Cisco Compute Hyperconverged with Nutanix ISM Field Guide

[beveritt]

To ensure the correct installation process is being followed for Cisco Compute Hyperconverged with Nutanix, please refer to the following flowchart to choose the correct documentation for the system being installed: 

This document covers the ISM deployment method.

To access the UMM deployment field guide, click here.

To access the IMM deployment filed guide click here.

A Field Guide covering the installation and initial configuration of Cisco Compute Hyperconverged with Nutanix systems using Intersight Standalone Mode for Cisco UCS C-Series servers. Intersight Standalone Mode (ISM) is the term for C-series rackmount servers that are not connected to Cisco UCS Fabric Interconnects, and are fully managed by Cisco Intersight. This guide is intended for technical training and educational purposes, for all who wish to install Nutanix on Cisco UCS based hardware following our supported hardware and software integrations and compatibility lists.

As a reference guide, the document presents our best practices and recommendations as of the time of publishing, and will be updated periodically. Please refer to the document history at the beginning to ensure you are using the most current version available.

The most recent revision, v2.2 adds information regarding role-based access control (RBAC) in Cisco Intersight and a few other minor improvements. Additionally, our guides are being renamed to help provide consistent naming related to the management modes used in each document.

Version v2.1 added information regarding migrating the Cisco Intersight Appliance from ESXi to AHV when it is used during greenfield deployments. Also added is information regarding the Cisco Intersight Device Connector for Prism Central, enabling inventory, monitoring and alerting of Nutanix clusters within Cisco Intersight. There are numerous other improvements to the document flow, updated screenshots, additional information for troubleshooting and network connectivity requirements, and dozens of edits for additional clarity. 

Version v2.0 added information regarding cluster expansions using Foundation Central, additional information when using a Cisco Intersight Private Virtual Appliance during installations, updated information for installing Prism Central on a non-Nutanix ESXi cluster, plus several minor edits.

Version v1.2 added information regarding using Nutanix Lifecycle Manager (LCM) to perform software upgrades and firmware upgrades on the Cisco rackmount servers.

The first revision v1.1 added information regarding the deployment of Witness VMs for 2-node clusters, additional information regarding deployment of Prism Central on ESXi clusters, and several other minor updates.

[mteixeira]

Thank you for this doc.

[magnus.von-rosen]

Magical! Great guide. I've just deployed (for three months running) a small Intersight non-managed cluster at a customer in Sweden, and with the ACI plugin to AHV. It's been a bit of a mess, versions wise. This gives a bit of hope. Looking forward to the "in two months" releases!

[s-hadi]

Hi @beveritt 

Thank you so much for this detailed guide.
We (netone: a Japan based company) are trying to use LCM to upgrade server firmware.

Can you please let me know, what firmware was installed, when you execute "Perform inventory" on page 112?
Page 113 is showing [4.3(4.240152)] as a recommended version. 

Our current firmware is [4.3(4.240152)] and I am thinking of downgrading the server firmware once and then using LCM to upgrade the firmware back to [4.3(4.240152)].

Thank you in advance,
Hadi

[magnus.von-rosen]

If you are deplyoing with Intersight, don’t worry. Intersight will deploy the correct firmware for you automatically, up or downgrade.

/Magnus

[s-hadi]

Hi @magnus.von-rosen 

Thank you for your kind response

Yes, I understand that Intersight is the magic lamp :D.
But My goal is to check what we can do with LCM when its about the server firmware.

Regards,
Hadi

[magnus.von-rosen]

Ah, I see.
My Nutanix Team advised me that LCM Updates is possible since July, LCM version 3.0.1 makes it possible to update AOS etc from LCM.
I'm not sure about firmware, haven't tested it myself yet.

BR
Magnus

[Mdu]

Thank you very @beveritt 

Very much helpful

[sunman123]

thanks for this guide! 

One question -- the guide seems to suggest deploying prism central on an existing hypervisor in the environment and then building the new Nutanix cluster on UCS. If we are building with AHV on the new cluster, should we use Nutanix Move to migrate prism central afterwards? Or build a new prism central instance on the AHV cluster and unregister from the old PC and re-register to new PC?

All existing ESXi hardware will be going away after the new cluster is online.

[beveritt]

The cluster being built is not automatically registered to Prism Central, so there is no need to unregister. Using Move to migrate Prism Central is not recommended. The Prism Central server used to deploy the first cluster is basically thrown away, then you would deploy Prism Central for long term use on the AHV cluster you just built, and register the cluster to it.  

[sunman123]

awesome, thanks so much for the clarification and the speedy response. Cheers!

[sunman123]

sorry, another question -- i noticed the guide has no mention of Hyper-V. Is it the same process if we're currently a Hyper-V shop? (Deploy 2022.6.0.7 for Hyper-V and follow the same instructions for modifying the network interfaces, host files, and launching cluster creation)

[beveritt]

I don't know the answer 100% because we did not test Prism Central and Foundation Central with Hyper-V, but I am assuming that the 2022.6.0.7 version works the same on ESXi or Hyper-V. The only way to know will be to try it out!

[sunman123]

thanks, will give it a shot and report back to others. Is the ISM guide the only way to deploy a Nutanix cluster on UCS? If it turns out that Hyper-V doesn't work, I'm thinking we may need to stand up a temporary ESXi host on VMW workstation unless there is another procedure.

[beveritt]

Yes, the process using Foundation Central is the only method for installing a cluster at this time.