Cassandra Design Patterns by Sanjay Sharma, Packt Publishings Book Review
Cassandra Design Patterns is a relatively short overview of both patterns and anipatterns for an IT professional entering the NoSQL/NewSQL world in retrospect to possible Cassanra database implementations in particular.
My experience with Cassandra was in versions prior to 2.0. And I needed to learn the hard way, now thanks to Sanjay and Packt there is a book to help decide where Cassandra would be a good fit, or not.
The book is easy to read, although, it exhibits very few lines of code. It may be seen as a disadvantage, but showing coding examples against Cassandra would require a separate setup.
My overall feel about the book however as if it remains somewhat incomplete, a tad hastily written with some proof-reading pending. Also it can be improved by bringing up some real-life, concrete usage examples. For example, there is no information on how to use the client code properly, e.g. why an RPC timeout is possible (query return result limits) and therefore paging on the client side is necessary. There was no hint on maximum recommended data node size (that stands at around 300 GB per node). As an aside, Cassandra requires a shift in IT operations. Suddenly a shop would need to deal with a cluster. Typically of several nodes. And it would be a totally new beast to manage, with its own challenges.
Likewise, there is no good explanation on the major differences between the Apache vs DataStax Cassandra distributions, Host OS choices and overall best setup practices.
On the positive side the book does cover major Cassandra database lifecycle events as Antientrophy, Snitches, Compaction, etc.
In my view, this book is targeting team leaders new to the NoSQL, BigData notion running fast growing in popularity webscale apps that require a new kind of data persistence layer and ease of development against, or managers looking into replacing their underperforming OLTP databases. This book would help narrowing down a list of candidate NoSQL/Key-Value store databases while you are is in pre-POC mode.
My verdict is a 3 out of 5 stars.