Terracotta had added native high-performance search to Ehcache, the open source de facto cache for enterprise Java, making it possible to do simple queries and real-time analysis on terabytes of in-memory data, overcoming in the process the latency of making queries to comparatively far distant databases.

Leveraging in-memory data stores in support of both transactional and analytical applications is supposed to make real-time query and analysis development costs cheaper because they both use the same native object level APIs. It could also save on hardware like Oracle's Exadata machines.

Used in conjunction with Terracotta's commercial BigMemory product, which bypasses Java garbage collection to enable terabyte caches, Ehcache Search is meant for both standalone and distributed cache users. Terracotta says distributed use is on the rise. According to a survey 15% of instances are distributed, and 70% are going that way.

Search is built into the core Ehcache API and can be downloaded at www.terracotta.org. It's both open source and commercial.

• The Growing Relevance of In-Memory Data Grids