When executing a query, SQL Anywhere chooses how to access each table. Indexes greatly speed up the access. When the database server cannot find a suitable index, it resorts to scanning the table sequentially—a process that can take a long time.

For example, suppose you need to search a large database for people, but you only know either their first or their last name, but not both. If no index exists, SQL Anywhere scans the entire table. If, however, you created two indexes (one that contains the last names first, and a second that contains the first names first), SQL Anywhere scans the indexes first, and can generally return the information to you faster.

Proper selection of indexes can make a large performance difference. Creating and managing indexes is described in Working with indexes.

Using indexes

Although indexes let SQL Anywhere locate information very efficiently, exercise some caution when adding them. Each index creates extra work every time you insert, delete, or update a row because SQL Anywhere must also update all affected indexes.

Consider adding an index when it allows SQL Anywhere to access data more efficiently. In particular, add an index when it eliminates unnecessarily accessing a large table sequentially. If, however, you need better performance when you add rows to a table, and finding information quickly is not an issue, use as few indexes as possible.

You may want to use the Index Consultant to guide you through the selection of an effective set of indexes for your database. See Index Consultant.

Clustered indexes

Using clustered indexes helps store rows in a table in approximately the same order as they appear in the index. See Indexes, and Using clustered indexes.