Hi, in this post I’ll show you how I created a three-node Nutanix CE 2.1 cluster nested on ESXi 8 in my lab, using a straightforward approach.
To download the software and start the cluster, you must be registered with Nutanix Community.
Sign in and open the following link, where you’ll find the files needed for installation. In my case, I used the CE 2.1 ISO.

Once downloaded, upload it to your ESXi host, as I did, so you can install the software.

Now let’s create the virtual machines that will become our three nodes.
Here are the specifications for the three VMs:

In the advanced configuration, I added a line so the virtual disks would have serial numbers:
In the VM settings, go to Edit > Options > Advanced > Edit Configuration and add the parameter.

Let’s continue with the screenshots and explanations.
One more detail: configure the vSwitch running the lab as shown below:
Enable promiscuous mode and forged transmits; otherwise, networking will fail.

Boot the CE installer on each VM and install the system, entering the requested settings. As you can see, the disks show serial numbers on all three nodes:



Accept the EULA and continue.



The installations are complete. Now reboot the last node, wait a little, and then create the cluster. Let’s get started:

Connect over SSH to one of the CVMs using one of the IP addresses we configured. From there, we’ll orchestrate the installation. The default username is nutanix and the password is nutanix/4u.

I’ve prepared the command. I’ll paste it below and explain it as clearly as I can:
cluster -s 192.168.101.21,192.168.101.22,192.168.101.23 --dns_servers=192.168.101.1,1.1.1.1 --cluster_name=NTX-LAB-1 --cluster_external_ip=192.168.101.20 create
This command creates a cluster named NTX-LAB-1 with external IP address 192.168.101.20 and the DNS servers shown. It will include the three CVMs ending in .21, .22, and .23.



After a while, we get this result: the cluster was created successfully!

Now let’s open Prism and complete the initial setup so we can access the cluster. In a browser, enter the IP address of one of the CVMs or the external cluster IP configured in the command above:




Done!
We’re in our cluster!
Keep in mind that this is a test and learning lab, so the cluster is running nested and uses quite a lot of resources.
Each node has 32 GB of RAM and 4 vCPUs. The disk uses thin provisioning, but it still takes up space.
Next, we’ll start tuning it before deploying our first VM and more.
The first change I like to make is to set the NTP server so everything has the correct time from the start:



Regards.
