Case Study: Modernizing a Legacy .NET App to Cloud-Native on AWS

“

Meerako — Dallas, TX experts specializing in complex legacy system modernization and AWS cloud migration.

A large Dallas-based manufacturing company relied on a mission-critical, internally-developed application built over a decade ago on an older Microsoft .NET Framework version, running on aging Windows servers in their on-premise data center.

The application worked, in the sense that it hadn't been replaced yet — but it was becoming a serious liability:

• Slow and unreliable, with frequent crashes and degraded performance during peak production hours.
• Difficult to update. Deploying new features took weeks and routinely broke unrelated functionality.
• Expensive to maintain, between aging server hardware, Windows licensing, and the shrinking pool of engineers still comfortable in the old codebase.
• A genuine security risk, running on an unsupported OS version with no realistic patching path.

They needed to modernize, but a complete rewrite was judged too risky and too expensive to attempt in one leap. They needed a phased approach — move to the cloud first, then gradually refactor without ever betting the whole business on a single "big bang" cutover. They chose Meerako, leveraging our experience in both legacy systems and modern AWS cloud-native architecture.

• The specific challenges of modernizing a legacy .NET monolith without a full rewrite.
• Meerako's phased migration strategy: rehost, replatform, refactor.
• Which AWS services are purpose-built for .NET workloads, and why we chose them.
• How we kept the business running throughout, with zero unplanned downtime.
• The measurable benefits: cost savings, agility, and reliability.

• Technology: .NET Framework 4.5, ASP.NET Web Forms, WCF services, SQL Server 2008.
• Infrastructure: physical Windows Server 2008 R2 machines, past end-of-support.
• Pain points: tightly coupled code with no clear module boundaries, no automated test suite to catch regressions, an entirely manual deployment process, and known, unpatched security vulnerabilities.

It's tempting to treat a legacy system like this as "just rewrite it in something modern." We advised against that as a first step, for a concrete reason: a decade-old, undocumented monolith encodes years of business logic and edge-case handling that nobody has fully mapped out anywhere except the running code itself. Attempting a ground-up rewrite risks silently dropping business-critical behavior that only surfaces as a production incident months later. A phased approach lets you de-risk the move to the cloud first, then refactor piece by piece with each change validated against real production behavior.

We used a pragmatic, multi-phase approach based on the standard cloud migration "R" strategies: rehost, replatform, refactor.

Phase 1: Rehost ("Lift and Shift" to EC2)

• Goal: get the application off aging on-premise hardware quickly and onto stable AWS infrastructure, without touching the application code yet.
• Action: we used AWS Application Migration Service (MGN) to replicate the existing Windows Server VMs directly into Amazon EC2 instances inside a secure VPC. The SQL Server database was similarly lifted and shifted onto an EC2 instance initially, with no schema changes.
• Result: the application was running on reliable AWS infrastructure within weeks, with an immediate drop in hardware maintenance costs — and, critically, this bought time to plan the next phase properly instead of rushing it.

Phase 2: Replatform (Move to AWS Managed Services)

• Goal: reduce operational overhead and improve scalability by shifting key components onto AWS managed services.
• Action:

  • Database: migrated SQL Server from EC2 to Amazon RDS for SQL Server, offloading patching, backups, and scaling to AWS instead of a manual DBA process.
  • Modernizing the runtime: we worked with the client's team to upgrade parts of the application from the old .NET Framework to modern, cross-platform .NET 8, and containerized those components with Docker.
  • Orchestration: deployed the containerized .NET services onto AWS ECS with Fargate (serverless containers), eliminating the need to manage EC2 instances for the application tier at all.
• Result: a significant drop in server management overhead, more reliable database operations, and basic auto-scaling for the application tier — all without a full rewrite.

Phase 3: Refactor and Re-architect (Ongoing)

• Goal: gradually break the monolith into smaller, independent services to improve agility and unlock genuine cloud-native benefits, rather than just "a monolith running in the cloud."
• Action (ongoing): we identify specific business domains within the application — inventory management, order processing — and refactor them individually into:

  • Microservices: independent .NET services running on ECS/Fargate, with Kubernetes on EKS as an option for domains that need it.
  • Serverless functions: event-driven logic migrated to AWS Lambda using the .NET runtime.
  • API Gateway: introduced AWS API Gateway as a stable front door, decoupling client applications from the backend as it continues to evolve underneath.
• Result: increased deployment frequency, better fault isolation (a bug in one service no longer risks the whole application), the ability to scale individual components independently, and adoption of modern CI/CD practices.

Through this phased approach, the Dallas manufacturing company achieved:

• A ~40% reduction in total infrastructure and maintenance costs.
• Improved application uptime, from roughly 98% to over 99.95%.
• Faster feature deployment, with release cycles shrinking from quarterly to bi-weekly sprints.
• Enhanced security posture, running on supported OS versions with continuous cloud security posture management.
• A real foundation for innovation, including early integration of AI/ML for predictive maintenance — something that simply wasn't feasible on the old on-premise architecture.

Modernizing legacy applications is a journey, not a single event. A phased approach — stabilize on the cloud first, then refactor gradually toward cloud-native — is consistently the most pragmatic and least risky path for a business-critical system nobody can afford to take offline.

Meerako's expertise in both legacy systems like .NET and modern AWS cloud-native patterns lets us guide enterprises like this Dallas manufacturer through a genuinely successful transformation, unlocking agility, reliability, and real cost savings along the way.

Is your legacy application holding your business back? Let's chart your path to the cloud.