Breaking down your data

This section of the chapter, like the others, assumes you are relatively new to NoSQL data modeling and may have some proficiency with modeling databases in RDBMS. Going from the RDBMS-based model to a NoSQL-based one, you may need to move your mindset from normalization to denormalization with potentially some duplication as well; however, the denormalization here could itself be done cleverly. Simply dumping all the columns/attributes into a single large JSON blob and retrieving that blob for any data access request that may not warrant reading the 100 other attributes within the blob is not quite what would work. If it does, that would not be efficient anyway.

This section will showcase a few design patterns and thought-provoking concepts that would help you think about the denormalization of data coming from the relational world itself, but also doing it cleverly, by identifying opportunities to break up the data as per the data access patterns, to be...