relational databases

I bought this book on the basis of the glowing recommendations here. As I have a number of servers to tune which execute some extremely complex SQL, and I need to be able to look inside with Perfmon and the profiler, I thought this book would be very useful. I particularly wanted help with sysmon.
This book gave me virtually nothing. Its coverage of tuning was shallow, information was repeated unnecessarily, text was copied almost verbatim from BOL, and it provided little or nothing that couldn't be found elsewhere and easily.
It tries to cover everything at the cost of giving real value. For example it provides 15 pages on data warehousing of which 12 are a description of data warehousing so cursory that if you don't know the subject you'll only be confused, and 3 pages on actual tuning which basically say that you should find out whether the bottleneck is CPU/disk/memory then add more CPU/disk/memory respectively.
Sizing and capacity planning are introduced with seven equations without justification. Okay, but completions C is given as the number of transactions that were completed during the observation period, but on the facing page C = 96 seconds [sic]. Did anyone proof-read this? With these and numerous other oddities (trunc. log on chkpt on SQL2000?) I don't know what I can trust.
The mathematics for this section is done and finished in 6 pages.
I was particularly looking for a comprehensive description of sysmon counters. Other than a quick rundown of the obvious ones there's a long list in the appendix of others, including such gems as "lock blocks allocated: the total number of allocated lock blocks". The whole point of buying this book was to find out how to use them, or indeed what they mean (Skipped Ghosted Records/Sec - means what?); merely giving me a list of them is redundant. This was the biggest letdown for me - I need this info!
There are other important omissions. I have spent literally weeks identifying and working round failures in the query plan optimiser. This serious issue is not properly addressed except for a chapter introducing query hints. A taxonomy of optimiser failures and ways of tackling each type might save others from the headaches I've had. Optimiser hints do not always suffice.
The book is rated on the back for user levels IT Implementer and Corporate Developer. That is far too generous.

Considering the difficulty of the topic, this book is a fairly easy read. The single best thing about the book is that the advice is actionable. You can read this book and immediately begin tuning.

Coverage is excellent - performance tuning, capacity planning, setting up disk drives, managing cpu, I/O, network, and memory, index tuning, backups, replication, OLTP versus OLAP, etc.

For each subject area, the authors explain the applicable concepts and SQL Server tools, and then systematically explain their application using practical examples.

Compared to other performance tuning books, it is an 80/20 book. By this I mean that the authors focus on what is most important and then move on to the next topic. They don't get carried away demonstrating how much they know about each concept or go into the minutia of the options of each SQL Server tool.

I hope they write more books.

Strong recommendations from others brought me to this book and I feel obliged to give back.
This is a masterpiece.Well researched and well written
I Got a hold of a couple of concepts missed in other books because of the lucidity of the examples and explanations.
Great book If you read it you will understand why....

Michael Tubuo Ngong