The Enterprise Migration Path to Serverless
The client really loved your previous work, especially the serverless part of "pay-per-use" infrastructure costs, easier maintenance, and faster time to market. Now, they would like you to migrate their own monolithic app to serverless.
From a greenfield perspective, building applications with serverless is quite easy, but a bigger concern is how do you approach an existing, brown-field project. As you already know, legacy projects aren't easy to maintain, not to mention migration.
Depending on the application architecture state, the path to serverless can be from being relatively easy to quite cumbersome, painful even. A nicely decoupled microservices application can appear easier to migrate, while a monolithic architecure might be a place where you might break your teeth.
In the case of our client, the application is an on-premise monolithic one, developed for years in COBOL. The client doesn't want to change his engineering team and architects, but to migrate it slowly, without impacting any of his business neither short nor long term.
The application architecture is in the following diagram
/// DIAGRAM
Your tasks in this step by step migration to serverless are:
- Draw a diagram how you would migrate client's API in a step by step approach
- How would you introduce AWS Lambda functions, in a step by step, without changing the database
- How would you migrate the database to DynamoDB
IMPORTANT NOTE: This exercise does not require any coding. Creating a diagram and explaining the architecture is enough. You can try to create a code for this solution as a bonus exercise at the end of the workshop or home.
Hints
...