GraphQL Live Querying with DynamoDB

Abstract: We present a method of implementing GraphQL live queries at the database level. Our DynamoDB simulation in Go mimics a distributed key-value store and implements live queries to expose possible pitfalls. Two key components for implementing live queries are storing fields selected in a live query and determining which object fields have been updated in each database write. A stream(key, fields) request to the system contains fields to include in the live query stream and on subsequent put(key, object) operations, the database asynchronously determines which fields were updated and pushes a new query view to the stream if those fields overlap with the stream() request. Following a discussion of our implementation, we explore motivations for using live queries such as simplifying software communication, minimizing data transfer, and enabling real-time data and describe an architecture for building software with GraphQL and live queries.