Traditional Cluster Setup Guide (Without vLCM)

Overview

This guide walks through creating a traditional vSphere cluster using baseline management instead of vLCM. This is the recommended approach for environments with:

  • Community-supported VIBs (like USB NIC drivers)
  • Mixed ESXi versions
  • Custom drivers or configurations

Step 1: Create Traditional Cluster

1.1 Start Cluster Creation

  1. Right-click on Homelab-DC in vCenter
  2. Select: New Cluster

1.2 Configure Cluster Settings

General Configuration:

  • Name: Compute-Cluster
  • DRS (Distributed Resource Scheduler):
    • ✅ Enable DRS
    • Automation Level: Fully Automated
    • Migration Threshold: 3 (Balanced) - Apply priority 1, 2, and 3 recommendations
    • Predictive DRS: ✅ Enable (optional)
  • HA (High Availability):
    • ✅ Enable HA
    • Host Monitoring: Enabled
    • Host Failure Response: Restart VMs
    • Host Isolation Response: Leave powered on
    • Datastore with PDL: Disabled
    • Datastore with APD: Disabled
    • VM Monitoring: VM and Application Monitoring

1.3 vSphere Lifecycle Manager Settings

IMPORTANT: Select the traditional option:

  • 🔘 Manage all hosts in the cluster collectively with baselines
  • ⭕ Manage all hosts in the cluster with a single image

1.4 Advanced Options

  • Enhanced vMotion Compatibility (EVC):
    • Leave Disabled for now (configure after adding hosts)
  • Virtual Machine Component Protection: Leave defaults
  • Proactive HA: Enable if desired

1.5 Complete Creation

  • Click OK to create the cluster

Step 2: Add Intel NUCs to Cluster

2.1 Move Hosts to Cluster

You have two options:

Option A: Drag and Drop (Easiest)

  1. In the vCenter inventory, find your Intel NUCs
  2. Select all three:
    • esxi-nuc-01.markalston.net
    • esxi-nuc-02.markalston.net
    • esxi-nuc-03.markalston.net
  3. Drag them onto Compute-Cluster
  4. Confirm the move

Option B: Right-Click Method

  1. Right-click on each host
  2. Select Move To…
  3. Choose Compute-Cluster
  4. Click OK

2.2 Verify Cluster Membership

After moving hosts:

  1. Click on Compute-Cluster
  2. Go to Hosts tab
  3. Verify all three Intel NUCs are listed
  4. Check that status shows as Connected

Since all Intel NUCs have the same CPU (Skylake i7-6770HQ):

  1. Right-click on Compute-Cluster
  2. Select: Settings
  3. Navigate to: VMware EVC
  4. Click: Edit
  5. Enable EVC for Intel Hosts
  6. Select Mode:
    • For Skylake CPUs: “Intel Broadwell” Generation or
    • “Intel Skylake” Generation if available
  7. Click: OK

Note: This ensures vMotion compatibility even with slight CPU differences.

Step 4: Configure Cluster Services

4.1 Verify DRS is Working

  1. Select Compute-Cluster → Monitor → DRS
  2. Check for DRS recommendations
  3. Verify DRS Score appears

4.2 Test HA Configuration

  1. Select Compute-Cluster → Monitor → vSphere HA
  2. Verify:
    • All hosts show as HA Agent Running
    • Cluster Status: Protected
    • Configured/Available Slots displayed

4.3 Configure Admission Control

  1. Cluster → Configure → vSphere Availability
  2. Edit Admission Control
  3. Recommended setting:
    • Define failover capacity by: Percentage of cluster resources
    • CPU: 25% (tolerates 1 of 4 hosts failing)
    • Memory: 25%

Step 5: Set Up Update Manager

Since we’re using traditional baselines instead of vLCM:

5.1 Access Update Manager

  1. Go to Menu → Lifecycle Manager
  2. Select Baselines tab (not Images)

5.2 Create Custom Baseline (Optional)

For your Intel NUCs with USB NICs:

  1. Actions → New → Baseline
  2. Name: Intel-NUC-ESXi-8.0-Updates
  3. Type: Host Patch
  4. Target: ESXi 8.0.x hosts

5.3 Attach Baselines to Cluster

  1. Select Compute-Cluster
  2. Updates → Baselines
  3. Attach → Select appropriate baselines
  4. Recommended baselines:
    • Critical Host Patches
    • Non-Critical Host Patches

Step 6: Configure Resource Pools (Optional)

To organize workloads:

  1. Right-click on Compute-Cluster
  2. New Resource Pool
  3. Create pools like:
    • Production-VMs (High share values)
    • Test-VMs (Normal share values)
    • Dev-VMs (Low share values)

Step 7: Verify Cluster Health

Final Checklist:

  • All 3 Intel NUCs showing in cluster
  • DRS enabled and showing recommendations
  • HA showing all agents running
  • EVC mode configured (optional)
  • Update baselines attached
  • No configuration issues in cluster summary

Test Basic Operations:

  1. Create a test VM on the cluster
  2. Test vMotion between hosts
  3. Verify DRS moves VMs automatically
  4. Check storage visibility across all hosts

Troubleshooting

Common Issues:

HA Agent Issues:

# Reconfigure HA on problematic host
ssh root@esxi-nuc-01.markalston.net
/etc/init.d/hostd restart
/etc/init.d/vpxa restart

DRS Not Working:

  • Verify vMotion network configured
  • Check for VM affinity rules
  • Ensure shared storage accessible

EVC Warnings:

  • May see warnings about future CPU support
  • Safe to ignore for homelab use

Management Best Practices

For Your Mixed Environment:

Intel NUC Cluster (ESXi 8.0.3):

  • ✅ Managed by this traditional cluster
  • ✅ Full DRS/HA functionality
  • ✅ Use Update Manager for patches
  • ✅ USB NIC drivers preserved

Mac Pro (ESXi 7.0.3):

  • ❌ Keep separate from cluster
  • 📌 Different ESXi version
  • 📌 Manage independently
  • 📌 Perfect for vCenter VM

Update Strategy:

  1. Use Update Manager baselines
  2. Stage updates to one host first
  3. Let DRS evacuate VMs
  4. Update remaining hosts

Next Steps

With your cluster configured:

  1. Configure networking (standard or distributed switches)
  2. Set up storage policies
  3. Create VM folders for organization
  4. Deploy production VMs

Summary

You now have a fully functional traditional cluster that:

  • ✅ Works with USB NIC drivers
  • ✅ Supports your mixed environment
  • ✅ Provides HA/DRS functionality
  • ✅ Uses proven baseline management
  • ✅ Avoids vLCM compatibility issues

This approach gives you all the benefits of clustering without the complexity of image management!

This project is for educational and home lab purposes.