Implementation Guide › Optimizing Identity Manager › Runtime Components Tuning › Tuning Identity Manager Databases

Tuning Identity Manager Databases

When executing tasks, Identity Manager uses the following databases:

• Task persistence

  Maintains information about Identity Manager tasks and events over time. This allows Identity Manager to restore the last known state of events and tasks in the case of system failure.

  Note: This database has the most significant impact on Identity Manager performance because the task and its events are saved and retrieved from the database during state transitions.

• Audit

  Provides a historical record of operations that occur in an Identity Manager environment.

• Workflow

  Stores workflow process definitions, jobs, scripts, and other data required by the workflow engine.

• Reporting

  Stores snapshot data, which reflects the current state of objects in Identity Manager at the time the snapshot is taken.

Identity Manager communicates with each database through a JDBC connection pool. You create and configure a JDBC connection pool in the application server that hosts Identity Manager. When you configure the JDBC connection pool, note the following:

• Consider the number of concurrent tasks that will execute at any one time.
• Consider the other runtime components when you configure the JDBC connection pool size. Each runtime component works in conjunction with the other runtime components.

  Note: CA recommends setting the initial connection pool value to 128.

• For the task persistence database, the number of database connections in the pool must allow each executing task to retrieve and update task and event data throughout the lifetime of the task.
• The task persistence database uses prepared statements. Be sure to configure the prepared statement cache for the database that you are using to store task persistence data.

  Note: See the documentation for the database that you are using for task persistence for information on configuring the prepared statement cache.