WebSphere logo Classic Federation Server for z/OS, Version 9.1
WebSphere logo Classic Replication Server for z/OS, Version 9.1
WebSphere logo Classic Data Event Publisher for z/OS, Version 9.1
WebSphere logo Data Integration Classic Connector for z/OS, Version 9.1


Record array definitions for federation and change capture

Federated queries can read record array data if you create a separate table for each array. You can insert, update, delete, and capture changes to record array data by flattening the column structure.

Record array definitions contain column definitions, and possibly additional record array definitions. To read, manipulate, or capture changes to record array data, create the required tables or column structure.

To query record array data, map a separate table for each record array in the data definition. To insert, update, or delete array data, flatten the array structure by choosing the array processing option Expand occurrences when you map the table in the New Table wizard.

A flattened structure provides a separate column for each array instance and field. For example, you map a record array for employee dependents that contains five fields:

  1. DEP_LAST_NAME
  2. DEP_FIRST_NAME
  3. DEP_GENDER
  4. DEP_ID
  5. DEP_DOB

If you want to support up to ten dependents, you must map 50 columns with names that uniquely identify each instance and field. In this example, the column names range from DEP_LAST_NAME_1 to DEP_DOB_10.

Restriction: You cannot insert, update, or delete array data when you map the structure as an array because of mismatched insert and update logic. Adding a new dependent updates a single record in the source database, but the change inserts a new row into the logical table on the data server. Flatten the structure of the table to perform inserts, updates, or deletes.

Change capture

Flatten the structure of a table to capture changes to array data.

You cannot map record array definitions for change capture. An ALTER TABLE statement on a table containing array structures mapped as arrays will fail, because change capture must send one notification per change. The mapping must return exactly one row per physical record to provide an accurate notification to consuming applications.

Related concepts
Record arrays
Performance considerations with multiple record arrays
Related tasks
Creating a separate table for each record array in a table definition
Mapping tables and views for redefined data
Related reference
Array definition examples
Related information
Mapping data for Classic federation
Mapping data for change capture
Record types in data definitions examples


Feedback

Update icon Last updated: 2007-10-09