Cassandra Data Model

As you know from my previous post, database backend of Trendocean is Apache Cassandra. If you have relational database background, data model of Cassandra is difficult to work with. Things get more difficult if nature of your application’s data is highly relational. You have to manage every relation yourself. You have to implement cascaded deletes, and foreign keys. When we have started to use Cassandra there were no secondary index support so we had to generate lots of extra tables to for queries and keeping record of mappings for some UUID fields. Although it was fun for us to experiment with a new technology, with the current state of the tools it was difficult to work with when compared to relational database world.

Data model of our application was simple and it was still difficult to work with, so do not even thing about using  Cassandra something like JC3IEDM.