BizDataX masking toolbox is equipped with various tools you need to mask all your sensitive data. For standard types (strings, numbers, dates) there are many out-of-the-box generators dealing with random generation, intervals, conditions and distributions. For all special cases specific to your database, there is a simple extension mechanism, so you can plug in your logic.
Data masking/data anonymization performs transformation of data in your database in 3 steps:
1. Reading data from database to BizDataX application memory.
2. Process data in memory.
3. Writing data from application memory to database.
Actual performance may vary and is usually determined by writing performance. The best real-world measurement at client’s site is just a bit above 1 billion records per hour. This depends on many factors like database technology, database configuration, network configuration, physical distances between servers and so on. Parallel writing into one or more tables, bulk inserting, tweaking the triggers and indexes or reducing the scope of the data in the final result are some of the options available and supported by BizDataX. Even when databases contain billions of records, processing duration measures in hours.
It is quite common for data masking to interfere with database referential integrity. For example when you need to change national identification numbers or account numbers that are used as primary and foreign keys in many tables. Also when a person’s sensitive data is scattered across multiple tables. BizDataX is guarding referential integrity and masks scattered data consistently.