Training project for learning how to build a serverless application in AWS using some of the most trending technologies.
The following diagram exemplifies the complete flow, described below.
stateDiagram-v2 state "S3 Bucket Raw" as S3A state "S3 Bucket Input" as S3B state "S3 Bucket Failures" as S3C state "Lambda A" as LambdaA state "Lambda B" as LambdaB [] --> S3A S3A --> [] S3A --> LambdaA LambdaA --> S3B LambdaA --> S3C S3B --> LambdaB LambdaB --> DynamoDB DynamoDB --> [*]
The idea is that:
New JSON files, compressed with tar.gz are stored on S3 Bucket Raw
There is a Lambda A, that picks up those files, decompresses them, and splits the objects into smaller JSONs, storing them on S3 Bucket Input. Invalid schemas are sent to S3 Bucket Failures.
There is another Lambda B, that picks up files from S3 Bucket Input, transforms the objects to DynamoDB records and writes them into the database.
## Requisites
Lambdas must be written in Scala using the Feral framework (https://github.com/typelevel/feral)
All the infrastructure needs to be written in CDK using Typescript
Ability to test it locally with Docker (using CDK in this case will be a nice to have)