The Mobilize Phase · Wave Planning & Lighthouse Pilot

In this module, you'll learn to:

• Organize the partner's and the client's engineers as one team with clear ownership
• Group servers into waves by dependencies, criticality, change windows, and tooling
• Run a pilot cutover with a tested rollback, and use what it measures

Before you start: read Financial Engineering and Target State Blueprint. Mobilize starts once that business case is approved, and builds that landing zone.

From "why" to "how"

Once leadership approves the business case, the Assess phase is done and Mobilize begins.

Assess answered "why should we migrate, and what will it cost?" Mobilize answers "how do we build the platform, prove the cutover works, and get our teams ready to run it?" It has three workstreams:

1. Build the landing zone: the accounts, guardrails, and hybrid connectivity designed in Assess (Module 04).
2. Plan the waves: group servers into small, ordered batches.
3. Run the pilot: move one real application end to end, to prove the process before anything else moves.

One team, not two

A migration can't be done to an organization by outside consultants. It works when the partner's people and the client's people work as one team, often called a cloud center of excellence (CCoE):

• Program leads, one from each side: pace, scope, and MAP reporting.
• Architects, one from each side: the landing zone, networking, and security.
• Migration engineers: replication tooling (MGN and DMS), automation, and infrastructure as code.
• Application owners and testers: cutover testing, data checks, and final sign-off.

Agree early who does what. For example, migration engineers run the replication, security engineers approve IAM roles and SCPs, and application owners decide whether a cutover is accepted. A weekly planning meeting keeps the landing zone, replication, and connectivity work moving together.

Grouping servers into waves

Moving everything in one "big bang" weekend is far too risky. Instead, servers move in waves, small batches (often a few dozen servers), grouped by four things:

1. What talks to what. Servers that constantly call each other, such as an application and its database, move in the same wave. Split them, and every call crosses the link between the datacenter and AWS, which slows the application down. Network flow data from discovery shows these links.
2. How critical it is. Development and test environments go first, to practise with no business risk. Internal tools come next. Customer-facing systems move later, once the process is well proven.
3. When change is allowed. Avoid month-end close, peak trading seasons, and other freeze periods.
4. Which tool moves it. Group servers that move the same way: whole servers with AWS Application Migration Service (MGN), databases with AWS Database Migration Service (DMS) or native replication.

Wave 0

The most important milestone in Mobilize is the pilot, sometimes called the lighthouse or Wave 0. Pick an application that is:

• low risk, so a delay doesn't hurt customers or revenue;
• typical of the estate, for example a three-tier application with a web front end, an application tier, and a relational database;
• owned by someone keen to help, who will test it properly and give honest feedback.

A full dress rehearsal

Ready for the waves

Mobilize ends with:

1. A production-ready landing zone, with guardrails, central logging, and working hybrid connectivity.
2. The full wave plan for every remaining application.
3. A proven runbook, corrected after the pilot, that every later wave follows.
4. The MAP Mobilize reporting that AWS or your partner requires.

With the platform built, the runbook proven, and the internal teams trained, the remaining waves can move steadily and repeatably.

Key terms

• Mobilize: the phase after the business case is approved: build the landing zone, plan the waves, and run the pilot.
• CCoE: a cloud center of excellence, the partner's and client's people working as one team.
• Migration wave: a small, ordered batch of servers grouped by what talks to what, criticality, change windows, and tooling.
• MGN and DMS: AWS Application Migration Service moves whole servers; AWS Database Migration Service moves databases.
• Pilot (lighthouse, Wave 0): the first real application moved end to end, to prove the cutover and the rollback.
• Reverse replication: replication from AWS back to the original database, so a rollback doesn't lose data written after the switch.

Pop quiz

Five questions: three on the ideas in this module, and two scenarios where you apply them. The order changes every time you take it, and 4 out of 5 passes.