Author: Jothy Rosenberg & Arthur Mateos
Publisher: Manning, 2010
Aimed at: Business decision makers
Pros: A good, moderately technical, overview
Cons: Slightly biased
Reviewed by: Mike James
This is a management-style overview of the cloud. Before this makes you switch off and move to something more practical, let me say that it is surprisingly good.
It starts out with an overview of cloud computing and attempts to classify the different approaches. The section on data centers should be something of an eye-opener if you have heard about the way that these super-computing factories work. It goes into a lot of relevant theory about how persistent storage is organized and provides a comparison of the different types of services offered by the major suppliers - Amazon, Azure, App Engine, Ruby on Rails in a Cloud and Salesforce. It also considers the idea of setting up your own cloud service, but more or less suggests that this would be missing the point - giving you the disadvantages without any real advantages.
Chapter 3 is all about the business case and I was delighted to discover that, rather than vague talk of ROI and other jargon, it actually tackles the question "how much does it cost?". It does, however, make the point that the cloud is great for zero capital startups. This chapter may not be technical in the computing sense, but it gives you a clearer idea of why you might be interested in the cloud approach.
Chapter 4 looks at privacy and security - two very different problems when it comes to cloud hosting. Again there is an interesting section on data centers and physical security. Beyond this it deals with many of the standard security ideas - certificates, key pairs, cryptography and so on - and looks at the differences between private and public cloud infrastructures. More sophisticated is the idea of virtual private cloud, something that Amazon offers.
The next two chapters look at the problem of designing for the cloud. This is basically a very general overview of ideas such as data sharding and cloud burst computing. Chapter 6 deals with SOA, loose couping and MapReduce type algorithms and Chapter 7discusses testing and deployment.
Chapter 8 deals with some practical problems that haven been covered earlier in the book, in particular considerations you should bear in mind when picking a cloud provider and the final chapter looks to the future with ten predictions about how the cloud will evolve.
This isn't your typical management introduction to the cloud. It really is a good and moderately technical overview of the subject of the sort you should read before you start a project. It is a little bit too pro the cloud in the sense that, while it does point out the potential problems, it doesn't really do so strongly enough. Even so it is well worth the read and definitely something you should get upper management to read.