Data masking is the process of changing a copy of the production data into SIT and UAT environments in order to make it anonymous.
This serves different purposes, most importantly: Protecting the privacy of the production data while providing realistic data values and formats.
There are multiple methods for data masking, and it also depends on the .
- Randomization functions: for integer and float/numeric data values.
- Shuffling letters in string data values.
It is important in the data masking process to respect and retain the business rules and the data rules in the masked data. So for example if there are values that need to be of a certain format, range or length, or a Primary key, Identity, Foreign keys or other data level constraints, that all need to be respected.