A Feasible Schema Design Strategy for Amazon DynamoDB: A Nested Normal Form Approach

Amazon DynamoDB is a next-gen NoSQL (Not only SQL), key-value and document database that delivers single-digit millisecond performance at any scale. As more complex web-based applications adopt DynamoDB, it is imperative to develop sound design strategies for DynamoDB schemas. Large-scale web-based applications often exhibit conflicting requirements. Hence, the research problem is to design the DynamoDB schemas of a given application with conflicting requirements such that the database must satisfy certain predetermined performance goals. In the era of relational databases, normal forms were developed to guide the schema design process. Past knowledge and experiences, we believe, are also applicable to DynamoDB schema design. Specifically, based upon Nested Normal Form and XML databases, this paper demonstrates the feasibility of a design strategy on DynamoDB schemas, particularly with the access patterns of the data in mind. Simulation further substantiates the feasibility of our approach.