EXPLAIN PLAN

Purpose

Use the EXPLAIN PLAN statement to determine the execution plan Oracle Database follows to execute a specified SQL statement. This statement inserts a row describing each step of the execution plan into a specified table. You can also issue the EXPLAIN PLAN statement as part of the SQL trace facility.

This statement also determines the cost of executing the statement. If any domain indexes are defined on the table, then user-defined CPU and I/O costs will also be inserted.

The definition of a sample output table PLAN_TABLE is available in a SQL script on your distribution media. Your output table must have the same column names and data types as this table. The common name of this script is UTLXPLAN.SQL. The exact name and location depend on your operating system.

Oracle Database provides information on cached cursors through several dynamic performance views:

For information on the work areas used by SQL cursors, query V$SQL_WORKAREA.
For information on the execution plan for a cached cursor, query V$SQL_PLAN.
For execution statistics at each step or operation of an execution plan of cached cursors (for example, number of produced rows, number of blocks read), query V$SQL_PLAN_STATISTICS.
For a selective precomputed join of the preceding three views, query V$SQL_PLAN_STATISTICS_ALL.
Execution statistics at each step or operation of an execution plan of cached cursors are displayed in V$SQL_PLAN_MONITOR if the statement execution is monitored. You can force monitoring using the MONITOR hint.

See Also:

INSERT and SELECT for information on the privileges you need to populate and query the plan table

Syntax

explain_plan::=

Description of the illustration explain_plan.gif

Semantics

SET STATEMENT_ID Clause

Specify a value for the STATEMENT_ID column for the rows of the execution plan in the output table. You can then use this value to identify these rows among others in the output table. Be sure to specify a STATEMENT_ID value if your output table contains rows from many execution plans. If you omit this clause, then the STATEMENT_ID value defaults to null.

INTO table Clause

Specify the name of the output table, and optionally its schema and database. This table must exist before you use the EXPLAIN PLAN statement.

If you omit schema, then the database assumes the table is in your own schema.

The dblink can be a complete or partial name of a database link to a remote Oracle Database where the output table is located. You can specify a remote output table only if you are using Oracle Database distributed functionality. If you omit dblink, then the database assumes the table is on your local database. See "References to Objects in Remote Databases" for information on referring to database links.

If you omit INTO altogether, then the database assumes an output table named PLAN_TABLE in your own schema on your local database.

FOR statement Clause

Specify a SELECT, INSERT, UPDATE, DELETE, MERGE, CREATE TABLE, CREATE INDEX, or ALTER INDEX ... REBUILD statement for which the execution plan is generated.

Notes on EXPLAIN PLAN The following notes apply to EXPLAIN PLAN:

If statement includes the parallel_clause, then the resulting execution plan will indicate parallel execution. However, EXPLAIN PLAN actually inserts the statement into the plan table, so that the parallel DML statement you submit is no longer the first DML statement in the transaction. This violates the Oracle Database restriction of one parallel DML statement in a single transaction, and the statement will be executed serially. To maintain parallel execution of the statements, you must commit or roll back the EXPLAIN PLAN statement, and then submit the parallel DML statement.
To determine the execution plan for an operation on a temporary table, EXPLAIN PLAN must be run from the same session, because the data in temporary tables is session specific.

Examples

EXPLAIN PLAN Examples The following statement determines the execution plan and cost for an UPDATE statement and inserts rows describing the execution plan into the specified plan_table table with the STATEMENT_ID value of 'Raise in Tokyo':

EXPLAIN PLAN 
    SET STATEMENT_ID = 'Raise in Tokyo' 
    INTO plan_table 
    FOR UPDATE employees 
        SET salary = salary * 1.10 
        WHERE department_id =  
           (SELECT department_id FROM departments
               WHERE location_id = 1700);

The following SELECT statement queries the plan_table table and returns the execution plan and the cost:

SELECT id, LPAD(' ',2*(LEVEL-1))||operation operation, options,
       object_name, object_alias, qblock_name, position 
    FROM plan_table 
    START WITH id = 0 AND statement_id = 'Raise in Tokyo'
    CONNECT BY PRIOR id = parent_id AND statement_id = 'Raise in Tokyo'
    ORDER BY id;

The query returns this execution plan:

 ID OPERATION            OPTIONS              OBJECT_NAME          OBJECT_ALIAS         POSITION
--- -------------------- -------------------- -------------------- -------------------- --------
  0 UPDATE STATEMENT                                                                           4
  1   UPDATE                                  EMPLOYEES                                        1
  2     INDEX            RANGE SCAN           EMP_DEPARTMENT_IX    EMPLOYEES@UPD$1             1
  3       TABLE ACCESS   BY INDEX ROWID       DEPARTMENTS          DEPARTMENTS@SEL$1           1
  4         INDEX        RANGE SCAN           DEPT_LOCATION_IX     DEPARTMENTS@SEL$1           1

The value in the POSITION column of the first row shows that the statement has a cost of 4.

EXPLAIN PLAN: Partitioned Example The sample table sh.sales is partitioned on the time_id column. Partition sales_q3_2000 contains time values less than Oct. 1, 2000, and there is a local index sales_time_bix on the time_id column.

Consider the query:

EXPLAIN PLAN FOR
  SELECT * FROM sales 
     WHERE time_id BETWEEN :h AND '01-OCT-2000';

where :h represents an already declared bind variable. EXPLAIN PLAN executes this query with PLAN_TABLE as the output table. The basic execution plan, including partitioning information, is obtained with the following query:

SELECT operation, options, partition_start, partition_stop,
       partition_id
  FROM plan_table;