I have been using the DataStax Cassandra Java driver from Spark driver to insert data to Cassandra tables. It seems that Spark SQL can also be used for the same. Though for a NoSQL DB like Cassandra, it seems to be wrong to misuse the JOIN operation that is supported in Spark SQL, for NoSQL data modeling done in Cassandra.
From what I have understood, I would rather stick with Cassandra Java driver. However want to know if I have missed something in this.
Which is more efficient to use or better recommended to use with Cassandra.
Related question - Is SparkSQL RDBMS or NOSQL?
AFAIK the most popular way to work with Cassandra from a Spark app is using Spark Cassandra Connector: https://github.com/datastax/spark-cassandra-connector