Transactions and Serverless are Made for Each Other
If serverless platforms could wrap functions in database transactions, they would be a good fit for database-backed applications.
Serverless cloud offerings are becoming increasingly popular for stateless applications because they simplify cloud deployment. This article argues that if serverless platforms could wrap functions in database transactions, they would also be a good fit for database-backed applications. There are two unique benefits of such a transactional serverless platform: time-travel debugging of past events and reliable program execution with “exactly-once” semantics.
Serverless cloud platforms such as Amazon Web Services (AWS) Lambda and Azure Functions are increasingly popular for building production applications as varied as website front ends, machine-learning (ML) pipelines, and image-processing systems. These platforms radically simplify development by managing application deployment. Developers can deploy functions with the click of a button and the platform automatically hosts them, guarantees their availability, and scales them to handle changing loads.
Serverless platforms are primarily used for stateless operations such as image resizing or video processing. Here, we will argue they should also be used to deploy stateful applications, particularly database-backed applications whose business logic frequently queries and updates a transactional database such as Postgres or MySQL. Database-backed applications are ubiquitous in modern businesses; examples include e-commerce Web services, banking systems, and online reservation systems. They run primarily on server-based platforms such as Kubernetes. Thus, they form a massive opportunity for serverless offerings, including the back ends of most enterprise APIs and much of the modern Web.
