As an example, Cassandra has a 2 billion column limit for a "row key". Some high volume IoT apps could push that boundary and you should therefore, design accordingly, if using Cassandra.
Is there any size limitations similar to that with SnappyData's column table that I should "design" around as well?