StorHouse/RM

Providing scalable, cost-effective, direct access to relational data throughout its useful lifecycle

Innovative Technology - Patented Segmentation StorHouse/RM's innovative technology incorporates a patented segmentation scheme that enables unlimited, yet manageable, database structural growth. Segmentation makes it possible to use removable storage technologies or massive arrays of idle disks (MAID) for relational database storage

In StorHouse/RM, applications produce relational tables through batch loads that create segments, or partitions, consisting of database user file components (tables, indexes, and large object data). As Figure 2 illustrates, each segment consists of one table data partition file, one index file for each value index and hash index, and one or more LOB subsegment files for a LOB column.

Segmentation

Figure 2: Segmentation

Segment components can be stored separately throughout the virtualized hierarchy to optimize the performance of each storage layer.

Segmentation provides numerous benefits.

  • Segments easily map to StorHouse files, volume sets, and file sets, which StorHouse/SM automatically manages. This simplifies system and database administration tasks and keeps storage and administration costs down.


  • StorHouse/RM maps index files to individual table data files for faster load and delete processing. There's no need to build one large global index for all segments and then issue time-consuming updates against that index.


  • Users can run loads and queries against the same table in parallel.

Queries process more quickly because index files and the most recent table data files can reside on faster access media.