
Rehost, replatform, refactor, repurchase, retire and retain; a sample mix for 120 servers and a five-year TCO comparison.
The most common mistake in cloud migration projects is moving all servers to the cloud as they are. This "lift and shift" approach is sometimes a quick fix but often leads to a cloud bill higher than running on-premises. The 6R framework lets you choose the right strategy for each application.
What are the 6Rs?
- Rehost: the application is moved to the cloud as is. It is quick and needs the least change.
- Replatform: moved with small improvements; for example the database switches to a managed service.
- Refactor: the application is rewritten for a cloud-native architecture. It is costly but gives the biggest long-term gain.
- Repurchase: a ready SaaS solution replaces the application (for example email or CRM).
- Retire: unused or duplicate applications are switched off.
- Retain: latency-sensitive, regulated or legacy-hardware-dependent applications stay on-premises.
Example mix: a company with 120 servers
| Strategy | Servers | Note |
|---|---|---|
| Rehost | 55 | A fast wave, in three months |
| Replatform | 25 | Databases to a managed service |
| Refactor | 6 | Critical customer-facing applications |
| Repurchase | 10 | Email, file sharing, HR |
| Retire | 14 | Unused test and legacy systems |
| Retain | 10 | Production line and customs systems |
In the inventory stage alone, 14 servers were found to be unnecessary; that by itself means a yearly saving of about 11 percent.
Cost comparison
In five-year total cost of ownership (TCO): on-premises operation 18 million TRY; moving everything as is 17.2 million TRY (limited gain); a 6R strategy with FinOps 13 million TRY. The difference comes from right-sizing, reserved capacity and the systems retired. The figures are examples.
Tips for success
- Cloud governance (tagging, budget alerts, identity) should be in place before the migration.
- Migration should be done in waves, measuring cost and performance at the end of each wave.
- Data residency should be addressed at the start because of KVKK and sector regulation.
Omurga starts with a readiness assessment and ends with a business case report that can be presented to the board.
A security and compliance note
Identity, encryption and logging decisions should be made at the very start of the architecture during migration. The shared responsibility between cloud provider and customer should be made clear in writing; the location and backups of workloads containing personal data should be documented under KVKK.
Related Posts
Disaster Recovery Planning: From RPO/RTO to the Test Calendar
Business impact analysis, a three-tier architecture and two tests a year: five steps of a reliable DR plan and the cost…

