Skip Navigation Links | |
Exit Print View | |
![]() |
Upgrading to Oracle Java CAPS Master Index Java CAPS Documentation |
Upgrading to Oracle Java CAPS Master Index
About the Upgrade Paths to Oracle Java CAPS Master Index
Upgrading From eView Studio to Oracle Java CAPS Master Index (Repository)
To Upgrade to Oracle Java CAPS Master Index (Repository)
Migrating to Service-Enabled Oracle Java CAPS Master Index
Step 1: Create the Service-Enabled Master Index
To Create the Service-Enabled Master Index
Step 2: Migrate any Custom Plug-ins
Step 3: Update the Object Definition
Step 5: Update the Master Controller
Step 6: Update the Matching and Standardization Configuration
Step 7: Update Query Definitions
Step 8: Update the Survivor Calculator
Step 9: Update Field Validations
Step 10: Validate the Configuration Files
To Validate the Configuration Files
Step 11: Update Match and Standardization Engine Files
To Update Standardization Engine Files
Step 12: Upgrade the Database Connection and Security
To Upgrade the Database Connection and Security
Step 13: Finalizing the Upgrade
Step 14 (Optional): Update the Master Index API in Client Projects
To Update the Master Index API in Client Projects
If you create a master index project using Oracle Java CAPS Master Index (Repository) or if you upgrade a master index to the Repository version, you can migrate that project to the service-enabled version in order to take advantage of the new features and tools.
Follow these steps in the order given to migrate to a service-enabled master index:
Step 6: Update the Matching and Standardization Configuration
Step 14 (Optional): Update the Master Index API in Client Projects
The first step to migrating a master index project is to create a service-enabled master index project. This project should be similar to the project being upgraded, but does not need to be identical because you will copy over the configuration of the project you are upgrading before you generate the new master index application.
This is described in Upgrading to Oracle Java CAPS 6.3. You can export and import the Environments along with the projects.
These updates are described in the Sun SeeBeyond eView Studio Upgrade Guide, Release 5.1.3, available at http://download.oracle.com/docs/cd/E19336–01/820–0965/820/0965.pdf.
Tip - Follow these guidelines when creating the new master index application:
The application name and parent object name need to be the same for the new service-enabled master index as for the Repository–based master index. They also need to be the same as each other.
You can accept the defaults for the deployment environment in the wizard since these values will be populated when you update the configuration files.
To make it simpler, you can create the object structure using one of the predefined templates.
For the final step of the wizard, you do not need to generate all remaining files.
If any custom plug-ins were created for the project you are upgrading, you need to migrate those Java classes to the service-enabled project. You create the new custom plug-in files manually in the ProjectName-ejb project, and then copy the text of the existing custom plug-ins into the new files. Minor text changes are also required.
Note - In Repository-based master index projects, Java package names begin with “com.stc.eindex”. In the service-enabled projects, they begin with “com.sun.mdm.index”. You should update the custom plug-in package names accordingly.
Note - When you first create the custom plug-ins, there will be several context errors. They should be resolved once you generate and build the master index application later in the migration process.
To ensure that the object definition does not change between versions, you need to copy the text of the Object Definition file in the Repository-based project to the object.xml file in the service-enabled project and then make certain updates to the text.
The upgrade from the Enterprise Data Manager file in the Repository-based project to midm.xml in the service-enabled project is partially automated. A few updates to the text need to be performed manually, but the bulk of the update is done through the Configuration Editor.
These are the lines you need to change:
<edm xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:noNamespaceSchemaLocation="sbyn:SeeBeyond/eView/schema/EDM.xsd"> ... </edm>
When you are done, the lines should look like this:
<midm xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:noNamespaceSchemaLocation="schema/midm.xsd"> ... </midm>
The Configuration Editor appears.
Note - This change is just to enable the Save command so you can save midm.xml in the new format. For example, you can change the data type for one of the fields and then change it back to its original value.
Tip - An easy way to check for these values is to right-click in the text of the file and then click Validate XML. If you get errors similar to The value '–1' of element 'max-length' is not valid., click on the error and the cursor will move to the errant text in the file.
<reports> <root-object>Person</root-object> <tab-name>Reports</tab-name> <screen-id>6</screen-id> <display-order>5</display-order> <search-pages/> <search-result-pages/> <subscreen-configurations> <subscreen> <enable>true</enable> <root-object>Person</root-object> <tab-name>Potential Duplicate Report</tab-name> <report-name>Potential Duplicate</report-name> <screen-id>0</screen-id> <display-order>5</display-order> <search-pages/> <search-result-pages> <search-result-list-page> <search-result-id>0</search-result-id> <item-per-page>10</item-per-page> <max-result-size>2000</max-result-size> <field-group> <description/> <field-ref>Person.FirstName</field-ref> <field-ref>Person.LastName</field-ref> <field-ref>Person.SSN</field-ref> <field-ref>Person.DOB</field-ref> <field-ref>Person.Gender</field-ref> </field-group> </search-result-list-page> </search-result-pages> </subscreen> <subscreen> <enable>true</enable> <root-object>Person</root-object> <tab-name>Deactivated Record Report</tab-name> <report-name>Deactivated</report-name> <screen-id>1</screen-id> <display-order>1</display-order> <search-pages/> <search-result-pages> <search-result-list-page> <search-result-id>0</search-result-id> <item-per-page>10</item-per-page> <max-result-size>2000</max-result-size> </search-result-list-page> </search-result-pages> </subscreen> <subscreen> <enable>true</enable> <root-object>Person</root-object> <tab-name>Merged Transaction Report</tab-name> <report-name>Merged</report-name> <screen-id>2</screen-id> <display-order>0</display-order> <search-pages/> <search-result-pages> <search-result-list-page> <search-result-id>0</search-result-id> <item-per-page>10</item-per-page> <max-result-size>2000</max-result-size> <field-group> <description/> <field-ref>Person.FirstName</field-ref> <field-ref>Person.LastName</field-ref> <field-ref>Person.SSN</field-ref> </field-group> </search-result-list-page> </search-result-pages> </subscreen> <subscreen> <enable>true</enable> <root-object>Person</root-object> <tab-name>Unmerged Transaction Report</tab-name> <report-name>Unmerged</report-name> <screen-id>3</screen-id> <display-order>2</display-order> <search-pages/> <search-result-pages> <search-result-list-page> <search-result-id>0</search-result-id> <item-per-page>10</item-per-page> <max-result-size>2000</max-result-size> </search-result-list-page> </search-result-pages> </subscreen> <subscreen> <enable>true</enable> <root-object>Person</root-object> <tab-name>Updated Record Report</tab-name> <report-name>Update</report-name> <screen-id>4</screen-id> <display-order>3</display-order> <search-pages/> <search-result-pages> <search-result-list-page> <search-result-id>0</search-result-id> <item-per-page>10</item-per-page> <max-result-size>2000</max-result-size> <field-group> <description/> <field-ref>Person.FirstName</field-ref> <field-ref>Person.SSN</field-ref> <field-ref>Person.Address.AddressLine1</field-ref> </field-group> </search-result-list-page> </search-result-pages> </subscreen> <subscreen> <enable>true</enable> <root-object>Person</root-object> <tab-name>Activity Report</tab-name> <report-name>Transaction Summary</report-name> <screen-id>5</screen-id> <display-order>4</display-order> <search-pages> <simple-search-page> <screen-title>Weekly Activity</screen-title> <report-name>Weekly Transaction Summary Report</report-name> <search-result-id>0</search-result-id> <search-screen-order>1</search-screen-order> <field-group/> </simple-search-page> <simple-search-page> <screen-title>Monthly Activity</screen-title> <report-name>Monthly Transaction Summary Report</report-name> <search-result-id>0</search-result-id> <search-screen-order>2</search-screen-order> <field-group/> </simple-search-page> <simple-search-page> <screen-title>Yearly Activity</screen-title> <report-name>Yearly Transaction Summary Report</report-name> <search-result-id>0</search-result-id> <search-screen-order>3</search-screen-order> <field-group/> </simple-search-page> </search-pages> <search-result-pages> <search-result-list-page> <search-result-id>0</search-result-id> <item-per-page>10</item-per-page> <max-result-size>2000</max-result-size> <field-group/> </search-result-list-page> </search-result-pages> </subscreen> <subscreen> <enable>true</enable> <root-object>Person</root-object> <tab-name>Assumed Matches Report</tab-name> <report-name>Assumed Match</report-name> <screen-id>6</screen-id> <display-order>6</display-order> <search-pages/> <search-result-pages> <search-result-list-page> <search-result-id>0</search-result-id> <item-per-page>10</item-per-page> <max-result-size>2000</max-result-size> <field-group> <description/> <field-ref>Person.FirstName</field-ref> <field-ref>Person.LastName</field-ref> <field-ref>Person.SSN</field-ref> </field-group> </search-result-list-page> </search-result-pages> </subscreen> </subscreen-configurations> </reports>
<root-object>Person</root-object>
The Threshold file is upgraded by copying the text of the file to the service-enabled project and then updating certain portions of the file. There are only minor differences between the Threshold file in the Repository-based project and master.xml in the service-enabled project. In addition to updating the package name and XSD location, you can also update the file to take advantage of the new distributed transaction capabilities in Oracle Java CAPS Master Index.
LOCAL means transactions are not distributed. Using this option gives you the same behavior as in previous versions.
CONTAINER means transaction are distributed across applications.
BEAN means transaction are distributed within the master index application.
The Match Field file is updated by copying the text of the file to the service-enabled project and then updating certain portions of the file. To update the Match Field file in the Repository-based project to mefa.xml in the service-enabled project, you only need to change the location of the XSD file and the names of the Java packages accessed by the file.
To update the query definitions of the Repository-based project, you need to copy the text of the Candidate Select file in the Repository-based project to query.xml in the service-enabled project and then update certain portions of the file. For this file, you only need to change the location of the XSD file and the names of the Java packages accessed by the file.
To update the Best Record file in the Repository-based project to update.xml in the service-enabled project, you need to copy the text of the Repository-based file to the service-enabled file and then update certain portions of the file. You only need to change the location of the XSD file and the names of the Java packages accessed by the file. If custom update policies are defined for the master index, you also need to update the Java package and class names for the policies.
You only need to update the Field Validation file if you defined custom field validations in the Repository-based project. If you do need to update the file, you only need to copy the text to the service-enabled project and then change the location of the XSD file and the names of the Java packages accessed by the file.
Once you have updated all the configuration files and used the Configuration Editor to complete the process, you should validate each file against its corresponding schema file to ensure the project can be built.
The match engine configuration file has not changed since the previous release, and can be copied directly into the service-enabled project. Some of the standardization files were changed extensively. If you want to use the previous configuration for the person name standardization files, you will need to reformat the files.
File names have changed as indicated in the following table. Any Address or Business Name files not listed here have the same name in both versions. Make sure you when you update the files that you update the correct country set.
|
File names have changed for this release, as indicated in the following table. The following files are obsolete: personNamePatt.dat, personConstants.cfg, personFirstNameDash.dat, personRemoveSpecChars.dat, personThree*.dat, personTwo*.dat, personLastNamePrefix*.dat, and personConjon*.dat.
|
STEVE|STEPHEN STEWART|STUART STEW|STUART STU|STUART SUE|SUSAN SULLY|SULLIVAN
Note - For more information about the new standardization files, see FSM–Based Person Name Configuration, in Understanding the Master Index Standardization Engine.
When you migrate to service-enabled Master Index, you need to redefine the database connection pool in the application server Admin Console. You also need to set up security on the Admin Console.
![]() | Caution - As of Java CAPS Release 6 Update 1, you need to define two connection pools and two JDBC resources for Oracle Java CAPS Master Index. For more information and instructions, see Defining the Database Connection Pools in Oracle Java CAPS Master Index User’s Guide. |
Tip - For the connection pool, use the javax.sql.DataSource class for the same functionality as in previous versions. If you specified distributed transactions when you updated master.xml, use the javax.sql.XADataSource class. Once you create the Connection Pool, you can test the configuration by clicking Ping at the top of the Connection Pool page.
Note - If you created any custom user roles or groups for the Repository-based version, you can create corresponding roles in midm-security.xml.
Finalizing the project includes building the project, verifying the custom plug-ins, and finally deploying the project. You can also take advantage of the new query, matching, and SBR filtering feature described in Filtering Default Values From Master Index Processes in Oracle Java CAPS Master Index Configuration Guide. Note that adding this feature might cause inconsistencies with the way records were previously matched, but can also prevent the incorrect matching of records or flagging them as potential duplicates.
Building will compile the custom plug-ins you added previously.
Note - For more information about this process, see Broadcasting Master Index Updates to External Systems.
Once you build the master index project, a set of dynamic API methods are generated for use in web services, Collaborations, and Business Processes. In order to make the API available to Repository-based client projects, you need to generate a WSDL file and then import the generated file into the client project that will access the new service-enabled master index. Alternatively, you can upgrade the client projects using JBI components and call master index services from BPEL processes, web services, and Java clients.
A web service OTD for the ProjectNameEJBService is created. You can use this OTD to access the operations of the migrated master index application.
If you use the Configuration Editor to modify the master index application once all of the Repository-based information has been copied over, there might be invalid XML code in update.xml and mefa.xml.
Standardization and normalization structures might get removed from mefa.xml. If you use the Configuration Editor, be sure the check mefa.xml for any discrepancies.
The Configuration Editor might make changes to the survivor strategy in update.xml that prevent the project from being built. To fix the file, do the following:
Open update.xml.
If any candidate fields have a specific strategy defined, reformat the XML structure for that candidate field by removing the forward slash at the end of the candidate-field element and adding a closing candidate-field element after the survivor-strategy element.
For example, the following excerpt:
<candidate-field name="Person.Phone[*].*"/> <survivor-strategy> <strategy-class>com.sun.mdm.index.survivor.impl.WeightedSurvivorStrategy </strategy-class> <parameters> <parameter> <parameter-name>ConfigurationModuleName</parameter-name> <parameter-type>java.lang.String</parameter-type> <parameter-value>WeightedSurvivorCalculator</parameter-value> </parameter> </parameters> </survivor-strategy>
Would change to this:
<candidate-field name="Person.Phone[*].*"> <survivor-strategy> <strategy-class>com.sun.mdm.index.survivor.impl.WeightedSurvivorStrategy </strategy-class> <parameters> <parameter> <parameter-name>ConfigurationModuleName</parameter-name> <parameter-type>java.lang.String</parameter-type> <parameter-value>WeightedSurvivorCalculator</parameter-value> </parameter> </parameters> </survivor-strategy> </candidate-field>