17 Automatic SQL Tuning

17.1.1 Statistics Analysis

The query optimizer relies on object statistics to generate execution plans. If these statistics are stale or missing, the optimizer does not have the necessary information it needs and can generate poor execution plans. The Automatic Tuning Optimizer checks each query object for missing or stale statistics, and produces two types of output:

• Recommendations to gather relevant statistics for objects with stale or no statistics.

  Because optimizer statistics are automatically collected and refreshed, this problem may be encountered only when automatic optimizer statistics collection has been turned off. See "Automatic Optimizer Statistics Collection".

• Auxiliary information in the form of statistics for objects with no statistics, and statistic adjustment factor for objects with stale statistics.

This auxiliary information is stored in an object called a SQL Profile.

17.1.2 SQL Profiling

The query optimizer can sometimes produce inaccurate estimates about an attribute of a statement due to lack of information, leading to poor execution plans. Traditionally, users have corrected this problem by manually adding hints to the application code to guide the optimizer into making correct decisions. For packaged applications, changing application code is not an option and the only alternative available is to log a bug with the application vendor and wait for a fix.

Automatic SQL tuning deals with this problem with its SQL profiling capability. The Automatic Tuning Optimizer creates a profile of the SQL statement called a SQL Profile, consisting of auxiliary statistics specific to that statement. The query optimizer under normal mode makes estimates about cardinality, selectivity, and cost that can sometimes be off by a significant amount resulting in poor execution plans. SQL Profile addresses this problem by collecting additional information using sampling and partial execution techniques to verify and, if necessary, adjust these estimates.

During SQL Profiling, the Automatic Tuning Optimizer also uses execution history information of the SQL statement to appropriately set optimizer parameter settings, such as changing the OPTIMIZER_MODE initialization parameter setting from ALL_ROWS to FIRST_ROWS for that SQL statement.

The output of this type of analysis is a recommendation to accept the SQL Profile. A SQL Profile, once accepted, is stored persistently in the data dictionary. Note that the SQL Profile is specific to a particular query. If accepted, the optimizer under normal mode uses the information in the SQL Profile in conjunction with regular database statistics when generating an execution plan. The availability of the additional information makes it possible to produce well-tuned plans for corresponding SQL statement without requiring any change to the application code.

The scope of a SQL Profile can be controlled by the CATEGORY profile attribute. This attribute determines which user sessions can apply the profile. You can view the CATEGORY attribute for a SQL Profile in CATEGORY column of the DBA_SQL_PROFILES view. By default, all profiles are created in the DEFAULT category. This means that all user sessions where the SQLTUNE_CATEGORY initialization parameter is set to DEFAULT can use the profile.

By altering the category of a SQL profile, you can determine which sessions are affected by the creation of a profile. For example, by setting the category of a SQL Profile to DEV, only those users sessions where the SQLTUNE_CATEGORY initialization parameter is set to DEV can use the profile. All other sessions do not have access to the SQL Profile and execution plans for SQL statements are not impacted by the SQL profile. This technique enables you to test a SQL Profile in a restricted environment before making it available to other user sessions.

It is important to note that the SQL Profile does not freeze the execution plan of a SQL statement, as done by stored outlines. As tables grow or indexes are created or dropped, the execution plan can change with the same SQL Profile. The information stored in it continues to be relevant, even as the data distribution or access path of the corresponding statement change. In general, it is not necessary to refresh the SQL profiles. However, over a long period of time, its content can become outdated and may need to be regenerated. This can be done by running the SQL Tuning Advisor again on the same statement to regenerate the SQL Profile.

SQL Profiles apply to the following statement types:

• SELECT statements

• UPDATE statements

• INSERT statements (only with a SELECT clause)

• DELETE statements

• CREATE TABLE statements (only with the AS SELECT clause)

• MERGE statements (the update or insert operations)

A complete set of functions are provided for management of SQL Profiles. See "SQL Profiles".

17.1.3 Access Path Analysis

Indexes can tremendously enhance performance of a SQL statement by reducing the need for full table scans on large tables. Effective indexing is a common tuning technique. The Automatic Tuning Optimizer also explores whether a new index can significantly enhance the performance of a query. If such an index is identified, it recommends its creation.

Because the Automatic Tuning Optimizer does not analyze how its index recommendation can affect the entire SQL workload, it also recommends running the SQL Access Advisor utility on the SQL statement along with a representative SQL workload. The SQL Access Advisor looks at the impact of creating an index on the entire SQL workload before making any recommendations. See "Automatic SQL Tuning Features".

17.1.4 SQL Structure Analysis

The Automatic Tuning Optimizer identifies common problems with structure of SQL statements that can lead to poor performance. These could be syntactic, semantic, or design problems with the statement. In each of these cases the Automatic Tuning Optimizer makes relevant suggestions to restructure the SQL statements. The alternative suggested is similar, but not equivalent, to the original statement.

For example, the optimizer may suggest to replace UNION operator with UNION ALL or to replace NOT IN with NOT EXISTS. An application developer can then determine if the advice is applicable to their situation or not. For instance, if the schema design is such that there is no possibility of producing duplicates, then the UNION ALL operator is much more efficient than the UNION operator. These changes require a good understanding of the data properties and should be implemented only after careful consideration.

17.3.1 Enabling and Disabling Automatic SQL Tuning

Automatic SQL tuning runs as part of the automated maintenance tasks infrastructure.

To enable automatic SQL tuning, use the ENABLE procedure in the DBMS_AUTO_TASK_ADMIN package:

BEGIN
  DBMS_AUTO_TASK_ADMIN.ENABLE(
    client_name => 'sql tuning advisor', 
    operation => NULL, 
    window_name => NULL);
END;
/

To disable automatic SQL tuning, use the DISABLE procedure in the DBMS_AUTO_TASK_ADMIN package:

BEGIN
  DBMS_AUTO_TASK_ADMIN.DISABLE(
    client_name => 'sql tuning advisor', 
    operation => NULL, 
    window_name => NULL);
END;
/

You can pass a specific window name using the window_name parameter to enable or disable the task in certain maintenance windows only.

Setting the STATISTICS_LEVEL parameter to BASIC will disable automatic statistics gathering by the AWR and, as a result, also disable automatic SQL tuning.

17.3.2 Configuring Automatic SQL Tuning

The behavior of the automatic SQL tuning task can be configured using the DBMS_SQLTUNE package. To use the APIs, the user needs at least the ADVISOR privilege.

In addition to configuring the standard behavior of the SQL Tuning Advisor, the DBMS_SQLTUNE package enables you to configure automatic SQL tuning by specifying the task parameters using the SET_TUNING_TASK_PARAMETER procedure. Because the automatic tuning task is owned by SYS, only the SYS user can set the task parameters.

Table 17-2 lists the parameters that are specific to automatic SQL tuning which can be configured.

To configure automatic SQL tuning, run the SET_TUNING_TASK_PARAMETER procedure in the DBMS_SQLTUNE package:

BEGIN
  DBMS_SQLTUNE.SET_TUNING_TASK_PARAMETER(
    task_name => 'SYS_AUTO_SQL_TUNING_TASK',
    parameter => 'ACCEPT_SQL_PROFILES', value => 'TRUE');
END;
/

In this example, the automatic SQL tuning task is configured to automatically accept SQL profiles recommended by the SQL Tuning Advisor.

17.3.3 Viewing Automatic SQL Tuning Reports

The automatic SQL tuning report is generated using the DBMS_SQLTUNE.REPORT_AUTO_TUNING_TASK function and contains information about all executions of the automatic SQL tuning task. To run this report, you need the ADVISOR privilege and SELECT privileges on the DBA_ADVISOR views. Unlike the standard SQL tuning report generated using the DBMS_SQLTUNE.REPORT_TUNING_TASK function, which only contains information about a single task execution of the SQL Tuning Advisor, the automatic SQL tuning report contains information about multiple executions of the automatic SQL tuning task.

To view the automatic SQL tuning report, run the REPORT_AUTO_TUNING_TASK function in the DBMS_SQLTUNE package:

variable my_rept CLOB;
BEGIN
  :my_rept :=DBMS_SQLTUNE.REPORT_AUTO_TUNING_TASK(
    begin_exec => NULL,
    end_exec => NULL,
    type => 'TEXT',
    level => 'TYPICAL',
    section => 'ALL',
    object_id => NULL,
    result_limit => NULL);
END;
/

print :my_rept

In this example, a text report is generated to display all SQL statements that were analyzed in the most recent execution, including recommendations that were not implemented, and all sections of the report are included.

Depending on the sections that were included in the report, you can view information about the automatic SQL tuning task in the following sections of the report:

• General information

  The general information section provides a high-level description of the automatic SQL tuning task, including information about the inputs given for the report, the number of SQL statements tuned during the maintenance, and the number of SQL profiles that were created

• Summary

  The summary section lists the SQL statements (by their SQL identifiers) that were tuned during the maintenance window and the estimated benefit of each SQL profile, or their actual execution statistics after test executing the SQL statement with the SQL profile

• Tuning findings

  This section contains the following information about each SQL statement analyzed by the SQL Tuning Advisor:

  • All findings associated with each SQL statement

  • Whether the profile was accepted on the system, and why

  • Whether the SQL profile is currently enabled on the system

  • Detailed execution statistics captured when testing the SQL profile

• Explain plans

  This section shows the old and new explain plans used by each SQL statement analyzed by the SQL Tuning Advisor.

• Errors

  This section lists all errors encountered by the automatic SQL tuning task.

17.4.1 Input Sources

The input for the SQL Tuning Advisor can come from several sources. These input sources include:

• Automatic Database Diagnostic Monitor

  The primary input source is the Automatic Database Diagnostic Monitor (ADDM). By default, ADDM runs proactively once every hour and analyzes key statistics gathered by the Automatic Workload Repository (AWR) over the last hour to identify any performance problems including high-load SQL statements. If a high-load SQL is identified, ADDM recommends running SQL Tuning Advisor on the SQL. See "Overview of the Automatic Database Diagnostic Monitor".

• Automatic Workload Repository

  The second most important input source is the Automatic Workload Repository (AWR). The AWR takes regular snapshots of the system activity, including high-load SQL statements ranked by relevant statistics, such as CPU consumption and wait time.

  You can view the AWR and manually identify high-load SQL statements and run the SQL Tuning Advisor on them, though this is done automatically by Oracle Database as part of the automatic SQL tuning process. By default, the AWR retains data for the last eight days. Any high-load SQL that ran within the retention period of the AWR can be located and tuned using this method. See "Overview of the Automatic Workload Repository".

• Cursor cache

  The third likely source of input is the cursor cache. This source is used for tuning recent SQL statements that are yet to be captured in the AWR. The cursor cache and AWR together provide the capability to identify and tune high-load SQL statements from the current time going as far back as the AWR retention allows, which by default is at least 8 days.

• SQL Tuning Set

  Another possible input source for the SQL Tuning Advisor is the SQL Tuning Set. A SQL Tuning Set (STS) is a database object that stores SQL statements along with their execution context. An STS can include SQL statements that are yet to be deployed, with the goal of measuring their individual performance, or identifying the ones whose performance falls short of expectation. When a set of SQL statements are used as input, a SQL Tuning Set (STS) has to be first constructed and stored. See "SQL Tuning Sets".

17.4.2 Tuning Options

The SQL Tuning Advisor provides options to manage the scope and duration of a tuning task. The scope of a tuning task can be set to limited or comprehensive.

• If the limited option is chosen, the SQL Tuning Advisor produces recommendations based on statistics checks, access path analysis, and SQL structure analysis. SQL Profile recommendations are not generated.

• If the comprehensive option is selected, the SQL Tuning Advisor carries out all the analysis it performs under limited scope plus SQL Profiling. With the comprehensive option you can also specify a time limit for the tuning task, which by default is 30 minutes.

17.4.3 Advisor Output

After analyzing the SQL statements, the SQL Tuning Advisor provides advice on optimizing the execution plan, the rationale for the proposed optimization, the estimated performance benefit, and the command to implement the advice. You simply have to choose whether or not to accept the recommendations to optimize the SQL statements.

17.4.4 Running the SQL Tuning Advisor

The recommended interface for running the SQL Tuning Advisor is the Oracle Enterprise Manager. Whenever possible, you should run the SQL Tuning Advisor using Oracle Enterprise Manager, as described in the Oracle Database 2 Day + Performance Tuning Guide. If Oracle Enterprise Manager is unavailable, you can run the SQL Tuning Advisor using procedures in the DBMS_SQLTUNE package. To use the APIs, the user must be granted specific privileges.

Running SQL Tuning Advisor using DBMS_SQLTUNE package is a multi-step process:

1. Create a SQL Tuning Set (if tuning multiple SQL statements)

2. Create a SQL tuning task

3. Execute a SQL tuning task

4. Display the results of a SQL tuning task

5. Implement recommendations as appropriate

A SQL tuning task can be created for a single SQL statement. For tuning multiple statements, a SQL Tuning Set (STS) has to be first created. An STS is a database object that stores SQL statements along with their execution context. An STS can be created manually using command line APIs or automatically using Oracle Enterprise Manager. See "SQL Tuning Sets".

Figure 17-2 shows the steps involved when running the SQL Tuning Advisor using the DBMS_SQLTUNE package.

Figure 17-2 SQL Tuning Advisor APIs

Description of "Figure 17-2 SQL Tuning Advisor APIs"

This section covers the following topics:

• Creating a SQL Tuning Task

• Configuring a SQL Tuning Task

• Executing a SQL Tuning Task

• Checking the Status of a SQL Tuning Task

• Checking the Progress of the SQL Tuning Advisor

• Displaying the Results of a SQL Tuning Task

• Additional Operations on a SQL Tuning Task

DECLARE
 my_task_name VARCHAR2(30);
 my_sqltext   CLOB;
BEGIN
 my_sqltext := 'SELECT /*+ ORDERED */ * '                      ||
               'FROM employees e, locations l, departments d ' ||
               'WHERE e.department_id = d.department_id AND '  ||
                     'l.location_id = d.location_id AND '      ||
                     'e.employee_id < :bnd';

 my_task_name := DBMS_SQLTUNE.CREATE_TUNING_TASK(
         sql_text    => my_sqltext,
         bind_list   => sql_binds(anydata.ConvertNumber(100)),
         user_name   => 'HR',
         scope       => 'COMPREHENSIVE',
         time_limit  => 60,
         task_name   => 'my_sql_tuning_task',
         description => 'Task to tune a query on a specified employee');
END;
/

SELECT task_name FROM DBA_ADVISOR_LOG WHERE owner = 'HR';

BEGIN
  DBMS_SQLTUNE.SET_TUNING_TASK_PARAMETER(
    task_name => 'my_sql_tuning_task',
    parameter => 'TIME_LIMIT', value => 300);
END;
/

BEGIN
  DBMS_SQLTUNE.EXECUTE_TUNING_TASK( task_name => 'my_sql_tuning_task' );
END;
/

SELECT status FROM USER_ADVISOR_TASKS WHERE task_name = 'my_sql_tuning_task';

SELECT sofar, totalwork FROM V$ADVISOR_PROGRESS WHERE user_name = 'HR' AND task_name = 'my_sql_tuning_task';

SET LONG 1000
SET LONGCHUNKSIZE 1000
SET LINESIZE 100
SELECT DBMS_SQLTUNE.REPORT_TUNING_TASK( 'my_sql_tuning_task')
  FROM DUAL;

17.5.1 Creating a SQL Tuning Set

The CREATE_SQLSET procedure is used to create an empty STS object in the database. For example, the following procedure creates an STS object that could be used to tune I/O intensive SQL statements during a specific period of time:

BEGIN
  DBMS_SQLTUNE.CREATE_SQLSET(
    sqlset_name => 'my_sql_tuning_set', 
    description  => 'I/O intensive workload');
END;
/

where my_sql_tuning_set is the name of the STS in the database and 'I/O intensive workload' is the description assigned to the STS.

17.5.2 Loading a SQL Tuning Set

The LOAD_SQLSET procedure populates the STS with selected SQL statements. The standard sources for populating an STS are the workload repository, another STS, or the cursor cache. For both the workload repository and STS, there are predefined table functions that can be used to select columns from the source to populate a new STS.

In the following example, procedure calls are used to load my_sql_tuning_set from an AWR baseline called peak baseline. The data has been filtered to select only the top 30 SQL statements ordered by elapsed time. First a ref cursor is opened to select from the specified baseline. Next the statements and their statistics are loaded from the baseline into the STS.

DECLARE
 baseline_cursor DBMS_SQLTUNE.SQLSET_CURSOR;
BEGIN
 OPEN baseline_cursor FOR
    SELECT VALUE(p)
    FROM TABLE (DBMS_SQLTUNE.SELECT_WORKLOAD_REPOSITORY(
                  'peak baseline',
                   NULL, NULL,
                   'elapsed_time',
                   NULL, NULL, NULL,
                   30)) p;

    DBMS_SQLTUNE.LOAD_SQLSET(
             sqlset_name     => 'my_sql_tuning_set',
             populate_cursor => baseline_cursor);
END;
/

17.5.3 Displaying the Contents of a SQL Tuning Set

The SELECT_SQLSET table function reads the contents of the STS. After an STS has been created and populated, you can browse the SQL in the STS using different filtering criteria. The SELECT_SQLSET procedure is provided for this purpose.

In the following example, the SQL statements in the STS are displayed for statements with a disk-reads to buffer-gets ratio greater than or equal to 75%.

SELECT * FROM TABLE(DBMS_SQLTUNE.SELECT_SQLSET( 
   'my_sql_tuning_set',
   '(disk_reads/buffer_gets) >= 0.75'));

Additional details of the SQL Tuning Sets that have been created and loaded can also be displayed with DBA views, such as DBA_SQLSET, DBA_SQLSET_STATEMENTS, and DBA_SQLSET_BINDS.

17.5.4 Modifying a SQL Tuning Set

SQL statements can be updated and deleted from a SQL Tuning Set based on a search condition. In the following example, the DELETE_SQLSET procedure deletes SQL statements from my_sql_tuning_set that have been executed less than fifty times.

BEGIN
  DBMS_SQLTUNE.DELETE_SQLSET(
      sqlset_name  => 'my_sql_tuning_set',
      basic_filter => 'executions < 50');
END;
/

17.5.5 Transporting a SQL Tuning Set

SQL Tuning Sets can be transported to another system by first exporting the STS from one system to a staging table, then importing the STS from the staging table into another system.

To transport a SQL Tuning Set:

1. Use the CREATE_STGTAB_SQLSET procedure to create a staging table where the SQL Tuning Sets will be exported.

   The following example shows how to create a staging table named staging_table. Table names are case-sensitive.

   BEGIN
     DBMS_SQLTUNE.CREATE_STGTAB_SQLSET( table_name => 'staging_table' );
   END;
   /
   

2. Use the PACK_STGTAB_SQLSET procedure to export SQL Tuning Sets into the staging table.

   The following example shows how to export a SQL Tuning Set named my_sts to the staging table.

   BEGIN
     DBMS_SQLTUNE.PACK_STGTAB_SQLSET(
         sqlset_name  => 'my_sts',
         staging_table_name => 'staging_table');
   END;
   /
   

3. Move the staging table to the system where the SQL Tuning Sets will be imported using the mechanism of choice (such as datapump or database link).

4. On the system where the SQL Tuning Sets will be imported, use the UNPACK_STGTAB_SQLSET procedure to import SQL Tuning Sets from the staging table.

   The following example shows how to import SQL Tuning Sets contained in the staging table.

   BEGIN
     DBMS_SQLTUNE.UNPACK_STGTAB_SQLSET(
         sqlset_name  => '%',
         replace  => TRUE,
         staging_table_name => 'staging_table');
   END;
   /
   

17.5.6 Dropping a SQL Tuning Set

The DROP_SQLSET procedure is used to drop an STS that is no longer needed. For example:

BEGIN
  DBMS_SQLTUNE.DROP_SQLSET( sqlset_name => 'my_sql_tuning_set' );
END;
/

17.5.7 Additional Operations on SQL Tuning Sets

You can use the following APIs to manage an STS:

• Updating the attributes of SQL statements in an STS

  The UPDATE_SQLSET procedure updates the attributes of SQL statements (such as PRIORITY or OTHER) in an existing STS identified by STS name and SQL identifier.

• Capturing the full system workload

  The CAPTURE_CURSOR_CACHE_SQLSET function enables the capture of the full system workload by repeatedly polling the cursor cache over a specified interval. This function is a lot more efficient than repeatedly using the SELECT_CURSOR_CACHE and LOAD_SQLSET procedures to capture the cursor cache over an extended period of time. This function effectively captures the entire workload, as opposed to the AWR—which only captures the workload of high-load SQL statements—or the LOAD_SQLSET procedure, which accesses the data source only once.

• Adding and removing a reference to an STS

  The ADD_SQLSET_REFERENCE function adds a new reference to an existing STS to indicate its use by a client. The function returns the identifier of the added reference. The REMOVE_SQLSET_REFERENCE procedure is used to deactivate an STS to indicate it is no longer used by the client.

17.6.1 Accepting a SQL Profile

When the SQL Tuning Advisor recommends that a SQL Profile be used, you should accept the SQL Profile that is recommended. In cases where the SQL Tuning Advisor recommends that an index and a SQL Profile be used, both should be used. You can use the DBMS_SQLTUNE.ACCEPT_SQL_PROFILE procedure to accept a SQL Profile recommended by the SQL Tuning Advisor. This creates and stores a SQL Profile in the database. For example:

DECLARE
 my_sqlprofile_name VARCHAR2(30);
BEGIN
 my_sqlprofile_name := DBMS_SQLTUNE.ACCEPT_SQL_PROFILE ( 
    task_name    => 'my_sql_tuning_task',
    name         => 'my_sql_profile',
    force_match  => TRUE);
END;

In this example, my_sql_tuning_task is the name of the SQL tuning task and my_sql_profile is the name of the SQL Profile that you want to accept.

Typically, an accepted SQL Profile is associated with the SQL statement through a special SQL signature that is generated using a hash function. This hash function normalizes the SQL statement for case (changes the entire SQL statement to upper case) and white spaces (removes all extra whites spaces) before generating the signature. The same SQL Profile thus will work for all SQL statements that are essentially the same, where the only difference is in case usage and white spaces. However, by setting force_match to true, the SQL Profile will additionally target all SQL statements that have the same text after normalizing literal values to bind variables. This may be useful for applications that use literal values rather than bind variables, since this will allow SQL with text differing only in its literal values to share a SQL Profile. If both literal values and bind variables are used in the SQL text, or if this parameter is set to false (the default value), literal values will not be normalized.

If SQL plan management is used and there is already an existing plan baseline for the SQL statement, a new plan baseline will be added when a SQL profile is created. If SQL plan management is not used, a new plan baseline will not be added when a SQL profile is created. There is no strict relationship between the SQL profile and the plan baseline. When hard parsing a SQL statement, the optimizer will use the SQL profile to select the best plan baseline from the ones available. In different conditions, the SQL profile may cause the optimizer to select different plan baselines. For information about SQL plan management, see Chapter 15, "Using SQL Plan Management".

You can view information about a SQL Profile in the DBA_SQL_PROFILES view.

17.6.2 Altering a SQL Profile

You can alter the STATUS, NAME, DESCRIPTION, and CATEGORY attributes of an existing SQL Profile with the ALTER_SQL_PROFILE procedure. For example:

BEGIN
  DBMS_SQLTUNE.ALTER_SQL_PROFILE(
     name            => 'my_sql_profile', 
     attribute_name  => 'STATUS', 
     value           => 'DISABLED');
END;
/

In this example, my_sql_profile is the name of the SQL Profile that you want to alter. The status attribute is changed to disabled, which means the SQL Profile is not used during SQL compilation.

17.6.3 Dropping a SQL Profile

You can drop a SQL Profile with the DROP_SQL_PROFILE procedure. For example:

BEGIN
  DBMS_SQLTUNE.DROP_SQL_PROFILE(name => 'my_sql_profile');
END;
/

In this example, my_sql_profile is the name of the SQL Profile you want to drop. You can also specify whether to ignore errors raised if the name does not exist. For this example, the default value of FALSE is accepted.