Hyper-V + SCVMM POC Agenda & Execution Guidelines

Storage Models: Local Storage | SAN/NAS | Storage Spaces Direct (S2D)
Purpose: Validate host configuration, storage scenarios, networking, clustering, SCVMM integration, VM provisioning, and operational workflows.

POC Agenda Overview

Day 0 – Pre-POC Preparation (Internal + Customer)

• Confirm hardware, firmware, and OS readiness.

• Validate network readiness (VLANs, IPs, routing, MTU).

• Validate storage readiness (LUNs, SMB shares, S2D disks).

• Confirm SCVMM prerequisites & accounts.

• Freeze configuration changes during POC.

Day 1 – Kickoff & Requirements Alignment

Goals: Define success criteria, test cases, stakeholders.

Agenda:

1. POC kickoff meeting & success criteria

2. Review current environment (Hosts, Clusters, Storage, Networking)

3. Define storage models to be tested:

   • Local storage (standalone host or non-cluster VMs)

   • SAN/NAS shared storage (cluster CSV)

   • Storage Spaces Direct (S2D hyper-converged)

4. Define SCVMM use cases:

   • Fabric discovery

   • Logical networks

   • Storage classifications

   • VM templates

5. Confirm project timeline and reporting cadence

Deliverables:

• POC Success Criteria Document

• Environment Readiness Checklist

Day 2 – Host Preparation & Validation

Goal: Build clean, stable Hyper-V hosts with best practices applied.

Agenda & Steps:

1. Update BIOS, firmware, network, and storage drivers on all hosts.

2. Install Windows Server 2022/2019 with secure baseline.

3. Apply OS patches + reboot.

4. Join domain.

5. Install Hyper-V + Failover Clustering + Mgmt tools.

6. Configure power profile to High Performance.

7. Validate:

   • Virtualization extensions

   • NTP sync

   • DNS resolution

   • WinRM open

   • Firewall rules

Deliverables:

• Host Validation Report

Day 3 – Networking Setup (Critical for Zero Errors)

Goal: Build repeatable, error-free networking foundation.

Agenda & Best Practices:

1. Verify VLANs and IP segments with network team.

2. Confirm required networks:

   • Host Management

   • Cluster

   • Live Migration

   • VM/Tenant

   • Storage (SMB, iSCSI, FC)

3. Create SET vSwitch on hosts:

   • Minimum 2×10/25/40G NICs

   • No LBFO (deprecated)

4. Create vNICs for:

   • Mgmt

   • LM

   • Cluster

   • Storage

5. Verify Jumbo Frames (only where end-to-end supported).

6. Validate RDMA (if applicable).

Deliverables:

• Network Validation Checklist

• vSwitch and vNIC Config

Day 4 – Storage Testing: Local, SAN/NAS, S2D

Goal: Validate all storage scenarios with correct configuration.

Scenario A: Local Storage (Standalone Hyper-V)

Steps:

1. Create a standalone host with local SSD/NVMe.

2. Create local VM storage paths.

3. Deploy test VMs directly on the host.

4. Test:

   • VM deployment speed

   • Checkpoints

   • Dynamic vs fixed VHDX

   • Live Migration (cross-host impossible here)

Success Criteria:

• VMs deploy successfully

• Host performance matches expectations

Scenario B: Shared Storage – SAN/NAS (Clustered Hyper-V)

Prerequisites:

• LUNs masked/presented to all hosts

• Identical paths (MPIO)

• Same HBA firmware + driver

Steps:

1. Configure MPIO.

2. Validate LUN visibility across hosts.

3. Create cluster.

4. Add disks to cluster → Convert to CSVs.

5. Test cluster roles:

   • Live Migration

   • Quick Migration

   • Storage Migration

6. Validate SCVMM recognition of CSVs & classifications.

Success Criteria:

• Stable cluster

• Host drain/resume works

• SCVMM can provision to CSV

Scenario C: Storage Spaces Direct (S2D)

Prerequisites:

• Matching nodes (CPU, NICs, disks)

• At least 4 SSD/NVMe if using hybrid

• RDMA recommended

Steps:

1. Validate disks with:

   Get-PhysicalDisk

2. Run full cluster validation:

   Test-Cluster

3. Enable S2D:

   Enable-ClusterS2D

4. Create storage pool:

   New-Volume -FriendlyName "S2D-CSV1" -FileSystem CSVFS_ReFS -StoragePoolFriendlyName "S2D Pool" -Size 500GB

5. Validate:

   • ReFS integrity

   • CSV availability

   • Fault tolerance

6. Test performance with VM deployment + Live Migration.

Success Criteria:

• S2D cluster healthy (no warnings)

• SCVMM detects S2D storage pool

• ReFS accelerated VHDX creation works

Day 5 – SCVMM Integration & Fabric Build-Out

Goal: Build full SCVMM fabric model with no deployment errors.

Steps:

1. Install SCVMM on dedicated server.

2. Connect to SQL instance (local or remote).

3. Add Hyper-V hosts:

   • Verify VMM agent installation

   • Confirm WinRM & permissions

4. Configure fabric components:

   • Logical networks

   • IP pools

   • Uplink port profiles

   • Logical switch

   • Storage classifications

5. Deploy Logical Switch to hosts.

6. Test VM Template provisioning.

7. Test Live Migration via SCVMM.

8. Validate placement rules & host ratings.

Success Criteria:

• SCVMM deploys logical switch without errors

• SCVMM deploys VMs successfully

• Host status = “OK” in VMM

Day 6 – Operational Testing & Failover

Goal: Validate availability + automation.

Tests:

1. Host Pause / Drain roles

2. Live Migration under load

3. Storage Migration across CSVs

4. Cluster failover

5. SCVMM servicing baseline / updates

6. Fail-back testing

7. Logging & monitoring validation (Event Viewer, SCVMM Jobs, Cluster Log)

Success Criteria:

• Zero failed cluster migrations

• Hosts remain compliant after updates

• SCVMM job logs clean

Day 7 – Findings, Gaps, Recommendations

Goal: Provide a deliverable that leadership can approve for next phase.

Deliverables:

• Architecture diagram

• Storage performance findings

• Networking issues (if any)

• SCVMM configuration gaps & recommendations

• Best practice remediation list

• Ready-for-production checklist

Best-Practice Guardrails to Ensure ZERO Errors

Host

• All hosts on same OS version + patch level

• Identical NIC, HBA, and storage firmware

• Consistent BIOS settings

• Time synchronization validated

Networking

• VLANs confirmed & routed before deployment

• Jumbo frames consistent across switches

• Only SET (no LBFO)

• Unique IPs for Mgmt, LM, Cluster, Storage

Storage

• SAN zoning & masking done before clustering

• All hosts see identical LUN IDs

• S2D disks identical in:

  • Type

  • Size

  • Firmware

Cluster

• Run Test-Cluster and fix all errors

• Configure Witness (File Share or Cloud)

• Reboot hosts after major changes

SCVMM

• Use a dedicated service account

• Apply Logical Switch from SCVMM (not manually)

• Ensure Run As accounts have corrected rights

• Ensure SCVMM SQL latency < 10ms

• Always refresh host cluster after changes