HCL Workload Automation, Version 9.4

Structure of custom job-type plug-in

Provides an overview of the structure of a custom job-type plug-in.

When you create a custom job-type plug-in project in the Integration Workbench, you see something similar to the following figure:
Screen capture of the Integration Workbench displaying the contents of a new custom job-type project available in the Package Explorer view.
The following sections give some outline information about how to modify the project for your needs.
<ProjectName>Action.java

Screen capture of the <ProjectName>Action.java file
This file contains the classes that control the job validation. As can be seen from the screen capture, you can code validation to be performed when the job is defined (typically, semantic validation) and when the job is run (for example checking that a file exists).

<ProjectName>ApplicationDescriptor.java
In the application descriptor, you can manually add the category property to describe the type of plug-in.
The property must have the ID of a predefined category, or a user-defined value. The IDs of the predefined plug-in categories are:
  • native
  • ERP
  • database
  • file_transfer
  • cloud
  • system_management
  • business_analytics

If your plug-in does not fall into any of the predefined categories, you can enter a new category and it will be added to the list of plug-in types in the Workload Designer section of the Dynamic Workload Console.

The following example shows the application descriptor of a plug-in that falls within the database category:
EJB Application descriptor service provider
com.ibm.scheduling.agent.database.jobexecutor.DatabaseApplicationDescriptor
application=database
factory=com.ibm.scheduling.agent.database.jobexecutor.DatabaseJobExecutor
FactorysupportedWorkstations=agent,pool,d-pool
supportedOS=UNIX,WNT,OTHER,IBM_i
isJobStoppable=true
producesJobOutput=true
category=database
schemaLocation=com/ibm/scheduling/resources/xsd/JSDL-DATABASE.xsd
namespace=http://www.ibm.com/xmlns/prod/scheduling/1.0/jsdldatabase

version=${agent.executors.ver}
label=Database
Father of <ProjectName>Action.java
Contains various useful material, such as:
  • The logger
  • A method for discovering the job log name
  • The possibility of customizing the success message label
  • The possibility to change the exit codes
  • Methods for canceling the job
com.<CompanyName>.<ProjectName>.msg
Contains the properties file where you add the messages. Messages can be localized by adding the locale code.
JSDL-<PROJECTNAME>.xsd
Contains the XML schema of the panel you created. It is not normally necessary to edit the schema, but it can be a useful point of reference.
MANIFEST.MF
This file records the version of your plug-in project.
labels.properties
Where you assign the name of your job type, as you want it to be displayed on the Create Workload Definition ► New ► Job Definition drop-down in the Dynamic Workload Console. Localized versions can also be specified here, and are automatically used when the console is in use on systems with the corresponding locale set.

Developing the plug-in

During the process of development, you might decide to modify the input data panel. In this event, you can use the option to regenerate the plug-in using the modified panel. Any coding you added is treated in the following way:
  • If you removed a field from the panel, you must manually delete any associated coding that you added.
  • If you modified a field in the panel, you must check if any associated coding still applies and amend it if not.
  • If you added fields, supply any additional coding required.

Testing the plug-in

A test project is automatically created when you create the plug-in project. This contains a file called job.xml, which can be launched as a Java™ application to run a HCL Workload Automation job. Alternatively, the file can be used as a job definition template in composer.