What are the cloud migration strategies 6 Rs? AWS's current documentation lists seven: retire, retain, rehost, relocate, repurchase, replatform, and refactor or re-architect. AWS calls them the 7 Rs and does not mention a six-strategy version. For large migrations, AWS says rehost, replatform, relocate and retire are the common choices, and it does not recommend refactor because it is the most complex.
Why Does AWS Say 7 Rs and Not 6?
Many guides still refer to six strategies. AWS's migration strategies page says there are seven, known as the 7 Rs, and its server migration explainer lists the same seven. Neither page says when Relocate was added or why. If a list you find has six items, compare it against the seven below and see which one is missing.
The Cloud Migration Strategies 6 Rs and 7 Rs at a Glance
| Strategy | What AWS says it is |
|---|---|
| Retire | Decommission or archive applications and shut down their servers. |
| Retain | Keep applications in the source environment. |
| Rehost | Move applications to AWS without changing them, also called lift and shift. |
| Relocate | Move many servers at once to a cloud version of the same platform, or between VPCs, Regions or accounts. |
| Repurchase | Replace the application with a different version or product, often SaaS. Also called drop and shop. |
| Replatform | Move with some optimization and no change to the core architecture. Also called lift, tinker and shift. |
| Refactor or re-architect | Change the application architecture to use cloud-native features. |
Rehost, Relocate and Replatform
Rehost moves applications without changes. AWS says it can migrate many machines from physical, virtual or other cloud sources, and that apps are easier to re-architect once they run in the cloud. Applications keep serving users during the move, so downtime depends on the cutover strategy. AWS names AWS Transform MGN, AWS Cloud Migration Factory Solution and VM Import/Export as automation services. Its server migration explainer adds that organizations doing incremental migrations often choose rehost, because data and workloads can move quickly without redesigning anything first.
Relocate needs no new hardware, application rewrites or operational changes. AWS calls it the quickest way to run a workload in the cloud because the architecture does not change. Applications keep serving users during the move. Its example is moving an Amazon RDS DB instance to another VPC or account. The explainer says relocate preserves existing configurations and can run during off-peak hours to avoid scheduling conflicts.
Replatform adds some optimization. AWS's examples include moving Microsoft SQL Server to Amazon RDS for SQL Server, using AWS Graviton processors, and containerizing .NET and Java applications with AWS App2Container. It also lists upgrading the operating system for better security and compliance, and moving from Windows to Linux, with the Porting Assistant for .NET tool to help port .NET Framework apps to .NET Core.
Repurchase, Refactor, Retire and Retain
Repurchase replaces an application. AWS lists moving from traditional licensing to SaaS and upgrading to a vendor's latest version. AWS also lists replacing a custom application to avoid recoding. After the purchase, it says teams and users must be trained, data migrated, authentication services such as Microsoft Active Directory integrated and secure networking configured, and that the vendor typically helps with these.
Refactor or re-architect is the most complex and costly strategy. AWS says that for large migrations you should use it only when other strategies are not acceptable, and modernize after migration where possible. AWS lists legacy mainframe applications that cannot meet demand, monoliths that slow product delivery, and applications with no maintainers or unavailable source code as cases for refactor.
Retire applies when an application has no business value. AWS's migration strategies page treats an application with average CPU and memory below 5% as a zombie and 5% to 20% over 90 days as idle. It also lists eliminating hosting and maintenance costs and reducing the security risk of unsupported operating system versions as reasons to retire an application.
Retain keeps an application where it is. AWS lists data residency, high risk, dependencies on other applications and unresolved physical dependencies among the reasons. It also names recent upgrades that make postponing to the next technical refresh sensible, a planned move to SaaS once the vendor releases a SaaS version, and mainframe or non-x86 Unix applications that need careful assessment.
Which Strategies Suit Large Migrations?
AWS's large migration guide says common strategies for large migrations are rehost, replatform, relocate and retire. Because refactor involves modernizing the application during the move, AWS advises rehosting, relocating or replatforming first and modernizing after the migration is complete.
How to Choose Among the Cloud Migration Strategies 6 Rs Lists
Use the guidance above in this order:
- List each application and note its business value. Retire applications that have none.
- Mark applications that cannot move yet and retain them.
- Rehost or relocate the rest if speed matters most.
- Replatform where a managed service saves time or cost.
- Refactor only where the business case is strong, after the main move where possible.
This order is ours. AWS describes each strategy but does not rank them for every case.
What Do These Sources Not Say?
- They do not explain the difference between the older 6 Rs list and the current 7 Rs.
- They give no cost or time figures for any single strategy.
- They do not give a scoring method for choosing between strategies.
Frequently Asked Questions
Is it the 6 Rs or the 7 Rs?
AWS's current pages say seven. They do not mention a six-strategy version.
Which strategy is quickest?
AWS calls relocate the quickest way to operate a workload in the cloud.
What is lift and shift?
It is rehost: moving applications to AWS without changing them.
Should large migrations use refactor?
AWS does not recommend it for large migrations.
Key Takeaways
- The cloud migration strategies 6 Rs list is seven strategies in AWS's current pages.
- Rehost, replatform, relocate and retire are common for large migrations.
- Refactor is the most complex and costly option.
- Retire and retain are decisions about what not to migrate yet.
Sources and References
Facts follow AWS documentation.
- About the migration strategies (AWS Prescriptive Guidance) docs.aws.amazon.com
- What is Server Migration? (AWS) aws.amazon.com