My HCI Lab with Nutanix Community Edition: Part 2: Lab Setup

In last post of this series we discussed about what Nutanix CE is what does it offer and we talked about limitation of the product. In this post I will walk through my lab setup which I used to deploy the Nutanix CE.

Recommended Hardware

Before planning your deployment, make sure to refer to below spreadsheet which lists all necessary piece of hardware you need to have in place. 

In my lab, I don’t have a dedicated hardware server for deployment, so I went ahead with Nested virtualization technique. But if you are interested in deploying CE on dedicated piece of hardware then I would recommend reading this article.

Step 1: Download Nutanix CE bits

Before starting deployment, you need to first register register yourself on Nutanix Next Community . Fill in all the needed details and hit submit.

You will get an email with download links and Getting Started docs. 

Step 2: Prepare Disk files

Once you have downloaded the Nutanix CE binary, you will get a zipped file named ce-2018.05.01-stable.img.gz.

On extracting the zipped file, you will get a file named “ce-2018.05.01-stable”. Rename this file to ce-flat.vmdk. Now we need a disk descriptor file, which can be obtained from here . Rename the ce.txt file to ce.vmdk

Step 3: Prepare Cent OS VM for AHV + CVM deployment

Create a Cent OS 7 x64 VM with following specifications:

  • 20 GB RAM (atleast)
  • 6 vCPU with “Expose hardware assisted virtualization to the guest os” enabled.

  • Don’t add any disk in during VM creation (delete the one which comes automatically).
  • Change the network adapter type to E1000.

Once VM is created, browse to datastore where VM directory is created and upload ce.vmdk and ce-flat.vmdk files (created in step 2) to vm directory.

Step 4: Modify CentOS VM and Add Disks.

We need to add 3 disks to the Cent OS vm. One for booting AHV and other two for data. 

Edit VM settings and attach new disk by selecting “Existing Hard Disk” 

And select the ce.vmdk file which you uploaded in VM directory. 

Note: Some blog post says the first disk should be attached to SATA controller instead of SCSI controller, but I tested with SCSI and it works just fine. 

Make sure your centos VM have now 3 disks with disk 2 and 3 sized as per hardware recommendation. 

Step 5: Change Boot Order for CentOS VM

Boot the CentOS vm into bios and change the boot order and select HDD as first boot device with SCSI 0:0 on top (as we mapped ce.vmdk file here) and save the changes. 

Step 6: Changes in installation script.

Next step is to boot the VM and make some minor changes in installation script.

Login with user root and nutanix/4u as password and navigate to directory /home/install/phx_iso/phoenix and edit sysUtil.py file.

Modify the SSD IOPS requirements

SSD_rdIOPS_thresh = 4000

SSD_wrIOPS_thresh = 4000

Note: In my lab IOPS threshold 4000 was working fine. In your case you might have to lower down to 3000 or 1000 (depending upon how much iops your backend datastore can provide when installer do a performance test)

If your VM is unable to provide the threshold value which you set in sysUtil.py file, you might see below error.

At this point of time we are ready to go with deployment. To save time you can convert this VM as template and just deploy new instances from the template. 

I hope you enjoyed reading this post. Feel free to share this on social media if it is worth sharing 

Leave a ReplyCancel reply