Cassandra Community Webinar | Understanding How CQL3 Maps to Cassandra's Internal Data Structure

CQL3 is the newly ordained, canonical, and best-practices means of interacting with Cassandra. Indeed, the Apache Cassandra documentation itself declares that the Thrift API as "legacy" and recommends that CQL3 be used instead. But I've heard several people express their concern over the added layer of abstraction. There seems to be an uncertainty about what's really happening inside of Cassandra.

In this presentation we will open up the hood and take a look at exactly how Cassandra is treating CQL3 queries. Our first stop will be the Cassandra data structure itself. We will briefly review the concepts of keyspaces, columnfamilies, rows, and columns. And we will explain where this data structure excels and where it does not. Composite rowkeys and columnnames are heavily used with CQL3, so we'll cover their functionality as well.

We will then turn to CQL3. I will demonstrate the basic CQL syntax and show how it maps to the underlying data structure. We will see that CQL actually serves as a sort of best practices interface to the internal Cassandra data structure. We will take this point further by demonstrating CQL3 collections (set, list, and map) and showing how they are really just a creative use of this same internal data structure.

Attendees will leave with a clear, inside-out understanding of CQL3 and will be able use CQL with a confidence that they are following best-practices.