AWS
Java
Spring Boot
Kubernetes
PostgreSQL
Kafka
Docker
Terraform
GitHub ActionsBackground
Manual scaling before a peak season
A fast-growing e-commerce startup had to launch a large web platform in about six months, ready for a peak shopping season. It needed personalized recommendations and real-time inventory updates, and it had built its backend as Golang microservices.
As the date approached, the deployment's manual scaling and management became a bottleneck. The team chose AWS Lambda for automatic scaling and to remove always-on infrastructure.
Approach
Five stages
Step 01 of 05
Assessment and planning
Map each service's functionality, dependencies and performance characteristics, and flag migration risks early.
Step 02 of 05
Conversion tooling
Build tools that translate the Golang services into Lambda functions and adapt them to the serverless model.
Step 03 of 05
Performance tuning
Tune the functions for variable, bursty workloads, not only the happy path.
Step 04 of 05
Integration and testing
Integrate with the rest of the platform and test function, performance and scalability under load.
Step 05 of 05
Deployment and monitoring
Deploy to AWS with monitoring and alerting in place.
Modernization
Link to modernization
Assess, convert, test and deploy in stages is the shape of every Ace8 migration. See Modernization assessment, the 8 principles and Application modernization services. The documented project is the American National case study.