DataTurn for Automated Mainframe Database Migration

DataTurn is the FastTrack Factory technology for converting supported legacy data structures and records. When paired with CodeTurn, it helps Amdocs coordinate changes to a database and to the programs that read and update it.

Organizations should consider automated mainframe database migration when they need to move large volumes of data, translate complex record layouts into a relational schema, or rehearse the move several times before cutover.  

Amdocs DataTurn is a good fit when the database and the applications that use it need to move together. It lets engineers configure how legacy structures and records map to the target database and rerun the conversion as those mappings are refined. In supported migrations, DataTurn works with CodeTurn so converted applications can read and update the new database through a generated data access layer.

Key Takeaways

• DataTurn supports repeatable schema and data conversion for scoped source and target combinations.
• An input and output module can preserve data access behavior while changing how records are stored.
• DataMatch can compare mapped data and batch outcomes after conversion.

Map the source structure to the target schema

Start with the record definitions, keys, access paths and relationships applications rely on. A legacy database can represent a customer’s phone numbers as dependent records that programs reach by following a path. A relational target might use a separate phone table with a customer key. Choose keys, constraints and indexes to support both the logical relationships and the workload.

DataTurn lets teams inspect and adjust how source structures map to the target. It supports scripting for repeatable conversion runs. Relational targets including Oracle Database, SQL Server, Db2 and PostgreSQL. Validate the exact source, version and target combination during discovery. Some scope may call for retaining files in a supported format.

Connect converted applications to migrated data

Programs can depend on navigation order, record status and transaction boundaries. In supported conversions, generated input and output modules and shared data access libraries let programs use the target representation while preserving agreed behavior. Review the SQL access patterns, indexes and transaction semantics together. Large batch jobs can expose costs that a single interactive request will miss.

Rehearse extraction and loading at scale

Profile real records for encoding, packed decimal values, alternative layouts and unexpected control characters. Check that extraction captures all relevant records, then reconcile target counts, field values and business totals. Script repeated loads, log rejected records and establish how cutover will capture or freeze changes before the final run.

Data conversion creates a target dataset. Verification tests whether the dataset is complete and equivalent under the approved mapping. DataMatch can help compare mapped source and target results. TestMatch supports application behavior checks where configured. Use the evidence together to decide whether the new data and the application using it are ready.