Omnidex stores its indexes in files that are located in the Index Directory, which is specified in the CREATE DATABASE statement. The index files will all begin with the name of the database as declared in the CREATE DATABASE statement and will be followed by various names or numbers. Small databases will have only a handful of index files, while large databases may have hundreds of files.
The Omnidex index files are interdependent and cannot be individually copied or separated without corrupting the indexes. During backups or deployments, the entire index fileset should be copied with the data. Most commonly, administrators will back up or copy the entire index directory, alongside a copy of the data.
07/12/2011 03:00 PM 4,096 SIMPLE0000 07/12/2011 03:00 PM 20,480 SIMPLE0001 07/12/2011 03:00 PM 1,434 SIMPLE0002 07/12/2011 03:00 PM 8,192 SIMPLE0003 07/12/2011 03:00 PM 456 SIMPLE0004 07/12/2011 03:00 PM 8,192 SIMPLE0005 07/12/2011 03:00 PM 10 SIMPLE0006 07/12/2011 03:00 PM 114,688 SIMPLE0007 07/12/2011 03:00 PM 30,544 SIMPLE0008 07/12/2011 03:00 PM 262,144 SIMPLE0009 07/12/2011 03:00 PM 20,480 SIMPLE0010 07/12/2011 03:00 PM 80,000 SIMPLE0011 07/12/2011 03:00 PM 14,404 SIMPLESTATS
Omnidex indexes use an efficient index compression algorithm to reduce the physical footprint of Omnidex on the disk drive. In most cases, Omnidex requires about 2.5 bytes per keyword for its indexes. A keyword is an individual index entry. For standard Omnidex indexes, a keyword is an index value; for QuickText indexes, a keyword is a parsed word from the text field. Note that Omnidex Bitmap indexes require substantially less space, and FullText indexes require more space.
See also: