Reviews for Database Tuning: Principles, Experiments, and Troubleshooting Techniques (The Morgan Kaufmann Series in Data Management Systems)

1. Uniquely broad and valuable perspective on real-world tuning

   Rated 5 out of 5 stars, March 12th, 2003

   This book has a much broader perspective than the often highly-prescriptive, vendor-specific tuning guides that are available. Nevertheless, it is full with practical, real-world information and with concrete experimental results from Oracle, DB2 and SQL/Server.

   I think practitioners especially will like this book because it provides real insight into the problems that designers and administrators face when trying to make real-world databases and applications perform. You may even also discover a few unexpected and possibly unwelcome truths about the major commercial databases.

   The book is hard work to read because it is densely packed with information and it frequently requires you to stop and think. This is not a criticism - it reflects the nature and the value of the book because it is often the case that databases and applications need tuning precisely because nobody building them apparently had time to stop and think.

   You will get the most from this book quickly if you already have a firm grasp of database technology. However the book contains much valuable tutorial material and Chapter 2 provides an excellent crash course for those who are less experienced. Independent commentary on the back cover says chapter 2 alone is worth the price of the book. I would agree with that and in fact would add that if all application programmers had read chapters 2 and 5 my working life would probably be far less painful.

   The only topic I might like to have seen more coverage of was the use of transactional queuing facilities as a performance tool. To be fair the authors may well have considered this in the domain of TP monitors and not database tuning. This also reflects my bias - my background is in OLTP and moving legacy applications onto new database technology.

   I thoroughly recommend this book and have given it five stars because I have wanted something like it for a long time and because I suspect there is nowhere else you can get this information and this perspective off the shelf. My copy is already littered with page markers and annotations.

   If, rather than tuning, you are lucky enough to be designing and building a new high performance database or application (especially if it is to run on more than one of the dominant databases) then you will find this book equally valuable.