Splice Machine Debuts Relational Database for Big Data Storage

Splice Machine, which is building a scale-out database storage system for Hadoop that it said dwarfs traditional SQL database performance yet runs on commodity hardware, has now released version 1.0 of its relational database platform for big data.

The company first previewed its storage system last May, when it released the product in public-beta form. The software is now ready for prime time, according to Splice Machine, bolstered by features such as LDAP integration as well as analytic function capabilities the developers have added since the beta release in the spring.

The goal behind the new product is to provide a relational database tailored for Hadoop that can deliver performance increases up to 10 times faster than traditional relational databases, such as Oracle MySQL and IBM DB2, while running on commodity hardware. That strategy differentiates Splice Machine from competing vendors in the ongoing storage wars, where most other solutions for big data storage hinge upon either so-called NoSQL storage that discards the relational database model, or scale-out storage systems designed specifically for the cloud and clusters, such as Red Hat‘s GlusterFS.

Time will tell whether Splice Machine’s approach will succeed, but the company’s platform has already won endorsement from some segments of the channel. Cloudera is deploying the Splice Machine relational database for Hadoop as part of its integrated big data solution, and analyst Robin Bloor of the Bloor Group called the software “the natural choice for companies that are looking to move database workloads to Hadoop for the sake of scale.”

Version 1.0 of Splice Machine’s database is freely available for download (after registration) now for use on a single node. The mainline product, for clusters, requires the purchase of a license.