IBM® Business Process Manager V8.0 brings a new redesigned Process Portal, integration with Enterprise Content Management systems, searching and sharing of content between process centers, enhanced governance capabilities, and various other new features to the IBM Business Process Manager V8.0 product.
The following new features improve the functionality offered to process participants in Process Portal:
In IBM BPM V8.0, the Coaches are completely redesigned to contain Coach Views. Coach Views are reusable user interfaces that you can create and customize. Coach Views consist of one or more other Coach Views, data bindings, layout information, and behavior. Because Coach Views are reusable, you can create a library of common user interfaces and behavior that you can use to rapidly develop new Coaches.
For greater flexibility in creating a service flow, Coach Views can broadcast boundary events that you use to connect nodes in the service.
To maintain backwards compatibility, Coaches from previous releases are now called Heritage Coaches. You can continue to use and maintain existing Heritage Coaches, but use Coach Views when creating user interfaces for services.
In V8.0, you now have more options when throwing and catching errors using error events in BPDs, subprocesses, and services (including Advanced Integration Services). You can throw a specific error object by selecting a variable, and you can catch specific errors and map the caught error data to a variable. Improved error-handling capabilities include the option to specify an error code and map to an error type on errors thrown from the flow of a BPD or a service using an error end event. When catching errors, you have the option to filter the errors that are caught by selecting an error from a list of all thrown errors for the linked process, subprocess, or service, using error intermediate events. Also, you can map the error data into a variable by selecting a variable that was previously defined. If you are catching specific errors, you can select the error code, map the error data, or both. Models created in earlier versions follow the behavior of the previous version.
For process instances, you have more flexibility in defining the scope of a terminate end event. You can designate whether all activities in the process instance are ended, even the parent processes. In earlier versions, terminating the entire process instance was the only option. The behavior was not visible when you designed models with terminate end events, and the behavior could not be changed. A new check box that terminates the entire process instance is cleared by default in V8.0 for new models. Therefore, the terminate end event ends activities at the level of the process where you add it, including at the level of subprocesses with lower-level activities. For process models that were created in earlier versions and migrated to V8.0, the previous behavior of terminating all activities in the process instance is preserved, unless you clear the check box. Depending on your needs, you can clear or select the check box.
Enterprise Content Management systems help you manage documents of all types, such as records, images, and web pages. By incorporating the new Enterprise Content Management service into your business processes in IBM Business Process Manager, you can search, view, and store documents on Enterprise Content Management systems.
A business object can be identified as a shared business object, making the business object and its values accessible to other instances at run time.
You can apply a governance process that provides control over the installation of process application snapshots. When this governance is in place on a process application, all requests made from IBM Process Center to install a snapshot of that process application pass through the governance process. The process application snapshot is installed on a process server only after the approvals that are defined in that governance process are completed.
You also can create a governance process that reacts to the status change of a snapshot.
Process documentation now includes rich text content and reference links so that you can attach links to content sources or other sources. The following examples are possible reference links:
This linking capability helps you achieve traceability or provide details about the changes to a business object or service interface.
You can now find assets, such as toolkits, process applications, services, or business objects faster using specific syntax or tags using the Search field.
IBM Business Process Manager Advanced provides the same enterprise service bus capabilities that are available in WebSphere® Enterprise Service Bus. Several new features are added to the mediation flow component, and these features are available only when you deploy to IBM Process Server V8.0.
These mediation flow primitives provide elastic scalability with WebSphere eXtreme Scale, giving you cache content-enhancing connectivity for throttled back ends and large binary data. Key uses include response caching, policy caching, and request persistence. The eXtreme Scale mediation primitives can be used only if WebSphere eXtreme Scale is installed.
Additional invocation-style options are available to control the invocation style for a service without the need to specify additional parameters and, in general, without the need to consider the invocation style that invoked the mediation flow. The new invocation styles are Async with deferred response, Async with callback, and As target.
The XSL Transformation mediation primitive is renamed as the Mapping mediation primitive. You can switch easily between XSLT and Business Object Mapper transformation engines for improved functionality or performance.
You can now reconfigure the database password, as needed, after your database configuration is completed. Reconfiguring the database password provides flexibility if new users take on the database administrator role, or if your company has a policy of changing passwords regularly.
WebSphere optimized local adapter (WOLA) is a resource adapter that is added in V8.0. Using WOLA, you can create and access external services to exchange information with COBOL, PL/I, C, and C++ programs that run on Customer Information Control System (CICS®), IBM Information Management System (IMS™) transaction systems, and batch programs on z/OS®. From the New External Service wizard, you can create services using WOLA that can make outbound calls from Process Server for z/OS to programs that run in an external address space on the same z/OS system. Existing, unchanged CICS programs and IMS transactions are driven using the WOLA CICS link server and the WOLA-over-IMS Open Transaction Manager Access (OTMA) interface. You also can target CICS programs and IMS transactions using the native WOLA APIs Host Service or Receive Request.
You now have more control to design the system retries that you plan for your runtime environment. You can set the retry count on the properties page of your module, or you can change the retry count for more than one module using the Configure Asynchronous Retry Count wizard. When a system error occurs, asynchronous invocations retry until the specified retry count is reached. In earlier versions, modules were created with a retry count of 4. Now new modules are created with a retry count of zero. Modules from earlier versions keep existing retry settings during migration.
Unlike in previous versions, the retry behavior of mediation primitives overrides the asynchronous retry count, even if you do not specify retries. Before this release, the retry logic of these primitives was not integrated with the underlying asynchronous retry logic, so that retries could have happened when you defined no retries, or retries could have happened from both the mediation primitive and the service integration bus destination at the same time. Now the behavior defined in these mediation primitives is honored and overrides the service integration bus destination retry logic.
Because the mediation primitive overrides the asynchronous retry count, failover situations, such as when you have an issue with an application server or a messaging engine, might cause failed event manager messages. In previous releases, it was possible that these messages were handled by the service integration bus destination.
You can now set the map generation type to generate a business object map in addition to an XSLT map.
An interface operation with faults is supported in an Advanced Integration Service.