Building Scalable Web Sites
Author: Cal Henderson
Publisher: O'Reilly, 2006
Aimed at: Web developers
Pros: Well-written coverage of the fundamentals
Cons: Doesn’t solve the problem of scalability
Reviewed by: Mike James
Scalability – the holy grail of every programmer. It’s easy to create a program that will handle a few transactions but what happens when the few becomes many, becomes very very many? The truth is that most of us haven’t a real clue as to how our applications scale, we just hope that they do. So a book that describes how to build scalable websites has to be a best seller.
Well indeed it would be but this particular book isn’t it. Despite the impeccable credentials of its author, one of the developers of the very heavily used Flickr website, it really doesn’t seem to tackle its subject head on. Many pages are spent on describing basic theory and practice – how routers work, how memory and processor configuration affect performance and so on. It even digresses into the issues surrounding building large software projects – source control, issue tracking, automatic testing and so on – and idiosyncratic topics such as how to deal with email from mobile services within your application. If you know a reasonable amount about computer science then all of this will be revision. However it’s easily digestible revision and if by any chance you have missed out on the fundamentals of how things work this is a fairly painless and well motivated way of learning about it. When the author does get to the topic suggested by the title the information provided is surprisingly unsatisfying. It’s mostly common sense which boils down to “throw more hardware at it” but with the rider “make sure you have a design that benefits from more hardware”.
So is this a good book? From my comments you might think that the conclusion has to be no - but not so. After reading it you are going to have a false sense of confidence that you know what scaling is all about but in fact you are probably still going to be in the same boat as the rest of us and just hoping that your concerns and considerations have indeed resulted in an application that scales. On the way to this basic “no change” you will have encountered lots of interesting ideas, had a good time thinking about the problem of scalability and generally enjoyed yourself. Giving you the solution to the problem of scalability isn’t this book’s strong point but it's well written, fun to read and packed with good, but not specific, advice.
Last Updated ( Friday, 30 January 2009 )