
The ability to continuously analyze operational data using big data techniques unlocks the potential for organizations to extract important patterns and trends that otherwise cannot be seen as the data rapidly changes. Unfortunately, popular big data systems such as Hadoop, which employ file-based storage and batch processing techniques, are not well suited for this challenge. However, the technology of in-memory data grids (IMDGs) offers important breakthroughs that enable real-time analysis of operational data. Recent measurements have demonstrated that an IMDG can deliver a complete map/reduce analysis every four seconds across a terabyte data set which is continuously being updated at the rate of one gigabyte per second.
This article discusses how IMDGs differ from other big data systems and deliver this important new capability to analyze fast-changing, operational data.