Fauna has been building out its serverless database technology since 2015 as a developer-focused platform for helping organizations build a range of data-driven applications, including e-commerce, web development and gaming.
Fauna is a NoSQL database platform, and it has developed its own Fauna Query Language (FQL) for data queries. On Nov. 17 the vendor made generally available an intelligent routing capability to let organizations better manage and access Fauna database deployments around the world. The new capability builds on other upgrades the vendor has released in 2022, including the debut of real-time streaming support in March.
Fauna competes against several developer-focused databases, including Planetscale and CockroachDB, both of which offer serverless platforms.
Fauna has been led since 2020 by CEO Eric Berg, who has positioned the vendor's technology with a database-as-an-API approach. In this Q&A, Berg outlines the direction the serverless database platform is headed as it brings relational capabilities to its NoSQL platform.
What is the document relational model at the heart of the Fauna serverless database?
Eric Berg: Fauna is a different kind of database architecture; we call it document relational. It fundamentally is a document-based NoSQL store. But we bring to that the power of relationships and foreign keys to be able to create relationship across those documents. We also have a query language that looks a lot more like what you can do in a traditional relational database.
Fauna is also distributed by default. We use a Calvin-based read ahead log to be able to distribute and replicate data that provides responsiveness and bring data closer to your user, It gets you reliability, because you span across regions.
What does the term serverless database actually mean, and how is it different for Fauna?
Berg: People will call all kinds of things serverless. To some, it's just a pricing model where users can consume things on demand and pay as they go. Everything behind the scenes, however, is still physical, with vendors spinning up Kubernetes container clusters and virtual machines.
Eric BergCEO, Fauna
We think of serverless more like API service platforms like Stripe or Twilio, but for an operational database. There's no picking a machine or memory and having to know about the physicality of the hardware. You sign up, pick a region group, create a database, and then you just start going. It's an API delivery model.
Databases have gone from on-premises deployments to cloud. And what we're doing is database as an API. So it's a next level of abstraction.
Why not provide a SQL-based approach for Fauna?
Berg: Our query language is not SQL. We effectively have a structural programming language with FQL, which is a query language that you use to represent business logic with user-defined functions that run in the context of data in Fauna.
We're coming in with a developer-friendly approach with a modern language that we think is well suited for semi-structured, interactive applications. Our bet is that you can do better than SQL when it comes to developers today using semi-structured data for interactive web applications.
How do you solve challenges with SQL compatibility?
Berg: There are several ways to deal with the challenge of SQL compatibility. There are use cases where organizations need to get data out of Fauna to work on it from an analytical standpoint.
We recently released an Airbyte connector, in preview, which helps. Airbyte is a popular open source ETL [extract, transform, load] tool. So that's one way to do that if you wanted to get data out of Fauna and into a data warehouse like Amazon Redshift or Snowflake to work with SQL and data analytics use cases.
Over time, we'll also have that capability natively in Fauna if you need to be able to do some basic administration with business intelligence and SQL tools directly on the data. That's not something that we offer today. But I see that in the fullness of time.
Editor's note: This interview has been edited for clarity and conciseness