Retail Monolith to Microservices
ECS Migration Case Study
How Commerce Architects helped a leading retailer break free from its legacy system using Amazon ECS
The Challenge
As part of its transition from an ATG monolith to microservices architecture, a leading regional retailer wanted to identify opportunities for improving efficiencies and streamlining its business. AWS Advanced Tier Services Partner Commerce Architects helped the company adopt Amazon Elastic Container Services (ECS) to migrate its business functionality, eliminating time spent managing infrastructure.
The Solution
A Shift from Monolith to Microservices
The retailer was executing a long-term strategy to decompose its ATG monolith and shift to a microservices architecture in order to scale rapidly and provide business users an easier way to make needed changes. The company had engaged Commerce Architects (CA) as an embedded team for several years to support this effort. In the initial requirements gathering, one of the opportunities the CA team identified was using Amazon Elastic Container Services (ECS) to provide a secure and easy-to-use solution for managing SEO redirects and metadata workloads in the cloud versus storing them in the monolith.
Empowering Business Users with an Easier Process
Business users of the company’s existing ecommerce platform Oracle Commerce ATG were faced with a time-consuming process for making changes and getting them into production. The ATG platform had a slow iteration speed, and if a user wanted to adjust a metadata tag or make other changes, it required code deployment each time. The retailer wanted to provide an easier, more efficient process for making these changes.
Amazon Elastic Container Services (ECS) Offers a Dynamic Solution
​
Given the company’s requirements for how SEO data was retrieved—performant under high throughput and scalable depending on season traffic—CA created an ECS Fargate cluster of Spring instances to handle the business logic for SEO redirects and metadata and stored the data in DynamoDB with a Redis cache leveraging Amazon ElastiCache for performance optimization.
​
​​​​​The company had a pre-existing GraphQL service that frontend clients interfaced with that needed to be integrated with the SEO service, and an ingress routing application that needed to be aware of redirects. These applications integrated with the SEO service through a REST API.
Using Amazon ECS on AWS Fargate to run containers eliminated the need for the company to manage infrastructure, and decoupled the management of SEO data by business users from its development work. ECS was also a more attractive solution as the retailer wanted to avoid the overhead of hiring a dedicated Kubernetes DevOps team.
The Takeaway
​
Because containerized workloads are easier to scale, manage, and deliver, the decision to use ECS freed the company’s development team to focus on more important tasks. It also increased the efficiency and speed at which business users could set SEO for pages or redirects in the backend, allowing the system to load those requests dynamically. This represents part of a larger strategic shift to microservices that will benefit the company’s agility and business velocity into the future.
Are you planning a move from a monolithic legacy system? Schedule time to discuss our
Monolith Battle Plans