I have some concerns about Denormalize to Optimize.
"a single column family are used to answer each query. This sacrifices disk space (one of the cheapest resources for a server) in order to reduce the number of disk seeks and the amount of network traffic."
First, it will NOT reduce the amount of network traffic, compared with Normalized data when you get same number of columns and rows.
2nd, for the number of disk seeks, it depends. some times Normalized data increase the cache hit ratio with smaller memory footprint and reduce the number of disk seeks.
And last, could you add some data model examples to explain how to Denormalize to Optimize? it help us to learn and understand the data modeling in Cassandra.
Thanks and happy Chinese new year,