Follow Datanami:
November 19, 2014

Hadoop RDBMS Ready for Primetime

Targeting database architects and application developers, Splice Machine Inc. announced general availability this week of its Hadoop database that had been in beta testing since May. The Hadoop relational database management system is designed to reduce the costs of developing real-time applications by, for example, eliminating the need for expensive proprietary hardware.

The relational database for operational applications is being positioned as a replacement for Oracle and MySQL databases that might be running out of steam or are becoming too expensive for real-time application development.

San Francisco-based Splice Machine said Nov. 19 more than 20 charter customers helped wring out the Hadoop RDBMS, including performance and functionality testing. In response, the vendor said it incorporated authentication and authorization features along with analytic capabilities such as window functions.

The analytical window functions would help provide advanced SQL analytic capabilities based on the SQL-2003 standard, the company said. The authentication and authorization features are intended to deliver Lightweight Directory Access Protocol integration along with column-level security.

Other database enhancements added to the 1.0 version of the Hadoop RDBMS include: native backup and recovery to prevent data loss; integration with the Hadoop ecosystem to help streamline use with MapReduce; and a native ODBC (Open Database Connectivity) driver.

Splice Machine said the driver provides a custom database connector to improve performance and interoperability.

The resulting real-time relational database is said to combine Hadoop capabilities, particularly the ability to scale, with SQL and ACID capabilities required in an operational database.

Splice Machine partner Cloudera predicted in a statement that version 1.0 of the Hadoop RDBMS would help expand the Hadoop market by addressing use cases traditionally handled by operational databases from large vendors like Oracle. The Hadoop RDBMS is better suited to transactional workloads and can be scaled out on commodity hardware, the data management vendor noted.

While Splice Machine did not reveal other charter testers, CEO Monte Zweben told Datanami in May that prospective customers either had an urgent database problem solvable only by adopting a “scale-up” approach or had previously experimented with Hadoop.

“We have the Hadoop layer and Hbase underneath our platform and get to leverage the tremendous development community around the world, which is contributing and hardening Hadoop for the enterprise,” Zweben added “The open source community is building all the ancillary tools around Hadoop for transferring and transforming data like Sqoop and Kafka and Storm, or the machine learning packages like Mahout–all that development is being done by the open source community and we leverage that.”

Splice Machine said a standalone release of version 1.0 of its Hadoop RDBMS is available now for download. Its database is available for licensing on a per-node basis that includes technical support.

Recent items:

Hadoop-based RDBMS Now Available from Splice

Rethinking Real-Time Hadoop

 

Datanami