INTRODUCTION

To discuss fragmentation intelligently, it is first necessary to establish what we mean when we use the word computer. The simple definition of computer is:

 

An electronic machine for making calculations, storing and analyzing information fed into it, and controlling machinery automatically.

There is more to it than that, however. It is important to note as well that a computer does not think, that a computer has speed and power out of proportion to anything else in our day-to-day lives, and that a computer can be used to "enormously increase the production and income of an area." I really like the viewpoint expressed by this last quotation, as it tends to perk up the ears and interest people in the real potential of this marvelous machine. The quote is from an essay entitled What is a Computer? by L. Ron Hubbard, which is included in full in Appendix A .

The computers with which we are specifically concerned are those which run on the OpenVMS operating system, namely the VAX and the Alpha AXP.

VAX is defined as:

 

A computer made by Digital Equipment Corporation. VAX is a high-performance, multiprogramming computer system based on a. 32-bit architecture VAX stands for Virtual Address eXtension .

Alpha AXP is defined as:

 

A computer made by Digital Equipment Corporation. Alpha AXP is a RISC-based computer that uses a few simple, fast instructions in order to facilitate faster processing speed .

The Alpha AXP computer was introduced by Digital in November 1992. Although Alpha AXP processors use different sets of instructions than VAX processors, Digital developed a translation of the OpenVMS operating system for the new Alpha AXP computers. For reasons which will become clear later on in this book, this had the effect of transferring the very problem about which this book is written from the VAX computer to the Alpha AXP computer; hence our discussion applies equally to both computer systems.

Since we are going to be talking a lot about disks, let's agree on what a disk is. The maker of both the VAX and Alpha AXP computers, Digital Equipment Corporation, defines disk as:

 

A thin, round plate with a magnetic surface coating on which data can be stored by magnetic recording.

Disks come in many different sizes and architectures . And, even though we will limit our discussion to fixed (hard, not floppy) disks, there are still a lot of sizes and architectures to consider. Disk architecture is so fundamental to a discussion of fragmentation that an entire chapter (the first) is devoted to it. The chapter consists mostly of definitions of terms, but it contains a lot of pictures and is worth reviewing even if you are already familiar with basic disk concepts.

The purpose of a disk is to store information. This information, or data, is said to be stored in "files."

In computer terminology, any collection of data that is treated as a single unit on a storage medium (such as a disk) is referred to as a "file." Not unlike a manila folder designed to hold sheets of paper in concept, computer files are stored on a disk, with the disk acting as a filing cabinet.

A file can be accessed (found and retrieved), modified (changed in some way) and again stored on the disk. In this way, thousands upon thousands of pieces of information can be stored on a physically small disk, much more than can be stored in a regular filing cabinet.

Now we come to the real question: What exactly is fragmentation anyway?

The word fragmentation means "the state of being fragmented." The word fragment means "a detached, isolated or incomplete part." It is derived from the Latin fragmentum, which in turn is derived from frangere, meaning "break." So fragmentation means that something is broken into parts that are detached, isolated or incomplete.

There are two types of fragmentation with which we are immediately concerned: file fragmentation and free space fragmentation. File fragmentation concerns computer disk files that are not whole but rather are broken into scattered parts, while free space fragmentation means that the empty space on a disk is broken into scattered parts rather than being collected all in one big empty space. File fragmentation causes problems with accessing data stored in computer disk files, while free space fragmentation causes problems creating new data files or extending (adding to) old ones.

In Figure 1, a file consisting of ten records is shown with all ten records contiguous (immediately adjacent to each other) and again with the records scattered around in three different groups. The first arrangement is not fragmented. The second arrangement is fragmented.

Figure 1 File Fragmentation

In Figure 2, three files are arranged contiguously and again with the files scattered around in three different places. In the first arrangement, the free space is not fragmented; it is consolidated into one large area. In the second arrangement, the free space is fragmented.

Figure 2 Free Space Fragmentation

Taken together, we refer to the two types of fragmentation as disk fragmentation. It is important to note that, when talking about fragmentation, we are talking about the file as a container for data and not about the contents (data) of the file itself. People sometimes use the word fragmentation to describe the condition of a file which has its records (contents) scattered about within the file, separated by numerous small gaps. This type of fragmentation may be a problem with the application which maintains the file; it is not inherent in the operating system or disk file structure .

In Figure 3, three records are arranged first contiguously and then again with the records separated by empty record space. In the first arrangement, the record space is not fragmented; it is consolidated into one large area. In the second arrangement, the record space is fragmented.

Figure 3 Record Space Fragmentation

Since record space fragmentation is the concern of applications and not of the operating system or file system, this book does not address the subject any further.

The various solutions to fragmentation are collectively referred to as defragmentation, meaning something that gets rid of the fragmentation problem.

This book will tell you all about fragmentation and defragmentation, and do it in a way that is clear and understandable.

The material will be presented with care taken to define technical terms. Lots of diagrams and pictures will be used and the material will be presented on a gradient, bit by bit from the simplest material to the more complex, so you do not have to be an expert in later areas to understand earlier areas. For your part, take care to look up the definitions of any words you do not understand or are uncertain of, even non-technical words. If you are unfamiliar with the physical objects discussed, arrange to look at or touch them. For intangibles, of which there are many in this subject, try drawing pictures of them to develop a more solid concept.

Finally, if you are having trouble in an area, go back to where you were last doing well, and check to see if there were any words you did not fully understand. If so, look up the definition in a good (not complicated) dictionary. If you do this thoroughly, you will find the material much easier to read. (This works for any subject.)

The Approach

Here is the approach that will be used to present the story of fragmentation:

After explaining how a disk works, and taking a quick look at the OpenVMS file system, fragmentation will be explained in detail. Then you will learn how to tell whether your system suffers from fragmentation, and if so, how badly. We will then look at the effects of fragmentation on the VAX or Alpha AXP system - what's wrong with fragmentation? After that, we'll see what you can do about it, and how you can get your computer to take care of fragmentation by itself. This will include a view to the future, anticipating the extinction of fragmentation for OpenVMS computer systems. The final chapter is devoted to the ultimate solution to the fragmentation problem. Along the way, we will view how this problem came about, why it wasn't detected and prevented early on, Digital's viewpoint, some of the controversy over defragmentation, and some of my personal opinions on the whole subject.

Appendices are included with fill-in-the-blanks solutions for organizational problems that get in the way of solving the fragmentation problem: how to determine the cost of fragmentation, how to justify the cost of a solution and how to get your company to agree to spend the money you need to handle it.

The single most important thing to keep in mind is that a computer is used to get work done, so fragmentation is not a problem unless it interferes with doing useful work. Similarly, any solution must be one which allows us to get more useful work done with our computer system.

Actually, fragmentation does interfere with our use of the computer to do work. It slows things down - little by little, more and more. If left unhandled for a year or more, it can bring any computer running the OpenVMS operating system close to a complete standstill. I have personally experienced a VAX with a system disk so badly fragmented it took half an hour just to log in to a user account! After defragmenting, it took only a few seconds to log in - good as new.

Now let's start at the beginning and unravel the whole story of fragmentation.

[ TABLE OF CONTENTS ][ NEXT CHAPTER ]