News

Versant Releases Java-Based API for Big Data

• By John K. Waters
• October 24, 2012

Data management software provider Versant Corporation has announced the general availability of a Java-based API that supports high-performance database and analytics operations aimed at Big Data. The Versant JPA, launched last week, features "multiple firsts for NoSQL Big Data technologies," said Robert Greene, Versant's vice president of technology, including a new "application evolution" feature designed to ensures zero app downtime when handling changing data models.

In a nutshell, this release of the Versant JPA aims to reduce the cost and risks of migration to other proprietary NoSQL technologies by being fully compliant with the latest JPA specs. The JPA (Java Persistence API) was an outgrowth of JSR 220 (Enterprise JavaBeans 3.0). It's designed to conquer the process of writing code that manipulates data, and to simplify the programming model for entity persistence. Java developers have traditionally had to write complex SQL queries at the persistence layer.

"This release is about us trying to help companies evaluate this emerging set of technologies -- call it collectively Big Data -- and address issues around scalability and performance under large data volumes," Greene told ADTmag.

The latest version of the Versant JPA comprises of the Versant JPA server and the Versant JPA SDK. Together, they allow Java developers to use their existing skillsets to dive into the Big Data ocean without drowning, the company says.

"We're taking NoSQL technologies, putting standards on top of them that are already being used within the enterprise, and giving [companies] a low-risk way to kick the tires," Greene said.

The Versant JPA server expands markedly on the original Java Persistence API's capabilities with the new Application Evolution feature. Greene called it a "first-of-its-kind technology" designed to allow enterprises to change application data models without downtime and restarts.

"Fifteen years ago, we didn't have a standards-based interface," Greene said. "We felt the impact of that lack of standardization back then -- it seemed to feel risky to a lot of companies -- and we learned from it. Now we're also taking this standards-based JPA, adding extensions to it, and integrating with other technologies that we think are necessary for Big Data analytics."

The company also announced that the Versant JPA is now compliant with the JBoss Application Server. This summer, the company integrated the Versant JPA with the Spring framework. It is also reportedly working on a connector with Talend's Unified Integration Platform, which addresses the converging data and application integration markets.

Many companies continue to accept application downtime as an unavoidable side effect of migrating to Big Data technologies or updating applications, says IDC analyst Carl Olofson. And they "tend to equate Big Data with Big Money." The solution, Olofson offered in a statement: standards-based, open source NoSQL technologies. They're rare, he said, "but they are hugely important for combatting both of these challenges, and for allowing enterprises to handle Big Data's inherently uncertain future."

The Versant JPA server and SDK are available now for download from the company's Web site. The download includes the Versant's Hadoop Connector and an integration with the popular R analytics framework.