Enterprise open source software provider Red Hat has jumpedinto the emerging NoSQL market, releasing an in-memory data store as part of an update of its JBoss management platform.
Red Hat JBoss Data Grid 6 "is JBoss' big data solution. It will allow companies to scale out their applications and reduce the need for adding more relational databases," said Craig Muzilla, Red Hat's vice president and general manager of middleware, in a Webcast Wednesday announcing the release.
The software is designed to serve as a large data cache for high-volume, low-latency transactional workloads. As such, it may be of particular value to run trading, logistics and e-commerce applications, the company claims. It is built to be fault-tolerant and scalable: nodes can be easily added or removed.
Red Hat JBoss Data Grid 6 is an in-memory, key-value store that can be accessed using REST (Representational State Transfer), the memcache API (application programming interface) or Red Hat's own specific API for the data store, called HotRod. Either Java programs or non-Java programs, such as Python or the those running on the Microsoft .Net platform, can access the data store. The Data Grid is built from the Infinispan JBoss Community project, and includes pieces from the JBoss Enterprise Application Platform and JBoss Operations Network.
Red Hat has released the data grid software to accompany the latest update of its JBoss Enterprise Application Platform, version 6 of which was also released this week. The JBoss Enterprise Application Platform is the management software for Red Hat's open source JBoss application server software, used for running server-based enterprise Java programs. This version is the first to run Java Enterprise Edition 6, the latest version of the JEE standard. It also has been updated to support the latest versions of Java frameworks, including Spring, Struts, and Google Web Toolkit.
The JBoss Enterprise Application Platform architecture has been changed to make it more easily run in PaaS (platform as a service) cloud deployments. "It provides a high degree of abstraction and automation. One can develop an application and deploy it in a public or private cloud, or on premise," Muzilla said. "You can use the application platform as the container in the cloud. It uses very little memory."