Infinispan's differs from other Java data caches in that it was built specifically for scalable multi-terabyte deployments that can involve thousands of servers, Sharples said.

Infinispan borrows ideas from , which offered a technique to reliably store large amounts of information across multiple servers. It also features easy scalability and the ability to partition off multiple applications from one another on a single data grid.

Red Hat has launched a free in order to get customer feedback, which it will use to improve the reliability of the software and develop new enterprise-friendly features. The company hopes organizations will set up test deployments and offer feedback. The program will offer a working version of Infinispan as well as updates, Web tutorials and access to the open source working code.

A pre-beta of the working code should be out within a few months, with successive beta versions to follow. The company plans to release a production-ready version of the product around the end of the year, Sharples said.

The IDG News Service