SQL Optimizer for Oracle 10.0 - User Guide

Analyze Impact of Parameter Change

Step two. After selecting to analyze the impact of a database parameter change, specify the parameter changes.

To analyze the impact of a database parameter change

1. If you have already created a new session, skip to 3. Otherwise, select Analyze Impact.
2. Click .
3. Now enter a session name, and select a database and schema.
4. Enter or select parameter changes in the Compare with value column.
5. After specifying the parameter changes, click to start collecting the SQL to evaluate.

Go to Step three. See Collect SQL and Run an Impact Analysis.

Related Topics

About Analyze Impact

Collect SQL and Run an Impact Analysis

Compare Databases

Step two. If you selected to compare two different databases, specify the databases to compare.

To compare databases

1. If you have already created a new session, skip to 3. Otherwise, select Analyze Impact.
2. Click .
3. Select a database connection and schema to specify the Baseline Database Connection.
4. Select a database connection and schema to specify the database to compare to the baseline.
5. Click to start collecting the SQL to evaluate.

Go to Step three. See Collect SQL and Run an Impact Analysis.

Related Topics

About Analyze Impact

Collect SQL and Run an Impact Analysis

Use Custom Scripts to Analyze Impact

Step two. If you selected to use a custom script to make database changes, specify the custom script details.

To use custom scripts to create a change for an Analyze Impact session

1. If you have already created a new session, skip to 3. Otherwise, select Analyze Impact.
2. Click .
3. Now enter a session name, and select a database connection and schema.

4. In the Pre-Script section, click to browse to and select a file for the prescript. After the prescript runs, SQL Optimizer retrieves each SQL statement's execution plan to be used as the post-change plan in the impact analysis.

   Note: Alternatively, you can use the script editor to compose a script.

5. (optional) In the Post-Script section, click to browse to and select a file for the post-script. You can use the post-script to roll back the change created with the first script. Using a post-script is not required.

   Note: Alternatively, you can use the script editor to compose a script.

6. After specifying the custom script details, click to start collecting the SQL to evaluate.

   Note: If you are not using a post-script, you can click Next->Select SQL after inputting the pre-script.

Go to Step three. See Collect SQL and Run an Impact Analysis.

Note: In the script editor, separate multiple commands with a semi-colon.

Related Topics

About Analyze Impact

Collect SQL and Run an Impact Analysis

Collect SQL and Run Analysis

Step three, four, five. After specifying the criteria for the type of change you want to analyze, select the SQL workload or group of SQL statements on which to evaluate the change's impact. You can collect SQL from one of several different sources.

Step three.

Collect SQL

1. Before collecting SQL, you must create an Analyze Impact session and specify a change to be analyzed.

   • If you have not created an Analyze Impact session, see About Analyze Impact.
   • If you have created an Analyze Impact session and specified the change details, click Next->Select SQL on the Change Details page.

2. On the SQL page, select the source from which you want to collect SQL. Select from the following:

   • Oracle Automatic Workload Repository (AWR)
   • Oracle System Global Area (SGA)
   • Source code

3. Next, define the SQL workload by specifying criteria for collecting SQL. Click one of the following for instructions.

   Collect SQL from the AWR

   To collect SQL from the AWR

   1. Select a database connection with AWR access from which to collect SQL.

      Note: Top N SQL displays the Top N setting for the database—the setting which specifies the number of SQL retained in the AWR for each category. If you want to modify this setting for a database, use the MODIFY_SNAPSHOT_SETTINGS system procedure.

   2. Specify SQL filter criteria. Review the following for additional information:

      SQL filter criteria	Description
      Date	Select a date from which to choose snapshots. After you select a date, Optimize Indexes displays all the snapshots available for the selected 24 hour period.
      Select SQL for parsing schema	You must select a parsing schema before collecting SQL.
      Module, Action	Use these fields to filter the available SQL before selecting snapshots. Default = All

   3. Select the SQL snapshots from which you want to collect SQL statements. Click a snapshot in the graph to select it. Click the selected snapshot again to deselect it.

      Notes:

      • Use the cursor to hover over a snapshot to display snapshot details.
      • Click List snapshot data to display snapshots in a grid format. Right-click a column heading to select options for filtering, sorting, or displaying.

      • You can display additional data in the graph (or in the grid). To do this, select/deselect data options from the group of options at the bottom of the page.

      • Select Total in the data options to display the total activity (the activity for all modules, actions, etc.).

    

   Collect SQL from the SGA

   To collect SQL from the SGA

   1. Select a database connection from which to collect SQL.

   2. Specify SQL filter criteria. To modify filter criteria, click the Edit link beside each filter. The available options display as bars in the graph. Click an option (bar) in the graph to select it. The selected option then displays in the filter criterion field in the top pane. Review the following for additional information:

      SQL filter criteria pane	Description
      Collect SQL for parsing schema	You must select a parsing schema before collecting SQL.
      Module, Action	Use these fields to filter the available SQL statements. Default = All

      Notes:

      • Use the cursor to hover over an option (bar) in the graph to display option details.
      • Click List statistics data to display the options for the currently selected filter criterion in a grid format.

      • You can display additional data in the graph (or in the grid). To do this, select/deselect data options from the group of options at the bottom of the page.

    

   Scan Code to collect SQL

   To Scan Code to collect SQL

   1. Select a source database connection.
   2. (Optional) Click Scan Options to specify options before scanning SQL.
   3. Select a method for scanning and collecting SQL.
      • Scan from Clipboard—Select this option to paste the SQL that you have previously copied to the clipboard.
      • Scan from File—Select this option to browse to and select a SQL file.
      • Scan from DB Object—Select this option to browse to and select a database object.

   4. After the SQL is scanned, the SQL Selector dialog opens displaying the list of scanned SQL. Select the checkbox to the left of each listed SQL to include it in your workload. Click on a scanned SQL in the left pane to highlight its SQL text in the right pane.

      Note: Enter a value in the REF (Relative Execution Frequency) column for each SQL. The REF will be used as an approximation of the number of times a SQL statement is executed relative to the other statements.

   5. Click OK to close the dialog. The Scan Code window now displays the list of SQL statements in your workload. To add additional SQL, select an additional method and source.

      Note: Click Remove to remove a selected SQL from the workload.

    

4. After specifying SQL criteria, click to start the collection process.

   Note: Click Cancel to return to the previous page and select a new SQL source type.

5. SQL Optimizer collects the specified SQL statements. The SQL page opens displaying the list of SQL statements collected. You can review the SQL or run the analysis.

Step four.

Review collected SQL (optional)

1. You can review or modify the list of SQL statements collected in the previous steps. Review the following for additional information:

   Pane / Button	Description
   SQL Tuning Set Pane / SQL Workload Pane	This pane displays the list of SQL in the workload. Include—To include a SQL statement in the optimize indexes process, select the checkbox in the Include column.
   SQL Text Pane	This pane displays the SQL text of the selected SQL statement.
   Parsed Execution Plan Pane	This pane displays the execution plan of the selected SQL statement. SQL Statistics—Select this tab to display statistics of the selected statement.
   	Select to send your SQL to Optimize SQL or Batch Optimize SQL. Click the arrow for more options.

   Notes:

   • Click or Collected from Source to return to the SQL Collection page. (On the SQL Collection page, click Review SQL to go back to the SQL Review page.)

   • To collect a new SQL workload, return to the SQL Collection page and click Edit Criteria.

     Caution: If you click Edit Criteria, the SQL workload will be lost, as well as any analysis data already gathered for the current session.

Step five.

Run the impact analysis

1. When you are satisfied with the SQL collected, click on the SQL page to run the impact analysis.

2. When the analysis process is finished, one of the following will display, depending on the analysis results:

   • If an impact was found, the Impacts page displays. See Review Impact Analysis Results.

   • If the analysis determines that non of the specified SQL were impacted, the log page displays and reports "No impacted SQL found." You can click to run the impact analysis again (without modifying any criteria).

     Note: You can modify the change or the SQL collection criteria and then run a new analysis within the same session. To do this, select the SQL or Change Details tab and click Edit criteria.

     Caution: If you click Edit Criteria, the analysis results for the current session will be lost, as well as the SQL workload or the Change criteria for the current session.

Note: Click to stop the SQL collection or analysis process.

Go to Step six. Review Impact Analysis Results.

Related Topics

About Analyze Impact

Review Impact Analysis Results