A practical cloud migration readiness checklist

Map the workload and its dependencies

Identify the application components, external services, scheduled tasks, data stores and people who depend on the workload. Include integrations that are easy to miss, such as exports, email delivery or an internal system that connects using a fixed address.

Describe normal and peak operating conditions. A representative dependency map helps the team choose a migration sequence and understand what must be tested together.

• Business owner and technical owner.
• Application, data and integration dependencies.
• Access patterns and expected workload conditions.
• Deployment and environment configuration.

Agree how the environment will be operated

Decide who can make changes, who responds to alerts and who approves access. Document how configuration and secrets are managed. Choose a deployment process that another authorized team member can repeat.

The AWS Well-Architected Framework provides a reference for assessing architecture across operational excellence, security, reliability, performance efficiency, cost optimization and sustainability. Use the relevant questions to review your workload rather than treating a provider choice as an assessment result.

Test data movement and recovery

Agree how data will be copied, checked and reconciled during the transition. Clarify when writes may occur and how the team will avoid losing or duplicating updates. A backup is useful only when the intended recovery procedure can restore the required service.

Document rollback triggers and the point after which the previous environment may no longer be a valid fallback. Rehearse with representative conditions and involve the people responsible for business acceptance.

Make cost and service behaviour visible

Estimate the components that contribute to running the workload, including storage, transfer, managed services and supporting environments. Assign an owner for budget review and investigate changes in usage rather than assuming the initial estimate will remain accurate.

Agree which signals indicate that the service is working for users. Alerts should lead to a defined response. Review the information needed to investigate a failed request without exposing unnecessary sensitive data.

Before the cutover decision

A readiness discussion should end with owners for remaining gaps and a clear basis for proceeding.

Choose a rehearsal that resembles the actual cutover, including integration access, data volume and the people who will support it. Record how long verification and recovery take. If the evidence exposes an unresolved dependency, revisit sequencing before setting a final migration date.

• Dependencies and migration order have been reviewed.
• Access, configuration and secrets have owners.
• Representative application and integration tests pass.
• Data checks and a recovery rehearsal are complete.
• Rollback conditions and responsibilities are documented.
• Operating cost, monitoring and support are understood.

Frequently asked questions