Upgraded In-Memory Database System Adds Index Functions

The latest upgrade to McObject’s eXtremeDB real-time in-memory embedded database system includes index functions that enhance its data sorting and retrieval performance. One index added to eXtremeDB 3.1 is the Patricia Trie index, which is a “longest match” index aimed at telecom and network communications applications. For example, when an application must find all the IP addresses in the subnet 124.57.32, a Patricia Trie is most efficient. The other addition is the R-Tree index, which is geared toward spatial access methods. Based on hierarchically nested “bouncing boxes” or rectangles, the index is efficient for searching multi-dimensional information, making it useful for geospatial-, navigation-, and location-based searches.

The upgrade also includes hot synchronization and schema evolution features that make the eXtremeDB High Availability version 3.1 more powerful for mission-critical systems like communications equipment, military/aerospace systems, and industrial control. Using a rugged, time-cognizant, two-phase commit protocol, eXtremeDB-HA ensures that changes to a master database and identical standby databases succeed or fail together. This allows deployment of multiple, fully synchronized eXtremeDB databases within the same hardware device or across multiple, widely distributed systems. The schema evolution or “hitless upgrade” capability enables a replica database that has been upgraded with new fields or tables to be attached to, and provisioned by, the master database. It permits other system nodes to be updated in turn with the database schema changes, all while maintaining system availability.