Wave-Based Migration with Coexistence

Why migrate in waves

Migrating everything at once, the so-called big bang, concentrates risk in a single window and leaves little room for correction. Wave-based migration splits the work into manageable batches, lets you learn each cycle and keeps the operation running with controlled coexistence between source and target.

As a Microsoft Solutions Partner, RHC applies this method in datacenter exits and in Microsoft 365 and Azure workload migrations alike.

The method's principles

A wave migration rests on four principles:

1. Small, frequent batches reduce the blast radius of any failure.
2. Planned coexistence lets old and new systems operate together during the transition.
3. Success criteria defined before each wave prevent advancing over problems.
4. Incremental learning: each wave improves the next.

How to group the waves

Grouping is the most important decision. Good grouping dimensions:

• By dependency: systems that communicate migrate together, avoiding source-target latency.
• By business unit or geography: reduces the organizational impact per pass.
• By criticality: start with low-risk workloads to calibrate the process.
• By maintenance window: respects acceptable downtime periods.

Coexistence: the heart of the transition

During migration, the two worlds coexist. Coexistence must be designed, not improvised:

• Unified identity so users access resources on both sides frictionlessly.
• Hybrid connectivity (VPN or ExpressRoute) with adequate bandwidth and latency.
• Routing and DNS that direct each service to the correct location.
• Data synchronization when source and target must stay consistent.

In Microsoft 365, for example, email coexistence lets migrated and non-migrated mailboxes exchange messages and see calendar availability for weeks.

Success criteria per wave

Before starting each wave, define what completion means:

1. Functional validation: systems do what they should in the target.
2. Performance: latency and response time within acceptable bounds.
3. Integrations: flows with other systems work.
4. Business-owner sign-off: users confirm it is good.
5. Rollback available: there is a way back if something fails.

Without these criteria, the temptation to push forward over problems grows, and risk accumulates.

Change management and communication

Wave migration is as much human as technical. Each wave affects people:

• Communicate in advance who will be migrated and when.
• Train users on experience changes.
• Offer reinforced support in the first days after each cutover (hypercare).
• Collect feedback and adjust the following waves.

Well-handled adoption is what turns a technical migration into real value.

Reducing risk each cycle

The great benefit of waves is compounding learning. After each wave:

• Run a short retrospective: what worked, what stalled.
• Adjust runbooks and automation.
• Re-estimate timelines with real velocity data.
• Refine the grouping criteria.

Key takeaways

• Prefer waves over big bang to dilute risk and learn each cycle.
• Group by dependency to avoid source-target latency.
• Carefully design coexistence of identity, network and data.
• Define success criteria and rollback before each wave.
• Treat change management and hypercare as part of scope.
• Use retrospectives to improve the following waves.

RHC structures the wave plan, operates coexistence and runs hypercare, delivering predictable, low-risk migrations.