Azure AKS to AWS EKS Migration

A cloud migration case study that moved containerised workloads from Azure to AWS with zero downtime, driven by a post-merger consolidation requirement.

Traffic cutover

< 5 min

Operational overhead

−18%

Downtime

Zero

Problem

A merger required consolidating workloads from Azure into an existing AWS environment. The platform had been running on AKS, but the acquiring organisation standardised on EKS, creating a mandate to migrate without disrupting live services.

The risk was straightforward but the stakes were high: any uncontrolled downtime during cutover would affect production traffic, and identity and access patterns needed to be re-established across the new cloud boundary without leaving governance gaps.

Approach

Migrated container images from ACR to ECR and rebuilt the target infrastructure from scratch using Terraform, ensuring the AWS environment matched the operational baseline rather than inheriting legacy configuration assumptions from Azure.

Workloads were deployed to EKS using Helm charts, preserving release structure and making the new environment verifiable before any traffic was moved. A blue-green cutover strategy was used to shift traffic in a controlled window, keeping the AKS environment live as a fallback until the migration was confirmed stable.

Identity and access were rebuilt using AWS-native controls, treating the migration as an opportunity to tighten governance rather than simply replicate what had existed in Azure.

Tools

AWS, EKS, Amazon ECR, Terraform, Helm, GitHub Actions

Outcome

Production traffic shifted within a controlled five-minute window with no unplanned downtime.

Identity governance improved by rebuilding access controls natively in AWS rather than lifting and shifting Azure policies.

Operational overhead reduced by 18 percent through a cleaner, standardised EKS setup aligned to the organisation’s existing AWS practices.