- 1 General Questions
- 2 Supported Platforms
- 3 Bug Reporting & Contributing
- 4 Build, Dependencies & Source Code
- 5 Environment Configuration and Deployment
- 6 Internationalization (i18n)
- 7 Troubleshooting
- 8 Programming
- 9 Other
- 10 References
What is the Eclipse Process Manager / Stardust?
Stardust is Business Process Management (BPM) for Eclipse.
Stardust is a comprehensive Business Process Management Suites (BPMS) which supports the whole process life cycle of process modeling, simulation, execution, monitoring and reporting. It allows you to model business processes in Eclipse, integrate them with your application services and UI components, run them in the process engine and to access workflow functionality and document management capabilities in the browser portal.
Stardust is a mature BPMS who's code base has been developed over more than 10 years. It has a proven track record in the industry with more than 1,500 customers on all 5 continents and is embedded in many solutions in mission critical, partially high throughput environments.
Where did Stardust come from?
Stardust is a code contribution from SunGard based on the Infinity Process Platform (IPP).
Initially the Infinity Process Platform was developed by the CARNOT AG and sold under the name CARNOT. In November 2006 SunGard acquired the CARNOT AG and included the rebranded product in its Infinity technology stack. Since then the BPMS has been widely adopted across SunGard. It is integrated in many of the company's products and solutions and also offered in a SAAS deliver model named Infinity on Demand.
Please see the pages on how to get access and how to install.
Who is building commercial Products based on Stardust?
Stardust is an open source product under the Eclipse Public License and does not include any liability or commercial support. The Infinity Process Platform is Stardust under a commercial license including commercial support provided by SunGard. All of the source code was submitted to Eclipse as open source and the code bases are synchronized on a regular basis.
Offerings that are based on the Infinity Process Platform are currently sold by SunGard, its national distribution partners and by SunGard customers who are reselling IPP-based solutions to their customer base.
Offerings from other vendors which are based on Stardust can only be expected after the open sourcing is completed and the platform is fully availabe under the EPL.
How is Stardust licensed?
Stardust is licensed under the Eclipse Public License. The Infinity Process Platform is Stardust under a commercial license including commercial support provided by SunGard.
How can I get Help?
Please check the knowledge base, contact us via the forum or send an email to the Stardust developer mailing list.
Can I get commercial Support?
SunGard offers commercial support for the Infinity Process Platform which is Stardust under a commercial license including commercial support.
On which Java Versions can I use Stardust?
For details on the Java runtimes supported by the latest Stardust version refer to the Release Notes residing in the Product documentation.
Does Stardust work on IBM's / JRockit JDK?
IBM JDK is supported in combination with the supported IBM Websphere version.
JRockit is supported in combination with the supported Weblogic version.
Also see the answers regarding supported Java runtimes and supported Application Servers.
Which Application Server does Stardust support?
For details on the application servers supported by the latest Stardust version, refer to the Release Notes residing in the Product documentation.
Which Databases does Stardust support?
For details on the database servers and according versions supported by Stardus refer to the Release Notes residing in the Product documentation.
There is no official testing for other databases, but some project have reported that they are successfully using other DBMS or other version, e.g. PostgreSQL.
Which Environments can Stardust run in?
Stardust can be deployed in EJB and Spring environments.The EJB mode requires a full-blown application server with EJB 2 support (JBOSS, Weblogic, Websphere). Currently the Stardust EJB services are only available as EJB 2 services. EJB 3 support is planned. A Spring deployment is supported in all supported application servers and on Apache Tomcat.
Can I use Stardust from a .NET environement?
The complete Stardust Java API functionality is also accessible from .NET via Stardust's Web Service API. The Web Service API has been tested for .NET compatibility and several products and customers projects have implemented custom .NET clients based on this API. Example code is available in the Knowledge Base.
Bug Reporting & Contributing
How do I report a Bug in Stardust?
If you are not completely sure that it is a bug then please consult the forum first. Bugs can be reported via the Eclipse Bugzilla system for Stardust.
How can I search the existing List of Bugs in Stardust?
You can log into Eclipse Bugzilla system for Stardust and browse the reported issues. Also consider consulting the forum.
How can I contribute to Startdust?
Build, Dependencies & Source Code
How can I get the Source Code?
Please see: http://wiki.eclipse.org/Stardust/Source_Code
What Libraries does Stardust depend on?
The Stardust dependencies are managed via Apache Maven. Maven POM files describing the dependencies for your target environment and configuration are can be found in our Maven repository.
Please see Maven page in the Stardust Knowledge Base for more details.
Which Jars do I need for a Java Client?
Please check the product documentation chapter suitable for your environment (Spring, EJB). A Maven project for Spring and EJB client environments can be generated using the Maven archetype.
Environment Configuration and Deployment
In clustered environments, is there an advantage in clustering the JMS services for Stardust?
Stardust uses the internal JMS queues to queue its own work, for example when a process has parallel execution paths or an asynchronous path. In those cases the advantage of clustering the JMS services (i.e. using shared queues) is that another node can pick up the work, so the clustered queues achieve a backend load balancing. More importantly it addresses failover: When one node is processing a transaction that was started by an internal queue message and the node goes down, the message is placed back in the shared queue for another node to pick it up. With local queues the processing would only continue after the original node is brought back up. We definitely recommend using shared queues when you have more than one Stardust instance connected to the same AuditTrail DB.
What languages are supported?
The supported languages are English, German, and simplified Chinese.
Can I get Stardust in another language?
Stardust's portal, error messages and Eclipse plugins are fully i18n and can be translated by anyone into additional languages. One only has to create and translate the corresponding resource bundles. We found a convenient way to do this using the Java.net excelbundle project. A compiled version of excelbundle is available on Maven Central repository. We create a script to make the rountrip (translation-update-review-adjustment) easier. This way also non technical people can work on the translation in an Excel sheet.
The Troubleshooting chapter residing in the Product documentation provides a detailed list with tips, known issues and restrictions. The following sections show a selection.
How can I prevent OutOfMemoryErrors on Tomcat?
The default Tomcat settings (in a RAD scenario) do not allocate much memory. Hence the application may run often out of memory quite frequently. The console log will often throw a java.lang.OutOfMemoryError exception. To avoid this, just add a different memory settings in the Tomcat launch configuration:
Servers > Open Launch Configuration > Arguments > VM arguments > at the very bottom add >
-Xms128m -Xmx1024m -XX:MaxPermSize=256m
On a standalone server (outside eclipse RAD environment) start tomcat using:
set JAVA_OPTS=-Xms128m -Xmx1024m -XX:MaxPermSize=256m
Why is my Completion Method invoked twice?
You most probably defined the completion method once in the application element which is connected to your activity and a second time in the access point of an out data mapping.
To perform the intended operation the application's completion method is invoked first. After the processing is done, the results can be mapped into the process context via outdata mappings. In the access point fields of the outdata mappings one has to select the getter methods which will return the data. When the mappings are executed, the engine will invoke the (getter) methods specified in the access points of the outdata mappings to fetch the values. (If any, it will also evaluate the specified access paths on returned objects.) If the completion method has a return value which should be mapped to a process data, then the access point of the outdata mapping should not use the completion method (again). This would result in a second invocation. To get the reurn value of the completion method use the "return value" access point.
To which Directory is my Web Application published in the RAD Environment?
If you checked the Tomcat installation which you use for the RAD environment, you saw that Eclipse leaves this installation unmodified. The stardust dynamic web project is not published to the server's webapps folder. Instead it is published to a temporary folder which, if you don't change the configuration, is created inside your workspace under:
When troubleshooting it can be helpful to browser this folder to see what Eclipse actually published to the server.
Which activity instances states exist and how do they map to the values in the db?
|Activity Instance State||Value||Description|
|CREATED||0||The activity instance has just been created.|
|APPLICATION||1||The activity instance is currently performed either interactively or automatically.|
|COMPLETED||2||The activity instance has been completed.|
|INTERRUPTED||4||Completion of the activity instance has caused exceptions.|
|SUSPENDED||5||The (interactive) activity instance has been suspended to the worklist of a user, a role or an organization.|
|ABORTED||6||The activity instance has been aborted directly by a user or implicitly by aborting the process instance.|
|HIBERNATED||7||The activity instance has an asynchronous receiving part and is hibernated * to wait for an awakening event.|
|ABORTING||8||The activity instance is in progress of being aborted directly by a user.|
How to change the default Port on Oracle XE?
c:\> sqlplus '/ as sysdba'
SQL> EXEC DBMS_XDB.SETHTTPPOR SQL> SELECT dbms_xdb.gethttpport, dbms_xdb.getftpport FROM dual;T(8081);
How can I browse the Jackrabbit Repository Contents?
The easiest way to browse the Jackrabbit repository is to use WebDAV and configure the repository as a network drive under Windows. You will need to ensure that the WebDAV servlets (Simple WebDAV and JCR over WebDAV) are configured properly in the web.xml of your web application. You can then use a command like the following to map the repository as a network drive:
net use t: "http://<hostname>:<port>/<context-root>/jackrabbit/repository/default/ipp-repository" /User:<jcr-user> <jcr-password>
The <jcr-user> and <jcr-password> can be obtained by looking up the values specified in the jackrabbit-jcr-context.xml file.
To browse the Jackrabbit tree and look up the node property values you can use a tool like JCR Browser (http://sourceforge.net/projects/jcrbrowser). This tool provides options to connect to the repository over RMI/WebDAV etc.
How can I configure the DMS / Content Repository / Jackrabbit to ...?
For Jackrabbit configuration, tuning options and more please see the Jackrabbit wiki.