Oracle Workflow Guide
Release 2.6.2

Part Number A95265-02
Previous Next       Contents Index Glossary
         Previous  Next          Contents  Index  Glossary

Overview of the Workflow Engine

The Workflow Engine manages all automated aspects of a workflow process for each item. The engine is implemented in server-side PL/SQL and is activated whenever a call to a workflow procedure or function is made. Since the engine is embedded inside the Oracle database server, if the Workflow server goes down for any reason, the Oracle database server is able to manage the recovery and transactional integrity of any workflow transactions that were running at the time of the failure.

Additionally, Workflow Engines can be set up as background tasks to perform activities that are too costly to execute in real time.

The Workflow Engine performs the following services for a client application:

The state of a workflow item is defined by the various states of all activities that are part of the process for that item. The engine changes activity states in response to an API call to update the activity. The API calls that update activity states are:

Based on the result of a previous activity, the engine attempts to execute the next activity directly. An activity may have the following status:

Attention: The Workflow Engine traps errors produced by function activities by setting a savepoint before each function activity. If an activity produces an unhandled exception, the engine performs a rollback to the savepoint, and sets the activity to the ERROR status. For this reason, you should never commit within the PL/SQL procedure of a function activity. The Workflow Engine never issues a commit as it is the responsibility of the calling application to commit.

For environments such as database triggers or distributed transactions that do not allow savepoints, the Workflow Engine automatically traps "Savepoint not allowed" errors and defers the execution of the activity to the background engine.

Note: The Oracle database server release 8i and higher offers autonomous transactions. By embedding the pragma AUTONOMOUS_TRANSACTION in your procedure, you can perform commits and rollbacks independently of the main transaction. Oracle treats this as a separate session; as such, you will not have access to any database changes that were made in the main session but are not yet committed. Consequently, you are restricted from updating workflow-specific data in an autonomous transaction; for instance, you cannot set item attributes. You cannot access this data because the item itself has not yet been committed, and because you may have lock contentions with the main session.

Oracle Workflow will not support autonomous commits in any procedure it calls directly. If you need to perform commits, then embed your SQL in a subprocedure and declare it as an autonomous block. This subprocedure must be capable of being rerun. Additionally, note that Oracle Workflow handles errors by rolling back the entire procedure and setting its status to ERROR. Database updates performed by autonomous commits cannot be rolled back, so you will need to write your own compensatory logic for error handling. For more information, see: Autonomous Transactions, Oracle Database Concepts.

         Previous  Next          Contents  Index  Glossary

Copyright © 2002 Oracle Corporation.

All Rights Reserved.