DEV Community

Cover image for VMware Migration Pre-Flight Checklist: Everything to Prepare Before You Switch
Leo
Leo

Posted on

VMware Migration Pre-Flight Checklist: Everything to Prepare Before You Switch

A comprehensive preparation guide for organizations planning to move off VMware. Use this checklist to avoid the most common (and costly) migration mistakes.


Why a Checklist Matters

Gartner estimates migration costs at $300–$3,000 per VM, with enterprise-scale projects taking 18–48 months. The organizations that migrate fastest and cheapest share one trait: they did thorough pre-migration homework.

This checklist is organized into six phases. Each skipped item is a potential outage, budget overrun, or rollback.


Phase 1: Understand What You're Actually Running

Before you pick an alternative, you need a complete inventory. Most VMware shops discover surprises during migration because nobody kept a clean asset register.

1.1 Server Hardware Inventory

  • [ ] List every ESXi host: brand, model, CPU (vendor + cores + generation), RAM, NIC type/speed
  • [ ] Document storage adapters: Fibre Channel HBAs, iSCSI, NVMe-oF — model, firmware, driver versions
  • [ ] Check HCL compatibility against your target platform (Proxmox HCL, Hyper-V HCL, Nutanix HCL)
  • [ ] Flag any hardware approaching EOL: If a server needs replacement within 18 months, consider refreshing BEFORE migrating (migrate once, not twice)
  • [ ] GPU devices: passthrough, vGPU, NVIDIA GRID — verify driver/API compatibility on target platform
  • [ ] Measure current utilization: CPU overcommit ratio, memory consumption, storage IOPS/throughput — this determines target sizing

1.2 Virtual Machine Inventory

  • [ ] Complete VM list: name, OS, vCPU, RAM, disk count/size, provisioned vs actual usage
  • [ ] Identify VMs with RDM (Raw Device Mappings): These cannot be migrated with standard tools — they need separate storage migration first
  • [ ] Identify VMs with passthrough devices: PCIe, USB, serial ports — verify target platform support
  • [ ] Flag VMs with active snapshots: Migrating a VM with snapshots is slow and risky — consolidate first
  • [ ] Identify orphaned or powered-off VMs: Don't waste effort migrating dead workloads. Archive or delete them now

1.3 Dependency Mapping

  • [ ] Map VM-to-VM dependencies: Application servers → database servers, web frontends → APIs
  • [ ] Identify anti-affinity rules: VMs that must NOT run on the same host (e.g., redundant domain controllers, cluster quorum nodes)
  • [ ] Document boot order requirements: Infrastructure VMs (DNS, AD, NTP) must come up before application VMs
  • [ ] Map external integrations: Monitoring systems, SIEM, CMDB, CI/CD pipelines that talk to vCenter APIs

Phase 2: Understand Your VMware-Specific Dependencies

Some VMware features have no direct equivalent elsewhere. Discover these NOW — not during the migration window.

2.1 Networking

  • [ ] Document all distributed switch (vDS) configurations: port groups, VLANs, NIOC shares, LACP/LAG configs
  • [ ] NSX inventory: DFW rules, micro-segmentation policies, load balancers, NAT rules, VPN tunnels
  • [ ] NSX V → T migration consideration: NSX policies do NOT map 1:1 to any other SDN platform. Expect to manually rebuild firewalling and segmentation
  • [ ] Verify IP addressing strategy: Will VMs keep their IPs? (DHCP reservations, static IPs — document them all)
  • [ ] MAC address dependency check: Some license servers and legacy applications bind to MAC addresses — list them

2.2 Storage

  • [ ] Determine storage architecture: Standalone SAN/NAS → easiest migration. vSAN → need full-stack replacement or storage migration first
  • [ ] Document VMFS datastores: version, LUN mapping, multipathing config
  • [ ] vSAN-specific: disk groups, dedup/compression, erasure coding, stretched cluster — all need replatforming
  • [ ] Storage vendor lock-in: Some storage arrays use VMware-only plugins (VAAI, VASA, VVols). Verify your array supports the target platform

2.3 VMware-Only Features (No Native Equivalent)

VMware Feature Status on Alternatives Mitigation
Fault Tolerance (FT) No true equivalent on any platform Application-layer HA (DB replication, Keepalived, pacemaker)
DRS (fully automated) Manual or semi-auto at best Oversize initial placement, monitor and rebalance manually
vSAN stretched cluster Limited/different implementations Ceph stretch mode (Proxmox), SAN replication, or application-layer
NSX micro-segmentation Partial on Nutanix Flow, Open vSwitch Rebuild policies; expect significant effort
VM Encryption (native) Platform-specific alternatives dm-crypt, Ceph encryption, Hyper-V Shielded VMs
vSphere Update Manager (VUM) Varies by platform Patch management may shift from hypervisor to OS-level tools
  • [ ] Inventory every FT-protected VM — this is your single highest-risk item
  • [ ] Audit DRS affinity/anti-affinity rules — document the INTENT, not just the config, so you can re-implement
  • [ ] Document all vSphere tags and custom attributes used for automation, backup policies, or reporting

2.4 Backup & Disaster Recovery

  • [ ] Confirm backup vendor supports your target platform: Veeam now supports Proxmox, but older versions may not. Check your exact version
  • [ ] Audit backup policies: which VMs, what schedule, what retention — these need re-creating on the new platform
  • [ ] DR replication: SRM, Zerto, array-based replication — verify target platform support or plan replacement
  • [ ] Test restore procedures BEFORE migration: You need a fast rollback path. Validate that VMware backups restore correctly

Phase 3: Know Your Licensing and Financial Position

Timing your migration around licensing events determines your budget and schedule.

3.1 VMware Contract Audit

  • [ ] Find your VMware renewal date(s): Different contracts may have different end dates
  • [ ] Understand your current license model: perpetual (grandfathered) or subscription (VCF/VVF/vSphere Standard)
  • [ ] Request renewal quote from Broadcom NOW (if within 12 months of renewal): This is your baseline for TCO comparison
  • [ ] Calculate the cost of a "bridge" renewal: If migration takes 18 months but your renewal is in 6 months, you may need a short-term renewal. Budget for it
  • [ ] Per-core minimum exposure: 72-core minimum per contract. If you have 48 cores across 6 servers (8-core each), you pay for 72. This changes the math significantly

3.2 Target Platform Budget

  • [ ] License/subscription costs for target platform: Free (Proxmox, XCP-ng) vs paid (Nutanix, Red Hat) vs included (Hyper-V if already on Windows Server DC)
  • [ ] Support contract costs: Proxmox ~$1K/yr per cluster, Nutanix bundled, XCP-ng €340-1,020/yr per node
  • [ ] Migration tooling costs: Many tools are free (StarWind V2V, qemu-img, WAC) but enterprise tools (Veeam, Hystax, ZConverter) cost
  • [ ] Hardware refresh cost (if needed): Factor in if your servers are near EOL or incompatible with the target platform
  • [ ] Training budget: Budget 2-5 days of training per admin, plus a learning curve productivity dip

3.3 Build Your TCO Model

  • [ ] Model 3-year and 5-year scenarios: License + support + migration labor + training + hardware
  • [ ] Include the cost of doing nothing: Staying on VMware for 5 years at Broadcom pricing. This is often the most persuasive number
  • [ ] Calculate break-even point: In how many months does the migration pay for itself?

Phase 4: Assess Your Team's Readiness

The best alternative on paper can fail if your team can't operate it.

4.1 Skills Inventory

  • [ ] Survey the team: Is anyone already familiar with Linux/KVM, Hyper-V, or public cloud?
  • [ ] Identify your vCenter power users: Their workflows need the most translation effort
  • [ ] Assess scripting/automation skills: What's in PowerCLI today? Will it need rewriting in bash/Python/Ansible?
  • [ ] Identify your NSX/networking specialist: They'll have the heaviest lift — NSX migrations are the hardest part

4.2 Training Plan

  • [ ] Schedule training BEFORE migration starts: Teams that learn on the job make costly mistakes
  • [ ] Run a lab migration: Give each admin a sandbox to break and rebuild on the target platform
  • [ ] Identify platform certifications: If required by compliance or confidence-building (e.g., Nutanix Certified Professional, Proxmox VE Administrator)
  • [ ] Document the "vCenter-to-[new platform]" translation map: Common tasks and how they translate (e.g., "vMotion → Live Migration → XenMotion")

4.3 Operations Impact

  • [ ] Map monitoring tools: How will you monitor the new platform? Same tools or new ones?
  • [ ] Patching cadence: How does the target platform handle updates? More frequent? Less?
  • [ ] Backup workflow changes: If backup UI/API changes, update runbooks
  • [ ] On-call runbook update: The 3am page-out playbook must be rewritten for the new platform

Phase 5: Pick Your Migration Strategy

There are four common migration patterns. Choose before you start executing.

5.1 Migration Patterns

Pattern Description Best For
Lift and Shift VMs converted as-is to the target hypervisor Most migrations; lowest risk
Rebuild and Replace VMs rebuilt fresh on the new platform from config management Immature/appliance VMs or where you want to modernize
Hybrid Coexistence Some workloads on VMware, some on new platform, indefinitely Large enterprise; phased migration
Full Rip and Replace All workloads moved, VMware decommissioned Smaller environments; strong motivation to exit
  • [ ] Choose your pattern(s) per workload tier
  • [ ] Decide: one big migration event or staggered batches? (Batches = safer, slower. Big bang = faster, riskier)

5.2 Wave Planning

  • [ ] Define Wave 0 (POC): 20-50 non-critical VMs, validate tools + process + performance
  • [ ] Define Wave 1 (Low Risk): Dev/test environments, internal tools, non-customer-facing apps
  • [ ] Define Wave 2 (Medium Risk): Internal production, batch processing, reporting systems
  • [ ] Define Wave 3 (High Risk): Customer-facing production, databases, FT-protected systems
  • [ ] Define Wave 4 (Hardest): NSX-dependent, FT-protected, legacy OS, complex networked apps

5.3 Tool Selection

  • [ ] Evaluate migration tools against your wave plan: Free tools for simple VMs, enterprise tools for complex/large-scale
  • [ ] Test tool with a single representative VM first, then with your Wave 0 batch
  • [ ] Measure throughput: How many VMs per day/week can the tool realistically handle?

Phase 6: Risk Mitigation and Rollback Planning

Assume something will go wrong. Plan for it.

6.1 Pre-Migration Validation

  • [ ] Run a full backup of every VM the day before migration
  • [ ] Test that backup restores cleanly — the worst time to discover backup corruption is during a failed migration rollback
  • [ ] Snapshot source VMs immediately before migration (if the tool doesn't do this automatically)
  • [ ] Pre-stage target environment: Storage, networking, VLANs, IP pools all configured and tested
  • [ ] Validate target platform can boot a test VM before migrating any production workloads

6.2 During Migration

  • [ ] Define a hard stop-loss trigger: "If X consecutive VMs fail, stop all migrations and investigate"
  • [ ] Monitor source VM performance during replication: Migration tools that do block-level replication can impact source performance
  • [ ] Have a rollback procedure documented and communicated: How fast can you power the source VM back on and fail back?
  • [ ] Migrate during a defined maintenance window (don't rely on live migration for Wave 1)

6.3 Post-Migration Validation

  • [ ] Network connectivity test: Can the VM reach its dependencies? Can clients reach the VM?
  • [ ] Application health check: Does the application start? Do all services pass health checks?
  • [ ] Performance baseline comparison: Compare CPU, memory, disk latency to pre-migration baselines
  • [ ] Backup validation: Does the new platform's backup job succeed? Can you restore?
  • [ ] Monitoring check: Is the VM reporting to your monitoring system? Are alerts firing?
  • [ ] Wait 24-48 hours before deleting source VM/backup: Some issues only surface under real load

The One-Page Summary: Print This Before You Start

BEFORE YOU TOUCH A SINGLE VM, CONFIRM:

□ Complete VM inventory (OS, vCPU, RAM, disk, dependencies) — documented
□ Every FT-protected VM identified and application-layer HA plan ready
□ Every RDM and passthrough device identified and migration path confirmed
□ All NSX rules documented (they will not auto-convert)
□ VMware renewal date known and bridge cost budgeted
□ Target platform selected and POC tested (20-50 VMs)
□ Backup vendor confirmed compatible with target platform
□ Full backup completed and RESTORE TESTED for migration-candidate VMs
□ Team trained on target platform (NOT learning on the job)
□ Rollback procedure written, tested, and time-boxed
□ Migration tool tested at representative scale
□ Wave plan defined and approved by stakeholders
□ Maintenance windows scheduled and communicated
□ Post-migration validation checklist ready
Enter fullscreen mode Exit fullscreen mode

What Happens If You Skip This

Skipped Item Typical Consequence Real-World Example
FT inventory Critical VM has no HA on target platform; outage on first host failure Financial services firm: 6-hour trading system outage
RDM/passthrough audit VM fails to migrate; extended troubleshooting; missed migration window Healthcare org: PACS imaging system offline for 3 days
NSX policy documentation Security rules lost; VMs exposed after migration E-commerce: internal admin panel accidentally internet-accessible
Backup restore test Migration fails, backup is corrupt, no rollback path MSP: 3 customers lost data
Team training Misconfigurations cause repeated outages; team morale tanks Enterprise: migration abandoned halfway, stayed on VMware at higher cost
License renewal timing Forced into expensive short-term renewal; migration budget gone SMB: $80K unplanned Broadcom renewal, migration delayed 12 months

Next Steps

Once you've worked through this checklist:

  1. Select your top 2 candidate platforms from the VMware Alternatives Comparison Guide
  2. Run a POC with 20-50 non-critical VMs on each candidate
  3. Read the migration tutorial for your chosen platform:

Last updated: April 2026. Migration tool landscapes change quickly — verify all tool compatibility before starting your project.

Top comments (0)