Using Function Based Indexes
Mark Rittman

Using Function-Based Indexes

One of the projects I'm working on now involves tuning an Oracle 8i-based data warehouse, that uses Discoverer as the user query tool. We've been asked to improve the query response time for the warehouse, and one of the first things we picked up on was the fact that most of the user queries resulted in full table scans against the database tables.

Having looked at the SQL being produced by Discoverer, the DDL for the various tables, and the explain plans, it became apparent that whilst there were indexes on most of the columns used in the WHERE clauses, the indexes weren't being used because functions were being used on the columns. For example, if a table called EMPLOYEES had an index on the ENAME column and the following query

SELECT name, address, postcode
FROM   employees
WHERE upper(name)='SMITH'

was issued against the table, the index wouldn't get used as the UPPER() function had been applied to it.

In this instance, a new feature called 'Function-Based Indexes' that first appeared with Oracle 8i would help us out. According to this dba-support.com article,

"One of the many new features in Oracle 8i is the Function-Based Index . This allows the DBA to create indexes on functions or expressions; these functions can be user generated pl/sql functions, standard SQL functions (non-aggregate only) or even a C callout.

The function-based index has forced the optimizer to use index range scans (retuning zero or more rowids) on the surname column rather than doing a full table scan (non-index lookup). Optimal performance does vary depending on table size, uniqueness and selectivity of columns, use of fast full table scans etc. Therefore try both methods to gain optimal performance in your database.

It is important to remember that the function-based B*Tree index does not store the expression results in the index but uses an "expression tree". The optimizer performs expression matching by parsing the expression used in the SQL statement and comparing the results against the expression-tree values in the function-based index. This comparison IS case sensitive (ignores spaces) and therefore your function-based index expressions should match expressions used in the SQL statement where clauses. "

In this case, to create the Function-Based Index we need, we'd issue the following DDL;

CREATE INDEX fb_name_idx on employees (UPPER(name));

The article also points out a few prerequistes for working with Function-Based Indexes;

• After creating the Function-Based Index, you must ANALYZE it

ANALYZE INDEX fb_name_idx COMPUTE STATISTICS;

• You then need to ANALYZE the table

ANALYZE TABLE employees COMPUTE STATISTICS;

• The users who will use the index need the Execute priviledge on the particular function used in the index, if it's a custom function
• They also need the CREATE INDEX (obviously) and QUERY REWRITE priviledge

GRANT CREATE INDEX, QUERY REWRITE TO user;

• The database needs to have the QUERY_REWRITE_INTEGRITY = TRUSTED, QUERY_REWRITE_ENABLED = TRUE and COMPATIBLE = 8.1.0.0.0 (or higher) parameters set, by using ALTER SYSTEM or amending the init.ora file.

For details of enhancements to Function-Based Indexes in Oracle 9i, check out this additional article by Don Burleson. 

Once an FBI is created, you need to create CBO statistics, but beware that there are numerous bugs and issues when analyzing a function-based index.  See these important notes on statistics and function-based indexes.