Sub setting - good for data security or not worth the trouble?
Recently we have been speaking to a number of clients who question whether or not they need to do sub setting before masking their data. In a real sign of the times, many of these organizations are coming to the conclusion it is not worth the trouble. The bottom line is that while sub setting sounds nice it is rarely cost effective. I think there are two primary reasons sub setting is falling out of favor. The first is that data masking products like DMsuite are fast, masking 1, 5 or 10 Terabytes of data does not take much longer than just moving the data. Second, disk space is cheap. There are other reasons of course, performing detailed data analysis on complex databases can be quite time consuming and while this can be an interesting exercise it gets you no closer to securing your data. Also most masking products let you do "simple" sub setting such as by date or on a key value which can address the needs of some applications. This means for example, that a large fact table in a data warehouse can be sub setted and thenmasked while all the reference data can be simply be masked without impacting the referential integrity of the data. The bottom line - you don't really need to subset your data to do data masking, there may be other reasons to do it, but the lowering cost of disk storage and the higher cost of skilled labor should make you think twice.