7 Cloud Migration Strategies

Hi inner circle,

Welcome back!

Let’s talk about a topic that's often underrated but can be crucial for your cloud architect interview (and yes, even a junior cloud engineer might be asked about it!).

We're diving into cloud migration.

Migrating to the cloud isn’t a one-size-fits-all problem. Teams that get it right don’t just pick any strategy — they evaluate their workloads, prioritize outcomes, and apply one or more of these 7 proven cloud migration strategies.These strategies help you:

1. Lower costs and reduce technical debt

2. Modernize systems in a phased, risk-aware way

3. Answer tricky interview questions with clarity

This is a bit of a long one, so bear with me! :)

Here’s the breakdown, complete with examples, when to use each, and the trade-offs to keep in mind.

1. Retain (Revisit Later): Keep It Where It Is... For Now

• What: Keep certain applications or parts of your infrastructure on-premises, postponing migration.

• When to use: Strict regulatory compliance, complex legacy dependencies, or apps not yet ready/beneficial for cloud.

• Trade-offs:

  • Pros: No immediate migration cost/effort; maintains existing operations/compliance.

  • Cons: Misses cloud benefits (scalability, cost optimization); continued on-prem burden.

• Interview Questions & Answers (Scenario-based):

  • Scenario: Your company has an aging ERP system on-premises that is heavily customized and integrated with many internal systems. It's stable but costly to maintain. How would you approach its cloud migration?

    • Answer: Recommend Retain due to high customization, critical nature, and cost/disruption of re-platforming.

  • Scenario: You're leading a cloud migration for a healthcare provider. One application stores patient records and is subject to very strict data sovereignty laws. What's your strategy for this app?

    • Answer: Retain is the safest choice for strict data sovereignty. Thoroughly evaluate if cloud providers meet specific legal/compliance needs; otherwise, keep on-premises or use private cloud, revisiting as cloud options evolve.

2. Relocate (Many Providers): Move House, Minimal Changes

• What: Moving workloads to a different cloud provider or a different region/account within the same cloud, often without significant application changes.

• When to use: Consolidating cloud accounts, or migrating between cloud providers for strategic reasons.

• Trade-offs:

  • Pros: Keeps your application mostly intact for quicker shifts between cloud environments.

  • Cons: Doesn't leverage cloud-native services for optimization; can introduce new complexities.

• Interview Questions & Answers (Scenario-based):

  • Scenario: Your company recently acquired a smaller startup that hosts its main application on a different cloud provider than your standard. How would you integrate their application into your infrastructure strategy with minimal disruption?

    • Answer: Assess Relocate to move existing workloads 'as-is' from their cloud to ours. This minimizes re-architecture and disruption, consolidating our cloud footprint. Optimize or replatform components post-migration.

  • Scenario: You're running a global service, and due to geopolitical reasons, you need to quickly shift your entire presence from one cloud region to another in a different country within the same cloud provider. How would you approach this?

    • Answer: Prime for Relocate within the same cloud. Redeploy the existing environment in the new region with minimal changes, leveraging IaC and data replication. This is faster than refactor for urgent geopolitical needs.

3. Retire (Decommission): Cut the Cord

• What: Identifying applications or services that are no longer needed, are redundant, or provide minimal business value, and then completely shutting them down.

• When to use: Functionality is duplicated, no/few users, or high maintenance costs for low business value.

• Trade-offs:

  • Pros: Significant cost reduction; reduces operational complexity and security attack surface.

  • Cons: Requires thorough analysis to ensure no critical dependencies are missed.

• Interview Questions & Answers (Scenario-based):

  • Scenario: During a cloud readiness assessment, you discover an old, internally developed CRM system that hasn't been updated in five years. You also have a new SaaS CRM that the sales team actively uses. What should you do with the old system?

    • Answer: Classic Retire candidate. The old CRM is redundant given the new SaaS system. Confirm all critical data/functionality are moved or unneeded with stakeholders, then decommission to cut costs and overhead.

  • Scenario: Your company has several data warehousing instances, some of which are very old and seem to have no active queries running against them. How would you confirm if they can be shut down?

    • Answer: To confirm Retire, monitor query logs and access for 3-6 months. If no activity, restrict access, then temporarily shut down (e.g., 30 days). If no issues, proceed with full decommissioning to avoid missing dependencies.

4. Repurchase (Drop & Shop): Buy Off-the-Shelf

• What: Switching to a new, ready-made Software-as-a-Service (SaaS) solution instead of migrating an existing application.

• When to use: Existing app is non-strategic or provides common, standardized functionality; need quick, cost-effective setup.

• Trade-offs:

  • Pros: Very fast time-to-market; significantly reduced operational burden; leverages vendor expertise.

  • Cons: Vendor lock-in; limited customization; data migration can be complex.

• Interview Questions & Answers (Scenario-based):

  • Scenario: Your finance department is using an old, on-premises accounting software that requires constant patching and dedicated IT support. They're asking for cloud-based capabilities. What's your recommendation?

    • Answer: Repurchase is most sensible for standard functions like accounting. Evaluate leading SaaS accounting solutions to gain immediate cloud benefits, reduce IT overhead, and achieve faster deployment than custom migration.

  • Scenario: A small e-commerce startup is currently managing its customer support via emails and spreadsheets. They're growing rapidly and need a more robust solution. What cloud migration strategy would you advise for their customer support?

    • Answer: Repurchase a dedicated SaaS customer support platform (e.g., Zendesk) is ideal. It offers immediate features, scalability, and reduces in-house development, letting the startup focus on core business.

5. Refactor (Re-architect): Build it Better for the Cloud

• What: Redesigning and rewriting significant parts of an application to fully leverage cloud-native capabilities.

• When to use: Existing app has significant scalability/performance/resilience issues; to maximize cloud benefits; to enable faster feature development.

• Trade-offs:

  • Pros: Unlocks full cloud potential; significantly improved performance/scalability/resilience; better long-term cost optimization.

  • Cons: Highest effort, cost, and longest migration time; requires skilled developers; significant risk.

• Interview Questions & Answers (Scenario-based):

  • Scenario: You have a monolithic on-premises application that processes millions of transactions daily, but it frequently experiences bottlenecks during peak loads and is difficult to update. How would you approach its cloud migration?

    • Answer: Refactor is appropriate due to scalability issues and update difficulty. Re-architect to microservices on Kubernetes or serverless functions for independent scaling, faster deployments, and improved resilience/agility.

  • Scenario: Your development teams are struggling with long release cycles and dependencies within a large, tightly coupled application. How can cloud migration help, and what strategy would you pick?

    • Answer: Clear indicator for Refactor. Break tightly coupled app into independent microservices, allowing dedicated teams to enable independent development, testing, and deployment, addressing long release cycles and dependencies.

6. Rehost (Lift & Shift): Move As-Is

• What: Moving applications from on-premises to the cloud without any changes to the application code, configuration, or architecture.

• When to use: Need to quickly vacate a data center; for low-complexity apps; as a first step to gain cloud experience.

• Trade-offs:

  • Pros: Quickest and simplest path to the cloud; minimal application disruption; early cloud value.

  • Cons: May not fully leverage cloud benefits; potentially higher costs than optimized solutions; still requires managing underlying OS/application.

• Interview Questions & Answers (Scenario-based):

  • Scenario: Your company needs to shut down its primary data center in six months due to a lease expiring. You have hundreds of applications, and a full refactor isn't feasible for all of them. What's your immediate strategy?

    • Answer: With tight deadlines and many apps, Rehost is primary. Prioritize moving applications (VMs, databases) 'as-is' to cloud IaaS. This 'lift and shift' is fastest for immediate cloud presence; optimize in a second phase.

  • Scenario: A development team wants to move their test and development environments to the cloud to reduce local hardware costs and improve collaboration. The applications in these environments are not highly optimized. Which migration strategy fits best?

    • Answer: Rehost is best for test/dev environments. They don't need peak performance or full cloud-native optimization immediately. Lifting and shifting existing VMs offers quick cloud flexibility, reduced hardware costs, and better collaboration without re-architecture.

7. Replatform (Lift, Tinker, & Shift): Optimize a Little

• What: Moving applications to the cloud while making minor optimizations to take advantage of managed services.

• When to use: To improve performance or reduce operational overhead without a full re-architecture; to leverage managed services.

• Trade-offs:

  • Pros: Better performance/cost than rehost; reduces operational burden using managed services; less effort/risk than full refactor.

  • Cons: Requires some application changes/testing; might not fully optimize for all cloud-native benefits.

• Interview Questions & Answers (Scenario-based):

  • Scenario: Your company's primary customer-facing application relies on an on-premises Oracle database that requires constant patching and backup management by a dedicated DBA. You're moving to the cloud. What's a sensible migration path for this database?

    • Answer: Replatform makes sense for critical databases. Migrate to a cloud-managed DB service to offload patching, backups, replication, and scaling, reducing operational overhead and improving reliability without full re-architecture.

  • Scenario: You have a web application running on VMs on-premises, and you want to reduce the operational burden of managing the web servers (OS updates, patching) while gaining better scalability. The application code is stable. What strategy would you choose?

    • Answer: Ideal for Replatform. Move the web app to a managed application platform (e.g., Elastic Beanstalk, App Engine, App Service) or container service. This allows deploying existing code while the cloud provider handles infrastructure, OS updates, and scaling, reducing operational burden.

That’s a wrap! You survived this one!

See you next Thursday with another topic and more interview prep insights!