To become productive a model must be deployed to the Stardust runtime environment. Having finished your work creating the model you can deploy your model in the Stardust Process Workbench - Developer Edition Eclipse environment. For the purposes of this tutorial, the Stardust runtime environment will be limited to the Eclipse environment which includes a database for process audit trail. The audit trail contains all state changes relating to any workflow instance with a time stamp and a link to the resource that initiated the state change. So it stores for instance the information who executed which activity and when.
! Tip: If you are deploying to a production environment outside
of Eclipse, use the console command deploy as described in
chapter Using the Console Command of the
Stardust Command Line Tools.
Before you configure the audit trail make sure that you set up the Apache Tomcat server as described in the section Setting up Embedded Tomcat in Eclipse of the Getting Started chapter.
To configure the audit trail:
! Tip: If the HTTP port 8080 is already used on your computer you can change the value of HTTP in the Ports section to a new port value.


! Tip: The Database to be used drop-down box is used to select the audit trail matching the active project. The Manage existing Audit trail databases link allows you to add or delete databases as needed. If you do not keep the Audit Trail database synchronized to the project you are deploying, Stardust will warn you of inconsistencies later in deployment.
To deploy a model proceed as follows:



To avoid running out of PermGenSpace, we will increase the size of the permanent generation
memory by setting the -XX:MaxPermSize argument in the run configuration for the
server.
To set the server configuration:
-Xms128m -Xmx1024m -XX:MaxPermSize=256m
Figure: Setting Server Arguments for Tomcat
Please note that you might increase the maximum perm size according to your requirements.
Note that the usage of Stardust Portal via the internal Web browser is currently not supported. Set the browser to be used to your default Web browser:
Figure: Set to external Web Browser.
To start the Tomcat server:

With Tomcat now running, the ACMEWorkflowModel process model can be deployed. Most of the time, you will leave Tomcat running as you develop and deploy process models. If Tomcat stops (this is expected after some operations), simply start Tomcat again as described in the previous section.


The Stardust Process engine prompts when a model is introduced or updated. In this initial deployment of the model only the first version is presented. On later deployments, you would have a choice to introduce a new version of the model or overwrite the older version. In most production situations, the update, (not the overwrite) is the appropriate choice.
Note that you see an additional model called Predefined Model in case you deploy more versions of the model. This technical model gets deployed automatically at the time of the first deployment of any workflow model. For details on this model, refer to section Predefined Model of the Case - Grouping Processes chapter of the Concepts section.
In case the deployment is successful, a notification dialog opens to confirm that the model was deployed.
The ACMEWorkflowModel Process Model is now deployed and ready for use.
! Tip: Stardust will test for model consistency when deployment occurs. You may get messages warning of model inconsistencies or mismatches to the current audit trail. In these cases, recheck your process model for any problems in the Problems View. Also, check the current audit Trail database to ensure it matches the XPDL process model you are deploying.