Smile news

Cloud migration to AWS: steps, tools and best practices

  • Date de l’événement Sep. 10 2026
  • Temps de lecture min.

Audit, 6Rs, Terraform, FinOps, sovereignty: plan your cloud migration to AWS methodically. Experience feedback and recognized expertise from Smile.

Most organizations still have on-premises IT infrastructure. Transforming IT infrastructure to the cloud has become a strategic priority. But a poorly planned migration is costly: it generates service interruptions and creates technical debt.

A successful cloud migration requires preparation, is broken down into steps, and is managed methodically.

This guide provides you with the strategies, steps, and tools to migrate to AWS without any unpleasant surprises.

  • 52.7% of EU companies use at least one paid cloud service (Eurostat, 2025)
  • Organizations migrating to AWS achieve an average of 20% savings on their infrastructure costs, with the best performers achieving up to 47% (ESG/TechTarget study commissioned by AWS)

Why migrate to AWS?

AWS (Amazon Web Services) is one of the world's most widely used cloud platforms with over 200 managed services. Compute, storage, databases, AI, security: everything is available on demand.

 

The tangible benefits

Three main reasons drive organizations to migrate:

  • Scalability: you can add or remove computing resources in minutes. No need to buy hardware in advance.
  • Agility: teams deploy faster. New features reach production more quickly.
  • Cost reduction: you only pay for what you use. Fixed costs for data centers decrease.

The 6 cloud migration strategies (the 6Rs)

AWS has formalized six migration strategies. They are called the 6Rs. Each workload (application workload) deserves to be evaluated individually.

 

Rehost: Move without modifying

The application is moved to the cloud as is, without modification. This is the fastest strategy: it is ideal for stable applications that need to migrate quickly.

 

Replatform: Optimizing at the Margins

We migrate the application with a few minor optimizations. For example: moving from a self-hosted database to Amazon RDS (RDS is a managed database service for which AWS handles maintenance on your behalf).

 

Repurchase: replace with a SaaS solution

The application is replaced with a SaaS solution, for example, abandoning an on-premise CRM for Salesforce. It's a clean break but often an effective one.

 

Refactor: rewriting for cloud native

We rewrite the application to fully leverage native cloud services: lambda, containers, microservices. This is more expensive in the short term but the most optimized in the long term.

 

Remove: Turn off what is no longer needed

Unused applications are identified and then shut down. Between 10 and 20% of the application portfolio is often deprecated during a thorough audit.

 

Retain: keep what needs to stay on-premises

Some applications remain on-premises for reasons of compliance, latency, or technical complexity. We don't migrate everything: it's a conscious decision.

The steps to a successful AWS migration

Step 1: Audit and Discovery

We begin by mapping the existing infrastructure. What applications are there? What dependencies are there? What are the data volumes? AWS Migration Hub centralizes this discovery. We also identify the applications that are candidates for each 6R strategy and create a map of the existing resource management.

 

Step 2: Planning and Prioritization

Applications are prioritized. The least critical applications migrate first. The migration backlog is built in waves, and the deployment of the target environments is planned.

 

Step 3: Proof of Concept (PoC)

We migrate an initial application under real-world conditions, which allows us to validate the tools, the migration process, and the required skills. We measure the actual costs, then adjust the plan before scaling up.

 

Step 4: Wave Migration

Applications are migrated in coherent groups. Each wave learns from the previous one and refines the migration process. Teams gradually develop their skills. Cloud migration progresses in a controlled manner.

 

Step 5: Post-migration optimization

Migration is not the end goal: costs are analyzed, resources are adjusted, and new, more efficient cloud-native technologies and services are adopted. This is the FinOps phase: a discipline that continuously optimizes cloud spending.

Key AWS services for migration

AWS Migration Hub

Migration Hub centralizes the tracking of all ongoing migrations. It provides a consolidated view of progress and integrates with AWS discovery and migration tools.

 

AWS Database Migration Service (DMS)

DMS migrates databases to AWS with minimal disruption. It supports both homogeneous (Oracle to Oracle) and heterogeneous (Oracle to PostgreSQL) migrations. It is the leading tool for data migration.

 

EC2, S3, RDS, Lambda: the fundamentals

Four services form the basis of most AWS cloud infrastructures.

  • EC2: Virtual machines. The cloud equivalent of a physical server
  • S3: Object storage. Ideal for files, backups, and static assets.
  • RDS: Managed relational databases (PostgreSQL, MySQL, MariaDB)
  • Lambda: serverless code execution. You only pay for the execution time.

Infrastructure as Code on AWS

Terraform and AWS CloudFormation

Infrastructure as Code (IaC) is the practice of describing infrastructure using versioned code. On AWS, two tools are considered the gold standard.

Terraform is the most widely used multi-cloud tool. It works with AWS, Microsoft Azure, and Google Cloud. Its declarative language is readable, and its community is huge.

AWS CloudFormation is the native AWS tool. It integrates seamlessly with all AWS services and is recommended for 100% AWS organizations.

 

IaC best practices on AWS

  • version all infrastructure code in Git.
  • use reusable modules,
  • Never modify the infrastructure manually.
  • Any modification is made via an updated file in Git.
  • test the changes in a non-production environment before applying them in production.

Security, cost and sovereignty

IAM: Access Management

Identity and Access Management (IAM) manages who can do what on AWS. It's the first pillar of security. We apply the principle of least privilege: each user or service only has the strictly necessary rights.

 

FinOps: Mastering AWS Costs

Cloud computing can be expensive if not properly managed. A FinOps approach aligns technical, financial, and business teams around cloud costs. AWS Cost Explorer allows for expense analysis. Budget alerts are configured, and unused resources are identified.

 

Cloud Act and GDPR: the risks

AWS is an American company. Therefore, the US Cloud Act may require AWS to hand over sensitive data to US authorities, even if the data is hosted in Europe. This risk must be assessed in your GDPR impact assessment for sensitive data.

 

Sovereign alternatives

For organizations with strict sovereignty requirements, two options exist. Bleu is a sovereign hybrid cloud solution based on Microsoft Azure, which qualifies as SecNumCloud. For AWS, there is currently no SecNumCloud-qualified sovereign offering available in France. A cloud engineer or specialized cloud administrator can help you evaluate the alternatives best suited to your specific situation.

Smile and AWS: our expertise

At Smile, we have been supporting cloud migrations to AWS for many years. Our expertise covers application portfolio auditing, cloud infrastructure design, migration strategy definition (6Rs), deployment with Terraform and CloudFormation, and post-migration optimization with a FinOps approach.

We support both initial migrations and modernization projects for architectures already in the cloud.

Our approach is pragmatic. It always begins with an audit, followed by a recommendation for a strategy tailored to each workload. Migration is never an end in itself.

Looking to plan your cloud migration to AWS? Discover Smile's AWS expertise

Hugues_Marilleau

Hugues Marilleau

Head of Cloud Transformation & Management