In the current implementation, full-text indexes are different from classic SQL Server indexes. These differences give rise to the need for the administrative tasks outlined here. Following is a summary of the main differences.
Classic SQL indexes | Full-text indexes |
---|---|
Stored in and under the control of the database. | Stored in the file system, but administered through the database. |
Several indexes per table. | Only one full-text index definition per table. |
Automatically updated when the data upon which they are based is inserted, updated, or deleted. | Population must be requested (either as scheduled or on demand). |
Not grouped. | Grouped. |
Created and dropped using SQL statements. | Created, managed, and dropped using stored procedures (either directly of by means of GUIs). |
Full-Text Administration is carried out at several different levels:
At all of these levels, there are facilities to retrieve information on property, status, and volume (for example, size of the full-text catalog). These facilities take the form of both GUIs and stored procedures, and almost all tasks can be accomplished either way. This chapter concentrates on the stored procedures and gives only a few examples of the use of GUIs. The term “stored procedures” has been used to illustrate the level at which an administrator is communicating with SQL Server. In fact, a combination of stored procedures and scalar-valued property functions is used.