Follow BigDATAwire:

April 7, 2021

RapidAPI Adds Kafka Support, Centralized Interfaces

The adoption by Apache Kafka users of asynchronous APIs that return data depending on the availability of microservices and other resources is said to be on the rise. That trend has prompted API vendors to offer platforms that facilitate discovery of Kafka clusters as developers build event-driven architectures.

Among them is RapidAPI, which this week announced the beta launch of a browser-based API platform for finding Kafka instances and topics, or categories, along with viewing topic schemas and configurations. The integration with the distributed streaming platform also would allow developers to connect to microservices, REST and other APIs.

Demand for asynchronous APIs also reflects greater developer participation in a booming “API Economy,” with more than 60 percent of survey respondents telling RapidAPI they used more APIs last year. Seventy-one percent of those polled said they expected to use more app interfaces in 2021.

“APIs are getting siloed” and the growing ecosystem is becoming a “Wild West,” said Iddo Gino, RapidAPI’s founder and CEO.

“We’ve seen the adoption of these asynchronous types of APIs and flavors of APIs rising, and what we’re bringing is the ability to surface those APIs” and related documentation, he added. The platform helps “developers discover and connect those asynchronous APIs more easily,” Gino told Datanami.

Source: RapidAPI

The other factor is growing enterprise adoption of Kafka, which “is increasingly used as the message broker in event-driven architectures with asynchronous microservices,” Srivatsan Srinivasan, RapidAPI’s vice president for product, noted in a blog post announcing support for Kafka discovery and testing.

Being browser-based, the API platform eliminates the need to write code to determine API functions. Srinivasan said developers can simply assess the schema and interface definition, then test the send/receive function.

RapidAPI CEO Into Gino

“We see Kafka services as another important ‘API type’ for development teams,” Srinivasanadded. The message broker also serves as the communications interface for emerging asynchronous microservices architectures.

According to a RapidAPI survey of about 1,500 developers with varying degrees of coding experience, the number of AsynchAPIs in production tripled last year to 19 percent of deployments.

With enterprise adoption growing, Kafka clients are being used to develop distributed applications and microservices that process streams of events. Srinivasan added that rapid adoption has created a new set of challenges, including how to determine availability of Kafka clusters and what defined topics they contain.

To address the disconnect, RapidAI said Wednesday (April 7) it is adding support for Kafka as another API type along with REST, SOAP and GraphQL APIs on its enterprise hub and marketplace.

The AsynchAPI platform seeks to boost support for both Kafka providers and consumers alike. For production, the tool makes it easier to expose Kafka topics for consumption. Easier access to Kafka topics is promoted as streamlining design services, allowing developers to stream data faster.

AsynchAPIs and GraphQL saw the biggest increases in adoption and use in production in 2020, according to the vendor survey. RapidAPI added support for GraphQL last year.

The addition of Kafka and AsynchAPI support “is part of a mission of really being able to have every type of API…living in one centralize place,” said Gino.

The combination of the Kafka message broker and asynchronous APIs allows users to, for example, add a task to a data request queue, then be notified when a computing-intensive job like video encoding is completed.

In February, the San Francisco-based company acquired API design and collaboration tool vendor Paw. The deal aims to extend RapidAPI’s open platform across the app interface development lifecycle.

Recent items:

Confluent Moves to Boost Kafka Reliability

Step One in Kafka’s Metamorphosis Revealed

Will GraphQL Become a Standard for the New Data Economy?

BigDATAwire