Customer Reviews

Inside Microsoft SQL Server 2000

5 star:		(31)
4 star:		(10)
3 star:		(4)
2 star:		(15)
1 star:		(14)

 

 

Before Inside SQL Server 2000 came out, my favorite technical book on earth was Inside SQL Server 7. Up until today, there was no other book that I consulted with as much as I did with Inside 7. My copy is already in a very bad shape from overuse. I often consult with it whenever I am in doubt on any issue concerning SQL Server internals. I sometimes read from it to my students in class to prove a point. I had many expectations from inside SQL Server 2000 and I eagerly read every word in it. The book exceeded my expectations! Before it came out, I thought that it would have the same level of detail that Inside 7 had with additional discussions on SQL Server 2000s new features, but it goes far beyond that. In many areas it goes into smaller details than Inside 7 did discussing issues that I have never seen discussed in any other book or published document. For example, in the chapter about indexes DBCC PAGE is used extensively to actually traverse the B-Trees and examine the actual layout of the data in the index pages. By examining those, a lot of the points concerning the interesting index architecture that SQL Server uses become clearer. By getting to this level of detail, I feel that I have more adequate tools to make the right critical decisions in a system regarding which columns to index; on which column(s) to create the clustered index; space consumption of indexes on huge tables, and so on. The chapter also covers the new indexes on views and on computed columns. The chapter on locking discusses the locking architecture and the lock manager in detail but it also discusses internal lock structures such as Lock Blocks and Lock Owner Blocks in great depth. There is simply no way to get that information in any other source. Many internals related areas get more attention in Inside 2000 such as the storage engine and the relational engine. The query processor has a whole chapter of its own! The installation process is covered in great detail covering all the installation options, some of which are totally new, such as multi instances. It also covers hardware considerations such as RAID controllers, file system, memory and so on. The chapter on tables discusses internal page and row structures explaining every bit inside the row. T-SQL is not neglected either. Beyond programmatic aspects, Inside 2000 gets into the various internals and optimization aspects of the various constructs and handles query and performance tuning in depth. Among the topics that are covered are joins, subqueries, derived tables, user defined functions, stored procedures, transactions, referential integrity including the new cascading referential constraints, after and instead of triggers, views, cursors, large objects and more. Full-Text searching has a whole new section. Tough areas such as plan caching; auto parameterization and reuse of execution plans are also explained in detail. Join algorithms including nested loops, hash and merge are also explained. The coverage of data modification internals is outstanding! It includes coverage of page splits; the various internal update mechanisms such as in-place and non in-place updates; bulk inserts optimization and more. All these were just examples. The book is very rich in the areas that it covers and it is definitely a must for programmers and DBAs that really want to know their stuff as far as SQL Server is concerned. The book is a work of art.

This book is the same decent book the version 7 book was. That's really the problem -- they're too similar. As some of the other reviews have pointed out, you won't find coverage of XML in this book. Should you expect to? Well, I guess that's debatable. Me, I expected it and was really disappointed when I couldn't find anything on it at all. That said, there's still useful info here. Most of the internals between 7.0 and 2K didn't change much, so the book still applies. Some of the same errors that existed in the 7.0 book are still here, but, mostly, the book gets it right. I guess the final verdict is: if you already own the 7.0 book, don't waste your money. If you don't have the 7.0 book, this book is probably worth reading through.

I read the previous version of this book and thought it was pretty good. It was a valuable resource when I was learning all I could about SQL Server for the first time. As I said in my review of that book, I felt the book belonged on every SQL Server DBA's bookshelf.

This book repeats much more of the version 7 book than I'd like. If it didn't claim to be a SQL Server 2000 book, I suppose that would be OK. To me the 2000 book should have a lot more 2000 specific stuff than it does. It really seems like this is just the version 7 book with a few additions. For example, I would like to have seen info on how the releases differ, but so far I haven't found it. I would like to have seen info on SQL Server's new XML support, but it's not there.... This doesn't mean that the book isn't useful or that the info is bad, just that it doesn't reach its full potential.

The conclusion I came to on this book is simply that the book is still pretty good, but could have been so much better

Bob S.

First of all, let me strongly disagree with low rates that some other reviewers gave to this book. I don't think it's correct to lower the rating just because the contents didn't change much since "Inside SQL Server 7". SQL Server 2000 is very much like SQL Server 7.0 (adding XML is the only major difference). It's a companion to the product, so it should be just a next iteration of "Inside SQL Server" series.

Then it seems that some reviewers probably expected to find how to program SQL Server using ADO, XML, ODBC etc, and instead only found a description of internals. While I agree that the books should have covered XML implementation better, I'd remind that the title of the book is "INSIDE SQL Server"! And I was really pleased with how much of SQL Server internals were exposed to the general public.

Perhaps I liked the book so much because I'm working now on a propriatary database implementation, and for me it was really useful to learn how SQL Server is implemented. But any serious database developer should understand how query execution is planned, how records are locked, and what is a performance and storage cost of certain operations. And all these topics are presented in a best possible way.

If you're looking to really understand the core engine of SQL Server and how to optimize your code, this is a really good book. If you're a database architect, this book may be a little too microscopic for you, lacking in design strategy. Surprisingly, I think this is actually a great book for the DBA - the discussions on the underlying mechanics are second to none.

This book is named very appropriately. The author does an excellent job detailing what goes on under the hood of SQL Server. As an example, she exposes the details of the Bulk Change Map pages in database files and how they work in relationship with the Bulk Logged recovery mode, new to SQL Server 2000. It's one thing to read and memorize what can and can't be accomplished in Bulk Logged recovery mode, but it's a totally different feeling being enlightened on why it works the way it works. There are many core principles in how SQL Server operates that the author describes in detail.

Here what I wished to have seen more coverage on: 1. Replication - this book does not cover replication. There are other books that show you how to point and click (I guess those point-and-click picture books will be called "Outside SQL Server 2000"), but none I've seen go into the "Inside" level. 2. Distributed Partitioned Views - This book shows how to create a distributed partitioned views, but it stops there. There are very important design considerations such as knowing where to place your data so as to minimize joins across the network. 3. Indexed Views - Same as distributed partitioned views - ends at the "how to create". I'd like to see how it works under the hood. 4. One way to classify this book is that it is very "server-centric". Many of us work on systems of database servers that work in concert under the application layer. I'd like to see more inter-server ("system-centric") insights.

It really wouldn't be fair for me to ask for clustering or log shipping in this book - those really aren't developer issues. Clustering is much more of a Windows 2000 feature than a SQL Server feature. If you're interested in clustering or log shipping, check out the SQL Server 2000 Admin Companion and the SQL Server 2000 Resource Kit.

Given the microscopic details packed into about 1000 pages, I found myself relying on Books Online for supplement. Many times, however, this book is actually more in depth than Books Online. I do agree with a few other reviewers that there's a lot of similarity with the version 7.0 of this book; however, there are many aspects of the products that are the same as well. If you really read the two books, though, you'll find that the 2000 version of the book is actually more in depth than the 7.0 version.

This book is very comprehensive and discusses, without wasting any space, what you should know if you want to call yourself expert in your role as a SQL Server DBA or even developer. You can find most of the information by using other Microsoft sources, mainly Books Online (BOL), but choosing between the two, I prefer the book thanks to it's pedagogical approach. It's also my first choice as a reference so I try to keep it within arms reach at all times.

As others have claimed, the coverage of the new features in SQL Server 2000 isn't complete, but neither is coverage of the old features. This is just a matter of limited space and doesn't affect my rating of the book since it covers the most important stuff and does it really well. If you already own the SQL Server 7 version of this book, the 2000 version might not add much value to you though.

SQL Server is a huge subject, not possible to cover completely in 1,000 pages. By the end of the book, the author has included a thorough list of suggested additional reading and he stresses the importance of also reading the SQL Server documentation, cover to cover.

I found this book to be the best I've read about the inner workings of SQL Server 2000 and have enjoyed it together with Ken England's SQL Server 2000 Performance Optimization and Tuning Handbook which fills in some gaps for those performance-oriented.

Brilliant as this book is, you still need to complement it with experience and Books Online to master SQL Server and all it's nitty-gritty details.

The first thing you notice about this book is that it's very much like its predecessor. One could try to explain that away by saying that the two versions of SQL Server are very similar, but that's not as true as it might seem. Fact is, there were some major enhancements in the 2000 product and this book doesn't cover them.

The second problem I have with the book is that it doesn't cover the topics an Inside book should cover. For one thing, the T-SQL coverage is more anemic than ever. The only good book on T-SQL out there is Ken Henderson's Guru's Guide book. I have all the others - the Itzik Ben Gan book (a real stinker), the Amo book, the Teach Yourself books - all have major shortcomings. An Inside book on SQL Server should cover T-SQL extensively. This book fails to do that.

In addition, there's no XML in this book. No, I didn't expect the book to be about XML, but I did expect it to be covered. I expected the internals of how it works to be discussed. I expected examples showing how SQL Server can process XML to be demonstrated. I expected insider info on what we might expect in the future regarding XML to be covered. Problem is, XML isn't even mentioned in the book. It's as though it was not added to the product. This was the single biggest new feature in SQL Server 2000. Not to have it mentioned at all is simply inexcusable.

The third problem I have is the sheer number of errors in the book. I've taken to marking them in my copy. There are errors of fact, historical errors (regarding SQL Server's past), and just plain bad advice. The book is rife with academic-inspired thinking that has little basis in reality.

The fourth problem is...[t]he book reads like a... chemistry text book. There's nothing engaging about it. There's nothing to welcome you into the technology. Nothing draws me in. On the contrary, I find myself getting dozy every time I sit down with it. It's missing something, though I'm not sure just what.

Technical books *can* be interesting. They can engage you. Just read the aforementioned Guru's Guide book to see what I mean. It is solid technically, but also funny, engaging, and insightful. This book could use a bit of the Guru's Guide in it.

I'm giving this book a low review because I felt snookered when I discovered how similar it was to the previous version. As someone else here said, it should have pointed that out in the Preface...

So, I'll say what the author should have said herself: If you are a competent SQL Server professional and you have read the version 7.0 book, you don't need this one. Enough said.

Let me start by saying I've read 3 books in this series (6.5, 7.0 & 2000). I've been a DBA for 5 years (2 years Oracle and now 3 years MS SQL Server).

This book is a one-of-a-kind in the sea of SQL Server books available these days. It is targetted specifically at advanced developers and DBAs (and anybody else) who wants or needs to know the internals of SQL Server eg. How the SQL Server engine manages threads, memory, and on disk data structures.

There have been a lot of reviews about what this book lacks but it is called "Inside SQL Server" not "EVERYTHING SQL Server". It's not about peripheral components like DTS, XML, replication and a vast array of other areas that SQL Server offers. It's also not about how to write SQL and it's not an Administration manual. At the risk of sounding repetitive (and a bit silly) it also won't teach you how to become a Day Trader or balance your checkbook and there's no discussion on the mechanics of Quantum Theory. If you buy this book hoping it will solve all of life's problems, I'm afraid, like many others (based on previous reviews), you'll be sadly disappointed.

What it will show you is how SQL Server's Query Processor compiles and optimizes your SQL (good or bad). It will show you what the Lock Manager does and why. It will teach you how indexes and tables are stored on disk and how and why they are cached in memory. It discusses what fields are good candidates for indexing and when the Optimizer will use them. There's a chapter on efficient use of cursors and another on why stored procedures and batches are critical to good performance. There's much much more but at all times the focus is strictly on the core RDBMS.

In closing I believe that this is an exceptional book, but it's not for everybody. If you're a developer (VB, c++, ...) who has to write some SQL as well then you're probably better off taking Kalen's own advice and reading "Advanced Transact-SQL for SQL Server 2000". After you read that (or something similar) and you need to get more performance, scalability, concurrency and/or consistency from SQL Server and know WHY you are getting it then this is "The [SQL Server] Bible".

As someone who already has alot of experience with the product, it's very nice indeed to have a book that covers the internals as well as this book does.

I guess it's important to understand that this is not a book for newcomers to SQL Server, nor is it a "how-to" book. This book is useful when you want to know what happens to your database when you perform an INSERT on a table with a clustered key. It tells you how the TEXT fields are stored, which allows a DBA to make an intelligent decision about when and if to use such a field. I obviously could go on and on. As far as I can tell, this is the *only* book out there that will give you that kind of information.

Since this isn't a sequel or trilogy, I was expecting that there would be some redundancy between the 7.0 and 2000 version of the book simply because that same redundancy exists in the real product!

If you are already an experienced DBA and are looking for an understanding of the internals so as to make better tuning decisions, then by all means,get this book. If you're new to SQL Server and you stiil don't know the difference between WHERE and HAVING, then you may be disappointed in this book, because it's not going to teach you how to use the EM and QA.

This is really the 7.0 book with a pretty adequate find-replace job done on it. None of the new features in 2000 (XML, multi-instancing, server federations) is covered in this book. A real let-down.