At Cloudsoft, we started to build a project on our familiar tech stack, but to overcome some known issues with the stack we decided to migrate it to Spring. The new tech stack is made of mostly AWS services and the project is designed to improve certain aspects of the experience for our customers using AWS services.
I can’t tell you as much detail as I’d like about it as it’s currently in Alpha status and due for public release later this year. The final product is going to be available to clients via subscription, and takes a novel approach to solving a known AWS challenge. Thus, I cannot elaborate on the nature of the service provided, but I can share with you the following details:
- the architecture
- the migration details and reasons behind it
- the challenges
- the conclusions and solutions
The core of the migration was introducing Spring. The most difficult part was using Spring to interface DynamoDB, not because of Spring, but because of the AWS DynamoDB Java SDK and the very cumbersome way to write filter expressions for querying and scanning.
The presentation shows the table design, a moderately complex repository method to extract some data and as a bonus, how to reuse Spring code from the main project in a Lambda Step function.