I suppose this book would have been a fair read if I knew nothing at all, and was soon going to begin shopping for a load balancing/failover solution...
It explains things in very general concepts... Large pieces of information are left out of the explanations of how these systems really work, and some concepts are simply wrong.
Even forgiving the low-brow content, it is written horribly. What should be a one-paragraph explanation, turns into several pages, because the author chooses to introduce concepts, piece by piece, in the middle of any other concept that even vaguely applies. Imagine reading a novel where each paragraph had 2 trailing pages which defined each word.
Also, even if you just want to know the basics, it still leaves much out. In the Direct Server Return (DSR) section, the author explains each step, and then skips what happens between the time a packet hits the loopback interface, and when it is returned to the client.
The same section is a great example of another problem with the book. The author says that D.S.R. is wonderful because only about 1/8th the processing is done. Not only did the author screw up the ratio that he gave (essentially saying it would cause 8x the load, which contradicting himself in the following text) but he completely left out the fact that the 8x processing load is not gone, but rather shifted to the important servers, rather than the SLB box.
If you have any idea what load balancing is, what a heartbeat is, and what NAT is, you have nothing to learn from this book, except what not to do if you are an author.
If there is some way that I can get a refund for my book, which is in like-new condition, please contact me.
|