IBM WebSphere WebSphere Integration Developer Version 6.1

Migration Guide

Version 6 Release 1

Note

Before using this information and the product it supports, read the information in Notices.

This edition applies to version 6, release 1, of WebSphere Integration Developer.

Copyright International Business Machines Corporation 2005, 2007. All rights reserved.

Contents

Chapter 1. Migrating to WebSphere Integration Developer
Chapter 2. Migrating from previous versions of WebSphere Integration Developer
Considerations for the version to version migration process
Development and deployment version levels
Chapter 3. Migrating to WebSphere Process Server from WebSphere InterChange Server
Supported migration paths for WebSphere InterChange Server
Preparing for migration from WebSphere InterChange Server
Considerations for the WebSphere InterChange Server migration process
Considerations: General development
Considerations: Common code utilities
Considerations: Database connection pools
Considerations: Business objects
Considerations: Collaboration templates
Considerations: Maps
Considerations: Relationships
Considerations: Access framework clients
Considerations: Preventing database collisions
Considerations: Post-migration
Migrating WebSphere InterChange Server using the Migration wizard
Verifying the WebSphere InterChange Server migration
Working with migration failures from WebSphere InterChange Server
WebSphere InterChange Server artifacts handled by the migration tools
Supported WebSphere InterChange Server APIs
Mapping the WebSphere Process Sever DataObject from WebSphere InterChange Server XML
Chapter 4. Migrating to WebSphere Integration Developer from WebSphere MQ Workflow
Preparing for migration from WebSphere MQ Workflow
Migrating WebSphere MQ Workflow using the Migration wizard
Verifying the WebSphere MQ Workflow migration
Limitations of the migration process (from WebSphere MQ Workflow)
Chapter 5. Migrating source artifacts to WebSphere Integration Developer from WebSphere Studio Application Developer Integration Edition
Supported migration paths for migrating source artifacts
Preparing source artifacts for migration
Considerations for the source artifact migration process
Migrating service projects using the WebSphere Integration Developer Migration wizard
Migrating service projects using WSADIEServiceProjectMigration
Completing the migration of the application
Creating SCA Components and SCA Imports for the services in the application for rewiring
Migrating a Java service
Migrating an EJB service
Migrating a Business Process to Business Process Service Invocation
Migrating a Web Service (SOAP/JMS)
Migrating a Web Service (SOAP/HTTP)
Migrating a JMS service
Migrating a J2C-IMS service
Migrating a J2C-CICS ECI service
Migrating a J2C-CICS EPI service
Migrating a J2C-HOD service
Migrating a transformer service
The consumption scenario for service migration
Creating SCA Exports to access the migrated service
Migrating the EJB and the EJB process bindings
Migrating the JMS and the JMS process bindings
Migrating the IBM Web Service binding (SOAP/JMS)
Migrating the IBM Web Service binding (SOAP/HTTP)
Migrating the Apache Web Service binding (SOAP/HTTP)
Migrating to the SCA programming model
Migrating WSIFMessage API calls to SDO APIs
Migrating WebSphere Business Integration Server Foundation client code
Migrating WebSphere Business Integration Server Foundation BPEL Java snippets
Migrating interactions with WebSphere Business Integration Adapters
Migrating WSDL interfaces that have SOAP-encoded array types
Migrating WebSphere Business Integration EJB projects
Manually deleting 5.1 Web Services Invocation Framework (WSIF) definitions
Verifying the source artifact migration
Working with source artifact migration failures
Limitations of the migration process (for source artifact migration)
Notices
Terms of use

Chapter 1. Migrating to WebSphere Integration Developer

WebSphere Integration Developer Version 6.1 provides the necessary tools to migrate your existing environment.

Note: WebSphere Integration Developer 6.0.2 and 6.1 projects cannot be used in WebSphere Integration Developer 6.0.1. Once you upgrade to WebSphere Integration Developer 6.0.2 or 6.1, you cannot take projects back to WebSphere Integration Developer 6.0.1.x. Support is also not available if you are a 6.0.2 or 6.1 user checking your code into a repository or exporting projects and then sharing them with a WebSphere Integration Developer 6.0.1 user.

The following topics describe concepts, reference information, and step-by-step instructions for migrating to WebSphere Integration Developer:

Chapter 2. Migrating from previous versions of WebSphere Integration Developer

Migrating from previous versions of WebSphere Integration Developer to WebSphere Integration Developer 6.1 is supported. This is referred to as version-to-version migration.

Migrating to WebSphere Integration Developer 6.1 preserves the basic structure of your existing application with minimal required reconfiguration. Note that the version-to-version process does not have a migration wizard associated with it.

The following topics provide further guidance on the WebSphere Integration Developer version-to-version migration process:

Considerations for the version to version migration process

When migrating from a previous version of WebSphere Integration Developer to V6.1, most of the migration is done for you. However, there are a number of considerations to be aware of that may require some additional manual configuration.

The following considerations are intended to assist in the version-to-version migration process:

Adapters
Applications using a previous level of adapter can be quickly migrated to the current level with a utility. For more information, see the "Migrating applications using previous adapter levels" topic in the related link section below.

CEI events
The WebSphere Integration Developer 6.0.2 monitor models cannot be newly created but can be used in WebSphere Integration Developer 6.1 if they've already been created.

Event definition editor
This editor is deprecated in WebSphere Integration Developer 6.1.

XSD Federation
XSD Federation (that is, xsd-includes generation) has been removed in WebSphere Integration Developer 6.1. Therefore, any old PIs must have any xsd-includes needed included in the PI. This happens automatically for PIs exported in WebSphere Integration Developer 6.0.2.x. However, to export PIs from 6.0.1.2, you must explicitly enable the check box Include derived files when performing the export.

XSL Transformation primitives
In WebSphere Integration Developer 6.1, the XSL Transformation primitive has a new XML mapping editor. XML maps that were built in a previous version must be migrated to the new format before you can edit them. For more information, see the "Migrating an XSL Transformation primitive" topic in the related link section below.

Development and deployment version levels

Your decision about what version levels you need in your environment will depend on the version levels with which your applications were developed.

WebSphere Process Server 6.1 and WebSphere Integration Developer 6.1 are compatible with prior releases as follows:

Note: For i5/OS systems, there are no previous installed versions.

Chapter 3. Migrating to WebSphere Process Server from WebSphere InterChange Server

Migration from WebSphere InterChange Server to WebSphere Process Server is supported through the following functions in WebSphere Integration Developer.

Note: Refer to the release notes for information concerning limitations related to migration in this release of WebSphere Process Server.

Even though migration of source artifacts is supported, it is recommended that extensive analysis and testing be done to determine if the resulting applications will function as expected in WebSphere Process Server, or if they will need post-migration redesign. This recommendation is based on the following limitations in functional parity between WebSphere InterChange Server and this version of WebSphere Process Server. There is no support in this version of WebSphere Process Server that is equivalent to these WebSphere InterChange Server functions:

Supported migration paths for WebSphere InterChange Server

WebSphere Process Server migration tools support migration from WebSphere InterChange Server versions 4.2.2 or later.

Any WebSphere InterChange Server release prior to Version 4.2.2 will first need to migrate to version 4.2.2 or 4.3 before migrating to WebSphere Process Server.

Preparing for migration from WebSphere InterChange Server

Before migrating to WebSphere Process Server from WebSphere InterChange Server, you must first ensure that you have properly prepared your environment. WebSphere Process Server provides the necessary tools to migrate from WebSphere InterChange Server.

These migration tools can be invoked from:

Input to the migration tools is a repository jar file exported from WebSphere InterChange Server. Therefore, before accessing the migration tools through any of these options, you must first:

  1. ensure that you are running a version of WebSphere InterChange Server that can be migrated to WebSphere Process Server. See the topic "Supported migration paths for WebSphere InterChange Server".
  2. export your source artifacts from WebSphere InterChange Server into a repository jar file using the WebSphere InterChange Server repos_copy command as described in the documentation for WebSphere InterChange Server. The wizard requires as input a WebSphere InterChange Server repository JAR file. This JAR file should be self-contained with respect to the applications being migrated. That is, all artifacts referenced by any of the artifacts in the JAR file must also be contained in the JAR file. To ensure that the repository JAR file that will be generated is self-contained, run the repos_copy command with the -vr option before exporting the server repository (this validates the repository). If the repository is valid then repos_copy writes the following output to the console: Validation Succeeded. All Dependencies Resolved. If the repository is not valid then repos_copy prints a list of the dependencies that must be resolved. Resolve the dependencies prior to exporting the repository. Export the repository artifacts and create the repository JAR file, using the WebSphere InterChange Server repos_copy command with the -o option (See the WebSphere InterChange Server 4.3 documentation for more details, including how to export individual components).

Considerations for the WebSphere InterChange Server migration process

The following considerations are intended to assist in the development of integration artifacts for WebSphere InterChange Server. By adhering to these guidelines, you can ease the migration of WebSphere InterChange Server artifacts to WebSphere Process Server.

These recommendations are meant to be used only as a guide. There may be cases where it is necessary to deviate from these guidelines. In these cases care should be taken to limit the scope of the deviation to minimize the amount of rework required to migrate the artifacts. Note that the guidelines outlined here are not all general recommendations for the development of WebSphere InterChange Server artifacts. They are instead limited in scope to those considerations which may affect the ease in which artifacts can be migrated at a future time.

Considerations: General development

There are several considerations which apply broadly to most of the integration artifacts. In general, the artifacts which leverage the facilities provided by the tooling and conform to the metadata models enforced by the tooling will migrate most smoothly. Also, artifacts with significant extensions and external dependencies are likely to require more manual intervention when migrating.

The following list summarizes the considerations for general development of WebSphere InterChange Server based solutions to help ease future migration:

It is important for integration solutions to adhere to the programming model and architecture provided by WebSphere InterChange Server. Each of the integration components within WebSphere InterChange Server plays a well-defined role within the architecture. Significant deviations from this model will make it more challenging to migrate content to the appropriate artifacts on WebSphere Process Server.

Another general practice which will improve the success of future migration projects is to document the system design. Be sure to capture the integration architecture and design, including functional design and quality of service requirements, the interdependencies of artifacts shared across projects, and also the design decisions that were made during the deployment. This will assist in system analysis during migration, and will minimize any rework efforts.

For creating, configuring, and modifying artifact definitions, use only the development tooling provided. Avoid manual manipulation of artifact metadata (for example, editing XML files directly), which may corrupt the artifact for migration.

Follow these guidelines when developing Java code within collaboration templates, maps, common code utilities, and other components:

Use only the APIs published in the WebSphere InterChange Server product documentation for the artifacts. These are outlined in detail in the WebSphere InterChange Server development guides. Compatibility APIs will be provided in WebSphere Process Server for published WebSphere InterChange Server APIs. Although WebSphere InterChange Server has many internal interfaces which you might wish to use, this practice is discouraged because these interfaces are not guaranteed to be supported in the future.

When designing business logic and transformation rules in maps and collaboration templates, try to avoid field developed common code utility libraries, included as a Java archive (*.jar) file in the classpath of WebSphere InterChange Server, as these will need to be migrated manually.

Use the Activity Editor tool to the greatest extent possible. This will ensure that the logic is described through metadata which can more readily be converted to the new artifacts. For operations that you wish to reuse in the tooling, use the "My Collections" feature of the Activity Editor wherever possible.

In any Java code snippets that may need to be developed, the code be as simple and atomic as possible. The level of sophistication in the Java code should be on the order of scripting, involving basic evaluations, operations, and computations, data formatting, type conversions, and so forth. If more extensive or sophisticated application logic is required, consider using EJBs running in WebSphere Application Server to encapsulate the logic, and use web service calls to invoke it from WebSphere InterChange Server. Use standard JDK libraries rather than third party or external libraries which would need to be migrated separately. Also, collect all related logic within a single code snippet, and avoid using logic where connection and transaction contexts span multiple code snippets. With database operations, for example, code related to obtaining a connection, beginning and ending a transaction, and releasing the connection should be in one code snippet.

In general, ensure that code which is designed to interface with an Enterprise Information System (EIS) is placed within adapters, and not within maps or collaboration templates. This is generally a recommended practice for architecture design. Also, this will help avoid prerequisites for third party libraries and related considerations within the code, such as connection management and possible Java Native Interface (JNI) implementations.

Make the code as safe as possible by using appropriate exception handling. Also make the code compatible to run within a J2EE application server environment, even though it is currently running within a J2SE environment. Adhere to J2EE development practices, such as avoiding static variables, spawning threads, and disk I/O. While these are generally good practices to adhere to, they can improve portability.

Considerations: Common code utilities

Avoiding the development of common code utility libraries for use across integration artifacts within the WebSphere InterChange Server environment is recommended. Where code reuse across integration artifacts is necessary, leveraging the "My Collections" feature of the Activity Editor tool is recommended. Also, consider using EJBs running in WebSphere Application Server to encapsulate the logic, and use Web service calls to invoke them from WebSphere InterChange Server.

While it is possible that some common code utility libraries may execute appropriately on WebSphere Process Server, you will be responsible for the migration of the custom utilities.

Considerations: Database connection pools

A WebSphere InterChange Server database connection pool within a map or collaboration template will be rendered as a standard JDBC resource in WebSphere Process Server. However, the way connections and transactions are managed might differ between WebSphere InterChange Server and WebSphere Process Server, so avoid keeping database transactions active across Java snippets.

User-defined database connection pools are useful within maps and collaboration templates for simple data lookups and for more sophisticated state management across process instances. A database connection pool in WebSphere InterChange Server will be rendered as a standard JDBC resource in WebSphere Process Server, and the basic function will be the same. However, the way connections and transactions are managed may differ.

To maximize future portability, avoid keeping database transactions active across Java snippet nodes within a collaboration template or map. For example, code related to obtaining a connection, beginning and ending a transaction, and releasing the connection should be in one code snippet.

Considerations: Business objects

For the development of business objects, use only the tooling provided to configure artifacts, use explicit data types and lengths for data attributes, and use only the documented APIs.

Business objects within WebSphere Process Server are based on Service Data Objects (SDOs). SDOs use data attributes that are strongly typed. For business objects in WebSphere InterChange Server and adapters, data attributes are not strongly typed, and users sometimes specify string data types for non-string data attributes. To avoid issues in WebSphere Process Server, explicitly specify data types.

Because business objects within WebSphere Process Server might be serialized at runtime as they are passed between components, it is important to be explicit with the required lengths for data attributes to minimize utilization of system resources. For this reason, do not use the maximum 255 character length for a string attribute, for example. Also, do not specify zero length attributes which currently default to 255 characters. Instead, specify the exact length required for attributes.

XSD NCName rules apply to business object attribute names in WebSphere Process Server. Therefore, do not use any spaces or ":" in names for business object attributes. Business object attribute names with spaces or ":" are invalid in WebSphere Process Server. Rename business object attributes before migration.

If using an array in a business object, you cannot rely on the order of the array when indexing into the array in Maps and/or Relationships. The construct that this migrates into in WebSphere Process Server does not guarantee index order, particularly when entries are deleted.

It is important to use only the Business Object Designer tool to edit business object definitions, and to use only the published APIs for business objects within integration artifacts.

Considerations: Collaboration templates

Many of the guidelines that have already been described apply to the development of collaboration templates.

To ensure processes are described appropriately with metadata, always use the Process Designer tool for the creation and modification of collaboration templates, and avoid editing the metadata files directly. Use the Activity Editor tool wherever possible to maximize the use of metadata to describe the required logic.

To minimize the amount of manual rework that may be required in migration, use only the documented APIs within collaboration templates. Avoid the use of static variables. Instead, use non-static variables and collaboration properties to address the requirements of the business logic. Avoid the use of Java qualifiers final, transient and native in Java snippets. These cannot be enforced in the BPEL Java snippets that are the result of migrating the Collaboration Templates.

To maximize future portability, avoid using explicit connection release calls and explicit transaction bracketing (that is, explicit commits and explicit rollbacks) for User Defined Database Connection Pools. Instead, make use of the container-managed implicit connection clean-up and implicit transaction bracketing. Also, avoid keeping system connections and transactions active across Java snippet nodes within a collaboration template. This applies to any connection to an external system, as well as user-defined database connection pools. Operations with an external EIS should be managed within an adapter, and code related to database operation should be contained within one code snippet. This may be necessary within a collaboration which, when rendered as a BPEL business process component may be selectively deployed as an interruptible flow. In this case, the process may be comprised of several separate transactions, with only state and global variable information passed between the activities. The context for any system connection or related transaction which spanned these process transactions would be lost.

Name collaboration template property names in accordance with W3C XML NCName naming conventions. WebSphere Process Server accepts names conforming to those conventions. Any disallowed characters are invalid in BPEL property names that they will be migrated into. Rename properties to remove any disallowed characters before migrating to avoid syntactical errors in the BPEL generated by migration.

Do not reference variables using "this." For example, Instead of "this.inputBusObj" use just "inputBusObj"

Use class-level scoping on variables instead of scenario-scoped variables. Scenario-scoping is not carried forward during migration.

Initialize all variables declared in Java snippets with a default value: "Object myObject = null;" for example. Be sure all variables are initialized during declaration before migrating.

Ensure that there are no Java import statements in the user modifiable sections of your collaboration templates. In the definition of the collaboration template, use the import fields to specify Java packages to import.

Do not set incoming business object values to be stored in the triggeringBusObj variable. Within WebSphere InterChange Server, the triggeringBusObj is read-only and its values cannot be overwritten, so any incoming business object values will not be saved. If the triggeringBusObj is used as the receiving variable for an incoming business object on an inbound service call, then after migration the behavior of the inbound service call will be different: within the BPEL process, the incoming value from the inbound service call will overwrite the value stored in triggeringBusObj.

Considerations: Maps

Many of the guidelines that have already been described for collaboration templates also apply to maps.

To ensure maps are described appropriately with metadata, always use the Map Designer tool for the creation and modification of maps, and avoid editing the metadata files directly. Use the Activity Editor tool wherever possible to maximize the use of metadata to describe the required logic.

When referencing child business objects in a map, use a submap for the child business objects.

Avoid using Java code as the "value" in a SET since that is not valid in WebSphere Process Server. Use constants instead. For example, if the set value is "xml version=" + "1.0" + " encoding=" + "UTF-8" this will not validate in WebSphere Process Server. Instead, change it to "xml version=1.0 encoding=UTF-8" before you migrate.

To minimize the amount of manual rework that may be required in migration, use only the documented APIs within maps. Avoid the use of static variables. Instead, use non-static variables. Avoid the use of Java qualifiers final, transient and native in map custom code.

If using an array in a business object, do not rely on the order of the array when indexing into the array in maps. The construct that this migrates into in WebSphere Process Server does not guarantee index order, particularly when entries are deleted.

To maximize future portability, avoid using explicit connection release calls and explicit transaction bracketing (that is, explicit commits and explicit rollbacks) for User Defined Database Connection Pools. Instead, make use of the container-managed implicit connection clean-up and implicit transaction bracketing. Also, avoid keeping system connections and transactions active in custom map steps across transformation node boundaries. This applies to any connection to an external system, as well as user-defined database connection pools. Operations with an external EIS should be managed within an adapter, and code related to database operation should be contained within one custom step.

Do not use inner classes in your maps. The migration command (reposMigrate) does not migrate inner classes and you will receive errors if your maps contain them. In a WebSphere InterChange Server repository, an inner class could be defined in a node and referenced by other nodes within the same collaboration template. In WebSphere Process Server, an inner class defined in a BPEL component cannot be used by other components. Due to this limitation, inner classes are not translated and must be dealt with manually. Recommended changes include packaging the inner class code in a library as an external class, or removing the inner class declaration, resolving any errors, and placing the code as needed throughout the BPEL.

Considerations: Relationships

For relationships, remember that while relationship definitions will be able to be migrated for use in WebSphere Process Server, the relationship table schema and instance data may be reused by WebSphere Process Server, and also shared concurrently between WebSphere InterChange Server and WebSphere Process Server.

For relationships, use only the tooling provided to configure the related components, and use only the published APIs for relationships within integration artifacts.

Use only the Relationship Designer tool to edit relationship definitions. In addition, allow only WebSphere InterChange Server to configure the relationship schema, which is generated automatically upon deployment of relationship definitions. Do not alter the relationship table schema directly with database tools or SQL scripts.

If you must manually modify relationship instance data within the relationship table schema, be sure to use the facilities provided by the Relationship Manager.

Use only the published APIs for relationships within integration artifacts.

Considerations: Access framework clients

Do not develop any new clients adopting the CORBA IDL interface APIs. This will not be supported in WebSphere Process Server.

Considerations: Preventing database collisions

You can prevent database collisions from occurring by scheduling events to occur at least two seconds apart.

If a migrated application causes multiple events to occur at the same time to WebSphere Business Integration components, database collisions, or deadlocks, could occur. These occur when the WebSphere Process Server Application Scheduler (AppScheduler) schedules multiple events to occur at exactly the same time. When a deadlock occurs, the event that caused it is rolled back and attempted again as soon as possible. This cycle continues until each of the threads attempting to access the database successfully updates it.

For example:

AppScheduler E com.ibm.wbiserver.scheduler.AppSchedulerMB process CWLWS0021E: 
The AppSchedulerMB.process method has generated an exception.
WSRdbXaResour E DSRA0304E: XAException occurred. XAException contents and details are:
The DB2 Error message is : Error executing a XAResource.end(), Server returned
XA_RBDEADLOCK The DB2 Error code is : -4203
The DB2 SQLState is : null  

To prevent this from occurring, schedule the events to occur far enough apart so that deadlocks do not occur. Schedule events to occur at least two seconds apart, however, the amount of time you need will vary depending on other factors in your environment that affect performance such as database size, hardware, connection speed and other factors.

Considerations: Post-migration

When applications have been migrated from WebSphere InterChange Server to WebSphere Integration Developer or WebSphere Process Server, special attention is required in some areas to enable migrated applications to function in WebSphere Integration Developer and WebSphere Process Server consistently with their intended function due to differences with the architecture WebSphere InterChange Server.

For more information regarding post-migration considerations, see Post-migration considerations.

Migrating WebSphere InterChange Server using the Migration wizard

You can use the WebSphere Integration Developer Migration wizard to migrate your existing WebSphere InterChange Server artifacts.

To use the Migration wizard to migrate your WebSphere InterChange Server artifacts, follow these steps:

  1. Invoke the wizard by selecting File -> Import -> Business Integration -> WebSphere InterChange Server JAR File and click Next:
    Import selection for WebSphere InterChange Server JAR File
    OR you can also open the Migration wizard from the Welcome page by clicking the Returning Users icon Returning users to open the Returning Users page (note that you can always return to the Welcome page by clicking Help -> Welcome ):
    Returning users page
    Click Migration on the left side of the Returning Users page to open the Migration page. From the Migration page, select the Migrate a WebSphere ICS repository option Migration page with the Migrate a WebSphere ICS repository option selected.
  2. The Migration wizard opens. Enter the name of the source file in the Source selection field by clicking the Browse button and navigating to the file. Enter the library name in the relevant field. If the shared library does not currently exist in the workspace, it must be created by clicking New.... Click Next:
    ICS migration wizard
  3. The Migration Options window opens. From here you can accept the migration defaults or select a check box to change the option:
    Migration options
    The following table outlines the available migration options:
    Option Description
    Warn in case of Java parsing errors (Optional) By default, migration of an individual artifact will fail if a Java conversion problem is found. If this option is set, all Java conversion problems will be treated as warnings only, and the artifact will be migrated as well as possible.
    Stop the migration on the first failure (Optional) By default the Migration wizard will continue processing the remaining artifacts in the JAR file if an error occurs during the processing of a certain artifact. If this option is set, the processing will stop as soon as an error is detected. The artifact with the error, and all subsequent artifacts, are not processed.
    Use collaboration template loop unraveling (Optional) Requests that all loops present in a Collaboration Template be maintained. If this option is not present, the default is for migration to use loop unraveling.
    Enable event sequencing (Optional) Requests Event Sequencing be enabled for all Asynchronous WSDL methods. If this option is not present, the default is for migration to not enable event sequencing on any WSDL methods.
    Use default template (Optional) Requests that all assembly editor templates in the specified directory be loaded and used for XML to Java conversion. The default for this property is for only the Standard Assembly Editor Template v4.3.3 to be used for XML to Java conversion.
  4. Click Finish.

A progress bar at the bottom of the migration dialog indicates the progress of the migration. Once the migration process has completed, the migration dialog disappears and the Migration Results window opens.

Verifying the WebSphere InterChange Server migration

If no errors are reported during the migration of the WebSphere InterChange Server jar file, then the migration of the artifacts was successful. If the migration has not completed successfully, a list of errors, warnings, and/or informational messages will be displayed. You can use these messages to verify the WebSphere InterChange Server migration.

Note: Due to the complexity of migrating from WebSphere InterChange Server to WebSphere Process Server, you are strongly urged to perform extensive testing of the resulting applications running in WebSphere Process Server to ensure they function as expected, before putting them into production.

The following page appears if migration messages were generated during the migration process:

Migration results window

In the Migration Results window, you can see the migration messages that were generated during the migration process. By selecting a message from the upper Message list, you can find more information regarding that message in the lower Message Description window.

To keep all messages for future reference, click the Generate ToDo's button to create a list of "ToDo" tasks in the task view and/or click the Save as... button to save the messages in a text file in the file system. To see the generated To Do's, click Window -> Show View -> Other... -> General -> Tasks and click OK. The Tasks view opens with the list of generated To Do's from the migration process.

Working with migration failures from WebSphere InterChange Server

If your migration from WebSphere InterChange Server fails, there are a two ways in which to deal with the failures.

Note: You may prefer the first option as you will initially be more familiar with WebSphere InterChange Server. However, as you become more experienced with WebSphere Process Server and its new artifacts, you may choose to repair the migrated artifacts in WebSphere Integration Developer.
  1. If the nature of the error permits, you can adjust the WebSphere InterChange Server artifacts using the WebSphere InterChange Server toolset, and export the jar file again and retry the migration.
  2. You can fix any errors in the resulting WebSphere Process Server artifacts by editing the artifacts in WebSphere Integration Developer.

WebSphere InterChange Server artifacts handled by the migration tools

The migration tools can automatically migrate some of the WebSphere InterChange Server artifacts.

The following artifacts can be migrated:

The migration tools will create a Jython script that can be used with the wsadmin command line tool to configure resources in the WebSphere Process Server for the following WebSphere InterChange Server artifacts/resources:

The migration tools do not handle the following WebSphere InterChange Server artifacts:

Supported WebSphere InterChange Server APIs

In addition to the WebSphere InterChange Server source artifact migration tools provided in WebSphere Process Server and WebSphere Integration Developer, there is also support for many of the APIs that were provided in WebSphere InterChange Server. The migration tools work in conjunction with these WebSphere InterChange Server APIs by preserving your custom snippet code as much as possible when migrating.

Note: These APIs are provided only to support migrated WebSphere InterChange Server applications until they can be modified to use new Process Server APIs.

The supported WebSphere InterChange Server APIs in Process Server are listed below. These APIs provide functions in WebSphere Process Server similar to the function that they provide in WebSphere InterChange Server. See the WebSphere InterChange Server documentation for a functional description of these APIs.

CwBiDiEngine
AppSide_Connector/

JavaConnectorUtil
AppSide_Connector/

JavaConnectorUtilDH
datahandler/
wbi/
ibm/
com/

BusObj
Collaboration/

BusObjArray
Collaboration/

BaseDLM
DLM/

CwDBConnection
CwDBConnection/
CxCommon/

CwDBConstants
CwDBConnection/
CxCommon/

CwDBStoredProcedureParam
CwDBConnection/
CxCommon/

DataHandler (Abstract Class)
DataHandlers/
crossworlds/
com/

NameHandler (Abstract Class)
DataHandlers/
crossworlds/
com/

ConfigurationException (extends java.lang.Exception)
Exceptions/
DataHandlers/
crossworlds/
com/

MalformedDataException (extends java.lang.Exception)
Exceptions/
DataHandlers/
crossworlds/
com/

NotImplementedException (extends java.lang.Exception)
Exceptions/
DataHandlers/
crossworlds/
com/

BusinessObjectInterface
CxCommon/

CxObjectAttr
CxCommon/

CxObjectContainerInterface
CxCommon/

DtpConnection
Dtp/
CxCommon/

DtpDataConversion
Dtp/
CxCommon/

DtpDate
Dtp/
CxCommon/

DtpMapService
Dtp/
CxCommon/

DtpSplitString
Dtp/
CxCommon/

DtpUtils
Dtp/
CxCommon/

BusObjInvalidVerbException (extends InterchangeExceptions)
Exceptions/
CxCommon/

IdentityRelationship
relationship/
utilities/
crossworlds/
com/

MapExeContext
Dtp/
CxCommon/

Participant
RelationshipServices/
Server/

Relationship
RelationshipServices/
Server/

UserStoredProcedureParam
Dtp/
CxCommon/

BaseCollaboration
Collaboration/

CxExecutionContext
CxCommon/

CollaborationException
Collaboration/

Filter
crossworlds/
com/

Globals
crossworlds/
com/

SmartCollabService
crossworlds/
com/

StateManagement
crossworlds/
com/

EventKeyAttrDef
EventManagement/
CxCommon/

EventQueryDef
EventManagement/
CxCommon/

FailedEventInfo
EventManagement/
CxCommon/

Mapping the WebSphere Process Sever DataObject from WebSphere InterChange Server XML

If you use the Legacy Adapters to connect to WebSphere Process Server, the following algorithm will enable you to further understand how the WebSphere Process Sever DataObject was created from the WebSphere InterChange Server XML. This information shows where the data values have been placed, and also what data values have been chosen to replace the ones used in WebSphere InterChange Server.

General

Loading

Loading will load a WebSphere InterChange Server runtime XML into a WebSphere Business Integration BusinessGraph AfterImage instance.

Saving

Saving will save a WebSphere Business Integration BusinessGraph AfterImage instance to a WebSphere InterChange Server runtime XML. An exception will be thrown if the input BusinessGraph is not AfterImage.

Attribute processing

Chapter 4. Migrating to WebSphere Integration Developer from WebSphere MQ Workflow

WebSphere Integration Developer provides the necessary tools to migrate from WebSphere MQ Workflow.

The Migration wizard enables you to convert FDL definitions of business processes that you exported from the build time component of WebSphere MQ Workflow into corresponding artifacts in WebSphere Integration Developer. The generated artifacts comprise XML schema definitions for business objects, WSDL definitions, BPEL, import and component definitions, and TEL definitions.

The conversion tool requires a semantically complete FDL definition of a process model that you export from WebSphere MQ Workflow build time with the option export deep. This option ensures that all necessary data, program, and subprocess specifications are included. Also, ensure that any user defined process execution server definitions (UPES) referenced in your WebSphere MQ Workflow process model is also selected when you export FDL from the WebSphere MQ Workflow build time.

Note: The Migration wizard does not cover the migration of:

For more information on migrating using the FDL2BPEL conversion tool, see WebSphere MQ Workflow Support Site.

Preparing for migration from WebSphere MQ Workflow

Before migrating to WebSphere Integration Developer from WebSphere MQ Workflow, you must first ensure that you have properly prepared your environment.

The scope and completeness of mapping depends on how far you adhere to the following guidelines for migration:

The Migration wizard will produce syntactically correct business process editor constructs even for WebSphere MQ Workflow constructs that cannot be migrated (PEA or PES program activities, some dynamic staff assignments, and so on), which need to be manually adapted to executable business process editor artifacts.

The following table outlines the applied mapping rules:

Table 2. Mapping rules
WebSphere MQ Workflow WebSphere Integration Developer
Process Process with execution mode: longRunning; Partner links for inbound and outbound interfaces of process
Source and Sink Variables for process input and process output; Receive activity and reply activity
Program activity Invoke activity
Process activity Invoke activity
Empty activity FMCINTERNALNOOP activity
Block Scope with embedded BPEL activities
Exit condition of activity While activity (enclosing the actual activity)
Start condition of activity Join condition of activity
Staff assignment of activity Human task activity
Input container and output container of activity Variables used to specify the input/output of invoke activity
Control connector; Transition condition Link; Transition condition
Data connector Assign activity
Global data container Variable
Note: You should initially try the migration process with small projects, if possible. The Migration wizard will simplify the conversion of your WebSphere MQ Workflow process models into business process editor process models, but you should be aware that the processes cannot be mapped one-to-one as you are creating a new programming model. The semantic scopes of the underlying process specification languages (FDL and BPEL) share an area of intersection, but they do not overlap in total. Otherwise, you could not expect any new benefits from business process editor. Web services represent a promising new technology that claim replacing deprecated solutions by new ones.

In general, you should always review and possibly modify the generated artifacts. Additional effort may be necessary to either make a successful migration possible or to complete the migration task.

Migrating WebSphere MQ Workflow using the Migration wizard

The Migration wizard enables you to convert FDL definitions of business processes that you exported from the build time component of WebSphere MQ Workflow into corresponding artifacts in WebSphere Integration Developer. The generated artifacts comprise XML schema definitions for business objects, WSDL definitions, BPEL, import and component definitions, and TEL definitions.

Note: The Migration wizard does not cover the migration of:

To use the Migration wizard to migrate your WebSphere MQ Workflow artifacts, follow these steps:

  1. Invoke the wizard by selecting File -> Import -> Business Integration -> WebSphere MQ Workflow FDL File and click Next:
    Import selection for WebSphere MQ Workflow FDL File
    OR you can also open the Migration wizard from the Welcome page by clicking the Returning Users icon Returning users to open the Returning Users page (note that you can always return to the Welcome page by clicking Help -> Welcome ):
    Returning users page
    Click Migration on the left side of the Returning Users page to open the Migration page. From the Migration page, select the Migrate a WebSphere MQ Workflow process option Migration page with the Migrate a WebSphere MQ Workflow process option selected.
  2. The Migration wizard opens. Enter the absolute path and name of the FDL file into the Source selection field or select one from the file system by clicking the Browse button and navigating to the file. Enter the module name in the relevant field (you must enter a module name in the Module name field before you are able to proceed). Click Next:
    FDL migration wizard
  3. The Migration Options page opens. From here you can accept the migration defaults or select a check box to change the option. By selecting the Treat name conflicts as errors check box, you can prevent the automatic addition of suffixes which could result in interoperability errors. The Initialize predefined data members check box adds extra nodes to the process to initialize the predefined data members:
    WebSphere MQ Workflow Migration Options page
  4. Click Finish.

A progress bar at the bottom of the migration dialog indicates the progress of the migration. Once the migration process has completed, the migration dialog disappears and the Migration Results window opens:

WebSphere MQ Workflow migration results page

Verifying the WebSphere MQ Workflow migration

If the migration completes with a list of errors, warnings, and/or informational messages, they will be displayed in the Migration Results window. Otherwise, the wizard window will close if the migration completed successfully.

The following page appears if migration messages were generated during the migration process:

Migration results window

The Migration Results window lists the migration messages that were generated during the migration process. By selecting a message from the upper Message list, you can find more information regarding that message in the lower Message Description window.

To keep all messages for future reference, click the Generate ToDo's button to create a list of "ToDo" tasks in the task view and/or click the Save as... button to save the messages in a text file in the file system. To see the generated To Do's, click Window -> Show View -> Other... -> General -> Tasks and click OK. The Tasks view opens with the list of generated To Do's from the migration process.

Limitations of the migration process (from WebSphere MQ Workflow)

There are certain limitations involved with the WebSphere MQ Workflow migration process.

Chapter 5. Migrating source artifacts to WebSphere Integration Developer from WebSphere Studio Application Developer Integration Edition

Source artifacts can be migrated from WebSphere Studio Application Developer Integration Edition to WebSphere Integration Developer. Migrating the source artifacts in an application involves migrating them to the new WebSphere Integration Developer programming model so that new functionality and features can be used. The application can then be redeployed and installed to the WebSphere Process Server.

Many features available in WebSphere Business Integration Server Foundation 5.1 have moved down into the base WebSphere Application Server 6.x. See this topic for tips on migrating those features: Tips for migrating programming model extensions.

To fully migrate a WebSphere Studio Application Developer Integration Edition service project, there are three fundamental tasks to complete:

  1. Preparing source artifacts for migration. These actions may need to be performed in WebSphere Studio Application Developer Integration Edition.
  2. User either the Migration wizard or the WSADIEServiceProjectMigration command line script to automatically migrate the artifacts to the Business Integration Module project.
  3. Use WebSphere Integration Developer to manually complete the migration. This involves fixing any Java code that could not be automatically migrated and verifying the wiring of the migrated artifacts.

Note: The runtime migration (upgrade path) will not be provided in WebSphere Process Server 6.x, therefore, this source artifact migration path will be the only option for migrating WebSphere Studio Integration Edition service projects in 6.x.

Supported migration paths for migrating source artifacts

Before beginning to migrate source artifacts from WebSphere Studio Application Developer Integration Edition, you should review the supported migration paths that are supported by WebSphere Integration Developer.

The Migration wizard offers the ability to migrate one WebSphere Studio Application Developer Integration Edition Version 5.1 (or above) service project at a time. It will not migrate an entire workspace.

The Migration wizard does not migrate application binaries - it will only migrate source artifacts found in a WebSphere Studio Application Developer Integration Edition service project.

Preparing source artifacts for migration

Before migrating source artifacts to WebSphere Integration Developer from WebSphere Studio Application Developer Integration Edition, you must first ensure that you have properly prepared your environment.

The following steps describe how to prepare your environment before migrating source artifacts to WebSphere Integration Developer from WebSphere Studio Application Developer Integration Edition:

  1. Ensure that you have a backup copy of the entire 5.1 workspace before attempting to migrate.
  2. Review the migration section of the Rational Application Developer Information Center to determine the best way to migrate the non-WBI-specific projects in your workspace: Migrating from WebSphere Studio V5.1, 5.1.1, or 5.1.2
  3. Review the Web service section of the Rational Application Developer Information Center for background information on the Web service functionality provided by Rational Application Developer: Developing Web services
  4. Ensure that you have all of the appropriate WebSphere Integration Developer features enabled. If you don't have these features enabled, you may not see the menu options that will be discussed below. To enable the important features:
  5. Use a new workspace directory for WebSphere Integration Developer. It is not recommended that you open WebSphere Integration Developer in an old WebSphere Studio Application Developer Integration Edition workspace that contains service projects, as those projects must first be migrated to a format that can be read by WebSphere Integration Developer. The recommended steps to do this are:
    1. Copy all non-service projects from the old workspace to the new workspace. Do not copy the 5.1 EJB, Web, and EAR projects created when deploy code was generated for a 5.1 service project. The new 6.x deploy code will be regenerated automatically when the BI module is built.
    2. Open WebSphere Integration Developer in the blank workspace and import all non-service projects by clicking on File -> Import -> General -> Existing Projects into Workspace and select the projects that you have copied over to the new workspace.
      • If the project is a J2EE project, then you should migrate it to the 1.4 level by using the Rational Application Developer Migration wizard:
        1. Right-click the project and select Migration -> J2EE Migration Wizard....
        2. Review the warning statements on the first page and, unless otherwise indicated, click Next.
        3. Ensure that the J2EE project is checked in the Projects list. Leave Migrate project structure and Migrate J2EE specification level checked. Choose J2EE version 1.4 and Target Server WebSphere Process Server v6.1.
        4. Select any other options that might be appropriate for your J2EE project and click Finish. If this step completes successfully, you will see a message Migration finished successfully.
        5. If there are errors in the J2EE project after migration, you should remove all classpath entries that refer to v5 .jar files or libraries and add the JRE System Library and WPS Server Target libraries to the classpath instead (discussed below). This should resolve most, if not all, of the errors.
      • For WebSphere Business Integration EJB projects with Extended Messaging (CMM) or Container Managed Persistence over Anything (CMP/A), the IBM EJB Jar Extension descriptor files must be migrated after the 5.1 project has been imported into the 6.x workspace. See "Migrating WebSphere Business Integration EJB Projects" for more information.
      • Fix the classpath for each non-service project imported in to the workspace. To add the JRE and WebSphere Process Server libraries to the classpath, right-click on the imported project and select Properties. Go to the Java Build Path entry and select the Libraries tab. Then do the following:
        1. Select Add library -> JRE System Library -> Alternate JRE - WPS Server v6.1 JRE -> Finish.
        2. Then select Add library -> WPS Server Target -> Configure wps server classpath -> Finish.
  6. By default, WebSphere Integration Developer generates the deploy code during build time.
  7. In order to fully migrate the .bpel files within a service project, you must ensure that all .wsdl and .xsd files referenced by the .bpel files can be resolved in a business integration project in the new workspace:

You are now ready to begin the migration process.

Considerations for the source artifact migration process

There are a number of considerations for the WebSphere Studio Application Developer Integration Edition source artifact migration process.

The following practices show how to design WebSphere Studio Application Developer Integration Edition services to ensure that they will migrate successfully to the new programming model:

Migrating service projects using the WebSphere Integration Developer Migration wizard

The WebSphere Integration Developer Migration wizard enables the migration of service projects.

Note:

The Migration wizard does the following:

  1. Creates a new business integration module (the module name is defined by you)
  2. Migrates the service project's classpath entries to the new module
  3. Copies all WebSphere Business Integration Server Foundation source artifacts from the selected source project to this module
  4. Migrates the BPEL extensions in WSDL files
  5. Migrates the business processes (.bpel files) from BPEL4WS version 1.1 to the new level supported by WebSphere Process Server, which is built on BPEL4WS version 1.1 with major capabilities of the upcoming WS-BPEL version 2.0 specification
  6. Creates an SCA component for each .bpel process
  7. Generates a monitoring .mon file for each BPEL process to preserve the default monitoring behavior from WebSphere Studio Application Developer Integration Edition (if necessary)
  8. Creates imports and exports depending on the deploy options chosen in WebSphere Studio Application Developer Integration Edition
  9. Wires the BPEL component to its partnerlinks (imports, exports, and Java components)

To migrate service projects using the WebSphere Integration Developer Migration wizard, follow these steps:

  1. Invoke the wizard by selecting File -> Import -> Business Integration -> WebSphere Studio Application Developer Integration Edition Service Project and click Next:
    Import selection for source artifact migration
    OR you can also open the Migration wizard from the Welcome page by clicking the Returning Users icon Returning users to open the Returning Users page (Note that you can always return to the Welcome page by clicking Help -> Welcome ):
    Returning users page
    Click Migration on the left side of the Returning Users page to open the Migration page. From the Migration page, select the Migrate an Integration Edition 5.1 service project option Migration page with the Migrate an Integration Edition 5.1 service project option selected.
  2. The Migration wizard opens. Enter the path for the Source Selection or click the Browse button to find it. Also enter the Module name of the location of the WebSphere Studio Application Developer Integration Edition Service Project to migrate:
    Source artifacts Migration wizard
    Note: It is recommended that you choose the name of the Service Project as the module name because if there are other projects in the WebSphere Studio Application Developer Integration Edition workspace that are dependent on this project, you won't have to update the dependent projects' classpaths after importing them in to WebSphere Integration Developer.
  3. From the Migration options, select the Preserve original BPEL Java snippets in the comments check box:
    Migration options
    Click Finish.
  4. After the migration process has completed, the Migration Results window opens:
    Migration results window
    A log file containing these migration messages will automatically get generated to the 6.x workspace's .metadata folder. The log file will be named ".log".

After the Migration wizard has completed, build the Business Integration module that was created and try to resolve any build errors. Inspect all migrated .bpel files: ensure that they are fully migrated and can be opened in the WebSphere Integration Developer BPEL Editor. There are some BPEL Java snippets that can not be automatically migrated. If you see any errors in the BPEL Java snippets, see "Migrating to the SCA Programming Model" for steps needed to fix the errors. Also, if you used the Migration wizard to migrate a service project to a BI Module, open the module dependency editor to ensure that the dependencies are set correctly. To do this, switch to the Business Integration perspective and double click the business integration module project. From there you can add dependencies on business integration library projects, Java projects, and J2EE projects.

Migrating service projects using WSADIEServiceProjectMigration

The WSADIEServiceProjectMigration command enables the migration of service projects.

The migration command does the following:

Note: You must migrate the service projects in their dependency order. For example, if a BPEL in service project A makes a process-to-process call to a BPEL in service project B, then service project B must be migrated before service project A. Otherwise, the process-to-process call cannot be configured correctly.

To run the WSADIEServiceProjectMigration script, follow these steps:

  1. Locate the script by opening the shared folder specified during the install of WebSphere Integration Developer. For example, the script will be located under: SHARED_FOLDER_HOME/plugins/com.ibm.wbit.migration.wsadie_6.1.0
  2. Invoke the script as follows: WSADIEServiceProjectMigration -e eclipse_dir -s source_project_dir -d workspace [-t target_project_name] [-preserveSnippets true|false] [-debug]

    Parameter definitions:

    -e eclipse_dir
    The location of your Eclipse folder (Eclipse runtime).
    -s source_project_dir
    The full path to the WebSphere Studio Application Developer Integration Edition 5.1 service project.
    -d workspace
    The workspace where the new business integration module will be created.
    -t target_project_name
    The name of the new business integration module to be created. The default is the same as the WebSphere Studio Application Developer Integration Edition 5.1 project being migrated.
    -preserveSnippets flag
    Enables or disables the preserving of existing BPEL Java snippets as commented out. The default is true.
    -debug flag
    Enables debugging output.

    For example:

    WSADIEServiceProjectMigration -e WID_HOME\eclipse" -d "\myWIDworkspace" -s "
    \\MyServiceProject" -t "MyBIModuleName" -preserveSnippets false -debug
  3. After the command has completed, start the new workspace in WebSphere Integration Developer.
  4. Build the Business Integration module that was created and try to resolve any build errors. Inspect all migrated .bpel files: ensure that they are fully migrated and can be opened in the WebSphere Integration Developer BPEL Editor. There are some BPEL Java snippets that can not be automatically migrated. If you see any errors in the BPEL Java snippets, see "Migrating to the SCA Programming Model for steps needed to fix the errors.
  5. Open the module dependency editor to ensure that the dependencies are set correctly. To do this, switch to the business integration perspective and double click on the business integration module project. From there you can add dependencies on business integration library projects, Java projects, and J2EE projects.

Completing the migration of the application

After the Migration wizard has successfully migrated the artifacts to the new Business Integration module, the artifacts must be wired together to create an application that adheres to the SCA model. Note that even though the Migration wizard attempts to successfully migrate the artifacts, manual verification should also be done. The information in this section can be used to help ensure that the migration was correct.

  1. Open WebSphere Integration Developer and switch to the Business Integration perspective. You should see the module(s) that were created by the Migration wizard (one module for each service project that was migrated). The first artifact listed under the module project is the module's assembly file (it has the same name as the module).
  2. Double-click the assembly file to open it in the Assembly Editor where SCA components can be created and wired together to obtain similar functionality to the Version 5.1 application. If there were any BPEL processes in the WebSphere Studio Application Developer Integration Edition service project, the migration wizard should have created default SCA components for each of those processes and they will be in the Assembly Editor.
  3. Select a component and go to the Properties view where the Description, Details, and Implementation properties will be displayed and can be edited.

Some projects may require some rewiring after migration in order to reconnect the services the way they were in 5.1. The following information further describes how to manually rewire the application using the tools available in WebSphere Integration Developer:

Creating SCA Components and SCA Imports for the services in the application for rewiring

All migrated business processes must be wired to their business partners. An SCA Component or Import must be created for all other service types. For WebSphere Studio Application Developer Integration Edition service projects that interact with systems or entities external to the project, an SCA Import can be created in order for the migrated project to access those entities as services according to the SCA model.

Note: The migration utility attempts to do this automatically, however, you can refer to the following information to help verify what the tool did.

For WebSphere Studio Application Developer Integration Edition service projects that interact with entities within the project (for example, a business process, transformer service or Java class), an SCA Import can be created in order for the migrated project to access those entities as services according to the SCA model.

The following sections provide details on the SCA Import or SCA Components to create based on the type of service that must be migrated:

Migrating a Java service

You can migrate a Java service to an SCA Java Component.

If the WebSphere Studio Application Developer Integration Edition Service Project was dependent on other Java projects, copy the existing projects into the new workspace directory and import them into WebSphere Integration Developer using the File -> Import -> General -> Existing Projects into Workspace wizard.

In WebSphere Studio Application Developer Integration Edition when generating a new Java service from an existing Java class, the following options were given:

There are many new components that offer new functionality such as data mapping, interface mediation, business state machines, selectors, business rules, and more. First you should determine whether one of these new component types can replace the custom Java component. If that is not possible, follow the migration path described below.

Import the service project using the Migration wizard. This will result in the creation of a business integration module with the WSDL Messages, PortTypes, Bindings, and Services generated in WebSphere Studio Application Developer Integration Edition.

In the Business Integration perspective, expand the module to see its contents. Open the Assembly Editor by double-clicking the first item under the module project (it will have the same name as the project).

You have the following options:

Creating the custom Java component: option 1

The recommended migration technique is to use the WebSphere Integration Developer Java Component type that allows you to represent the Java service as an SCA component. During migration, custom Java code must be written to convert between the SCA Java interface style and the existing Java component's interface style.

To create the custom Java component, follow these steps:

  1. Under the module project, expand Interfaces and select the WSDL interface that was generated for this Java class in WebSphere Studio Application Developer Integration.
  2. Drag and drop this interface onto the Assembly Editor. A dialog will pop up asking you to select the type of component to create. Select Component with No Implementation Type and click OK.
  3. A generic component will appear on the Assembly diagram. Select it and go to the Properties view.
  4. On the Description tab, you can change the name and display name of the component to something more descriptive.
  5. On the Details tab you will see that this component has one interface - the one that you dragged and dropped onto the Assembly Editor.
  6. Ensure that the Java class that you are trying to access is on the classpath of the service project if it is not contained within the service project itself.
  7. Right-click on the module project and select Open Dependency Editor.... Under the Java, section ensure that the project containing the old Java class is listed. If it is not, add it by clicking Add....
  8. Back in the Assembly Editor, right-click the component that you just created and select Generate Implementation... -> Java Then select the package where the Java implementation will be generated. This creates a skeleton Java service that adheres to the WSDL interface according to the SCA programming model, where complex types are represented by an object that is a commonj.sdo.DataObject and simple types are represented by their Java Object equivalents.

The following code examples show:

  1. Relevant definitions from the 5.1 WSDL interface
  2. The WebSphere Studio Application Developer Integration Edition 5.1 Java methods that correspond to the WSDL
  3. The WebSphere Integration Developer 6.x Java methods for the same WSDL

The following code shows the relevant definitions from the 5.1 WSDL interface:

<types>
	<schema xmlns="http://www.w3.org/2001/XMLSchema" 
    			attributeFormDefault="qualified" 
    			elementFormDefault="unqualified"    
    			targetNamespace="http://migr.practice.ibm.com/" 
    			xmlns:xsd1="http://migr.practice.ibm.com/">

			<complexType name="StockInfo">
				<all>
					<element name="index" type="int"/>
					<element name="price" type="double"/>
					<element name="symbol" nillable="true" 
							    type="string"/>

				</all>
			</complexType>
	</schema>
</types>

<message name="getStockInfoRequest">
	<part name="symbol" type="xsd:string"/>
</message>
<message name="getStockInfoResponse">
	<part name="result" type="xsd1:StockInfo"/>
</message>

	<operation name="getStockInfo" parameterOrder="symbol">
			<input message="tns:getStockInfoRequest" 
							name="getStockInfoRequest"/>
			<output message="tns:getStockInfoResponse" 
 			 				name="getStockInfoResponse"/>
        </operation>

The following code shows the WebSphere Studio Application Developer Integration Edition 5.1 Java methods that correspond to the WSDL:

public StockInfo getStockInfo(String symbol)
	{
		return new StockInfo();
	}

	public void setStockPrice(String symbol, float newPrice)
	{
		// set some things
	}

The following code shows the WebSphere Integration Developer 6.x Java methods for the same WSDL:

public DataObject getStockInfo(String aString) {
		//TODO Needs to be implemented.
		return null;
	}

	public void setStockPrice(String symbol, Float newPrice) {
		//TODO Needs to be implemented.
	}

Now you will need to fill in code where you see the "//TODO" tags in the generated Java implementation class. There are two options:

  1. Move the logic from the original Java class to this class, adapting it to use DataObjects
  2. Create a private instance of the old Java class inside this generated Java class and write code to:
    1. Convert all parameters of the generated Java implementation class into parameters that the old Java class expects
    2. Invoke the private instance of the old Java class with the converted parameters
    3. Convert the return value of the old Java class into the return value type declared by the generated Java implementation method
    4. This option is recommended for consumption scenarios where the WSIF service proxies must be consumed by new 6.x style Java components.

Once you have completed one of the above options, you must rewire the Java service. There should not be any references, therefore you just need to rewire the Java component's interface:

Creating a Java Web service: option 2

An alternative option to consider is the Rational Application Developer Web services tooling that allows you to create a Web service around a Java class.

Note: See the information at the following site before attempting to migrate using this method: Creating a Web service from a Java bean
Note: This option requires that a Web service runtime be configured through WebSphere Integration Developer before invoking the Web service wizard.

If you had taken a bottom-up approach in WebSphere Studio Application Developer Integration Edition to generate WSDL around the Java class, then follow these steps:

  1. Create a new Web project and copy the Java class that you would like to build a service around to this Web project's Java source folder.
  2. Right-click on the enterprise application project that is the container for the Java class you are creating a service around.
  3. Select Properties, go to the Server properties and ensure that the Target runtime is set to WebSphere Process Server v6.1 and Default server is set to the installed WebSphere Process Server v6.1.
  4. Start the test server and deploy this application to the server and ensure that it starts successfully.
  5. Next, right-click on the Java class that you would like to create a service around and select Web Services -> Create Web service.
  6. For Web Service Type select Java bean Web Service and uncheck the Start Web service in Web project option unless you want to deploy the web service right away. You can optionally select to generate a client proxy as well. Click Next.
  7. The Java class that you right-clicked will be shown, click Next.
  8. You must now configure your service deployment options. Click Edit.... For the server type choose WPS Server v6.1 and for the Web service runtime choose IBM WebSphere and J2EE version 1.4. If you are not able to select a valid combination by doing this, see the section "Preparing for Migration" for information on migrating J2EE projects to the v1.4 level. Click OK.
  9. For the Service project, enter the name of the Web project. Also select the appropriate EAR project. Click Next. Note that you may have to wait for several minutes.
  10. On the Web Service Java Bean Identity panel, select the WSDL file that will contain the WSDL definitions. Choose the methods that you would like to expose on the Web service and choose the appropriate style/encoding (Document/Literal, RPC/Literal, or RPC/Encoded). Select the Define custom mapping for package to namespace option and select a namespace that is unique to the Java class being migrated for all Java packages used by this Java class's interface (the default namespace will be unique to the package name which may cause conflicts if you create another Web Service that uses the same Java classes). Complete the other parameters if appropriate.
  11. Click Next and on the Web Service package to namespace mapping panel, click Add and in the row that is created, enter the name of the package of the Java bean, then add the custom namespace that uniquely identifies this Java class. Continue to add mappings for all Java packages used by the Java bean interface.
  12. Click Next. Note that you may have to wait for several minutes.
  13. Click Finish. After completing the wizard, you should copy the generated WSDL file that describes the Java service to the business integration module project if the service project was a consumer of the Java service. It can be found in the generated router Web project under the folder WebContent/WEB-INF/wsdl. Refresh/rebuild the business integration module project.
  14. Switch to the Business Integration perspective and expand the module and then the Web Service Ports logical category.
  15. Select the port that was created in the previous steps and drag and drop it onto the Assembly Editor and select to create an Import with Web Service Binding. Select the Java class's WSDL interface if prompted. Now the SCA component that consumed the Java component in 5.1 can be wired to this Import to complete the manual rewiring migration steps.

Note that the interface may be slightly different than the 5.1 interface, and you may need to insert an Interface Mediation component in between the 5.1 consumer and the new Import. To do this, click on the wire tool in the Assembly Editor and wire the SCA source component to this new Import with Web Service Binding. As the interfaces are different, you will be prompted: Source and target nodes do not have matching interfaces. Choose to create an interface mapping between the source and target node. Double-click on the mapping component that was created in the Assembly Editor. This will open the mapping editor. See the Information Center for instructions on creating an interface mapping.

If you had taken a top-down approach in WebSphere Studio Application Developer Integration Edition, generating Java classes from a WSDL definition, then follow these steps:

  1. Create a new Web project and copy the WSDL file that you would like to Java skeleton to this Web project's source folder.
  2. Right-click on the WSDL file containing the PortType that you want to generate a Java skeleton from and select Web Services -> Generate Java bean skeleton.
  3. Choose the Web service typeSkeleton Java bean Web Service and complete the wizard.

After completing the wizard, you should have Java classes that implement the service interface and are not dependent on WSIF APIs.

Advantages and disadvantages for each of the Java service rewiring options

There are advantages and disadvantages for each of the Java service rewiring options.

The following list describes both options and the advantages and disadvantages of each:

Migrating an EJB service

You can migrate an EJB service to an SCA Import with stateless session bean binding.

If the WebSphere Studio Application Developer Integration Edition Service Project was dependent on another EJB, EJB client, or Java project, import those existing projects using the File -> Import -> General -> Existing Projects into Workspace wizard. This was usually the case when an EJB was referenced from a service project. If any WSDL or XSD files that are referenced from the service project exist in another type of project, create a new Business Integration Library with the same name as the old non-service project, and copy all of those artifacts to the library.

Import the service project using the Migration wizard. This will result in the creation of a business integration module with the WSDL Messages, PortTypes, Bindings, and Services generated in WebSphere Studio Application Developer Integration Edition.

In the Business Integration perspective, expand the module to see its contents. Open the Assembly Editor by double-clicking the first item under the module project (it will have the same name as the project).

You have the following options:

Creating the custom EJB component: option 1

The recommended migration technique is to use the WebSphere Integration Developer Import with Stateless Session Binding type that allows you to invoke a stateless session EJB as an SCA component. During migration, custom Java code must be written to convert between the SCA Java interface style and the existing EJB interface style.

Note: Even though the migration tool automatically handles this, any changes made after migration to the interfaces and data types (business objects) involved in the EJB interface will require manual updates to the conversion code mentioned here. Errors may be displayed in WebSphere Integration Developer depending on the type of change made.

To create the custom EJB component, follow these steps:

  1. Under the module project, expand Interfaces and select the WSDL interface that was generated for this EJB in WebSphere Studio Application Developer Integration.
  2. Drag and drop this interface onto the Assembly Editor. A dialog will pop up asking you to select the type of component to create. Select Component with No Implementation Type and click OK.
  3. A generic component will appear on the Assembly diagram. Select it and go to the Properties view.
  4. On the Description tab, you can change the name and display name of the component to something more descriptive. Choose a name like your EJB's name, but append a postfix such as "JavaMed" as this is going to be a Java component that mediates between the WSDL interface generated for the EJB in WebSphere Studio Application Developer Integration and the Java interface of the EJB.
  5. On the Details tab you will see that this component has one interface - the one that you dragged and dropped onto the Assembly Editor.
  6. Back in the Assembly Editor, right-click the component that you just created and select Generate Implementation... -> Java Then select the package where the Java implementation will be generated. This creates a skeleton Java service that adheres to the WSDL interface according to the SCA programming model, where complex types are represented by an object that is a commonj.sdo.DataObject and simple types are represented by their Java Object equivalents.

The following code examples show:

  1. Relevant definitions from the 5.1 WSDL interface
  2. The WebSphere Studio Application Developer Integration Edition 5.1 Java methods that correspond to the WSDL
  3. The WebSphere Integration Developer 6.x Java methods for the same WSDL

The following code shows the relevant definitions from the 5.1 WSDL interface:

<types>
	<schema xmlns="http://www.w3.org/2001/XMLSchema" 
    			attributeFormDefault="qualified" 
    			elementFormDefault="unqualified"    
    			targetNamespace="http://migr.practice.ibm.com/" 
    			xmlns:xsd1="http://migr.practice.ibm.com/">

			<complexType name="StockInfo">
				<all>
					<element name="index" type="int"/>
					<element name="price" type="double"/>
					<element name="symbol" nillable="true" 
							    type="string"/>

				</all>
			</complexType>
	</schema>
</types>

<message name="getStockInfoRequest">
	<part name="symbol" type="xsd:string"/>
</message>
<message name="getStockInfoResponse">
	<part name="result" type="xsd1:StockInfo"/>
</message>

	<operation name="getStockInfo" parameterOrder="symbol">
			<input message="tns:getStockInfoRequest" 
							name="getStockInfoRequest"/>
			<output message="tns:getStockInfoResponse" 
 			 				name="getStockInfoResponse"/>
        </operation>

The following code shows the WebSphere Studio Application Developer Integration Edition 5.1 Java methods that correspond to the WSDL:

public StockInfo getStockInfo(String symbol)
	{
		return new StockInfo();
	}

	public void setStockPrice(String symbol, float newPrice)
	{
		// set some things
	}

The following code shows the WebSphere Integration Developer 6.x Java methods for the same WSDL:

public DataObject getStockInfo(String aString) {
		//TODO Needs to be implemented.
		return null;
	}

	public void setStockPrice(String symbol, Float newPrice) {
		//TODO Needs to be implemented.
	}

Eventually you need to fill in real code where you see the "//TODO" tags in the generated Java implementation class. First you need to create a reference from this Java component to the actual EJB so that it can access the EJB according to the SCA programming model:

  1. Keep the Assembly Editor open and switch to the J2EE perspective. Locate the EJB project containing the EJB that you are creating a service for.
  2. Expand its Deployment Descriptor: <project-name> item and locate the EJB. Drag and drop it onto the Assembly Editor. If warned about project dependencies needing to be updated, select the Open the module dependency editor... check box and click OK.
  3. Under the J2EE section ensure that the EJB project is listed and if it is not, add it by clicking Add....
  4. Save the module dependencies and close that editor. You will see that a new Import was created in the Assembly Editor. You can select it and go to the Properties view on the Description tab to change the import's name and display name to something more meaningful. On the Binding tab you will see that the import type is automatically set to Stateless Session Bean Binding and the JNDI name of the EJB is already set appropriately.
  5. Select the Wire tool from the palette in the Assembly Editor.
  6. Click on the Java component and release the mouse.
  7. Next click on the EJB Import and release the mouse.
  8. When asked A matching reference will be created on the source node. Do you want to continue?, click OK. This creates a wire between the two components.
  9. Select the Java component in the Assembly Editor and in the Properties view under the Details tab, expand References and select the reference to the EJB that was just created. You can update the reference's name if the generated name is not very descriptive or appropriate. Remember the name of this reference for future use.
  10. Save the Assembly diagram.

You must use the SCA programming model to invoke the EJB from the generated Java class. Open the generated Java class and follow these steps to write the code that will invoke the EJB service. For the generated Java implementation class:

  1. Create a private variable (whose type is that of your remote EJB interface):
    private YourEJBInterface ejbService = null;
  2. If there are complex types in your EJB interface, then also create a private variable for the BOFactory:
    private BOFactory boFactory = (BOFactory) 
    	ServiceManager.INSTANCE.locateService("com/ibm/websphere/bo
    	/BOFactory");
  3. In the constructor of the Java implementation class, use the SCA APIs to resolve the EJB reference (remember to fill in the name of the EJB reference that you wrote down a few steps back) and set the private variable equal to this reference:
    // Locate the EJB service
    	this.ejbService = (YourEJBInterface) 
    	ServiceManager.INSTANCE.locateService("name-of-your-ejb-reference");

For each "//TODO" in the generated Java implementation class:

  1. Convert all parameters into the parameter types that the EJB expects.
  2. Invoke the appropriate method on the EJB reference using the SCA programming model, sending the converted parameters.
  3. Convert the return value of the EJB into the return value type declared by the generated Java implementation method
/**
	 * Method generated to support the implementing WSDL port type named
	 * "interface.MyBean".
	 */
	public DataObject getStockInfo(String aString) {
		DataObject boImpl = null;

		try {

			// invoke the EJB method
			StockInfo stockInfo = this.ejbService.getStockInfo(aString);

			// formulate the SCA data object to return.
			boImpl = (DataObject) 
					this.boFactory.createByClass(StockInfo.class);

			// manually convert all data from the EJB return type into the 
			// SCA data object to return
			boImpl.setInt("index", stockInfo.getIndex());
			boImpl.setString("symbol", stockInfo.getSymbol());
			boImpl.setDouble("price", stockInfo.getPrice());
		} catch (RemoteException e) {
			e.printStackTrace();
		}
		return boImpl;
	}

	/**
	 * Method generated to support the implementing WSDL port type named
	 * "interface.MyBean".
	 */
	public void setStockPrice(String symbol, Float newPrice) {
		try {
			this.ejbService.setStockPrice(symbol, newPrice.floatValue());
		} catch (RemoteException e) {
			e.printStackTrace();
		}

	}
Creating an EJB Web service: option 2

An alternative option to consider is the Rational Application Developer Web services tooling that allows you to create a Web service around an EJB.

Note: See the information at the following site before attempting to migrate using this method: Creating a Web service from an enterprise bean (EJB) using the WebSphere run-time environment
Note: This option requires that a Web service runtime be configured through WebSphere Integration Developer before invoking the Web service wizard.

To create a Web service around an EJB, follow these steps:

  1. Right-click on the enterprise application project that is the container for the EJB that you are creating a service around.
  2. Select Properties, go to the Server properties and ensure that the Target runtime is set to WebSphere Process Server v6.1 and Default server is set to the installed WebSphere Process Server v6.1.
  3. Start the test server and deploy this application to the server and ensure that it starts successfully.
  4. In the J2EE perspective, expand the EJB project in the Project Explorer view. Expand the Deployment Descriptor then the Session Beans category. Select the bean that you want to generate the Web service around.
  5. Right-click and select Web Services -> Create Web service.
  6. For Web Service Type select EJB Web Service and uncheck the Start Web service in Web project option unless you want to deploy the Web service right away. Click Next.
  7. Ensure that the EJB that you right-clicked is selected here and click Next.
  8. You must now configure your service deployment options. Click Edit.... For the server type choose WPS Server v6.1 and for the Web service runtime choose IBM WebSphere and J2EE version 1.4. If you are not able to select a valid combination by doing this, then see the section "Preparing for Migration" for information on migrating J2EE projects to the v1.4 level. Click OK.
  9. For the Service project, enter the name of the EJB project containing the EJB. Also select the appropriate EAR project. Click Next. Note that you may have to wait for several minutes.
  10. On the Web Service EJB Configuration panel, select the appropriate router project to use (choose the name of the router Web project you would like to be created and this project will be added to the same enterprise application as the original EJB. Select the desired transport (SOAP over HTTP or SOAP over JMS). Click Next.
  11. Select the WSDL file that will contain the WSDL definitions. Choose the methods that you would like to expose on the Web service and choose the appropriate style/encoding (Document/Literal, RPC/Literal, or RPC/Encoded). Select the Define custom mapping for package to namespace option and select a namespace that is unique to the EJB being migrated for all Java packages used by the EJB (the default namespace will be unique to the package name which may cause conflicts if you create another Web Service that uses the same Java classes). Complete the other parameters if appropriate. There are limitations around each style/encoding combinations. See the limitations for more information: Limitations of Web services
  12. Click Next and on the Web Service package to namespace mapping panel, click Add and in the row that is created, enter the name of the package of the your EJB, then the custom namespace that uniquely identifies this EJB. Continue to add mappings for all Java packages used by the EJB interface.
  13. Click Next. Note that you may have to wait for several minutes.
  14. Click Finish. After completing the wizard, you should copy the generated WSDL file that describes the EJB service to the business integration module project if the service project was a consumer of the EJB service. It can be found in the generated router Web project under the folder WebContent/WEB-INF/wsdl. Refresh/rebuild the business integration module project.
  15. Switch to the Business Integration perspective and expand the migrated module and then the Web Service Ports logical category.
  16. Select the port that was generated in the previous steps and drag and drop it onto the Assembly Editor and select to create an Import with Web Service Binding. Select the EJB's WSDL interface if prompted. Now the SCA component that consumed the EJB in 5.1 can be wired to this Import to complete the manual rewiring migration steps.

If you had taken a top-down approach in WebSphere Studio Application Developer Integration Edition, generating an EJB skeleton from a WSDL definition, then follow these steps:

  1. Create a new Web project and copy the WSDL file that you would like to generate the EJB skeleton from to this Web project's source folder.
  2. Right-click on the WSDL file containing the PortType that you want to generate an EJB skeleton from and select Web Services -> Generate Java bean skeleton.
  3. Choose the Web service typeSkeleton EJB Web Service and complete the wizard.

After completing the wizard, you should have an EJB that implements the service interface and is not dependent on WSIF APIs.

Note that the interface may be slightly different than the 5.1 interface, and you may need to insert an Interface Mediation component in between the 5.1 consumer and the new Import. To do this, click on the wire tool in the Assembly Editor and wire the SCA source component to this new Import with Web Service Binding. As the interfaces are different, you will be prompted: Source and target nodes do not have matching interfaces. Choose to create an interface mapping between the source and target node. Double-click on the mapping component that was created in the Assembly Editor. This will open the mapping editor. See the Information Center for instructions on creating an interface mapping.

Once you have completed this, you must rewire the EJB service. There should not be any references, therefore you just need to rewire the Java component's interface:

Advantages and disadvantages for each of the EJB service rewiring options

There are advantages and disadvantages for each of the EJB service rewiring options.

The following list describes both options and the advantages and disadvantages of each:

Migrating a Business Process to Business Process Service Invocation

This scenario applies to a business process that invokes another business process, where the second business process is invoked using a WSIF Process Binding. This section shows how to migrate a BPEL to BPEL service invocation using a wire or an Import/Export with SCA Binding.

To migrate a process (BPEL) binding service project for an outbound service, follow these steps:

  1. In the Business Integration perspective, expand the module to see its contents. Open the Assembly Editor by double-clicking the first item under the module project (it will have the same name as the project).
  2. There are several scenarios where a BPEL process can invoke another BPEL process. Find the scenario below that applies to your application:

Migrating a Web Service (SOAP/JMS)

You can migrate a Web Service (SOAP/JMS) to an SCA Import with Web Service binding.

To migrate a SOAP/JMS service project for an outbound service migration, follow these steps:

  1. First, you will need to import the service project using the Migration wizard. This will result in the creation of a Business Integration module with the WSDL Messages, PortTypes, Bindings, and Services generated in WebSphere Studio Application Developer Integration Edition. Note that if the IBM Web Service (SOAP/JMS) that this application will invoke is also a WebSphere Studio Application Developer Integration Edition Web service that will be migrated, there may have been updates to that Web service during migration. If this is the case, you should use that Web service's migrated WSDL files here.
  2. In the Business Integration perspective, expand the module so that you can see its contents. Open the Assembly Editor by double-clicking the first item under the module project (it will have the same name as the project).
  3. Next, add an Import that will allow the application to interact with the IBM Web Service (via SOAP/JMS) according to the SCA programming model. Ensure that the WSDL interface, binding, and service definitions are present in the migrated module or in a library that the migrated module is dependent on.
  4. In the Business Integration perspective, expand the migrated module and open its Assembly Diagram in the Assembly Editor.
  5. Expand the Web Service Ports logical category and drag and drop the port that corresponds to the service you want to invoke onto the Assembly Editor.
  6. Choose to create an Import with Web Service Binding.
  7. After creating the import, select it in the Assembly Editor and go to the Properties view. Under the Binding tab you will see the port and service that the import is bound to.
  8. Save the assembly diagram.

Once you have completed this, you must rewire the service:

Migrating a Web Service (SOAP/HTTP)

You can migrate a Web Service (SOAP/HTTP) to an SCA Import with Web Service binding.

To migrate a SOAP/HTTP service project for an outbound service migration, follow these steps:

  1. First, you will need to import the service project using the Migration wizard. This will result in the creation of a Business Integration module with the WSDL Messages, PortTypes, Bindings, and Services generated in WebSphere Studio Application Developer Integration Edition. Note that if the IBM Web Service (SOAP/HTTP) that this application will invoke is also a WebSphere Studio Application Developer Integration Edition Web service that will be migrated, there may have been updates to that Web service during migration. If this is the case, you should use that Web service's migrated WSDL files here.
  2. In the Business Integration perspective, expand the module so that you can see its contents. Open the Assembly Editor by double-clicking the first item under the module project (it will have the same name as the project).
  3. Next, add an Import that will allow the application to interact with the IBM Web Service (via SOAP/HTTP) according to the SCA programming model. Ensure that the WSDL interface, binding, and service definitions are present in the migrated module or in a library that the migrated module is dependent on.
  4. In the Business Integration perspective, expand the migrated module and open its Assembly Diagram in the Assembly Editor.
  5. Expand the Web Service Ports logical category and drag and drop the port that corresponds to the service you want to invoke onto the Assembly Editor.
  6. Choose to create an Import with Web Service Binding.
  7. After creating the import, select it in the Assembly Editor and go to the Properties view. Under the Binding tab you will see the port and service that the import is bound to.
  8. Save the assembly diagram.

Once you have completed this, you must rewire the service:

Migrating a JMS service

You can migrate a JMS service to an SCA Import with JMS binding.

Note: If the JMS message is being sent to a WebSphere Business Integration Adapter, then see the section "Migrating Interactions with WebSphere Business Integration Adapter" in the link below.

To migrate a JMS service project for an outbound service migration, follow these steps:

  1. First, you will need to import the service project using the Migration wizard. This will result in the creation of a Business Integration module with the WSDL Messages, PortTypes, Bindings, and Services generated in WebSphere Studio Application Developer Integration Edition.
  2. In the Business Integration perspective, expand the module so that you can see its contents. Open the Assembly Editor by double-clicking the first item under the module project (it will have the same name as the project).
  3. Next, add an Import that will allow the application to interact with a JMS queue according to the SCA programming model.
  4. In the Assembly Editor expand the migrated module project and expand the Interfaces category and find the WSDL PortType that describes the Web service that the application will invoke. Drag and drop it onto the Assembly Editor.
  5. A Component Creation dialog will allow you to select they type of component to create. Choose Import with No Binding.
  6. You will see that a new Import was created in the Assembly Editor and if you select it and go to the Properties view, on the Description tab you can change the import's name and display name to something more meaningful.
  7. You can refer to the 5.1 WSDL binding and service files to find details about the JMS service that you are migrating and use them to fill in the details of the 6.x "Import with JMS Binding". Locate the 5.1 JMS binding and service WSDL files within the 5.1 service project (they are usually named *JMSBinding.wsdl and *JMSService.wsdl). Inspect the binding and service information captured there. From the binding, you can determine whether text or object messages were used and whether any custom data format bindings were used. If there were any, you should consider writing a custom data binding for your 6.x "Import with JMS Binding" as well. From the service, you can find the initial context factory, JNDI connection factory name, JNDI destination name, and destination style (queue).
  8. Right-click the import and select Generate Binding then JMS Binding. You will be prompted to enter the following parameters:
    Select JMS messaging domain:
    • Point-to-Point
    • Publish-Subscribe
    • Domain-Independent
    Select how data is serialized between Business Object and JMS Message:
    • Text
    • Object
    • User-supplied
    If User-supplied is selected then:
    Specify fully qualified name of com.ibm.websphere.sca.jms.data.JMSDataBinding implementation class. You should specify a user-defined data binding if your application needs to set any JMS header properties that are not normally available in the JMS Import Binding. In this case, you can create a custom data binding class that extends the standard JMS data binding "com.ibm.websphere.sca.jms.data.JMSDataBinding" and add custom code to access the JMSMessage directly. See the JMS examples in "Creating and modifying bindings for import and export components" from the link below.
    Inbound connectivity is using default JMS function selector class:
    <selected> or <deselected>
  9. Select the import that you just created. In the Properties view, go to the Binding tab. You can manually fill in all the binding information listed there to the same values that you specified before in WebSphere Studio Application Developer Integration Edition. The binding information that you may specify is:

Once you have completed this, you must rewire the service:

Migrating a J2C-IMS service

You can migrate a J2C-IMS service to an SCA Import with EIS Binding or SCA Import with Web Service Binding.

Do not use any of the WebSphere Studio Application Developer Integration Edition artifacts that were generated for this IMS service. You will need to recreate the service using the wizards available in WebSphere Integration Developer and manually rewire the application.

Note: Turn Auto-Build on or build the module manually.

You have the following options:

Note: For both options, note that if a BPEL service invokes this IMS service, the BPEL will need to change slightly, as the interface exposed by the EIS service will be slightly different than the old 5.1 interface. To do this, open the BPEL editor and adjust the partner link that corresponds to the EIS service and use the new interface (WSDL file) generated when performing the steps above. Make any necessary changes to the BPEL activities for the new WSDL interface of the EIS service.

Creating an SCA Import to invoke the IMS service: option 1

You can create an SCA Import with EIS Binding that will use DataObjects to store the message/data to communicate with the IMS system.

To create an SCA Import to invoke the IMS service, follow these steps:

  1. Create a new business integration module project to house this new IMS service.
  2. To recreate the EIS service, go to File -> New -> Other -> Business Integration -> External Service.
  3. This wizard allows you to import a service from an EIS system. It is very similar to the WebSphere Studio Application Developer Integration Edition wizard that created the WSIF-based EIS service in 5.1. You can import the new J2C IMS resource adapter in this wizard. You should browse to the directory where WebSphere Integration Developer is installed and drill down to Resource Adapters -> ims15 -> imsico9102.rar.
    Note: See the Information Center for more information on completing the saving properties and operations panels. During the External Service wizard, when you add an operation you will also be able to create business objects for the input or output data type of the operation. This requires that you have the C or COBOL source file that you used in the WebSphere Studio Application Developer Integration Edition wizard. These files should have been copied to the old service project so that you can point to the source files there. You can also import the business objects using the separate wizard File -> New -> Other -> Business Integration -> External Data.
  4. Once you have completed the wizard, open the Business Integration perspective and expand the module so that you can see its contents. You should see new business objects listed under the module's Data Types and new interfaces listed under Interfaces.
  5. Open the Assembly Editor by double-clicking the first item under the module project (it will have the same name as the project). You should see that an Import exists on the canvas, this Import has an EIS Binding and it represents the service that you just created.

Now see the section entitled "Creating SCA Exports to access the migrated service" for instructions on how to expose this service to consumers.

Creating a Web service around the J2C service: option 2

You can create a J2C Web service and if the consumer of the service is an SCA component, consume the service as an IBM Web Service (SOAP/HTTP or SOAP/JMS).

To create a Web service around the J2C service, follow these steps:

  1. Create the J2C Java Bean by clicking File -> New -> J2C -> J2C Java Bean
  2. Choose the 1.5 version of the IMS Connector for Java and click Next.
  3. Check Managed Connection and enter the JNDI lookup name. Click Next.
  4. Specify the project, package, and name for the new Java bean. The bean consists of an interface and an implementation class. Click Next.
  5. Add a Java method for each function or service you want to access from the EIS. Additional methods can be added later in the Java source editor through the Snippets View. When you click the Add... button, choose the name for the method and click Next.
  6. Now you can choose Browse... to reuse existing types or New... to launch the CICS/IMS Java Data Binding Wizard (where you can refer to a COBOL or C source file) for the input and output data types.
  7. Once you are finished creating Java methods, Click Next.
  8. Complete the remaining steps in this wizard to create your J2C Java Bean.
  9. Create the Web Service by clicking File -> New -> J2C -> Web Page, Web Service, or EJB from J2C Java Bean to create the Web service around your J2C Java Bean.
  10. Complete the wizard.

The consumers of this service can now use the WSDL service that is generated by this wizard to invoke the IMS service.

Advantages and disadvantages for each of the J2C-IMS service rewiring options

There are advantages and disadvantages for each of the J2C-IMS service rewiring options.

The following list describes both options and the advantages and disadvantages of each:

Migrating a J2C-CICS ECI service

You can migrate a J2C-CICS ECI service to an SCA Import with EIS Binding or SCA Import with Web Service Binding.

Follow the instructions in the topic "Migrating a J2C-IMS service", but ensure to import the following RAR file instead of the IMS RAR file:

If you follow the second option to create a J2C Web service, then choose the v1.5 ECIResourceAdapter on the second panel of the J2C Java Bean creation wizard.

Also, see the topic "Migrating a J2C-IMS service".

Migrating a J2C-CICS EPI service

There is no direct support for the J2C-CICS EPI service in WebSphere Integration Developer. In order to access this service from an SCA module, you will need to migrate using the consumption scenario.

See the topic "The consumption scenario for service migration" for instructions on migrating this service type to WebSphere Integration Developer.

Migrating a J2C-HOD service

There is no direct support for the J2C-HOD service in WebSphere Integration Developer. In order to access this service from an SCA module, you will need to migrate using the consumption scenario.

See the topic "The consumption scenario for service migration" for instructions on migrating this service type to WebSphere Integration Developer.

Migrating a transformer service

You can migrate a transformer service to an SCA Data Map and Interface Map where possible. You can also use the consumption scenario to access this service from an SCA module.

The data map and interface map components are new in version 6.0. They offer similar function to the transformer service from 5.1 but they do not have the full XSL transform capability. If you are not able to replace your transformer service with one of these components, then you must migrate using the consumption scenario as there is no direct support for the transformer service in WebSphere Integration Developer. Follow the steps documented in the "The consumption scenario for service migration" section to access this service from an SCA module.

The consumption scenario for service migration

In the cases where there is no direct counterpart for a WebSphere Studio Application Developer Integration Edition service type, a consumption scenario is needed to consume the old WebSphere Studio Application Developer Integration Edition service as-is when redesigning the application in WebSphere Integration Developer.

Here are the steps to perform in WebSphere Studio Application Developer Integration Edition before invoking the Migration wizard:

  1. Create a new Java project to hold this client proxy code. Do not put this client proxy code in the service project because the 5.1-style generated messages and Java bean classes will be skipped by the automatic Migration wizard that migrates service projects.
  2. Open WebSphere Studio Application Developer Integration Edition and right-click the WSDL file containing the transformer binding and service and select Enterprise Services -> Generate Service Proxy. You will be asked what type of proxy to create, but only Web Services Invocation Framework (WSIF) will be available. Click Next.
  3. You can now specify the package and name of the service proxy Java class to create (you will create the proxy in the current service project). Click Next.
  4. You can now specify the proxy style, choose Client Stub, select the desired operations to include in the proxy, and click Finish. This creates a Java class that exposes the same methods as the WebSphere Studio Application Developer Integration Edition service, where the arguments to the Java methods are the parts of the source WSDL message.

You can now migrate to WebSphere Integration Developer:

  1. Copy the client proxy Java project to the new workspace and import it by going to theFile -> Import -> Existing Project into Workspace.
  2. Import the service project using the Migration wizard. This will result in the creation of a Business Integration module with the WSDL Messages, PortTypes, Bindings, and Services generated in WebSphere Studio Application Developer Integration Edition.
  3. In the Business Integration perspective, expand the module so that you can see its contents. Open the Assembly Editor by double-clicking the first item under the module project (it will have the same name as the project).
  4. To create the custom Java component, under the module project, expand Interfaces and select the WSDL interface that was generated for this transformer service in WebSphere Studio Application Developer Integration Edition.
  5. Drag and drop this interface onto the Assembly Editor. A dialog will pop up asking you to select the type of component to create. Select Component with No Implementation Type and click OK.
  6. A generic component will appear on the Assembly diagram. Select it and go to the Properties view.
  7. On the Description tab, you can change the name and display name of the component to something more descriptive (in this case name it something like your EJB's name but append a postfix such as "JavaMed" as this is going to be a Java component that mediates between the WSDL interface generated for the transformer service in WebSphere Studio Application Developer Integration Edition and the Java interface of the transformer client proxy).
  8. On the Details tab you will see that this component has one interface - the one that you dragged and dropped onto the Assembly Editor.
  9. Back in the Assembly Editor, right-click the component that you just created and select Generate Implementation... -> Java Then select the package where the Java implementation will be generated. This creates a skeleton Java service that adheres to the WSDL interface according to the SCA programming model, where complex types are represented by an object that is a commonj.sdo.DataObject and simple types are represented by their Java Object equivalents.

Now you will need to fill in code where you see the "//TODO" tags in the generated Java implementation class. There are two options:

  1. Move the logic from the original Java class to this class, adapting it to the new data structure.
  2. Create a private instance of the old Java class inside this generated Java class and write code to:
    1. Convert all parameters of the generated Java implementation class into parameters that the old Java class expects
    2. Invoke the private instance of the old Java class with the converted parameters
    3. Convert the return value of the old Java class into the return value type declared by the generated Java implementation method

Once you have completed the above options, you must rewire the client proxy. There should not be any "references", therefore you just need to rewire the Java component's interface:

Creating SCA Exports to access the migrated service

An SCA Export must be created to make the migrated service available to external consumers according to the SCA model for all services that deployment code was generated for in the WebSphere Studio Application Developer Integration Edition service project. This includes all services invoked by clients external to the application. Note: The migration utility attempts to do this automatically, however, you can refer to the following information to help verify what the tool did.

If from WebSphere Studio Application Developer Integration Edition, you right-clicked the BPEL process or other Service WSDL and selected Enterprise Services -> Generate Deploy Code , you must perform the manual migration steps below. Note that WebSphere Integration Developer is different from WebSphere Studio Application Developer Integration Edition in that it stores all of the deployment options. When the project is built, the deployment code is automatically updated in the generated EJB and Web projects so there is no option to manually Generate Deploy Code anymore.

Five binding options were given under the Interfaces for Partners section of the Generate BPEL Deploy Code wizard. The following inbound BPEL service migration information provides more details on the Export type and properties to create based on the deployment binding type(s) that were selected in WebSphere Studio Application Developer Integration Edition:

Migrating the EJB and the EJB process bindings

The EJB and EJB process bindings can be migrated to the recommended SCA construct.

In WebSphere Studio Application Developer Integration Edition this binding type enabled clients to communicate with a BPEL process or other service type by invoking an EJB. Note that this binding type was not optional for microprocesses - it was always selected as the generated EJB was used internally by the other binding types.

The JNDI name of the generated EJB was automatically generated as a combination of the BPEL's name, target namespace, and valid-from timestamp. For example, these attributes can be found by examining the BPEL process's properties in the BPEL editor on the Description and Server content tabs:

Table 3. Generated namespace
Process name MyService
Target namespace http://www.example.com/process87787141/
Valid From Jan 01 2003 02:03:04

The generated namespace for this example is then com/example/www/process87787141/MyService20030101T020304.

In WebSphere Studio Application Developer Integration Edition when the EJB binding was selected as the deployment type, there were no options given.

There are four options for migrating the WebSphere Studio Application Developer Integration Edition process binding. The type of client(s) that access the service will determine which migration option(s) below to perform:

Note: After the manual migration steps have been completed, the client must be migrated to the new programming model as well. See the appropriate topic for the following client types:

Table 4. Further information for migrating clients
Client type For further information see
EJB client that invokes the generated session bean. Such a client would invoke an EJB method corresponding to the BPEL operation to invoke "Migrating the EJB client"
WSIF client that uses the EJB process binding "Migrating the EJB process binding client"
Generic business process choreographer EJB API "Migrating the business process choreographer generic EJB API client"
Generic business process choreographer Messaging API "Migrating the business process choreographer generic Messaging API client"
Another BPEL process in the same module N/A: Wire BPEL components together using Assembly Editor
Another BPEL process in a different module N/A: Create an Import with SCA Binding in the referencing module, and configure its binding to point to the Export with SCA Binding that you create below in Option 1

Migration option 1 for the EJB and EJB process binding

The first migration option for the WebSphere Studio Application Developer Integration Edition EJB process binding is to make business processes accessible to another component in the same module.

In the Assembly Editor, to wire this other component to the BPEL component, follow these steps:

  1. Select the Wire item from the toolbar.
  2. Click on the other component to select it as the source of the wire.
  3. Click the BPEL SCA component to select it as the target of the wire.
  4. Save the assembly diagram.
Migration option 2 for the EJB and EJB process binding

The second migration option for the WebSphere Studio Application Developer Integration Edition EJB process binding is to make business processes accessible to other SCA modules and clients.

Note: These steps are mandatory if the generic business process choreographer APIs will be used to invoke the business process.

The Export with SCA Binding makes an SCA component accessible by other SCA modules. To create an Export with an SCA binding, follow these steps:

  1. Open the Assembly Editor for the module created by the migration wizard.
  2. Create an Export with SCA Binding for each BPEL process interface that had an EJB binding generated for it in WebSphere Studio Application Developer Integration Edition:
    1. Right-click the BPEL component in the Assembly Editor.
    2. Select Export....
    3. Select SCA Binding.
    4. If there are multiple interfaces for the process, select the interface(s) to export with this binding type.
    5. Once the SCA Export is created, select the export in the Assembly Editor and in the Properties view, select the Description content pane. The Export's name and description are listed and may be modified as necessary.
    6. Save the assembly diagram.
Migration option 3 for the EJB and EJB process binding

The third migration option for the WebSphere Studio Application Developer Integration Edition EJB process binding is to make modules accessible by a non-SCA entity (for example, a JSP or a Java client).

The Standalone Reference makes an SCA component accessible by any external client. To create a Standalone Reference, follow these steps:

  1. Open the Assembly Editor for the module created by the Migration wizard.
  2. Create a Standalone Reference for each BPEL process interface that had an EJB binding generated for it in WebSphere Studio Application Developer Integration Edition:
    1. Select the Standalone References item from the toolbar.
    2. Click the canvas of the Assembly Editor to create a Standalone References SCA entity.
    3. Select the Wire item from the toolbar.
    4. Click the Standalone References entity to select it as the source of the wire.
    5. Click the BPEL SCA component to select it as the target of the wire.
    6. You will see an alert Matching reference will be created on the source node. Would you like to continue?, click OK.
    7. Select the Standalone References entity that was just created and in the Properties view select the Description content pane.
    8. Expand the References link and select the reference that was just created. The reference's name and description are listed and may be modified as necessary.
    9. If there are multiple interfaces for the process, select the interface(s) to export with this binding type.
    10. Save the assembly diagram.
Migration option 4 for the EJB and EJB process binding

The fourth migration option for the WebSphere Studio Application Developer Integration Edition EJB process binding is to make business processes accessible by a Web Services client.

The Export with Web service binding makes an SCA component accessible by an external web services client. To create an Export with Web Service binding, follow these steps:

  1. Open the Assembly Editor for the module created by the migration wizard.
  2. Create an Export with SCA Binding for each BPEL process interface that had an EJB binding generated for it in WebSphere Studio Application Developer Integration Edition:
    1. Right-click the BPEL component in the Assembly Editor.
    2. Select Export... .
    3. Select Web Service Binding .
    4. If there are multiple interfaces for the process, select the interface(s) to export with this binding type.
    5. Select the transport: soap/http or soap/jms.
    6. Once the Web services Export has been created, select the export in the Assembly Editor and in the Properties view, select the Description content pane. The Export's name and description are listed and may be modified as necessary.
    7. Save the assembly diagram.

Migrating the JMS and the JMS process bindings

The JMS and JMS process bindings can be migrated to the recommended SCA construct.

In WebSphere Studio Application Developer Integration Edition, this binding type gave clients the ability to communicate with a BPEL process or other service type by sending a message to an MDB. Note that this binding type was not optional for long-running processes and it was always selected. In fact, this binding type was the only binding type allowed for request-response interfaces of long-running processes. For the other service types, an MDB would be generated and it would invoke the appropriate service.

The JNDI name used by the JMS binding was a combination of the BPEL's name, target namespace, and valid-from timestamp.

In WebSphere Studio Application Developer Integration Edition when the JMS binding was selected as the deployment type for a BPEL process, the following options were given:

There are five options for migrating the WebSphere Studio Application Developer Integration Edition JMS process binding. The type of client(s) that access the service will determine which migration option(s) below to perform:

Note: After the manual migration steps have been completed, the client must be migrated to the new programming model as well. See the appropriate topic for the following client types:

Table 5. Further information for migrating clients
Client type For further information see
WSIF Client that uses the JMS process binding "Migrating the business process choreographer generic Messaging API client and the JMS process binding client"
Generic business process choreographer EJB API "Migrating the business process choreographer generic EJB API client"
Generic business process choreographer Messaging API Migrating the business "Migrating the business process choreographer generic Messaging API client"
Another BPEL process in the same module N/A: Wire BPEL components together using Assembly Editor
Another BPEL process in a different module N/A: Create an Import with SCA Binding in the referencing module, and configure its binding to point to the Export with SCA Binding that you create below in Option 1.

Migration option 1 for the JMS and JMS process binding

The first migration option for the WebSphere Studio Application Developer Integration Edition JMS process binding is to make business processes accessible to another component in the same module.

In the Assembly Editor, to wire this other component to the BPEL component, follow these steps:

  1. Select the Wire item from the toolbar.
  2. Click on the other component to select it as the source of the wire.
  3. Click the BPEL SCA component to select it as the target of the wire.
  4. Save the assembly diagram.
Migration option 2 for the JMS and JMS process binding

The second migration option for the WebSphere Studio Application Developer Integration Edition JMS process binding is to make business processes accessible to other SCA modules and clients.

The Export with SCA Binding makes an SCA component accessible by other SCA modules. To create an Export with an SCA binding, follow these steps:

  1. Open the Assembly Editor for the module created by the migration wizard.
  2. Create an Export with SCA Binding for each BPEL process interface that had a JMS binding generated for it in WebSphere Studio Application Developer Integration Edition:
    1. Right-click the BPEL component in the Assembly Editor.
    2. Select Export....
    3. Select SCA Binding.
    4. If there are multiple interfaces for the process, select the interface(s) to export with this binding type.
    5. Once the SCA Export is created, select the export in the Assembly Editor and in the Properties view, select the Description content pane. The Export's name and description are listed and may be modified as necessary.
    6. Save the assembly diagram.
Migration option 3 for the JMS and JMS process binding

The third migration option for the WebSphere Studio Application Developer Integration Edition JMS process binding is to make business processes accessible by a non-SCA entity (for example, a JSP or a Java client).

The Standalone Reference makes an SCA component accessible by any external client. To create a Standalone Reference, follow these steps:

  1. Open the Assembly Editor for the module created by the migration wizard.
  2. Create a Standalone Reference for each BPEL process interface that had a JMS binding generated for it in WebSphere Studio Application Developer Integration Edition:
    1. Select the Standalone References item from the toolbar.
    2. Click the canvas of the Assembly Editor to create a Standalone References SCA entity.
    3. Select the Wire item from the toolbar.
    4. Click the Standalone References entity to select it as the source of the wire.
    5. Click the BPEL SCA component to select it as the target of the wire.
    6. You will see an alert Matching reference will be created on the source node. Would you like to continue?, click OK.
    7. Select the Standalone References entity that was just created and in the Properties view select the Description content pane.
    8. Expand the References link and select the reference that was just created. The reference's name and description are listed and may be modified as necessary.
    9. If there are multiple interfaces for the process, select the interface(s) to export with this binding type.
    10. Save the assembly diagram.
Migration option 4 for the JMS and JMS process binding

The fourth migration option for the WebSphere Studio Application Developer Integration Edition JMS process binding is to make business processes accessible by a Web services client.

The Export with Web service binding makes an SCA component accessible by an external web services client. To create an Export with Web service binding, follow these steps:

  1. Open the Assembly Editor for the module created by the migration wizard.
  2. Create an Export with SCA Binding for each BPEL process interface that had a JMS binding generated for it in WebSphere Studio Application Developer Integration Edition:
    1. Right-click the BPEL component in the Assembly Editor.
    2. Select Export... .
    3. Select Web Service Binding .
    4. If there are multiple interfaces for the process, select the interface(s) to export with this binding type.
    5. Select the transport: soap/http or soap/jms.
    6. Once the Web services Export has been created, select the export in the Assembly Editor and in the Properties view, select the Description content pane. The Export's name and description are listed and may be modified as necessary.
    7. Save the assembly diagram.
Migration option 5 for the JMS and JMS process binding

The fifth migration option for the WebSphere Studio Application Developer Integration Edition JMS process binding is to make business processes accessible by a JMS client.

The Export with JMS binding makes an SCA component accessible by an external JMS client. To create an Export with JMS binding, follow these steps:

  1. For BPEL services, you will need to create and reference new queue resources, as the 5.1 JMS process binding was quite different from the standard 5.1 JMS binding. For non-BPEL services, you can find the values you selected for the JMS deployment code in WebSphere Studio Application Developer Integration Edition 5.1 by finding the WSDL file named JMSBinding.wsdl and JMSService.wsdl in the appropriate package underneath the generated EJB project's ejbModule/META-INF folder and inspecting the binding and service information captured there. From the binding, you can determine whether text or object messages were used and whether any custom data format bindings were used. If there were any, you should consider writing a custom data binding for your 6.x Export with JMS Binding as well. From the service, you can find the initial context factory, JNDI connection factory name, JNDI destination name, and destination style (queue).
  2. Open the Assembly Editor for the module created by the migration wizard.
  3. Create an Export with JMS Binding for each BPEL process interface that had a JMS binding generated for it in WebSphere Studio Application Developer Integration Edition by right-clicking the BPEL component in the Assembly Editor.
  4. Select Export... .
  5. Select JMS Binding .
  6. If there are multiple interfaces for the process, select the interface(s) to export with this binding type.
  7. On the next panel (JMS Export Binding attributes), select JMS messaging domain. Define this attribute as Point-to-Point.
  8. Select how data is serialized between Business Object and JMS Message and enter the following values (it is recommended that you select Text instead of Object because text, which is usually XML, is independent of the runtime and enables service integration between disparate systems):
    1. For Text, select to use the Default JMS function selector or enter the fully qualified name of the FunctionSelector implementation class.
    2. For Object, select to use the Default JMS function selector or enter the fully qualified name of the FunctionSelector implementation class.
    3. For User Supplied, enter the fully-qualified name of the JMSDataBinding implementation class. You will need to select User Supplied if your application needs access to any JMS header properties that are not readily available in the JMS Import Binding. In this case, then you must create a custom data binding class that extends the standard JMS data binding com.ibm.websphere.sca.jms.data.JMSDataBinding and add custom code to access the JMSMessage directly. Then you will provide the name of your custom class for this field. See the JMS examples in "Creating and modifying bindings for import and export components" from the link below.
    4. For User Supplied, select to use the Default JMS function selector or enter the fully qualified name of the FunctionSelector implementation class.
  9. Once the Export with JMS Binding has been created, select the export in the Assembly Editor and in the Properties view, select the Description content pane. The Export's name and description are listed and may be modified as necessary.
  10. Select the Binding content pane to see many more options.
  11. Save the assembly diagram.

Migrating the IBM Web Service binding (SOAP/JMS)

The IBM Web Service binding (SOAP/JMS) for a BPEL process or other service type can be migrated to the recommended SCA construct.

In WebSphere Studio Application Developer Integration Edition, this binding type gave clients the ability to communicate with a BPEL process or other service type by invoking an IBM Web Service, where the communication protocol was JMS and the message adhered to the SOAP encoding rules.

The following is an example of the conventions used when generating an IBM Web Service (SOAP/JMS) for a 5.1 BPEL service. The JNDI name of the generated IBM Web Service was a combination of the BPEL's name, target namespace, and valid-from timestamp, as well as the name of the interface (WSDL port type that the deployment code was generated for). For example, these attributes can be found by examining the BPEL process' properties in the BPEL editor on the Description and Server content tabs:

Table 6. Generated namespace
Process name MyService
Target namespace http://www.example.com/process87787141/
Valid From Jan 01 2003 02:03:04
Interface ProcessPortType

The generated namespace for this example is then com/example/www/process87787141/MyService20030101T020304_ProcessPortTypePT.

In WebSphere Studio Application Developer Integration Edition when the IBM Web Service binding (SOAP/JMS) was selected as the deployment type for a BPEL process or other service type, the following options were given:

A WSDL file specifying the IBM Web Service SOAP/JMS binding and service is created in the generated EJB project but not in the service project itself. This means that you must manually locate that file and copy it to your business integration module project if it is important that the IBM Web Service client code must not change. By default, this WSDL file was created in the EJB project at ejbModule/META-INF/wsdl/<business process name>_ <business process interface port type name>_JMS.wsdl

The WSDL PortType and Messages of the business process interface are actually copied to this WSDL file as well rather than referencing the existing WSDL PortType and Messages defined in the service project.

If it is important that the IBM Web Service client code remain unchanged after migration, then the information in this file will be needed for the manual migration steps below.

There are two options for migrating the WebSphere Studio Application Developer Integration Edition SOAP/JMS process binding. The choice will have to be made whether to migrate the client to the SCA programming model or to leave it as a web services client:

Note: After the manual migration steps have been completed, the client must be migrated to the new programming model as well. See the appropriate topic for the following client types:

Table 7. Further information for migrating clients
Client type For further information see
IBM Web service client "Migrating the IBM Web service (SOAP/JMS) client"

Migration option 1 for the IBM Web Service binding (SOAP/JMS)

The first migration option for the WebSphere Studio Application Developer Integration Edition SOAP/JMS binding is to make the service accessible to a Web services client.

The Export with Web Service Binding makes an SCA component accessible by an external Web services client. To create an Export with Web Service Binding, follow these steps:

  1. Open the Assembly Editor for the module created by the migration wizard.
  2. Create an Export with SCA Binding for each service interface that had an IBM Web Service (SOAP/JMS) binding generated for it in WebSphere Studio Application Developer Integration Edition:
    1. Right-click the SCA component in the Assembly Editor.
    2. Select Export....
    3. Select Web Service Binding.
    4. If there are multiple interfaces for the component, select the interface(s) to export with this binding type.
    5. Select the transport soap/jms.
  3. Once the Web Services Export is created, select the export in the Assembly Editor and in the Properties view, select the Description content pane. The Export's name and description are listed and may be modified as necessary.
  4. Save the assembly diagram.
  5. Select the Binding content pane and you will see that an IBM Web Service WSDL Binding and Service has been generated directly in the module's project folder. It will be named component-that-was-exported Export WSDL PortType name Jms_Service.wsdl. If you inspect that file, you will find that the Document/Literal wrapped binding is used by default, as it is the preferred style in 6.x. This is the WSDL that IBM Web Service clients will use to invoke the service.
  6. Follow these steps to generate a new web service binding and service if preserving client code is desired:
    1. Copy the 5.1 WSDL file from the 5.1 generated EJB project at ejbModule/META-INF/wsdl/business process name/business process interface port type nameJMS.wsdl to the business integration module project.
    2. After copying over the file and rebuilding the module, you may see error messages because the XML schema types, WSDL messages, and WSDL port types used by the Web service are duplicated in the IBM Web Service WSDL file in 5.1. To fix this, delete those duplicate definitions from the IBM Web Service binding/service WSDL and in their place add a WSDL import for the real interface WSDL. Note: It is important to note that when WebSphere Studio Application Developer Integration Edition generated the IBM Web Service deployment code, it did modify the schema definitions in some cases. This could cause inconsistencies for existing clients that use the IBM Web Service WSDL. For example, the "elementFormDefault" schema attribute was set to "qualified" in the inline schema generated in the IBM Web Service WSDL even if the original schema definition was not qualified. This would cause the following error to be generated during runtime: WSWS3047E: Error: Cannot deserialize element.
    3. Right-click on this WSDL file you just copied to the business integration module and select Open With then WSDL Editor.
    4. Go to the Source tab. Delete all WSDL PortTypes and Messages defined in this file.
    5. Now you will see the error: The '<portType>' port type specified for the '<binding>' binding is undefined. To fix this, in the WSDL editor in the Graph tab, right-click in the Imports section and select Add Import.
    6. In the Properties view on the General tab, click the ... button to the right of the Location field. Browse to the interface WSDL where the WSDL message and port type definitions are located and click OK to import the interface WSDL into the service/binding WSDL.
    7. Save the WSDL file.
    8. Refresh/rebuild the project. Switch to the Business Integration perspective. Open the module's Assembly Diagram in the Assembly Editor.
    9. In the project explorer view, expand the module that you are migrating and expand the Web Service Ports logical category. You should see the port that exists in the binding/service WSDL listed. Drag and drop it on to the Assembly Editor.
    10. Choose to create an Export with Web Service Binding and select the appropriate port name. This will create the Export that uses the old binding/service such that existing Web service clients do not have to change. If you select the export you just created in the Assembly Editor and go to the Properties view, on the Binding tab you should see that the 5.1 port and service names have been filled in for you.
    11. Save all changes.
    12. Just before deploying the application, you can change the generated Web project's configuration to match the 5.1 service address (you have to make these changes every time you make any changes to the SCA module that cause this file to be regenerated). If you look at the IBM Web Service WSDL Service definition that you are reusing from 5.1 you will see the service address that the 5.1 client is coded to:<wsdlsoap:address location="http://localhost:9080/MyServiceWeb/services/MyServicePort"/>
    13. In order to make the 6.x generated Web project artifacts match this old service address, you should modify the generated Web project's deployment descriptor. Open the deployment descriptor in WebSphere Integration Developer and on the Servlets tab, add an additional URL Mapping that is very similar to the existing URL mapping for that export, with the same servlet name but a different URL pattern.
    14. Also, if you need to modify the context root of this web project such that it matches the context root in the original service address (in this example the context root is "MyServiceWeb"), then you can open the deployment descriptor for the J2EE Enterprise Application that this web project is in and change the context root of that web module to match the old service address's context root. You may see the following error which you can ignore: CHKJ3017E: Web Project: <WEB PROJ NAME> is mapped to an invalid Context root: <NEW CONTEXT ROOT> in EAR Project: <APP NAME>.
Migration option 2 for the IBM Web Service binding (SOAP/JMS)

The second migration option for the WebSphere Studio Application Developer Integration Edition SOAP/JMS process binding is to make business processes accessible to a non-SCA entity (for example, JSP or a Java client).

The Standalone Reference makes an SCA component accessible by any external client. To create a Standalone Reference, follow these steps:

  1. Open the Assembly Editor for the module created by the migration wizard.
  2. Create a Standalone Reference for each BPEL process interface that had an IBM Web Service (SOAP/JMS) binding generated for it in WebSphere Studio Application Developer Integration Edition:
    1. Select the Standalone References item from the toolbar.
    2. Click the canvas of the Assembly Editor to create a Standalone References SCA entity.
    3. Select the Wire item from the toolbar.
    4. Click the Standalone References entity to select it as the source of the wire.
    5. Click the BPEL SCA component to select it as the target of the wire.
    6. You will see an alert Matching reference will be created on the source node. Would you like to continue?, click OK.
    7. Select the Standalone References entity that was just created and in the Properties view select the Description content pane.
    8. Expand the References link and select the reference that was just created. The reference's name and description are listed and may be modified as necessary.
    9. If there are multiple interfaces for the process, select the interface(s) to export with this binding type.
    10. Save the assembly diagram.

Migrating the IBM Web Service binding (SOAP/HTTP)

The IBM Web Service binding (SOAP/HTTP) for a BPEL process or other service type can be migrated to the recommended SCA construct.

In WebSphere Studio Application Developer Integration Edition, this binding type gave clients the ability to communicate with a BPEL process or other service type by invoking an IBM Web Service, where the communication protocol was HTTP and the message adhered to the SOAP encoding rules.

The following is an example of the conventions used when generating an IBM Web Service (SOAP/HTTP) for a 5.1 BPEL service. The JNDI name of the generated IBM Web Service was a combination of the BPEL's name, target namespace, and valid-from timestamp, as well as the name of the interface (WSDL port type that the deployment code was generated for). For example, these attributes can be found by examining the BPEL process' properties in the BPEL editor on the Description and Server content tabs:

Table 8. Generated namespace
Process name MyService
Target namespace http://www.example.com/process87787141/
Valid From Jan 01 2003 02:03:04
Interface ProcessPortType

The generated namespace for this example is then com/example/www/process87787141/MyService20030101T020304_ProcessPortTypePT.

In WebSphere Studio Application Developer Integration Edition when the IBM Web Service binding (SOAP/HTTP) was selected as the deployment type for a BPEL process or other service type, the following options were given:

A WSDL file specifying the IBM Web Service SOAP/HTTP binding and service is created in the generated Web and EJB projects but not in the service project itself. This means that you must manually locate that file and copy it to your business integration module project if it is important that the IBM Web Service client code must not change. By default, this WSDL file was created in the Web project at WebContent/WEB-INF/wsdl/<business process name>_<business process interface port type name>_HTTP.wsdl

The WSDL PortType and Messages of the business process interface are actually copied to this WSDL file as well rather than referencing the existing WSDL PortType and Messages defined in the service project.

If it is important that the IBM Web Service client code remain unchanged after migration, then the information in this file will be needed for the manual migration steps below.

There are two options for migrating the WebSphere Studio Application Developer Integration Edition SOAP/HTTP process binding. The choice will have to be made whether to migrate the client to the SCA programming model or to leave it as a Web services client:

Note: After the manual migration steps have been completed, the client must be migrated to the new programming model as well. See the appropriate topic for the following client types:

Table 9. Further information for migrating clients
Client type For further information see
IBM Web service client "Migrating the IBM Web service (SOAP/HTTP) client"

Migration option 1 for the IBM Web Service (SOAP/HTTP) binding

The first migration option for the WebSphere Studio Application Developer Integration Edition SOAP/HTTP process binding is to make business processes accessible to a Web services client.

The Export with Web Service Binding makes an SCA component accessible by an external Web services client. To create an Export with Web Service Binding, follow these steps:

  1. Open the Assembly Editor for the module created by the Migration wizard.
  2. Create an Export with SCA Binding for each BPEL process interface that had an IBM Web Service (SOAP/HTTP) binding generated for it in WebSphere Studio Application Developer Integration Edition by right-clicking the BPEL component in the Assembly Editor.
  3. Select Export....
  4. Select Web Service Binding.
  5. If there are multiple interfaces for the component, select the interface(s) to export with this binding type.
  6. Select the transport soap/http.
  7. Once the Web Services Export is created, select the export in the Assembly Editor and in the Properties view, select the Description content pane. The Export's name and description are listed and may be modified as necessary.
  8. Save the assembly diagram.
  9. Follow these steps to generate a new web service binding and service if preserving client code is desired:
    1. Copy the 5.1 WSDL file from the 5.1 generated EJB project at ejbModule/META-INF/wsdl/business process name/business process interface port type name_HTTP.wsdl to the business integration module project.
    2. After copying over the file and rebuilding the module, you may see error messages as the XML schema types, WSDL messages, and WSDL port types used by the Web service are duplicated in the IBM Web Service WSDL file in 5.1. To fix this, delete those duplicate definitions from the IBM Web Service binding/service WSDL and in their place add a WSDL import for the real interface WSDL. Note: It is important to note that when WebSphere Studio Application Developer Integration Edition generated the IBM Web Service deployment code, it did modify the schema definitions in some cases. This could cause inconsistencies for existing clients that use the IBM Web Service WSDL. For example, the "elementFormDefault" schema attribute was set to "qualified" in the inline schema generated in the IBM Web Service WSDL even if the original schema definition was not qualified. This would cause the following error to be generated during runtime: WSWS3047E: Error: Cannot deserialize element.
    3. Right-click on this WSDL file you just copied to the business integration module and select Open With then WSDL Editor.
    4. Go to the Source tab. Delete all WSDL PortTypes and Messages defined in this file.
    5. Now you will see the error: The '<portType>' port type specified for the '<binding>' binding is undefined. To fix this, in the WSDL editor in the Graph tab, right-click in the Imports section and select Add Import.
    6. In the Properties view on the General tab, click the ... button to the right of the Location field. Browse to the interface WSDL where the WSDL message and port type definitions are located and click OK to import the interface WSDL into the service/binding WSDL.
    7. Save the WSDL file.
    8. Refresh/rebuild the project. Switch to the Business Integration perspective. Open the module's Assembly Diagram in the Assembly Editor.
    9. In the project explorer view, expand the module that you are migrating and expand the Web Service Ports logical category. You should see the port that exists in the binding/service WSDL listed. Drag and drop it on to the Assembly Editor.
    10. Choose to create an Export with Web Service Binding and select the appropriate port name. This will create the Export that uses the old binding/service such that existing Web service clients do not have to change. If you select the export you just created in the Assembly Editor and go to the Properties view, on the Binding tab you should see that the 5.1 port and service names have been filled in for you.
    11. Save all changes.
    12. Just before deploying the application, you can change the generated Web project's configuration to match the 5.1 service address (you have to make these changes every time you make any changes to the SCA module that cause this file to be regenerated). If you look at the IBM Web Service WSDL service definition that you are reusing from 5.1, you will see the service address that the 5.1 client is coded to:<wsdlsoap:address location="http://localhost:9080/MyServiceWeb/services/MyServicePort"/>
    13. In order to make the 6.x generated Web project artifacts match this old service address, you should modify the generated Web project's deployment descriptor. Open the deployment descriptor in WebSphere Integration Developer and on the Servlets tab, add an additional URL Mapping that is very similar to the existing URL mapping for that export, with the same servlet name but a different URL pattern.
    14. Also, if you need to modify the context root of this web project such that it matches the context root in the original service address (in this example the context root is "MyServiceWeb"), then you can open the deployment descriptor for the J2EE Enterprise Application that this web project is in and change the context root of that web module to match the old service address's context root. You may see the following error which you can ignore: CHKJ3017E: Web Project: <WEB PROJ NAME> is mapped to an invalid Context root: <NEW CONTEXT ROOT> in EAR Project: <APP NAME>.
Migration option 2 for the IBM Web Service (SOAP/HTTP) binding

The second migration option for the WebSphere Studio Application Developer Integration Edition SOAP/HTTP process binding is to make business processes accessible to a non-SCA entity (for example, JSP or a Java client).

The Standalone Reference makes an SCA component accessible by any external client. To create a Standalone Reference, follow these steps:

  1. Open the Assembly Editor for the module created by the migration wizard.
  2. Create a Standalone Reference for each interface that had an IBM Web Service (SOAP/HTTP) binding generated for it in WebSphere Studio Application Developer Integration Edition:
    1. Select the Standalone References item from the toolbar.
    2. Click the canvas of the Assembly Editor to create a Standalone References SCA entity.
    3. Select the Wire item from the toolbar.
    4. Click the Standalone References entity to select it as the source of the wire.
    5. Click the SCA component to select it as the target of the wire.
    6. You will see an alert Matching reference will be created on the source node. Would you like to continue?, click OK.
    7. Select the Standalone References entity that was just created and in the Properties view select the Description content pane.
    8. Expand the References link and select the reference that was just created. The reference's name and description are listed and may be modified as necessary.
    9. If there are multiple interfaces for the process, select the interface(s) to export with this binding type.
    10. Save the assembly diagram.

Migrating the Apache Web Service binding (SOAP/HTTP)

The Apache Web Service binding (SOAP/HTTP) for a BPEL process or other service type can be migrated to the recommended SCA construct.

In WebSphere Studio Application Developer Integration Edition, this binding type gave clients the ability to communicate with a BPEL process or other service type by invoking an Apache Web Service.

In WebSphere Studio Application Developer Integration Edition when the Apache Web Service binding was selected as the deployment type for a BPEL process or other service type, the following options were given:

A WSDL file specifying the Apache SOAP binding and service is created in the service project. By default it is created in the same directory as the service it is wrapping with the name <business process name>_<business process interface port type name>_SOAP.wsdl. The WSDL PortType and Messages of the business process interface are used by this binding and service directly. After migration, you should not use this WSDL for anything aside from perhaps using the same namespace, port, and service names in the new WSDL that will be generated for you in Version 6.x.

There are two options for migrating the WebSphere Studio Application Developer Integration Edition Web Service process binding. The choice will have to be made whether to migrate the client to the SCA programming model or to leave it as an IBM Web Services programming model. There is no binding that is equivalent to the Apache Web Service (SOAP/HTTP) binding type anymore in the 6 SCA programming model.

You should migrate this Apache Web service to use the IBM Web Service engine. See the topic "Migrating the IBM Web Service (SOAP/HTTP) binding" for instructions on how to perform this migration and create an IBM Web Service (SOAP/HTTP).

Migrating to the SCA programming model

For any free-form Java code that interacts with a WebSphere Studio Application Developer Integration Edition service, this section will show how to migrate from the WSIF programming model to the new SCA programming model where the data flowing through the application is stored in Eclipse Service Data Objects (SDOs). This section will also show you how to manually migrate the most common client types to the new programming model.

For any BPEL processes that contain Java snippets, this section explains how to migrate from the old Java snippet API to the new Java snippet API where the data flowing through the application is stored in Eclipse Service Data Objects (SDOs). Whenever possible, the snippets are migrated automatically by the migration wizard but there are snippets that the migration wizard can not fully migrate, meaning manual steps are required to complete the migration.

Here is a summary of the programming model changes:

V5.1 Programming Model
  1. WSIF and WSDL based
  2. Generated proxies for services
  3. Beans and format handlers for types
V6.x Programming Model (more Java-centric)
  1. SCA services based on SDOs with doclet tags
  2. Interface bindings for services
  3. SDOs and Databindings for types

Migrating WSIFMessage API calls to SDO APIs

The following section details how to migrate from the old WebSphere Business Integration Server Foundation Version 5.1 programming model where the data flowing through the application is represented as WSIFMessage objects with a generated interface that was strongly-typed to the new WebSphere Process Server programming model where the data is represented as Service Data Objects (SDOs) and no strongly-typed interface is generated.

Table 10. Changes and Solutions for migrating WSIFMessage API calls to SDO APIs
Change Solution
WSIFMessage-based wrapper classes are no longer generated for WSDL message types, nor are the Java bean helper classes generated for complex schema types. When writing code that interacts with SCA services, the generic SDO APIs must be used to manipulate the commonj.sdo.DataObject messages that hold the data that flows through the application.

WSDL message definitions that have a single simple-typed part will now be represented by a simple Java type that directly represents the part instead of having a wrapper around the actual data. If the single message part is a complex type, then the data is represented as a DataObject that adheres to the complex type definition.

WSDL message definitions that have multiple parts now correspond to a DataObject that has properties for all of the message parts, where complexTypes are represented as "reference-type" properties of the parent DataObject, accessible via the getDataObject and setDataObject methods.

Strongly-typed getter methods for WSIFMessage parts and generated Java beans should not be used. Weakly-typed SDO API should be used to get the DataObject properties.
Strongly-typed setter methods for BPEL variables' message parts are no longer available. Weakly-typed SDO API must be used to set the DataObject properties.
Weakly-typed getter methods for WSIFMessage properties should no longer be used. Weakly-typed SDO API must be used to set the DataObject properties.
Weakly-typed setter methods for WSIFMessage properties should no longer be used. Weakly-typed SDO API must be used to set the DataObject properties.
All WSIFMessage API calls should be migrated to the SDO API where possible. Migrate the call to an equivalent SDO API call where possible. Redesign logic if not possible.

Migrating WebSphere Business Integration Server Foundation client code

This section shows how to migrate the various client types that were possible for the WebSphere Business Integration Server Foundation 5.1 service types.

Migrating the EJB client

This topic shows how to migrate clients that use an EJB interface to invoke a service.

  1. Drag and drop the Export with SCA Binding from the migrated module onto this new module's Assembly Editor. This will create an Import with SCA Binding. In order for a client to obtain a reference to this import, a Standalone Reference must be created.
  2. On the palette, select the Standalone References item. Click the Assembly Editor canvas once to create a new standalone reference for this new module.
  3. Select the wire tool and click the service reference and then click Import.
  4. Click OK when alerted that a matching reference will be created on the source node.
  5. You will be asked: It is easier for a Java client to use a Java interface with this reference - would you like to convert the WSDL reference to a compatible Java reference?:
    1. Answer Yes if you would like the client to look up this service and cast it as a Java class to invoke it using a Java interface. This new Java interface takes the name of the WSDL PortType, where the package of the interface is derived from the namespace of the WSDL PortType. There is a method defined for each operation defined on the WSDL PortType, and each WSDL message part is represented as an argument to the interface methods.
    2. Answer No if you would like the client to look up this service and use the generic com.ibm.websphere.sca.Service interface to invoke it using the invoke operation as a generic SCA service.
  6. Rename the standalone reference to a more meaningful name if appropriate by selecting the Standalone References component in the Assembly Editor. Go to the Properties view, to the Details tab, drilling down to and selecting the reference that was just created, and modifying the name. Remember the name you chose for this reference because the client will need to use this name when invoking the locateService method of the com.ibm.websphere.sca.ServiceManager instance.
  7. Click Save to save the Assembly diagram.

The client must have this new module on its local classpath in order to access the migrated EJB module running on the server.

The following shows what the client code looks like for a service of type "CustomerInfo":

// Create a new ServiceManager
ServiceManager serviceManager = ServiceManager.INSTANCE;

// Locate the CustomerInfo service
CustomerInfo customerInfoService = (CustomerInfo) serviceManager.locateService 
("<name-of-standalone-reference-from-previous-step");

	// Invoke the CustomerInfo service
	System.out.println("	[getMyValue] getting customer info...");
	DataObject customer = customerInfoService.getCustomerInfo(customerID);

The client must change how the message is constructed. In the messages were based on the WSIFMessage class but now they should be based on the commonj.sdo.DataObject class.

Migrating the EJB process binding client

This topic shows how to migrate clients that use the WSIF EJB process binding to access a BPEL service.

Clients that used the EJB Process Binding to invoke a business process should now use either the SCA API to invoke the service (the migrated business process must have an Export with SCA Binding) or the IBM Web Service Client API to invoke the service (the migrated business process must have an Export with Web Service Binding).

See the topics "Migrating the EJB Client", "Migrating the IBM Web Service (SOAP/JMS) client", or "Migrating the IBM Web Service (SOAP/HTTP) client" for more information on generating such clients.

Migrating the IBM Web Service (SOAP/JMS) client

This topic shows how to migrate clients that use Web Service APIs (SOAP/JMS) to invoke a service.

No migration is needed for existing clients during migration. Note that you must manually modify the generated Web project (create a new servlet mapping) and sometimes have to modify the Web project's context root in the enterprise application deployment descriptor to publish the service to the exact same address that it was published to on WebSphere Business Integration Server Foundation. See the topic "Migrating the IBM Web Service binding (SOAP/JMS)".

It is important to note that unlike 5.1 where a WSIF or RPC client proxy could be generated, in 6.x the tools only support RPC client generation because RPC is the 6.x preferred API over the WSIF API.

Note: To generate new client proxy from WebSphere Integration Developer, you must have a WebSphere Process Server or WebSphere Application Server installed.

  1. Ensure that you have a WebSphere Process Server or WebSphere Application Server installed.
  2. In the Resources or Java perspective, find the WSDL file corresponding to the Export with Web Service Binding then right-click and select Web Services -> Generate Client (Note that this wizard is very similar to the 5.1 wizard).
  3. For Client Proxy Type choose Java proxy and click Next.
  4. The location of the WSDL should be filled in. Click Next.
  5. Next you must select the appropriate options to specify your client environment configuration including the Web service runtime and server, J2EE version, client type (Java, EJB, Web, Application Client). Click Next.
  6. Finish the remaining steps to generate the client proxy.
Migrating the IBM Web Service (SOAP/HTTP) client

This topic shows how to migrate clients that use Web Service APIs (SOAP/HTTP) to invoke a service.

No migration is needed for existing clients during migration. Note that you must manually modify the generated Web project (create a new servlet mapping) and sometimes have to modify the Web project's context root in the enterprise application deployment descriptor to publish the service to the exact same address that it was published to on WebSphere Business Integration Server Foundation. See the topic "Migrating the IBM Web Service binding (SOAP/HTTP)".

If design changes have occurred and you would like to generate a new client proxy, the following steps will show you how to do that. It is important to note that unlike 5.1 where a WSIF or RPC client proxy could be generated, in 6.x the tools only support RPC client generation because RPC is the 6.x preferred API over the WSIF API.

Note: To generate new client proxy from WebSphere Integration Developer, you must have a WebSphere Process Server or WebSphere Application Server installed.

  1. Ensure that you have a WebSphere Process Server or WebSphere Application Server installed.
  2. Select the WSDL file corresponding to the Export with Web Service Binding then right-click and select Web Services -> Generate Client (Note that this wizard is very similar to the 5.1 wizard).
  3. For Client Proxy Type choose Java proxy and click Next.
  4. The location of the WSDL should be filled in. Click Next.
  5. Next you must select the appropriate options to specify your client environment configuration including the Web service runtime and server, J2EE version, client type (Java, EJB, Web, Application Client). Click Next.
  6. Finish the remaining steps to generate the client proxy.

Migrating the Apache Web Service (SOAP/HTTP) client

The Apache Web Service client APIs are not appropriate for invoking a WebSphere Integration Developer service. Client code must be migrated to use the IBM Web Service (SOAP/HTTP) client APIs.

See the topic, "Migrating the IBM Web Service (SOAP/HTTP) client" for more information.

In 5.1 if a client proxy was automatically generated, that proxy used WSIF APIs to interact with the service. In 6.x the tools only support RPC client generation because RPC is the 6.x preferred API over the WSIF API.

Note: To generate new client proxy from WebSphere Integration Developer, you must have a WebSphere Process Server or WebSphere Application Server installed.

  1. Ensure that you have a WebSphere Process Server or WebSphere Application Server installed.
  2. Select the WSDL file corresponding to the Export with Web Service Binding then right-click and select Web Services -> Generate Client (Note that this wizard is very similar to the 5.1 wizard).
  3. For Client Proxy Type choose Java proxy and click Next.
  4. The location of the WSDL should be filled in. Click Next.
  5. Next you must select the appropriate options to specify your client environment configuration including the Web service runtime and server, J2EE version, client type (Java, EJB, Web, Application Client). Click Next.
  6. Finish the remaining steps to generate the client proxy.

Migrating the JMS client

Clients that communicated with a 5.1 service via the JMS API (sending a JMS message to a queue) may require manual migration. This topic shows how to migrate clients that use JMS APIs (sending a JMS message to a queue) to invoke a service.

You must ensure that the Export with JMS Binding that you created in a previous step will be able to accept this text or object message with no changes. You may need to write a custom data binding to accomplish this. See the section "Migrating the JMS and the JMS process bindings" for more information.

The client must change how the message is constructed. The messages were previously based on the WSIFMessage class but now they should be based on the commonj.sdo.DataObject class. See the section "Migrating WSIFMessage API calls to SDO APIs" for more details on how to do this manual migration.

Migrating the business process choreographer generic EJB API client

This topic shows how to migrate clients that use the 5.1 process choreographer generic EJB API to invoke a BPEL service.

There is a new version of the Generic EJB API that uses DataObjects as its message format. The client must change how the message is constructed. The messages were previously based on the WSIFMessage class but now they should be based on the commonj.sdo.DataObject class. Note that the Generic EJB API has not changed significantly, as the ClientObjectWrapper still provides a message wrapper around the particular message format.

Ex: DataObject dobj = myClientObjectWrapper.getObject();
String result = dobj.getInt("resultInt");

The JNDI name of the old Generic EJB that takes WSIFMessage objects is:

GenericProcessChoreographerEJB
JNDI Name: com/ibm/bpe/api/BusinessProcessHome
Interface: com.ibm.bpe.api.BusinessProcess

There are two generic EJBs in 6.0 as the human task operations are now available as a separate EJB. The 6.0 JNDI names of these Generic EJBs are:

GenericBusinessFlowManagerEJB
JNDI Name: com/ibm/bpe/api/BusinessFlowManagerHome
Interface: com.ibm.bpe.api.BusinessFlowManager

HumanTaskManagerEJB
JNDI Name: com/ibm/task/api/TaskManagerHome
Interface: com.ibm.task.api.TaskManager
Migrating the business process choreographer generic Messaging API client and the JMS process binding client

There is no generic messaging API in WebSphere Process Server. See the section "Migrating the JMS and JMS process bindings" to choose a different way to expose the business process to consumers and rewrite the client according to the chosen binding.

Migrating the business process choreographer Web client

This topic shows how to migrate the 5.1 process choreographer Web client settings and custom JSPs.

The Migration wizard preserves the 5.1 Web client settings and in the Human Task Editor, you may not edit these settings. You should create new Web client settings and JSPs using the WebSphere Integration Developer 6.x.

Migrating Web Client modifications
In 5.1 you could modify the look and feel of the Struts-based Web client by modifying its JSP Header.jsp and style sheet dwc.css.

Since the 6.x Web client (renamed to the Business Process Choreographer Explorer) is based on Java Server Faces (JSF) instead of Struts, automatic migration of Web client modifications is not possible. Therefore, it is recommended that you see the "Business Process Choreographer Explorer" documentation for details on customizing the 6.x version of this application.

User-defined JSPs could be defined for business processes and for staff activities. The Web client uses these JSPs to display input and output messages for the process and activities.

These JSPs are particularly useful when:

  1. Messages have non-primitive parts to enhance the usability of the message's data structure.
  2. You want to extend the Web client's capabilities.

There are more and different options available when specifying Web client settings for a 6.x process, so you will have to use WebSphere Integration Developer to redesign the Web client settings for migrated processes and activities:

  1. Select the process canvas or an activity in the process.
  2. In the Properties view, select the Client tab to redesign the Web client settings.
  3. Manually migrate any user-defined JSP:
    1. See the "Migrating to the SCA programming model" section for programming model changes.
    2. The Web client uses the Generic APIs to interact with business processes. See the sections that show how to migrate calls to these generic APIs..
  4. Specify the name of the new JSP in the 6.x Web client settings for the process
Note: Mapping JSPs are not needed with the 6.x Business Process Choreographer Explorer because DataObjects do not need any custom mapping.

Migrating WebSphere Business Integration Server Foundation BPEL Java snippets

For any BPEL processes that contain Java snippets, this section details how to migrate from the old Java snippet API to the new Java snippet API where the data flowing through the application is stored as Eclipse Service Data Objects (SDOs).

See the section "Migrating from the WSIFMessage API calls to SDO APIs" for migration steps to perform specific to the WSIFMessage to SDO transition.

Whenever possible, the snippets are automatically migrated by the migration wizard but there are snippets that the migration wizard can not fully migrate. This requires extra manual steps to complete the migration. See the Limitations topic for details on the types of Java snippets that must be migrated manually. Whenever one of these snippets is encountered, the Migration wizard will explain why it can not be automatically migrated and issue a warning or error message.

The following table detail the changes in the BPEL Java snippet programming model and API from Process Choreographer version 5.1 to 6.x:

Table 11. Changes and solutions for migrating WebSphere Business Integration Server Foundation BPEL Java snippets
Change Solution
WSIFMessage-based wrapper classes are no longer generated for WSDL message types, nor are the Java bean helper classes generated for complex schema types. BPEL variables can be directly accessed by name. Note that for BPEL variables whose WSDL message definition has a single-part, these variables will now directly represent the part instead of having a wrapper around the actual data. Variables whose message type has multiple parts will have a DataObject wrapper around the parts (where the wrapper in WebSphere Application Developer Integration Edition was a WSIFMessage).

Because BPEL variables can be used directly in 6.x snippets, there is less need for local variables than there was in 5.1.

The strongly-typed getters for the BPEL variables implicitly initialized the WSIFMessage wrapper object around the message parts. There is no 'wrapper' object for BPEL variables whose WSDL message definition has only a single part: in this case the BPEL variables directly represent the part (In the case where the single part is an XSD simple type, the BPEL variable will be represented as the Java object wrapper type such as java.lang.String, java.lang.Integer, etc). BPEL variables with multi-part WSDL message definitions are handled differently: there is still a wrapper around the parts and this DataObject wrapper must be explicitly initialized in the 6.x Java snippet code if it has not already been set by a previous operation.

If any local variables from the 5.1 snippets had the same name as the BPEL variable there may be conflicts so try to remedy this situation if possible.

WSIFMessage objects are no longer used to represent BPEL variables. If any custom Java classes invoked from the Java snippets have a WSIFMessage parameter it will need to be migrated such that it accepts/returns a DataObject.
Strongly-typed getter methods for BPEL variables are no longer available. The variables can be directly accessed by name. Note that for BPEL variables whose WSDL message definition has a single-part will now directly represent the part instead of having a wrapper around the actual data. Variables whose message type has multiple parts will have a DataObject wrapper around the parts (where the wrapper in WebSphere Application Developer Integration Edition was a WSIFMessage).
Strongly-typed setter methods for BPEL variables are no longer available. The variables can be directly accessed by name. Note that for BPEL variables whose WSDL message definition has a single-part, these variables will now directly represent the part instead of having a wrapper around the actual data. Variables whose message type has multiple parts will have a DataObject wrapper around the parts (where the wrapper in WebSphere Application Developer Integration Edition was a WSIFMessage).
Weakly-typed getter methods for BPEL variables that return a WSIFMessage are no longer available. The variables can be directly accessed by name. Note that for BPEL variables whose WSDL message definition has a single-part, these variables will now directly represent the part instead of having a wrapper around the actual data. Variables whose message type has multiple parts will have a DataObject wrapper around the parts (where the wrapper in WebSphere Application Developer Integration Edition was a WSIFMessage).

Note that there were two variations of the getVariableAsWSIFMessage method:

getVariableAsWSIFMessage(String variableName)
getVariableAsWSIFMessage(String variableName, boolean forUpdate)

For a Java snippet activity, the default access is read-write. You can change this to read-only by specifying @bpe.readOnlyVariables with the list of names of the variables in a comment in the snippet. For example, you could set variable B and variable D to read only as follows:

variableB.setString("/x/y/z", variableA.getString("/a/b/c")); 
// @bpe.readOnlyVariables names="variableA"
variableD.setInt("/x/y/z", variableC.getInt("/a/b/c")); 
// @bpe.readOnlyVariables names="variableC"

Additionally, if you have a Java snippet in a condition, the variables are read-only by default, but you can make them read-write by specifying @bpe.readWriteVariables...

Weakly-typed setter methods for BPEL variables are no longer available. The variables can be directly accessed by name. Note that for BPEL variables whose WSDL message definition has a single-part, these variables will now directly represent the part instead of having a wrapper around the actual data. Variables whose message type has multiple parts will have a DataObject wrapper around the parts (where the wrapper in WebSphere Application Developer Integration Edition was a WSIFMessage).
Weakly-typed getter methods for BPEL variables message parts are not appropriate for single-part messages and have changed for multi-part messages. Migrate to the weakly-typed getter method for BPEL variables (DataObject's) properties.

Note that for BPEL variables whose WSDL message definition has a single-part, the BPEL variable directly represents the part and the variable should be accessed directly without using a getter method.

There were two variations of the getVariablePartAsObject method:

getVariablePartAsObject(String variableName, String partName)
getVariablePartAsObject(String variableName, String partName,boolean 
forUpdate)

For multi-part messages, equivalent functionality is provided by this method in 6.x:

getVariableProperty(String variableName, QName propertyName);  		

In 6.x there is no notion of using a variable for read-only access (which was the case in 5.1 for the first method above as well as the second method with forUpdate='false'). The variable is directly used in the 6.x snippet and it is always able to be updated.

Weakly-typed setter methods for BPEL variables' message parts are not appropriate for single-part messages and have changed for multi-part messages. Migrate to the weakly-typed setter method for BPEL variables' (DataObject's) properties.

Note that for BPEL variables whose WSDL message definition has a single-part, the BPEL variable directly represents the part and the variable should be accessed directly without using a setter method.

Calls to the following method must be migrated:

setVariableObjectPart(String variableName, String partName, 
Object data)  

For multi-part messages, equivalent functionality is provided by this method in 6.x:

setVariableProperty(String variableName, QName propertyName,
Serializable value);
Strongly-typed getter methods for BPEL partner links are no longer available. Migrate to the weakly-typed getter methods for BPEL partner links.
Strongly-typed setter methods for BPEL partner links are no longer available. Migrate to the weakly-typed setter methods for BPEL partner links.
Strongly-typed getter methods for BPEL correlation sets are no longer available.
V5.1 Snippet:
String corrSetPropStr = 
getCorrelationSetCorrSetAPropertyCustomerName();
int corrSetPropInt = 
getCorrelationSetCorrSetBPropertyCustomerId();
V6.x Snippet:
String corrSetPropStr = (String) getCorrelationSetProperty
("CorrSetA", new QName("CustomerName"));
int corrSetPropInt = ((Integer) getCorrelationSetProperty 
("CorrSetB", new QName("CustomerId"))).intValue();
Additional parameter needed for the weakly-typed getter methods for BPEL activity custom properties.
V5.1 Snippet:
String val = getActivityCustomProperty("propName");
V6.x Snippet:
String val = getActivityCustomProperty
("name-of-current-activity", "propName");
Additional parameter needed for the weakly-typed setter methods for BPEL activity custom properties.
V5.1 Snippet:
String newVal = "new value";
setActivityCustomProperty("propName", newVal); 
V6.x Snippet:
String newVal = "new value";
setActivityCustomProperty("name-of-current-activity", 
"propName", newVal);
The raiseFault(QName faultQName, Serializable message) method no longer exists. Migrate to the raiseFault(QName faultQName, String variableName) where possible; otherwise migrate to the raiseFault(QName faultQName) method or create a new BPEL variable for the Serializable object.

Migrating interactions with WebSphere Business Integration Adapters

If the JMS Client is a WebSphere Business Integration Adapter, you may need to use the External Service tooling to create the Import with JMS Binding. This Import uses a special data binding in order to serialize the SDO to the exact format that the WebSphere Business Integration Adapter expects.

To access the External Service tooling, follow these steps:

  1. Go to File -> New -> Other -> Business Integration and select External Service. Click Next.
  2. Choose Adapters. Click Next.
  3. Enter the path of the WebSphere Business Integration Adapter's configuration (.cfg) file and the directory that contains the XML schema of the business objects that the adapter uses. Click Next.
  4. Examine the query that is generated for you, and if it is correct click Run Query. In the Objects discovered by query list, select the objects that you want to add (one by one) and click the >> Add button.
  5. Accept the configuration parameters for the business object and click OK.
  6. Repeat for each business object.
  7. Click Next.
  8. For the Runtime business object format select SDO. For the Target project select the module you just migrated. Leave the Folder field blank.
  9. Click Finish.

This tool will migrate the old XSDs to the format expected by the special data binding so remove the old WebSphere Business Integration Adapter's XSDs from the module and use the new XSDs. If the module will not receive messages from the adapter, delete the Exports generated by this tool. If the module will not send any messages to the adapter, delete the Import. See the information center for more information on this feature.

Migrating WSDL interfaces that have SOAP-encoded array types

This section shows how to migrate or handle XML schemas that have SOAP-encoded array types.

Soap-encoded array types that have the RPC style will be treated as unbounded sequences of a concrete type in 6.x. It is not recommended that you create any XSD types that reference the soapend:Array types in any way, as the programming model is moving towards the Document/Literal wrapped style instead of the RPC style (although this could change).

There will be cases when an SCA application must invoke an external service that does use the soapend:Array type. There is no way to avoid this in some cases, so the following shows how to handle this situation:

Sample WSDL code:

		<xsd:complexType name="Vendor">
			<xsd:all>
				<xsd:element name="name" type="xsd:string" />
				<xsd:element name="phoneNumber" type="xsd:string" />
			</xsd:all>
		</xsd:complexType>
	</xsd:schema>

		<xsd:complexType name="Vendors">
			<xsd:complexContent mixed="false">
				<xsd:restriction base="soapenc:Array">
					<xsd:attribute wsdl:arrayType="tns:Vendor[]" ref="soapenc:arrayType" 
					xmlnxsd:wsdl="http://schemas.xmlsoap.org/wsdl/" />
				</xsd:restriction>
		</xsd:complexContent>

		<xsd:complexType name="VendorsForProduct">
			<xsd:all>
				<xsd:element name="productId" type="xsd:string" />
				<xsd:element name="vendorList" type="tns:Vendors" />
			</xsd:all>
		</xsd:complexType>

		<xsd:complexType name="Product">
			<xsd:all>
				<xsd:element name="productId" type="xsd:string" />
				<xsd:element name="productName" type="xsd:string" />
			</xsd:all>
		</xsd:complexType>

	<message name="doFindVendorResponse">
		<part name="returnVal" type="tns:VendorsForProduct" />
	</message>

	<operation name="doFindVendor">
		<input message="tns:doFindVendor" />
		<output message="tns:doFindVendorResponse" />
	</operation>

Sample code for a client of this Web Service:

  // Locate the vendor service and find the doFindVendor operation 

Service findVendor=(Service)ServiceManager.INSTANCE.locateService("vendorSearch");
OperationType doFindVendorOperationType=findVendor.getReference().getOperationType("doGoogleSearch");
            
 // Create the input DataObject
 DataObject doFindVendor=DataFactory.INSTANCE.create(doFindVendorOperationType.getInputType());
 doFindVendor.setString("productId", "12345");
 doFindVendor.setString("productName", "Refrigerator");

            // Invoke the FindVendor service
DataObject FindVendorResult = (DataObject)findVendor.invoke(doFindVendorOperationType, doFindVendor);
           
            // Display the results
            int resultProductId=findVendorResult.getString("productId");
            
            DataObject resultElements=findVendorResult.getDataObject("vendorList");
            Sequence results=resultElements.getSequence(0);
            for (int i=0, n=results.size(); i
            for (int i=0, n=results.size(); i

Here is another example where the data object's root type is a soapenc:Array. Note how the sampleElements DataObject is created using the second schema listed above. The DataObject's type is first obtained, and then the property for sampleStructElement is obtained. This is really a placeholder property and used only to get a valid property to use when adding the DataObjects to the sequence. A pattern like this can be used in your scenario:

Sample WSDL code:

<s:schema elementFormDefault="qualified" targetNamespace="http://soapinterop.org/xsd">
			<s:import namespace="http://schemas.xmlsoap.org/soap/encoding/" />
			<s:import namespace="http://schemas.xmlsoap.org/wsdl/" />
			<s:complexType name="SOAPStruct">
				<s:sequence>
					<s:element minOccurs="1" maxOccurs="1" form="unqualified" name="varInt" type="s:int" />
					<s:element minOccurs="1" maxOccurs="1" form="unqualified" name="varString" type="s:string" />
					<s:element minOccurs="1" maxOccurs="1" form="unqualified" name="varFloat" type="s:float" />
				</s:sequence>
			</s:complexType>

			<s:complexType name="ArrayOfSOAPStruct">
				<s:complexContent mixed="false">
					<s:restriction base="soapenc:Array">
						<s:attribute wsdl:arrayType="s0:SOAPStruct[]" ref="soapenc:arrayType" />
					</s:restriction>
				</s:complexContent>
			</s:complexType>
		</s:schema>

	<wsdl:message name="echoStructArraySoapIn">
		<wsdl:part name="inputStructArray" type="s0:ArrayOfSOAPStruct" />
	</wsdl:message>
	<wsdl:message name="echoStructArraySoapOut">
		<wsdl:part name="return" type="s0:ArrayOfSOAPStruct" />
	</wsdl:message>

		<wsdl:operation name="echoStructArray">
			<wsdl:input message="tns:echoStructArraySoapIn" />
			<wsdl:output message="tns:echoStructArraySoapOut" />
		</wsdl:operation>

	<schema targetNamespace="http://sample/elements"
		xmlns="http://www.w3.org/2001/XMLSchema"
		xmlns:tns="http://sample/elements">

	<element name="sampleStringElement" type="string"/>
	
	<element name="sampleStructElement" type="any"/>

</schema>

Sample code for a client of this Web Service:

// Create the input DataObject and get the SDO sequence for the any 
// element
DataFactory dataFactory=DataFactory.INSTANCE;
DataObject arrayOfStruct = dataFactory.create("http://soapinterop.org/xsd","ArrayOfSOAPStruct");
Sequence sequence=arrayOfStruct.getSequence("any");
            
// Get the SDO property for the sample  element that we want to use 
// here to populate the sequence
// We have defined this element in an XSD file, see SampleElements.xsd
DataObject sampleElements=dataFactory.create("http://sample/elements", 
"DocumentRoot");
Property property = sampleElements.getType().getProperty("sampleStructElement");
            
            // Add the elements to the sequence
DataObject item=dataFactory.create("http://soapinterop.org/xsd", "SOAPStruct");
item.setInt("varInt", 1);
item.setString("varString", "Hello");
item.setFloat("varFloat", 1.0f);
sequence.add(property, item);
item=dataFactory.create("http://soapinterop.org/xsd", "SOAPStruct");
item.setInt("varInt", 2);
item.setString("varString", "World");
item.setFloat("varFloat", 2.0f);
sequence.add(property, item);

// Invoke the echoStructArray operation
System.out.println("[client] invoking echoStructArray operation");
DataObject echoArrayOfStruct = (DataObject)interopTest.invoke("echoStructArray", arrayOfStruct);
            
// Display the results
if (echoArrayOfStruct!=null) {
    sequence=echoArrayOfStruct.getSequence("any");
    for (int i=0, n=sequence.size(); i<n; i++) {
      item=(DataObject)sequence.getValue(i);
      System.out.println("[client] item varInt = "+ 
          item.getInt("varInt")+" 
          varString="+item.getString("varString")+" 
          varFloat="+item.getFloat("varFloat"));

Migrating WebSphere Business Integration EJB projects

In WebSphere Studio Application Developer Integration Edition, EJB projects could have special WebSphere Business Integration features such as Extended Messaging (CMM) and CMP/A (Component-managed persistence anywhere). The deployment descriptors for such projects must be migrated and this section shows how to perform that migration.

To perform this migration, follow these steps:

  1. Copy the WebSphere Business Integration EJB project to the new workspace and import it from WebSphere Integration Developer using the File -> Import -> Existing Project into Workspace wizard. Optionally, you can also run the J2EE Migration wizard.
  2. Close all instances of WebSphere Integration Developer running in the 6.x workspace.
  3. Run the following script which will migrate the WebSphere Business Integration deployment descriptors in the EJB project:
    On Windows:
    SHARED_FOLDER_HOME/plugins/com.ibm.wbit.migration.wsadie_6.1.0/WSADIEEJBProjectMigration.bat
    On Linux:
    SHARED_FOLDER_HOME/plugins/com.ibm.wbit.migration.wsadie_6.1.0/WSADIEEJBProjectMigration.sh
    The following parameters are supported, where the workspace and project name are mandatory:
    Usage:      WSADIEEJBProjectMigration.bat
                [-e eclipse-folder] -d workspace -p project
    
    eclipse-folder: The location of your eclipse folder -- usually it's the 'eclipse'
                    found under your product installation folder.
    
    workspace:  The workspace containing the WSADIE EJB project to be migrated.
    
    project:    The name of the project to migrate.
    For example,
     WSADIEEJBProjectMigration.bat -e "C:\IBM\WID6\eclipse" -d "d:\my60workspace" -p "MyWBIEJBProject"
  4. When you open WebSphere Integration Developer you will need to refresh the EJB project to get the updated files.
  5. Search for the file ibm-web-ext.xmi in the EJB project. If one is found, ensure that the following line is present in the file under the element:
    <webappext:WebAppExtension> element:  
    <webApp href="WEB-INF/web.xml#WebApp"/>
  6. Remove the old deployment code that was generated in 5.1. Regenerate the deployment code by following the WebSphere Application Server guidelines for doing so.

Manually deleting 5.1 Web Services Invocation Framework (WSIF) definitions

After you have completed migrating your source artifacts, you should delete all 5.1 WSIF Binding and Service WSDL definitions from your 6.x projects that are no longer being used. The consumption scenario for service migration is the only case where a WSIF Binding or Service would still be in use.

The following WSDL namespaces indicate that a binding or service definition is a 5.1 WSIF Service and can be discarded if no longer used:

EJB WSIF Namespace:
http://schemas.xmlsoap.org/wsdl/ejb/
Java WSIF Namespace:
http://schemas.xmlsoap.org/wsdl/java/
JMS WSIF Namespace:
http://schemas.xmlsoap.org/soap/jms/
Business Process WSIF Namespace:
http://schemas.xmlsoap.org/wsdl/process/
Transformer WSIF Namespace:
http://schemas.xmlsoap.org/wsdl/transformer/
IMS WSIF Namespace:
http://schemas.xmlsoap.org/wsdl/ims/
CICS-ECI WSIF Namespace:
http://schemas.xmlsoap.org/wsdl/cicseci/
CICS-EPI WSIF Namespace:
http://schemas.xmlsoap.org/wsdl/cicsepi/
HOD WSIF Namespace:
http://schemas.xmlsoap.org/wsdl/hod3270/

Verifying the source artifact migration

If the migration completes with a list of errors, warnings, and/or informational messages, they will be displayed in the Migration Results window. Otherwise, the wizard window will close if the migration completed successfully.

The following page appears if migration messages were generated during the migration process:

Migration results window

In the Migration Results window, you can see the migration messages that were generated during the migration process. By selecting a message from the upper Message list, you can find more information regarding that message in the lower Message Description window.

To keep all messages for future reference, click the Generate ToDo's button to create a list of "ToDo" tasks in the task view and/or click the Save as... button to save the messages in a text file in the file system. Examine each message to see if any action needs to be taken to immediately fix an artifact that couldn't be fully migrated. To see the generated To Do's, click Window -> Show View -> Other... -> General -> Tasks and click OK. The Tasks view opens with the list of generated To Do's from the migration process.

To verify that a portion of the migration is complete, switch to the Business Integration perspective and ensure that all processes and WSDL interfaces from the old service project appear in the new module. Build the project and fix any errors that prevent the project from building.

After performing the manual migration steps required to complete the migration of the business integration application, export the application as an EAR file and install it to a WebSphere Process Server, configuring the appropriate resources.

Perform the manual migration steps required to migrate any client code or generate new client code using WebSphere Integration Developer. Ensure that the client can access the application and that the application exhibits the same behavior it did on the previous runtime environment.

Working with source artifact migration failures

If your source artifact migration from WebSphere Studio Application Developer Integration Edition fails, there are a number of ways in which to deal with the failures.

The following are some of the possible source artifact migration failures:

If the Migration wizard completes without this message, a list of info, warning, and error messages will be displayed. These signify that some portion of the service project could not be automatically migrated and that manual changes must be performed to complete the migration.

Limitations of the migration process (for source artifact migration)

There are certain limitations involved with the WebSphere Studio Application Developer Integration Edition source artifact migration process.

The following lists detail some of the limitations of the migration process for source artifact migration:

General limitations

SCA Programming Model limitations

BPEL migration process technical limitations

Notices

U.S. Government Users Restricted Rights - Use, duplication or disclosure restricted by GSA ADP Schedule Contract with IBM Corp.

This information was developed for products and services offered in the U.S.A. IBM may not offer the products, services, or features discussed in this documentation in other countries. Consult your local IBM representative for information on the products and services currently available in your area. Any reference to an IBM(R) product, program, or service is not intended to state or imply that only that IBM product, program, or service may be used. Any functionally equivalent product, program, or service that does not infringe any IBM intellectual property right may be used instead. However, it is the user's responsibility to evaluate and verify the operation of any non-IBM product, program, or service.

IBM may have patents or pending patent applications covering subject matter described in this documentation. The furnishing of this documentation does not give you any license to these patents. You can send license inquiries, in writing, to:

IBM Director of Licensing 
IBM Corporation 
North Castle Drive 
Armonk, NY 10504-1785 
U.S.A.

For license inquiries regarding double-byte (DBCS) information, contact the IBM Intellectual Property Department in your country or send inquiries, in writing, to:

IBM World Trade Asia Corporation
Licensing
2-31 Roppongi 3-chome, Minato-ku 
Tokyo 106, Japan 

The following paragraph does not apply to the United Kingdom or any other country where such provisions are inconsistent with local law: INTERNATIONAL BUSINESS MACHINES CORPORATION PROVIDES THIS PUBLICATION "AS IS" WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESS OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OR CONDITIONS OF NON-INFRINGEMENT, MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. Some states do not allow disclaimer of express or implied warranties in certain transactions, therefore, this statement may not apply to you.

This information could include technical inaccuracies or typographical errors. Changes are periodically made to the information herein; these changes will be incorporated in new editions of the publication. IBM may make improvements and/or changes in the product(s) and/or the program(s) described in this publication at any time without notice.

Any references in this information to non-IBM Web sites are provided for convenience only and do not in any manner serve as an endorsement of those Web sites. The materials at those Web sites are not part of the materials for this IBM product and use of those Web sites is at your own risk.

IBM may use or distribute any of the information you supply in any way it believes appropriate without incurring any obligation to you.

Licensees of this program who wish to have information about it for the purpose of enabling: (i) the exchange of information between independently created programs and other programs (including this one) and (ii) the mutual use of the information which has been exchanged, should contact:

Intellectual Property Dept. for WebSphere Integration Developer
IBM Canada Ltd.
8200 Warden Avenue
Markham, Ontario L6G 1C7
Canada

Such information may be available, subject to appropriate terms and conditions, including in some cases, payment of a fee.

The licensed program described in this documentation and all licensed material available for it are provided by IBM under terms of the IBM Customer Agreement, IBM International Program License Agreement or any equivalent agreement between us.

Any performance data contained herein was determined in a controlled environment. Therefore, the results obtained in other operating environments may vary significantly. Some measurements may have been made on development-level systems and there is no guarantee that these measurements will be the same on generally available systems. Furthermore, some measurements may have been estimated through extrapolation. Actual results may vary. Users of this document should verify the applicable data for their specific environment.

Information concerning non-IBM products was obtained from the suppliers of those products, their published announcements or other publicly available sources. IBM has not tested those products and cannot confirm the accuracy of performance, compatibility or any other claims related to non-IBM products. Questions on the capabilities of non-IBM products should be addressed to the suppliers of those products.

All statements regarding IBM's future direction or intent are subject to change or withdrawal without notice, and represent goals and objectives only.

This information contains examples of data and reports used in daily business operations. To illustrate them as completely as possible, the examples may include the names of individuals, companies, brands, and products. All of these names are fictitious and any similarity to the names and addresses used by an actual business enterprise is entirely coincidental.

COPYRIGHT LICENSE:

This information contains sample application programs in source language, which illustrates programming techniques on various operating platforms. You may copy, modify, and distribute these sample programs in any form without payment to IBM, for the purposes of developing, using, marketing or distributing application programs conforming to the application programming interface for the operating platform for which the sample programs are written. These examples have not been thoroughly tested under all conditions. IBM, therefore, cannot guarantee or imply reliability, serviceability, or function of these programs. You may copy, modify, and distribute these sample programs in any form without payment to IBM for the purposes of developing, using, marketing, or distributing application programs conforming to IBM's application programming interfaces.

Each copy or any portion of these sample programs or any derivative work, must include a copyright notice as follows:

(C) (your company name) (year). Portions of this code are derived from IBM Corp. Sample Programs. (C) Copyright IBM Corp. 2000, 2008. All rights reserved.

If you are viewing this information softcopy, the photographs and color illustrations may not appear.

Programming interface information

Programming interface information is intended to help you create application software using this program.

General-use programming interfaces allow you to write application software that obtain the services of this program's tools.

However, this information may also contain diagnosis, modification, and tuning information. Diagnosis, modification and tuning information is provided to help you debug your application software.

Warning: Do not use this diagnosis, modification, and tuning information as a programming interface because it is subject to change.

Trademarks and service marks

IBM, IBM Logo, WebSphere, Rational, DB2, Universal Database DB2, Tivoli, Lotus, Passport Advantage, developerWorks, Redbooks, CICS, z/OS, and IMS are trademarks or registered trademarks of International Business Machines Corporation in the United States or other countries or both.

UNIX is a registered trademark of The Open Group in the United States, other countries, or both.

Java and all Java-based trademarks and logos are trademarks or registered trademarks of Sun Microsystems, Inc. in the United States, other countries, or both.

Microsoft and Windows are trademarks or registered trademarks of Microsoft Corporation in the United States, other countries, or both.

Linux is a trademark of Linus Torvalds in the United States, other countries, or both.

Adobe is either a registered trademark or trademark of Adobe Systems Incorporated in the United States, other countries, or both.

Other company, product and service names may be trademarks or service marks of others.

Terms of use

Permissions for the use of publications is granted subject to the following terms and conditions.

Personal Use: You may reproduce these publications for your personal, non commercial use provided that all proprietary notices are preserved. You may not distribute, display or make derivative work of these publications, or any portion thereof, without the express consent of IBM.

Commercial Use: You may reproduce, distribute and display these publications solely within your enterprise provided that all proprietary notices are preserved. You may not make derivative works of these publications, or reproduce, distribute or display these publications or any portion thereof outside your enterprise, without the express consent of IBM.

Except as expressly granted in this permission, no other permissions, licenses or rights are granted, either express or implied, to the publications or any information, data, software or other intellectual property contained therein.

IBM reserves the right to withdraw the permissions granted herein whenever, in its discretion, the use of the publications is detrimental to its interest or, as determined by IBM, the above instructions are not being properly followed.

You may not download, export or re-export this information except in full compliance with all applicable laws and regulations, including all United States export laws and regulations.

IBM MAKES NO GUARANTEE ABOUT THE CONTENT OF THESE PUBLICATIONS. THE PUBLICATIONS ARE PROVIDED "AS-IS" AND WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING BUT NOT LIMITED TO IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE.

(C) Copyright IBM Corporation 2005, 2007. All Rights Reserved.