/sysconfig/network-scripts/ifcfg-eth0″ 
• Make sure the file contains the following lines:
  sudo vi /etc/sysconfig/network-scripts/ifcfg-eth0
  ONBOOT=”yes”
  NM_CONTROLLED=”no”
  NETMASK=”subnet_mask”
  IPADDR=”cvm_ip_addr”
  DEVICE=”eth0″ 
  TYPE=”ethernet”GATEWAY=”gateway_ip_addr”
  BOOTPROTO=”none”
• Reboot the controller VM. After boot of the CVM, start the network validation again.
• Happy configuring the Nutanix cluster.

After receiving this procedure from Nutanix support, it turned out to be useful for the four other clusters also. I asked Nutanix support about any faults in the manufacturing process, and it was already a known error. Nutanix support also confirmed that the problem was solved. So hopefully you won’t need this procedure.