> The cost was high because they modeled and queried their data like they were still using a SQL database which was incredibly inefficient. That's literally ev

> The cost was high because they modeled and queried their data like they were still using a SQL database which was incredibly inefficient. That's literally every Cassandra database I've ever encountered in the wild. If you use Cassandra, you WILL need to duplicate data across tables for lookups. Don't use Cassandra if you can't stomach that fact (and the disk bills that come with it).