Azure Tables: best practices for choosing partition/row keys
What would be best practices for choosing partition/row keys for entities in Azure Tables? The common advice is to magically balance between partition size and number of partitions. But no one seems to have a good definition of how it can be accomplished in 3 easy steps. Is there a general approach for choosing keys so that everything then just works?
Answer
There is a detailed article on this very subject up on MSDN: Designing a scalable partitioning strategy for Windows Azure Storage.