The business is growing, and this means more users, more virtual desktops, and you’re likely going to need more RAM. You could add a node or two, but what if rack space isn’t readily available?  The existing nodes may still have room to grow. How do we add memory to nodes in a Nutanix cluster?

The process for upgrading memory is covered in this KB: HW: Upgrading Physical Memory

You should understand the restrictions and have a plan before purchasing DIMMs. If you need to review supported memory configurations, such as what DIMM capacities are supported and which slots to fill, check the NX Series Hardware Administration Guide for details.

There are restrictions on mixing memory in the same node. You can’t just add any DIMM that fits into an existing node, but if each node has some slots available there are options to move forward. You could purchase new DIMMs to populate one node and complete this replacement first, and then use the removed DIMMs to add memory capacity to other nodes in the cluster.

So long as only one node is offline at a time, and you check cluster resiliency before taking a node down, you can keep the cluster and hosted VMs running. The steps for a rolling non-disruptive upgrade are covered in the KB I linked earlier. 

The restrictions on mixing DIMMs in a node are also covered in the NX Series Hardware Administration Guide, or take a look at this post about hardware mixing guidelines.

To identify supported DIMM models for your existing hardware, check with your Nutanix Systems Engineer or Account Manager. If you want to purchase memory through a third party please review the section “WHAT IS THE THIRD PARTY HARDWARE OR SOFTWARE SUPPORT POLICY?” in the following Nutanix policy reference before installation. www.nutanix.com/support-services/product-support/faqs/