My home lab helps me learn the technology I am assisting customers with.  I use it to:

• Test updates and configurations
• Walk through steps & screens with customers
• Write blog posts
• Experiment and learn  – a place where I can break things without impacting anyone.

So far the family CFO has been very understanding of the money and time I have invested.  I try to be a good steward of that credit!

Current Home Lab ( Mark V – November 2022 )

This version of my lab has a Management node and Compute cluster.

Management Node

MacBook Pro Retina (16gb RAM, 1 TB SSD) – VMware Fusion

• • Used to run DNS/DHCP/Mail Relay services
  • Using this device helps isolate the most critical services from the lab – ensuring continued services while the lab undergoes change.

Primary Compute Cluster

This is a 4-node cluster running vSphere 7

3 nodes are:

• • ASUS Prime H570-PLUS motherboard
  • Intel Core I-5 10400  CPU
  • 128 GB RAM
  • HP NC382T dual-port NIC
  • Intel 520x 10gb NIC

4th node – HP z620 Workstation

• • Intel Xeon CPU E5-2620
  • 96 GB RAM
  • HP NC382T dual-port NIC
  • Intel 520x 10gb NIC

Secondary Compute Cluster

I have repurposed one of the older hosts from a prior lab iteration to create a single platform that can support 2 nested ESXi hosts. This is primarily to allow me to experiment with Site Recovery Manager and vRealize Automation blueprints that deploy to multiple clusters.

Networking

• Core Switch – Dell PowerConnect 6248
• 10gb switch – MicroTIK CRS309-1G-8S+IN for vMotion and possibly vSAN in the future
• Firewall – Ubiquiti UDM SE
• Wifi – Ubiquiti AP-AC-LR

Storage

Synology DS1512+

• • iSCSI and NFS
  • Seagate IronWolf – 4TB (x3) – capacity
  • Samsung 128GB SSD (x2) – read/write cache

Synology DS920+

• • iSCSI
  • 2.6 TB Flash

Workloads

• vRealize Automation
• vRealize Identity Manager
• VMware Horizon
• vRealize Operations Manager
• vRealize Log Insight
• Plex

Other Applications

• Veeam Backup & Replication Community Edition – Veeam continues to be an excellent product!  I am very thankful to Veeam for keeping a free edition of their product for vExperts and IT Pros to learn with.
• Nagios Core – This is running on a Raspberry PI. My time is limited, so I only use it for basic up/down notifications of lab components.   It can certainly do much more!

Notes

• Using the Macbook Pro as a management node may seem like an unconventional choice.  It is an older Macbook but has enough power to run the bare necessities.  It has a built-in UPS, and KVM  as well as easy remote access.  Most importantly – it is very easy for my family to power it back on after a power outage if I am away!
• I have used whitebox-based ESX hosts for a long time.  It does require some research and careful planning using the HCL for the various components.  The upside for me has been – the lab is quiet, saves power, and was easier to adapt to requirement updates.  One of the changes with vSphere 7 is the evolution of Update Manger to Lifecycle Manager.    Lifecycle Manager has become aware of the host make & model.  It uses this capability to check the HCL as part of the remediation process. This has great potential to identify upgrade problems before a customer implements the upgrade.  This may eventually  impact vSphere Homelabs using whitebox-based ESXi hosts; but the capabilities will be a great benefit to production environments.

Home Lab .NEXT

The memory capacity of the cluster is close to 480gb, I finally have enough compute to run the full VMware stack at a moderate activity level and still have some headroom.    Some items on my planning list:

• Replacing the z620 workstation with a matching host in the Compute Cluster.    The CPU in this system is reaching the end of the support line.  While I believe I will be able to get it to work with vSphere 8, I will eventually need to replace it.
• Replace the Dell PowerConnect Core switch.  I am quite firmly entrenched in the Ubiquiti ecosystem.  Their products work quite well for what I need and provide enough “nerd nobs” to keep me occupied.  If I can find the budget, it would be nice to replace this aging switch with a model that is managed in the same system as the other Ubiquiti products.
• vSAN – With the 10gb MicroTIK switch, I can implement an all-flash vSAN datastore.  I will use the on-board NVME slot on 3 hosts and add a PCI-E NVME card for the 4th.
• Public Cloud integration with VMware solutions and Azure or AWS.
• I have resisted going the route of used Enterprise gear.  This is mostly due to the noise/power factor.  However if I have to rebuild my white-boxes significantly it may be more cost effective to go that route at some point (A single large host that I can use to nest environments may be a good option)

Previous Generations

Mark I  – January 2010

This was a single host (core i-7 and 12 gb of ram).  This was my first lab, built a few years after I started working with VMware products.  It served me quite well for a long time.

Mark II – May 2015

• 3 ESX hosts
  • core i7 from previous generation with ram increased to 24gb
  • 2  – HP DL 380’s with 32gb RAM each

Mark III – March 2017

• 3-node Compute cluster (whitebox nodes, each has 32gb RAM, core I5 CPU, and hybrid vSAN)
• Single-node Management cluster (MacBook Pro Retina running Fusion)
• Synology DS1512+ (12TB raw with R/W cache)
• Networking – Dell PowerConnect 6428, Ubiquiti Wifi APs and Edge Gateway

Mark IIIa – April 2020

• 4-node Compute Cluster (3 whitebox nodes with 32gb RAM and Core i5 CPU, 1 HP z620 Workstation – 96gb RAM)
• Single-node Management cluster (MacBook Pro Retina running Fusion
• Synology DS1512+ (12TB raw with R/W cache)
• Networking – Dell PowerConnect 6428, Ubiquiti Wifi APs and Edge Gateway

Mark IV – June 2021

• 4-node Compute Cluster (3 whitebox nodes with 64gb RAM and Core i5 CPU, 1 HP z620 Workstation – 96gb RAM)
• Single-node Management cluster (MacBook Pro Retina running Fusion
• 2 Synology Storage arrays
  • 1512+ – 12 TB spinning disk with R/W cache
  • 920+ – 2.6 TB Flash
• Networking – Dell PowerConnect 6428, Ubiquiti Wifi APs and Edge Gateway