Database Replication

Being unable to share data across the enterprise makes it difficult to develop new products, understand real costs, and cultivate new business. Replacing a legacy system altogether can be expensive, risky, and could require developer re-training.

Our Data Replication solution offers an alternative approach to migration, detailed below using IDMS as an example. IDMS data is held in a rigid, context-sensitive format which makes business intelligence queries and ETL very complex. Our solution unlocks data into a queryable format while maintaining the data context and business logic. This flexible, automated, and full-featured solution enables companies to achieve integration goals with zero footprint on the mainframe.

Our Data Replication solution consists of three primary modules: The Data Dictionary, Initial Loader, and Ongoing Replication.

Solution implementation begins with data modeling and schema mapping leveraging the Data Dictionary module. The Data Dictionary module includes our System Dictionary, a complete metadata repository, Windows UI navigation, and data mapping rules. This Data Dictionary system generates the Initial Loader configuration and translation engine, target relational database scripts, begins table creation, and creates SQL load scripts. XML definitions of schemas may be used to generate programs and documentation as well. The Data Dictionary system manages all IDMS constructs in this example, including set membership and sequence, redefines / occurs. The Data Dictionary module also handles the next step, establishing conversion rules. IDMS schema and segments are imported via DDDL and the relational schema is automatically generated from the system dictionary. There is a one-to-one relationship between IDMS and relational models by default, and relational schema can be adjusted within the system dictionary. Data mapping rules can also be adjusted within the system dictionary, which also contains views of IDMS, the relational database, and the relationship between the two. Lastly, a relational data model is generated.