AWS Migration Strategies: The 7Rs Master Study Guide

Determining the right migration path is the cornerstone of the AWS Certified Solutions Architect - Professional exam. This guide explores the seven common migration strategies (the 7Rs) used to evaluate applications and workloads for cloud transition.

---

Learning Objectives

After studying this guide, you should be able to:

• Identify and define each of the 7Rs of migration.
• Evaluate application workloads to determine the most cost-effective and efficient migration strategy.
• Differentiate between similar strategies such as Relocate vs. Rehost and Replatform vs. Refactor.
• Analyze database migration paths (homogeneous vs. heterogeneous) within the context of the 7Rs.

Key Terms & Glossary

• 7Rs: A framework for categorizing migration strategies (Retire, Retain, Relocate, Rehost, Repurchase, Replatform, Refactor).
• TCO (Total Cost of Ownership): A financial estimate intended to help buyers determine the direct and indirect costs of a product or system.
• Wave Planning: The process of grouping applications into logical migration groups based on dependencies and business priority.
• Lift and Shift: A colloquial term for Rehost, where an application is moved to the cloud without modifications.
• Homogeneous Migration: Migrating a database to the same engine (e.g., Oracle to Oracle on EC2).
• Heterogeneous Migration: Migrating a database to a different engine (e.g., Oracle to Amazon Aurora).

The "Big Idea"

Migration is not a "one-size-fits-all" process. The objective is to balance Time-to-Value with Cloud-Native Optimization. Organizations typically start with faster strategies (Rehost/Relocate) to exit data centers quickly and then perform more complex optimizations (Replatform/Refactor) once they are already running in the AWS environment.

Formula / Concept Box

Hierarchical Outline

1. Phase 1: Assess & Mobilize
   • Inventory applications and assess business value.
   • Evaluate TCO and technical dependencies.
   • Tools: AWS Migration Hub, Application Discovery Service.
2. Phase 2: Strategy Selection (The 7Rs)
   • Non-Migration Paths: Retire (Decommission), Retain (Keep on-prem).
   • Rapid Migration Paths: Relocate (Hypervisor level), Rehost (OS level).
   • Transition Paths: Repurchase (Switch to SaaS).
   • Optimization Paths: Replatform (Modify platform), Refactor (Rearchitect code).
3. Phase 3: Database Specifics
   • WQF (Workload Qualification Framework): Categorizing database workloads.
   • Homogeneous: Like-for-like (Rehost/Relocate).
   • Heterogeneous: Cross-platform (Refactor).

Visual Anchors

Migration Decision Logic

Effort vs. Optimization Curve

Definition-Example Pairs

• Relocate: Moving a workload at the hypervisor level.
  • Example: Moving a VMware VM from a local data center to VMware Cloud on AWS without changing the VM configuration.
• Replatform: Changing the platform or underlying infrastructure to a managed service without changing core code.
  • Example: Moving a self-managed MySQL database on a Linux VM to Amazon RDS for MySQL.
• Repurchase: Moving to a different product, typically a SaaS model.
  • Example: Replacing an on-premises self-hosted CRM with Salesforce or a legacy email server with Amazon WorkMail.

Worked Examples

Example 1: The Urgent Data Center Exit

Scenario: A company has a data center lease expiring in 3 months. They have 200 web applications. Strategy: Rehost. Reasoning: Rehosting is the fastest way to migrate. Changes to architecture (Refactoring) or platform (Replatforming) would take too long to meet the 3-month deadline. They can modernize "in-cloud" later.

Example 2: Modernizing a Proprietary DB

Scenario: A company has an Oracle database using specific proprietary features (Category 4 workload) but wants to reduce license costs. Strategy: Refactor. Reasoning: Since the workload is tied to proprietary features, a heterogeneous migration to Amazon Aurora is required. This involves re-architecting the database schema and application code, which defines the Refactor strategy.

Checkpoint Questions

1. Which of the 7Rs specifically involves the use of VMware Cloud on AWS?
2. What is the main difference between Rehost and Relocate?
3. True or False: Replatforming requires a complete rewrite of the application's source code.
4. Which strategy is most appropriate for an application that is no longer used but still costs money to run?

[!TIP] Answers: 1. Relocate. 2. Relocate moves the hypervisor/VM state; Rehost moves the workload to AWS-native instances (EC2). 3. False (that is Refactor). 4. Retire.

Muddy Points & Cross-Refs

• Rehost vs. Relocate: This is a common exam trap. Relocate is specifically for hypervisor-to-hypervisor moves (like VMware). Rehost moves the workload to a new platform (like EC2) even if the source was a VM.
• Replatform vs. Refactor: Think of Replatform as "tinkering at the edges" (changing the OS or the database service) and Refactor as "changing the heart" (changing the code from monolithic to microservices).

Comparison Tables

Detailed Strategy Comparison

---