This is pretty interesting, though casual readers may find the jargon tough sledding. The article describes a Red Hat product, called Data Grid 6, that manages large bodies of data more efficiently than traditional databases. It does this first by adopting a no-SQL approach, which creates data dynamically, rtaher than in predefined tables, and by creating a peer-to-peer grid, so data management is distributed, with no mater control. Each transaction basically looks after itself. A series of interfaces, such as a REST-API, allows the database to communicate with web-based applications. There's a lot of overlap between this and OpenSplice DDS, which I was looking at this morning. It supports "content-based subscriptions (also known as continuous queries in the CEP domain), distributed data persistence, and transparent fault-management." This kind of cold steel in the background will provide the basis for real-time and almost magical web applications.
Today: 7 Total: 100 [Direct link] [Share]
