Application developers always want production like data for their testing requirements. Although benefitial, uncontrolled exposure of production data to developers leads to antitrust liabilities and lawsuits from customers.
As the main function of RC/Extract is to fetch data from a particular set of tables and load it into another set of similar tables. It is a preferred tool of customers to generate data for testing in development environment from production databases. To make this generation of test data secure and reliable, RC/Extract has an inbuild feature which masks the extracted data and converts them into random values. The data retains the same attributes (character text is still character text, numerics remain numerics, dates and times are valid dates and times), but the sensitive data values are protected. You can keep data secure after it has been extracted from DB2 to tape or sequential files.
Cascading data masking is provided for primary and foreign key fields, so you can protect your data without compromising the relational integrity of the data. When you specify data masking for a primary or foreign key column, it is automatically applied to all referentially related columns in an RI structure when the extract object is processed. This applies to all tables in an RI structure, regardless of which table you specifically mask. For example, if you mask the foreign key in a child table, the primary key in the parent table is also masked.
Procedure:
>> Specify 'E' against any extracted dataset in the RC/Extract command centre.
>> Specify a one- to eight-character name, such as your TSO user ID to be used as the high-level qualifier of the masked file.
>> Specify a one- to eight-character name for the low level qualifier of the masked file. The default is EXTOBJM.
>> Specify I in the C column to select column of an object for masking. Using D against a selection a row can be removed from masking set.
>> Then pressing PF3 or type in END command processes the request. If the mode was selected as 'B' a JCL will be created, if the mode was 'O' the masking engine will be triggered automatically and the masked dataset will be created and shown in Command centre.