JL Computer Consultancy

Tales of the Oak Table

(In association with Amazon.com and Amazon.co.uk)

 

Pages: 456

Cover Price US: $39.99

Cover Price UK: £25.00

I’ve just had a note from the publisher about this book telling me it’s about to go into its second print run. This reminded me that I hadn’t put an advert, or even a comment, on my website about the book. So I’ve finally got around to rectifying that oversight.

This book is a combined effort from (largely) members of the Oak Table – each individual writing one chapter in their own style about their own choice of topic. The result is a mixture of history, philosophy, science, drama, laughter and tears. Some of it is technical, some of it is simply background color on the development, use, and abuse of Oracle databases.

If you read this book, it will probably help you understand how to use the database better – even though it contains no deep “internals” information, no amazing quick tips, and only a relatively straightforward description of where Oracle came from,  mistakes (lots of mistakes) people have made with it, and an education in how to investigate what is going on.

Chapter Author, Title, and Synopsis:

Mogens Norgaard

Introduction: Oracle and the Oak Table

Dave Ensor

A brief history of Oracle.

In this chapter, Dave, the grand old man of performance optimization methodology in the Oracle world, gives a personal perspective on the key events that have shaped the way in which the Oracle database has developed, and the way in which it is used (and abused). It is packed with wisdom and insight but is also rather long. Life’s like that sometimes.

 

Mogens Norgaard (again)

You probably don’t tune right

Why does history repeat itself? Who will live, who will die? What is the true meaning of life? Why do Danish towns have such strange names? The answer to all these questions, apart from the last one, can be found here. In this chapter, Mogens yearns for a quiet, peaceful life where database systems are stable and predictable, patches are a thing of the past, and people don’t all try to get on the fast lane of the highway during traffic jams. He argues that the only way to experience the first two is to build systems that are fully instrumented and tested, and that use response time-based performance optimization techniques. The only way to experience the third one is to move to Middelfart, in Denmark.

Connor McDonald

Waste Not, Want Not

Connor explains why most database performance problems are caused by unnecessary waste, by making the database perform more work than is necessary to achieve a task. This often arises when people try to build “generic” (e.g. “database-independent”) applications. Once these design mistakes have been made they can’t always be undone, but Connor shows that, with a bit of lateral thinking, you can successfully tune a database to acceptable levels even with one hand tied behind your back (and the other holding a stubby). [Ed: “Stubby” is Australian for “a bottle of weak, flavorless lager” (allegedly).]

Anjo Kolk

Why I invented YAPP (Yet another Performance Profiling Methodology

While stationed in Japan for Oracle Development, Anjo was trying to find a way to convince the Japanese developers on the project that it was a good idea to use fewer servers in the system. They said he was crazy, and perhaps there was some truth in that, but out of this was born the YAPP method, which was to revolutionize the way the Oracle world think about performance optimization.

Cary Millsap

Extended SQL Trace Data

On a dark and rainy Texan night in 1999, Jeff Holt and Cary were faced with a serious performance problem during a customer visit. They had heard rumors about an extended SQL trace facility in Oracle and decided to investigate. The rest is history. This is the story of how to turn a “simple,” yet starkly overlooked, feature of Oracle into a global business, in the process forcing most competitors to change their ways. [Ed: No box turtles were killed or injured during the making of this chapter.]

Kyle Hailey

Direct Memory Access

One fateful day in 1994, Kyle met a man called Roger Sanders, and was sucked into the twilight world of Direct Memory Access. Some years later he emerged, battered and bruised (but with perfect hair), clutching his own fully operational SGA attach program. In this chapter he explains how and, more importantly, why.

Gaja Vaidyanatha

Compulsive Tuning Disorder

For the knock-down price of $39.99 (the cost of this book) the master Standup Database Comedian himself provides the first total therapy session for those people who don’t know how or when to stop tuning. Includes psychological analysis, palliative treatment, and outlines your (long) path to redemption. All future sessions will be charged at market rates.

James Morle

New Releases and Big Projects

In 1993, James cut short his honeymoon to join a landmark project that would turn out to have widespread consequences for the customer and for several future members of the OakTable network. Armed only with endless determination and a winning smile, James fought against the dark forces of OPS and eventually prevailed. Peace was restored, happy customers wandered into the sunset clutching their rented toilet brushes. Aaahh ...

David Ruthven

Testing and Risk Management

Software quality and testing is something David knows about. As head of Oracle’s European DDR group, which makes the decisions about what bugs to fix, or not, his acute observations and insights are more useful than ever in today’s multi-component, multi-tier IT environments. David tackles his subject with a calm and measured approach, providing a refreshing antidote to the chaos and confusion around him.

Jonathan Lewis

Dreadful Design

According to Jonathan, an Oracle project is like a “duck faced with a 12-bore shotgun.” If it’s quick on its feet it might survive, with a just few grazes to show for it. In this chapter, however, Jonathan tells the heartbreaking tale of the duck that took every pellet. Not for the faint-hearted.

Tim Gorman

Bad CaRMa  (sic)

Sometimes the cleverest people make the most fundamental mistakes. And when they do, the results can be horrifying. In this chapter, Tim describes a really bad case: a vision and design for an entire OE/CRM application housed in one table (named, appropriately enough, DATA). It stored data just fine, but retrieval was a problem that was never resolved. The company that tried to make the vision a reality paid with its life.

Mogens Norgaard (yet again)

and James Morle

Join the BAARF party (or not)

The Battle Against Any Raid Five/Four/Free Party was founded in Birmingham because Mogens and James had had enough fruitless discussions about the inferior RAID-F technologies. This chapter is definitely the last word they’ll ever mutter on the subject.

Reviews from Amazon and Bookpool

 Review

Comment

This book is intended for Oracle internals experts who want a deep, deep drill down into the guts of Oracle to look for optimizations. It's well written and very in-depth, but you should have a look at the table of contents to make sure that you can get anything out of this book before you buy it.

 

If you aren't the target audience then you are likely to get little or nothing out of it.

I’m not sure that this review is correct about “deep, deep” material. There is a little high-tech stuff, principally from Kyle Hailey and his direct attach to the SGA and maybe a little bit from Cary about how queue theory is relevant; but most of the technical stuff is about how some very ordinary features of Oracle are badly applied.

 

This was a four star review, but I’m not entirely convinced that this reviewer read much more than Kyle’s chapter – but the advice to check the table of contents is good, and I think I’ll suggest to Apress that they make the descriptions of the chapters (see above) part of the ‘look-inside’ on Amazon.

I have 14 years of Oracle Experience with some of the busiest and largest transaction systems in the world. This book really hit home for me and brought back a lot of memories of painful times as well as gave me new insights. I have re-read this book twice since getting it a month ago. It is that good.

The Tales of the Oak Table is a skeptic's work of technological history that is funny and hard to put down, but which also provides experienced database professionals roadmaps to solve their pressing problems (or even see that they have a problem.)

The authors take an empirical, rational approach to diagnosing and discovering the most serious problems while providing amusing revelations about the people and organizations they have worked with. Along the way they lift Oracle's skirts and take us out back to show us the dirty laundry and other junk in Oracle's back yard. They provide methods to diagnose and repair problems in oracle performance as well as enumerate the known pitfalls in project management and database design. Seasoned IT types will groan and laugh during these chapters.

They also look at larger architectural, economic, psychological, and philosophical issues which have a direct impact on databases and large information systems. Norgaard's history of computing is quite depressing. And I agree that many of the "new" blood want to focus on .NET and J2EE - when the real heart and soul is still the data.

Oracle Insights DOES require both deep thought and deep, hard-won knowledge of Oracle in order to fully enjoy it. So, if at first you don't like it, then you don't know what you don't know.

This is a pretty thorough review – the reader seems to have gone through every chapter (three times), so it’s a good indication of the range and depth you’re going to get if you buy the book.

 

The point about “you don’t know what you don’t know” was interesting. It’s a common problem with Oracle. How do you find out which things are going to give you problems if you have no idea what sort of things might give you a problem.  There are two or three chapters (I’d like to think that my chapter is one of them) that will give you a rapid insight into where a lot of the traps are and how to pre-empt them.

This book is not designed to teach you anything about programming, database administration, or architecture. It is an excellent series of essays about real life professional experiences.

 

Two essays stand out above all the others. The first essay in the book about the evolution of the Oracle database provides excellent insights into how the database has changed over time.  The other article that stands out is the last one by Tim Gorman about the worst project he has ever been on. The project was run so badly it put a young company out of business

I’d agree with this one. The book is not a “how-to” book – although it has got quite a lot of “how not to” in it.

The book is aimed squarely at an existing Oracle user. It offers idiosyncratic viewpoints devoted to filling in gaps in the official Oracle documentation.

The authors are longtime Oracle experts who offer advice on how to optimally use various Oracle versions, and how to avoid potential pitfalls in usage.

Being a mostly MySql user myself, there was little here germane to my database operations.

But the first chapter stands well apart from the rest of the book. A behind the scenes history. Presumably unfettered by Oracle's corporate lawyers. A fascinating warts and all technical commentary on the development of the Oracle database. [And there are plenty of warts.] It can and probably should be read by anyone in the database field. Enough technical details are given to illustrate the best and worst features of each major Oracle release.

 

Yet during all this 25 years, Oracle rose from nothing to being the world's largest database company. They must have been getting crucial capabilities implemented correctly. The first chapter is a good complement to other books on Oracle that are written for a general audience. Those books describe more of the business/corporate side of Oracle. Of necessity, they had to go easy on the technical details. This chapter helps fill in those gaps.

 

Also makes one wonder what a similar description of IBM's dB2 history would say.

I like that word “idiosyncratic”. This is one of the more unusual books that you will find about the Oracle RDBMS.

You will certainly get very useful insights into managing Oracle RDBMS (up to version 10g). This book is worth every penny you are going to pay for it. I would say it is like oracles of Oracle speak! The list of authors (every one of them wrote a separate chapter) speaks for itself.

 

Enjoy! (Your spouse might not like this book because unless you read every page of it, you won't put it down).

I think there is one really important word in this review. If you are interested in your work with the Oracle database system, you will probably enjoy this book.

 

You may struggle to find anything specific about 10g in this book – despite the comment in the review. But nearly everything in this book is about nearly every version of Oracle – even the ones not yet defined.

Declaration of Interest: I was lucky to have worked alongside one of the authors. So that's why I bought it.

Glad I did. Read it on the beach over Easter weekend. Dipped in and out. Learned and laughed. Certain to re-read some chapters, and to refer back to it. The very different styles of the contributors make it all the more readable - you just don't know what’s coming next.

It’s worth the price for pretty much any 3 of the 11 chapters, even if you chose them at random. And you can hit your SAN manager over the head with it, he probably won't feel a thing anyway.

Thank you OakTable.

But it’s not in your shopping cart yet is it, because I've marked it down a star. Look, there are some minor bugbears. Don't be put off, just setting expectations:

1) There's some heavy name-dropping, perhaps in keeping with the project. Perhaps I'm jealous my brain isn't the size of a planet; these chaps' are.

2) You can see from (1) that HHGTTG was funny once, about 20 years ago. My dad has been citing it as the answer to everything ever since; it’s worn a bit thin. I'm sure he would love this book for exactly that reason.

3) I'm the sure the sequel will be even better.

You can put me down for 2 copies of "Seeing Double: Tales from Under the Oak Table" while I'm here.

For the uninitiated, HHGTTG is the Hitchhikers Guide to the Galaxy, by Douglas Adams. (First published 25 years ago, and as I write still ranked at 245 on Amazon.com).

 

In the guide we learn that the Earth an organic computer was built for the white mice to find the question to the answer to Life, the Universe, and Everything.  (An earlier computer had deduced that the answer was 42, but had not been able to reveal the question).

 

We also meet Marvin, the paranoid android with the latest “genuine people personality” circuits, a pain in all the diodes down his left hand side, a brain the size of a planet, and no job satisfaction.

 

 

The phrase to keep in mind is: “learned and laughed”. This is a book about Oracle that you really could read on the beach.

 

This Book is Not Yet Released. But like all other Oracle Professionals I am eagerly awaiting for this. The Contents page is enough to guess how good the book is going to be.

I approve of the sentiment because I am one of the authors – but I can’t help thinking of Richard Feynman’s dictum: “Question Authority”. The authors may have a track record of excellence but that doesn’t prove that they can write a good book.

 

(Interestingly, the bookshop sales have been much better than the on-line sales – possibly because the book isn’t obviously a ‘how-to’ book so people want to browse it before they buy it).

This book was shamelessly promoted on all the Oracle newsgroups as one of the finest Oracle collaborations of all time. Some material makes sense but much of the content is rubbish, like where the authors challenge simple concepts like buffer utilisation with circular logic.

 

The optimisation chapter is dated with Oracle 10g Automatic Session History and the OWI wait interface. The only things that was clear after reading the book was the great self-love between the authors, most of which I've never heard of. I would recommend the Oracle Press book on wait event tuning instead.

 

If I wanted a book on tuning, then I think the Oracle Press book on the Wait Interface (Kirtikumar Deshpande, Richmond Shee, and K Gopalakrishnan) is one that I’d choose as well. But I’m not sure which chapter this reviewer intended by “the optimisation chapter” as there isn’t one.

 

I searched comp.databases.oracle.* for “Oracle Insights” and “Tales of the Oak Table” and got only two hits for the book, by the way.  But there you go – I’m quite well known for spoiling a good story by insisting on checking the facts. I would have checked the bit about circular logic as well, except the reviewer failed to mention the chapter and page.

Anybody who creates and delivers software understands that the best times are when you have smart & opinionated people. When these people get the opportunity to informally sit down, discuss and knock ideas and opinions around, the outcome is usually the best for all. This book is just that. The short history on Oracle versions alone will help you understand the need to keep your team current and progressing.

Sitting down and knocking ideas and opinions around – that’s a pretty good description of what goes on when a handful of Oak Table members get together. Followed, of course, by hours of checking whether the ideas and opinions are right or wrong.

Very disappointing. Waited a long time and it's just a rehash of the Oracle manuals which you can get online.

I started with a reviewer that may not have read the whole book – so I’ve ended with one that may not have read any of it. There isn’t much of the manuals in this book.

 

Back to Main Index of Topics