SQL statements that are executed in a production system are typically not manually inputted by the users, but instead come from one or more applications running on an application server that is connected to the database server. There is usually a finite number of such SQL statements in an application. Even if different bind values are used in every execution of a particular statement, its SQL text essentially remains the same.
Depending on the user input to the application, a code path is executed that includes one or more SQL statements submitted to the database in a given order as defined by the application code. Frequent user actions correspond to application code paths that are regularly executed. Such frequently executed code paths generate a frequent pattern of SQL statements that are executed by the database in a given order. By analyzing a captured workload, Workload Intelligence discovers such patterns and associates them with related execution statistics. In other words, Workload Intelligence uses the information stored in capture files to discover patterns that are generated by significant code paths of applications running in the production system during workload capture. Workload Intelligence does this without requiring any information about the applications.
Using Workload Intelligence to discover significant patterns:
Enables you to better visualize what was running in the database during workload capture.
Provides more information that can be used for optimizations.
Offers a better context because SQL statements are not isolated, but are combined.