Hasura has announced the third version (aka V3) of its GraphQL platform. Key developments in Hasura V3 are promised to enable a ‘modern approach’ to data and API integration, a claim meant to be validated by the ability to allow API consumers to flexibly compose data on demand – across any number of domains – with predictable performance and reliability at any scale, speeding time to market on new ideas.
As data engineering fans will know, GraphQL is a query language for Application Programming Interfaces (APIs). It is also a runtime for fulfilling API queries with an organization’s existing data. GraphQL.org reminds us that this technology provides a ‘complete and understandable description’ of the data in an API to give software developers the power to ask for exactly what they need of them.
Supergraph modeling framework
Describing itself as a data API and GraphQL company, Hasura’s latest release promises to deliver on-demand API composability with a new domain-centric supergraph modeling framework, a distributed supergraph execution engine and an extensible ecosystem of open source connectors to address the challenges faced during integration of data and APIs.
As a term on its own, a supergraph modeling framework can be explained by noting that a supergraph is an architecture and operating model to build and scale multiple data domains as a single graph of composable entities and operations.
The new platform also features a new build system with API previews and rollback features. There is also a redesigned console featuring a graphical supergraph viewer, API analytics, observability and auto-generated API documentation. With data and service sprawl happening in so many enterprise IT environments, fast, secure and reliable API access to data is a common bottleneck in modern application development.
Hasura says its supergraph takes this challenge head-on.
“While the data layer that feeds modern applications has grown increasingly complex, API strategies needed to deliver data to the applications have not evolved to keep up,” said Torsten Volk, managing research director, Enterprise Management Associates (EMA). “As a result, teams are spending more time finding and understanding data, composing APIs, and building bespoke integration layers than shipping new features.”
A powerful paradigm, possibly
Volk suggests that the supergraph is a ‘powerful paradigm’ that has the potential to solve these challenges by providing a unified semantic graph that connects all data – no matter the source – and making it available via one composable endpoint.
“Platforms like Hasura that not only make it easier for organisations to create and manage supergraphs, but layer new functionality that makes it easier to visualise and utilise data on top of that, are providing strong value at every level of the org chart today.”
The Hasura team point to the fact that applications today are powered by a complex data layer composed of multiple databases, services and APIs, and new feature development is often bottlenecked while teams work to determine how to aggregate and deliver necessary data from this complex data backend to the frontend/client. The Hasura Data Delivery Network (DDN) is claimed to be capable of ‘completely dissolving this pain’ by specifying how data domains are connected, meaning that users can create a performant API that handles any complex, composite query.
“We built the Hasura Data Delivery Network to answer a couple of hard questions,” said Tanmai Gopal, co-founder and CEO of Hasura. ‘‘One, ‘What if you could fetch just the data you need from any number of enterprise sources and deliver it to any client or user worldwide quickly, securely, and reliably?’ Two, ‘What if you had rich self-serve composability via a unified API?’ Last, ‘What if the effort of building and operating such a delivery network for data was trivial?’ Hasura V3 answers these questions and gives teams the power of a supergraph to streamline and simplify the dev and ops workflows that go into serving data from any source to the end user via fast, secure, flexible, and reliable APIs. In today’s world of accelerating data and backend sprawl, the supergraph is a must-have.”
Declaring domains
Developments in Hasura V3 include self-service composability i.e complex data aggregations, like nested joins and nested filters are hard to do and make consuming data from multiple sources onerous. Hasura-powered supergraphs delivers composability in an API through the act of declaring how domains are related. This is made possible by the Hasura connector design and a new engine architecture.
This supergraph modeling framework is a way to create and evolve a supergraph. Defining a supergraph is as easy as specifying how the entities and domains are related. It eliminates the need to create GraphQL servers or services to enable a supergraph. The modeling framework is data agnostic, allowing users to easily onboard new domains or swap out underlying data sources (e.g., PostgreSQL → ClickHouse) without breaking the supergraph.
The Hasura supergraph modeling framework is backed by a rearchitected serverless runtime engine (written in RUST) that is built to operate the supergraph at scale with reliability. With a query compiler and router, a distributed edge network and edge routing and edge caching (coming soon), users get global low latency and high availability in their APIs without any extra operational overhead.
Every iteration of the Hasura supergraph (changes to relations, permissions, data sources, etc.) is instantly available as an immutable build and testable via a preview API. Any build can be promoted to production (or rolled back) with zero downtime.
What does Hasura mean?
Asure – The Sanskrit word for demon, so a reference to ‘daemons’ in computer programming.
H – From Haskell, the language the company’s GraphQL to SQL compiler is written in.