Data integration provider Hasura has added a visual component to its data API platform, offering developers a handy visualization of the complex topologies that may be wrangling in their integration efforts.
When fully released by the end of this month, Hasura v3 comes with a redesigned console featuring a graphical “supergraph” viewer, displaying, in a spider web-esque graph format, the tables of all the sources of data.
Using a graph format, the supergraph shows all the data sources available from data stores and APIs, along with how they are connected.
“Teams are spending more time finding and understanding data, composing APIs, and building bespoke integration layers than shipping new features,” said Torsten Volk, managing research director for Enterprise Management Associates, in a statement. “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.”
Today, developers are wrestling with aggregating multiple sources of data, including data from databases enterprise resource planning tools, Software as a Service offerings and other sources. They need to aggregate and filter the results for their particular use.
Say you want to build a sales intelligence application that summarizes the top billing issues faced by enterprise customers in the U.S. within the first 30 days of their onboarding: You need billing issue data, that then needs to be filtered by a U.S. customers list, then filtered again within the 30-day window.
“It is not just joining data from multiple places,” noted Tanmai Gopal, CEO and co-founder of Hasura, in an interview with TNS. Logic needs to be added into the set of queries itself.
The recent addition of LLM vector data for Generative AI features only compounds this cognitive complexity of understanding how it all fits together,
Hasura engine provides a single “federated” API integration point, via GraphQL, for multiple databases and data stores. The Supergraph visualizes how multiple sources of data relate to one another.
So for instance, vector search can be combined with structured search into a single query.
The open source Hasura GraphQL Engine, and the commercially-available the Hasura Data Delivery Network (DDN), provide a workspace to build queries to executethese complex, composite queries.
Hasura can derive data types from databases and data sources, so users don’t have to do that work manually. But before v3 of Hasura, users only had text summaries of tables and other data sources of the federated data set to work with.
“The v2 architecture was a platform that could be used by a single team,” Gopal said. The compiler was monolithic. The new distributed diversion can federate all the different sources across an enterprise, including not only databases but API-based sources such as TypeScript or Java functions.
This allows multiple teams to publish to a single supergraph independently.
How the Supergraph Runs
The supergraph was made possible through a number of innovations in version 3 of the Hasura engine, including a new engine architecture and improvements in the Hasura connector design.
Users can define a supergraph with a new modeling framework, by specifying how the entities and domains are related — No need to spin up a new GraphQL server or a service. It is data agnostic, and users can swap in and out resources without restarting.
A serverless runtime engine, written in the Rust programming language, runs the supergraph. The company said the engine offers sub-millisecond start times and can scale with usage spikes.
Queries are translated for the source database through connectors, all of which are available as open source through the Hasura Connector Hub. Those ready to connect include:
- AlloyDB
- AWS Aurora PostgreSQL
- AzureDB
- Citus
- CockroachDB
- CosmosDB PostgreSQL
- Google Cloud SQL
- Neon
- PostgreSQL
- Timescale DB
- Yugabyte
Custom connectors can also be built via a Software Development Kit offered by Hasura.
The build system for the Supergraph offers immutable versioning, with versions instantly testable by way of a preview API.
Users can change the graph by specifying relations, permissions and data sources. Each change is validated through checks for potential bugs, stylistic issues, expected contracts, syntax, and type errors.
Pricing for version 3 of the commercial service has been changed as well. It went from usage-based pricing to pricing based on the number of data entities in your domain. This approach will give users far more predictability in their monthly bills, Gopal said.
Today, about 44% of companies in the Fortune 100 are users of Hasura, the company asserts.
The post Hasura Visualizes Data API Integration into a ‘Supergraph’ appeared first on The New Stack.
Data integration provider Hasura has added a visual component to its data integration API platform, offering developers a handy visualization of the complex topologies that they wrangle with.