John wiley sons storage networks explained basics and application of fibre channel san nas iscsi and infiniband 0470861827

Storage Networks Explained
Basics and Application of Fibre Channel SAN,
NAS, iSCSI and InfiniBand

Ulf Troppens, Rainer Erkens
IBM TotalStorage Interoperability Center, Mainz, Germany

Wolfgang Muller
¨
IBM Software Development Open Systems, Mainz, Germany
Translated by Rachel Waddington, Member of the Institute of Translating and
Interpreting


Storage Networks Explained



Storage Networks Explained
Basics and Application of Fibre Channel SAN,
NAS, iSCSI and InfiniBand

Ulf Troppens, Rainer Erkens
IBM TotalStorage Interoperability Center, Mainz, Germany

Wolfgang Muller
¨
IBM Software Development Open Systems, Mainz, Germany
Translated by Rachel Waddington, Member of the Institute of Translating and
Interpreting

First published under the title Speichernetze. Grundlagen und Einsatz von Fibre Channel SAN, NAS, iSCSI und
InfiniBand, ISBN: 3-89864-135-X by dpunkt.verlag GmbH
 dpunkt.verlag GmbH, Heidelberg, Germany, 2003
Copyright  2004

for the English translation: John Wiley & Sons Ltd
The Atrium, Southern Gate, Chichester,
West Sussex PO19 8SQ, England
Telephone (+44) 1243 779777

Email (for orders and customer service enquiries):
Visit our Home Page on www.wileyeurope.com or www.wiley.com
All Rights Reserved. No part of this publication may be reproduced, stored in a retrieval system or
transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning or
otherwise, except under the terms of the Copyright, Designs and Patents Act 1988 or under the terms of a
licence issued by the Copyright Licensing Agency Ltd, 90 Tottenham Court Road, London W1T 4LP, UK,
without the permission in writing of the Publisher. Requests to the Publisher should be addressed to the
Permissions Department, John Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex PO19
8SQ, England, or emailed to , or faxed to (+44) 1243 770620.
This publication is designed to provide accurate and authoritative information in regard to the subject matter
covered. It is sold on the understanding that the Publisher is not engaged in rendering professional services. If
professional advice or other expert assistance is required, the services of a competent professional should be
sought.
Other Wiley Editorial Offices
John Wiley & Sons Inc., 111 River Street, Hoboken, NJ 07030, USA
Jossey-Bass, 989 Market Street, San Francisco, CA 94103-1741, USA
Wiley-VCH Verlag GmbH, Boschstr. 12, D-69469 Weinheim, Germany
John Wiley & Sons Australia Ltd, 33 Park Road, Milton, Queensland 4064, Australia

John Wiley & Sons (Asia) Pte Ltd, 2 Clementi Loop #02-01, Jin Xing Distripark, Singapore 129809
John Wiley & Sons Canada Ltd, 22 Worcester Road, Etobicoke, Ontario, Canada M9W 1L1
Wiley also publishes its books in a variety of electronic formats. Some content that appears
in print may not be available in electronic books.

British Library Cataloguing in Publication Data
A catalogue record for this book is available from the British Library
ISBN 0-470-86182-7
Typeset in 10/12pt Times by Laserwords Private Limited, Chennai, India
Printed and bound in Great Britain by Antony Rowe Ltd, Chippenham, Wiltshire
This book is printed on acid-free paper responsibly manufactured from sustainable forestry
in which at least two trees are planted for each one used for paper production.


For Silke, Hannah, Nina, and Julia
You keep showing me what really matters in life.
For Christina, Edith, and Heribert
For the love and constant support you have given me.
For Christel
Only your patience and your understanding have made my contribution
to this book possible.



Contents

About the Authors

xv

Foreword by Tom Clark

xvii

Foreword to the German Edition by Hermann Strass

xix

Preface by the Authors

xxi

List of Figures and Tables

xxv

1 Introduction
1.1 Server-Centric IT Architecture and its Limitations
1.2 Storage-Centric IT Architecture and its Advantages
1.3 Case Study: Replacing a Server with Storage Networks
1.4 The Structure of the Book

1
1
3
4
7

PART I Technologies for Storage Networks

11

2 Intelligent Disk Subsystems
2.1 Architecture of Intelligent Disk Subsystems
2.2 Hard Disks and Internal I/O Channels
2.3 JBOD: Just a Bunch of Disks
2.4 Storage Virtualization using RAID
2.5 Different RAID Levels in Detail
2.5.1 RAID 0: block-by-block striping
2.5.2 RAID 1: block-by-block mirroring
2.5.3 RAID 0+1/RAID 10: striping and mirroring combined
2.5.4 RAID 4 and RAID 5: parity instead of mirroring

13
14
16
19
20
22
23
24
24
28


viii

CONTENTS

2.5.5 RAID 2 and RAID 3

2.5.6 A comparison of the RAID levels
Caching: Acceleration of Hard Disk Access
2.6.1 Cache on the hard disk
2.6.2 Write cache in the controller of the disk subsystem
2.6.3 Read cache in the RAID controller
Intelligent Disk Subsystems
2.7.1 Instant copies
2.7.2 Remote mirroring
2.7.3 LUN masking
Availability of Disk Subsystems
Summary

33
34
35
35
36
36
37
37
39
43
45
46

3 I/O Techniques
3.1 The Physical I/O Path from the CPU to the Storage System
3.2 SCSI
3.2.1 SCSI basics
3.2.2 SCSI and storage networks

3.3 The Fibre Channel Protocol Stack
3.3.1 Links, ports and topologies
3.3.2 FC-0: cables, plugs and signal encoding
3.3.3 FC-1: 8b/10b encoding, ordered sets and link control
protocol
3.3.4 FC-2: data transfer
3.3.5 FC-3: common services
3.3.6 Link services: login and addressing
3.3.7 Fabric services: name server and co
3.3.8 FC-4 and ULPs: application protocols
3.4 Fibre Channel SAN
3.4.1 Point-to-point topology
3.4.2 Fabric topology
3.4.3 Arbitrated loop topology
3.4.4 Hardware components for Fibre Channel SAN
3.4.5 InterSANs
3.4.6 Interoperability of Fibre Channel SAN
3.5 IP Storage
3.5.1 IP storage standards: iSCSI, iFCP, mFCP, FCIP and iSNS
3.5.2 TCP/IP and Ethernet as an I/O technology
3.5.3 Migration from SCSI and Fibre Channel to IP storage
3.6 InfiniBand
3.7 Virtual Interfaces and Remote Direct Memory Access (RDMA)
3.8 RDMA over TCP, Socket Direct Protocol (SDP) and iSCSI
Extensions for RDMA (iSER)
3.9 Summary

49
50
52

52
54
56
57
60

2.6

2.7

2.8
2.9

62
65
71
71
74
75
77
77
78
81
84
88
89
91
91
97
102

103
106
108
109


CONTENTS

ix

4 File Systems and Network Attached Storage (NAS)
4.1 Local File Systems
4.1.1 File systems and databases
4.1.2 Journaling
4.1.3 Snapshots
4.1.4 Volume manager
4.2 Network File Systems and File Servers
4.2.1 Basic principle
4.2.2 Network Attached Storage (NAS)
4.2.3 Performance bottlenecks in file servers
4.2.4 Acceleration of network file systems
4.2.5 Case study: the Direct Access File System (DAFS)
4.3 Shared Disk File Systems
4.3.1 Case study: the General Parallel File System (GPFS)
4.4 Comparison: NAS, Fibre Channel SAN and iSCSI SAN
4.5 Summary

111
111
112

112
113
113
115
116
117
118
120
120
124
126
129
132

5 Storage Virtualization
5.1 Once Again: Virtualization in the I/O Path
5.2 Limitations and Requirements
5.2.1 Architecture-related limitations of non-virtualized storage
networks
5.2.2 Implementation-related limitations of storage networks
5.2.3 Increasing complexity in the administration of large storage
networks
5.2.4 Proposed solution: storage virtualization
5.3 Definition of Storage Virtualization
5.4 Implementation Considerations
5.4.1 Realization of the virtualization entity
5.4.2 Exchange of storage devices
5.4.3 Efficient use of resources by dynamic storage allocation
5.4.4 Efficient use of resources by data migration
5.4.5 Performance increase

5.4.6 Availability due to the introduction of redundancy
5.4.7 Back-up and archiving
5.4.8 Data sharing
5.4.9 Privacy protection
5.5 Storage Virtualization on Block or File Level
5.6 Storage Virtualization on Various Levels of the Storage Network
5.6.1 Storage virtualization in the server
5.6.2 Storage virtualization in storage devices
5.6.3 Storage virtualization in the network

133
135
139
139
141
143
143
144
146
146
147
147
148
148
148
149
149
149
149
152

153
153
154


x

CONTENTS

5.7 Symmetric and Asymmetric Storage Virtualization in the Network
5.7.1 Symmetric storage virtualization
5.7.2 Asymmetric storage virtualization
5.8 Summary

155
155
158
162

PART II Application and Management of Storage Networks

163

6 Application of Storage Networks
6.1 Definition of the Term ‘Storage Network’
6.1.1 Layering of the transmission techniques and protocols
6.1.2 Networks in the I/O path
6.1.3 Data networks, voice networks and storage networks
6.2 Storage Sharing
6.2.1 Disk storage pooling

6.2.2 Dynamic tape library sharing
6.2.3 Data sharing
6.3 Availability of Data
6.3.1 Failure of an I/O bus
6.3.2 Failure of a server
6.3.3 Failure of a disk subsystem
6.3.4 Failure of virtualization in the storage network
6.3.5 Failure of a data centre based upon the case study
‘protection of an important database’
6.4 Adaptability and Scalability of IT Systems
6.4.1 Clustering for load distribution
6.4.2 Web architecture
6.4.3 Web applications based upon the ‘travel portal’
case study
6.5 Summary

165
165
166
167
168
169
169
170
172
175
176
179
181
184

7 Network Back-up
7.1 General Conditions for Back-up
7.2 Network Back-up Services
7.3 Server Components
7.3.1 Job scheduler
7.3.2 Error handling
7.3.3 Metadata database
7.3.4 Media manager
7.4 Back-up Clients
7.5 Performance Gains as a Result of Network
Back-up
7.6 Performance Bottlenecks of Network Back-up
7.6.1 Application-specific performance bottlenecks

205
206
206
208
209
209
209
209
212

184
187
188
192
195

203

213
213
214


CONTENTS

7.7

7.8

7.9

7.10

7.11
7.12

7.6.2 Performance bottlenecks due to server-centric IT
architecture
Limited Opportunities for Increasing Performance
7.7.1 Separate LAN for network back-up
7.7.2 Several back-up servers
7.7.3 Back-up server and application server on the same
physical computer
Next Generation Back-up
7.8.1 Server-free back-up
7.8.2 LAN-free back-up

7.8.3 LAN-free back-up with shared disk file systems
7.8.4 Tape library sharing
7.8.5 Back-up using instant copies
7.8.6 Data protection using remote mirroring
Back-up of File Systems
7.9.1 Back-up of file servers
7.9.2 Back-up of file systems
7.9.3 Back-up of NAS servers
7.9.4 The Network Data Management Protocol (NDMP)
Back-up of Databases
7.10.1 Operating method of database systems
7.10.2 Classical back-up of databases
7.10.3 Next generation back-up of databases
Organizational Aspects of Back-up
Summary

8 Management of Storage Networks
8.1 System Management
8.2 Requirements of Management Systems
8.3 Support by Management Systems
8.4 Management Interfaces
8.5 Standardized and Proprietary Mechanisms
8.5.1 Standardized mechanisms
8.5.2 Proprietary mechanisms
8.6 In-band Management
8.6.1 In-band management in Fibre Channel SAN
8.7 Out-band Management
8.7.1 Use of SNMP
8.7.2 CIM and WBEM
8.7.3 Storage Management Initiative Specification

(SMI-S)
8.7.4 CMIP and DMI
8.8 Operational Aspects of the Management of Storage Networks
8.9 Summary

xi

214
215
215
217
218
219
219
221
222
223
224
227
227
227
228
229
231
238
238
240
242
243
244

247
247
248
249
251
252
253
253
254
256
258
259
264
270
271
272
275


xii

9 Removable Media Management
9.1 The Significance of Removable Media
9.2 Removable Media
9.2.1 Tapes
9.2.2 CD, DVD and magneto-optical media
9.2.3 Management features of removable media
9.3 Libraries and Drives
9.3.1 Drives
9.3.2 Media changers

9.3.3 Control of the media changer
9.4 Problems and Requirements Related to Removable Media
Management
9.4.1 Efficient use of the available resources
9.4.2 Access control
9.4.3 Access synchronization
9.4.4 Access prioritization and mount request queuing
9.4.5 Media tracking
9.4.6 Grouping, pooling
9.4.7 Monitoring
9.4.8 Reporting
9.4.9 Life cycle management
9.4.10 Vaulting
9.5 The IEEE 1244 Standard for Removable Media Management
9.5.1 Media management system architecture
9.5.2 The IEEE 1244 data model
9.5.3 Media Management Protocol (MMP)
9.5.4 Library and drive management protocol (LMP/DMP)
9.6 Summary

10 The SNIA Shared Storage Model
10.1 The Model
10.1.1 The functional approach
10.1.2 Graphical representations
10.1.3 An elementary overview
10.1.4 The components
10.1.5 The layers
10.1.6 The file/record layer
10.1.7 The block layer
10.1.8 Combination of the block and file/record layers

10.1.9 Access paths
10.1.10 Caching
10.1.11 Access control
10.1.12 Clustering
10.1.13 Storage, data and information

CONTENTS

277
277
278
278
279
279
280
281
281
282
283
286
287
289
289
290
292
294
295
296
298
298

299
302
302
306
306

309
310
310
311
311
312
313
314
316
317
318
319
321
322
322


CONTENTS

10.2

10.3

10.4

10.5

10.1.14 Resource and data sharing
10.1.15 The service subsystem
Examples of Disk-Based Storage Architectures
10.2.1 Direct attached block storage
10.2.2 Storage network attached block storage
10.2.3 Block storage aggregation in a storage device: SAN
appliance
10.2.4 Network attached block storage with metadata server:
asymmetric block services
10.2.5 Multi-site block storage
10.2.6 File server
10.2.7 File server controller: NAS heads
10.2.8 Asymmetric file services: NAS/file server metadata
manager
10.2.9 Object-based storage device (OSD)
Extension of the SNIA Shared Storage Model to Tape Functions
10.3.1 Logical and physical structure of tapes
10.3.2 Differences between disk and tape
10.3.3 Extension of the model
Examples of Tape-Based Back-Up Techniques and Architectures
10.4.1 File back-up
10.4.2 File system volume back-up
10.4.3 Volume back-up
10.4.4 File back-up to virtual tape
10.4.5 Direct attached tape
10.4.6 LAN attached tape
10.4.7 Shared tape drive

10.4.8 Partitioned tape library
10.4.9 Virtual tape controller
10.4.10 Virtual tape controller with disk cache
10.4.11 Data mover for tape
10.4.12 File server with tape drive
10.4.13 File server with external tape
10.4.14 File server with data mover
Summary

xiii

322
324
326
326
326
327
328
329
330
331
333
334
336
336
337
338
339
340
341

341
343
343
344
344
345
346
347
348
349
350
351
352

Glossary

353

Annotated Bibliography

373

Appendix A: Proof of the Calculation of the Parity Block
of RAID 4 and 5

379

Appendix B: Checklist for the Management of Storage Networks
B.1 Applications

381
382


xiv

B.1.1 Monitoring
B.1.2 Availability
B.1.3 Performance
B.1.4 Scalability
B.1.5 Efficient use
B.2 Data
B.2.1 Availability
B.2.2 Performance
B.2.3 Data protection
B.2.4 Archiving
B.2.5 Migration
B.2.6 Data sharing
B.2.7 Security/access control
B.3 Resources
B.3.1 Inventory/asset management and planning
B.3.2 Monitoring
B.3.3 Configuration
B.3.4 Resource use
B.3.5 Capacity
B.3.6 Efficient resource utilization
B.3.7 Availability
B.3.8 Resource migration
B.3.9 Security
B.4 Network

B.4.1 Topology
B.4.2 Monitoring
B.4.3 Availability
B.4.4 Performance
Index

CONTENTS

382
382
382
382
382
383
383
383
383
383
383
384
384
384
384
384
384
385
385
385
385
385

386
386
386
386
386
386
387


About the authors

Ulf Troppens (left) and Rainer Erkens (centre) are employed at IBM TotalStorage
Interoperability Centre in Mainz, a testing, development and demonstration laboratory for
storage products and storage networks. Both authors work at the interface between technology and customers. Their duties include the testing of new products and the validation
of concepts on the basis of customer environments set up in the laboratory. They present


xvi

ABOUT THE AUTHORS

the latest hardware and software products in the field of storage networks to customers
and discuss the set up test environment with them.
Ulf Troppens studied Computer Science at the University of Karlsruhe until 1995.
Since 1989 he has been primarily involved in the administration of Unix systems, storage
systems, data and storage networks and distributed applications. In April 2004 Ulf joined
Wolfgang’s team to help with the roll-out of a software for tape library virtualization.
Rainer Erkens studied Mathematics at the University of Mainz until 2000. His experience in the management of computers and the management of distributed applications
goes back to 1992. Since 2000 he has been working primarily with storage systems and
storage networks. In February 2004 he was appointed to Chairman of the SNIA Europe

Solutions Committee.
Wolfgang Muller
(right) is currently working as a software architect in the Storage
¨
Software Development Department at IBM in Mainz, Germany, where the focus is on
software development projects supporting open standards such as SMI-S/CIM/WBEM
and IEEE 1244. He received his Dipl.-Inform.(FH) degree in computer science from the
University of Applied Sciences, Darmstadt, Germany, in 1993.


Foreword by Tom Clark

It is gratifying to see Storage Networks Explained appear in this new English translation.
The fact that this work originally appeared in German and is now available in English
translation is a testimony to the global adoption of SAN technology and the proliferation
of SAN expertise internationally. Ulf Troppens, Rainer Erkens, and Wolfgang M¨uller
have created an invaluable resource for understanding and implementing efficient storage
networking solutions. Although this work does not claim to provide detailed blueprints
for SANs, it provides a wealth of practical information for leveraging the benefits that
SANs provide for storage data, including storage consolidation, high availability access
to data, and data protection via data copy and tape backup.
The migration from fixed, direct-attached storage to storage networking has had a
major impact on the IT community over the past decade. The first-generation Fibre Channel SANs that were initially adopted by large enterprises have now spread to a much
wider market of enterprise branches and small and medium businesses. At the same time,
entirely new technologies have emerged within storage networking such as IP SANs, storage virtualization, and comprehensive SAN management based on CIM and SMI-S. These
new SAN technologies are helping to drive storage networking into the mainstream while
delivering more productive applications based on SAN intelligence and storage-process
automation. Although it has taken several years to overcome basic transport, interoperability, and management issues in storage networking, the ideal of a storage utility seems
now within reach.
Like other technologies before it, SAN technology is quickly evolving towards its

own demise as a separate discipline. This will not occur through disappearance from
the market, but through universal and ubiquitous adoption across the entire market. Just
as Gigabit Ethernet over long distances is gradually obscuring the difference between
LANs and WANs, the melding of Fibre Channel and IP storage technologies will obscure
the difference between LANs, WANs, and SANs. Eventually, there will just be ‘the
network’ that will offer extremely high performance and resiliency where required to


xviii

FOREWORD BY TOM CLARK

service a wide variety of applications, including storage data transport and high availability
access.
Storage Networks Explained provides the basic knowledge to understand the various
technical components of both conventional and new SAN solutions, as well as practical
guidelines for aligning technical solutions with the business objectives of data availability
and preservation. Troppens, Erkens, and M¨uller apply technical explanations as needed
to assist the reader in differentiating between the many options available for SANs but
avoid burdening the work with excessive granular detail. This helps the reader to focus on
what is most relevant for making SAN technology decisions. Customers do not, after all,
deploy technology for technology’s sake but to solve real and pressing business problems.
By concentrating on the practical benefits of SANs for applications and business processes, this book is an essential resource for managers, administrators, and SAN architects
who have day-to-day responsibility for aligning the appropriate technologies to specific
business problems.
Tom Clark
Seattle, Washington
Former Board Member, Storage Networking Industry Association (SNIA)
Author: Designing Storage Area Networks Second Edition and IP SANs: A Guide to iSCSI,
iFCP and FCIP Protocols for IP Storage Area Networks

Foreword to the German
Edition by Hermann Strass

The subject of storage networks is particularly topical at the moment and for a number
of reasons will remain so for a very long time to come. Storage networking technology
brings with it fundamental new structures and procedures that will continue to be of
great importance for the foreseeable future regardless of incremental differences and
changes in products. A book on this subject is therefore of particular importance in these
fast-moving times.
This book is based upon the experience of the two authors (Wolfgang joined the other
two authors for the English edition of this book), who work with the subject matter on a
daily basis. It provides system administrators and system architects in particular, with the
tools for the optimal selection and the cost-effective use of this partially new technology,
the use and operation of which currently seem indispensable in view of the ever-growing
storage quantities in companies. The technology of networked storage is associated with
demonstrable decisive cost savings. Therefore, growth continues even in an unfavourable
economic climate.
Storage quantities are growing because we are currently tending to work in colour, in
three-dimension and in digital format to a much greater extent than was the case a few
years ago. Furthermore, there are now legal regulations in the European Union and in
other countries that make the electronic/digital storage of all business data compulsory.
The archiving of old business events in files in printed format is no longer sufficient to
comply with the law. And data quantities continue to increase in both good times and
bad: even lost orders and the related data have to be stored electronically. These legal
regulations thus ensure that a certain amount of growth in data quantities is inevitable.
In the past, data was stored upon disk and tape drives connected directly to the server.
Storage was operated as a peripheral to the computer. Access rights, virus protection
and other functions could thus be performed on the computer (server) in question. For

xx

FOREWORD

reasons that will be described in detail in this book, this way of working is no longer
practical today. Storage is therefore detached from the servers and brought together to
form a dedicated storage network. This gives rise to a fundamentally different way of
working. The new procedures that this requires have been and will continue to be further
developed and introduced now and in the future. Data storage thus has its own value. It
is no longer a question of attaching a further disk drive to a server.
Nowadays, stored data and the information that it contains are the prized possessions
for a company. The computers (servers) necessary for data processing can be purchased
at any time, by the dozen or in greater quantities, in the form of individual blade servers
or packed in cabinets, they can be integrated into a LAN and defective units can be
exchanged. However, if stored data is lost, getting it back is infinitely more expensive
and time-consuming – if it is indeed possible to restore some or all of the data. Data
generally also has to be available around the clock. Data networks must, therefore, be
designed with built-in redundancy and for high availability.
These and related topics are considered in detail in this book. The approach is based
only to a certain degree upon the current state of the art. More important in this context is
the description of the underlying topics and their interdependencies. This extends beyond
the scope of even more lengthy magazine articles and will continue to be topical even in
the future.
The requirements imposed upon storage networks are fundamentally different from
the requirements made of the local networks (LANs) used in the past. Therefore, storage
networks have up until now almost exclusively used the Fibre Channel technology that was
specially developed for them as a connection technology. However, storage networking is
no short-term fad and efforts are currently under way to use existing network technologies

(e.g. Ethernet/LAN/TCP-IP) and emerging technologies (e.g. InfiniBand). Under certain
circumstances this is a completely worthwhile alternative. The book highlights which
selection criteria play a role in this context. It is not usually technical details or prejudices
that are decisive here, but usage requirements, the available infrastructure and devices and
a careful estimate of the future development in companies. This book will therefore be a
valuable aid in the structural planning and selection of devices and software.
Overall, this book is an excellent work, which explains the subject comprehensively,
in detail and with a good technical foundation. It is to be hoped that the book gains a
wide circulation, particularly as it corrects a great many half-truths and prejudices.
Hermann Strass
Hermann Strass is an author and a consultant on new technologies, in particular bus architectures, mass storage, industrial networks and automation. He is a member of national
and international standardization committees and Technical Coordinator of the VMEbus
International Trade Association (VITA) in Europe.


Preface by the Authors

This Preface answers the following main questions:
•
•
•
•

What does this book deal with?
Who should read this book?
How should this book be read?
Who has written this book?

What does this book deal with?
The technology of Storage Area Networks (SANs) fundamentally changes the architecture

of IT systems. In conventional IT systems, storage devices are connected to servers by
means of SCSI cables. The idea behind storage networks is that these SCSI cables are
replaced by a network, which is installed in addition to the existing LAN. Server and
storage devices can exchange data over this new network using the SCSI protocol. Storage
networks have long been a known quantity in the world of mainframes. Fibre Channel,
iSCSI and Network Attached Storage (NAS) are now also taking storage networks into
the field of Open Systems (Unix, Windows NT/2000/2003, OS/400, Novell Netware,
MacOS). The term ‘Unix’ for us also encompasses the Linux operating system, which is
sometimes presented separately in such itemizations.
Storage networks are becoming a fundamental technology like databases or LANs.
According to market research, in 2004 over 70% of external storage will be realized in
the form of storage networks. The term ‘external storage’ is used here to mean storage
that is accommodated in a different enclosure to the server itself. Consequently, anyone
who is involved in the planning or the operation of IT systems requires a basic knowledge


xxii

PREFACE BY THE AUTHORS

of the principles and the use of storage networks. They are thus becoming as omnipresent
as SCSI, but are more complex than LANs and TCP/IP.
The book is divided into two parts. Part I deals with fundamental technologies relating
to storage networks. It guides the reader from the structure and operating method of
storage devices through I/O techniques and I/O protocols to the file systems and storage
virtualization.
Part II introduces applications that utilize the new functions of storage networks and
intelligent storage subsystems. The focus is upon the sharing of resources that can be
accessed via the storage network and upon server and application clusters, web applications and data protection (backup). Further focal points are the administration of storage
networks and the administration of removable media. Last but not least, the SNIA Shared

Storage Model provides a reference model to describe storage networks.
At the end of the book we have added a glossary, an index and an annotated bibliography, which in addition to further literature also highlights numerous freely available
sources on the Internet.
Section 1.4 sets out in detail the structure of the book and the relationships between
the individual chapters. Figure 1.7 on page 8 illustrates the structure of the book. At
this point, it is worth casting a glance at this illustration. Note that the illustration also
describes the subjects that we will not be covering.

Who should read this book?
Our approach is, first, to explain the basic techniques behind storage networks and, secondly, to show how these new techniques help to overcome problems in current IT
systems. The book is equally suitable for beginners with basic IT knowledge and for old
hands. It is more an introduction to the basic concepts and techniques than a technical
reference work. The target group thus includes:
• system administrators
• system architects
• decision makers
• students.
After reading the whole book you will be familiar with
• the concepts of storage networks and their basic techniques
• usage options for storage networks
• proposed solutions for the support of business processes with the aid of storage networks
• the advantages of storage networks
• new possibilities opened up by storage networks.


PREFACE BY THE AUTHORS

xxiii

How should this book be read?

There are two possible ways of reading this book. Those who are interested exclusively
in the concepts and the possibilities for the use of storage networks should read Chapter 1
(‘Introduction’), Chapter 5 (‘Virtualization’) and the Part II; any missing technical fundamentals can be looked up in the first part as required. Anyone who is also interested in
the technical foundations of storage networks should read the book from start to finish.

Who has written this book?
Two of us, Ulf Troppens and Rainer Erkens, work at the IBM TotalStorage Interoperability
Centre in Mainz, a testing and demonstration centre for storage networks. In our daily work
we install and configure hardware and software for storage networks. We tell customers
about suitable products and explain the concepts that underlie the products. We also
discuss with customers how storage networks can help solve problems in their current
IT systems. We are therefore very well acquainted with the questions that customers ask
about storage networks, both those with experience in storage networks and beginners. We
therefore believe that our daily work has helped us to structure the content of this book
and to select the subjects that are important to the readers of a book on storage networks.
Wolfgang M¨uller joined the authors for the English edition of this book. Wolfgang is
the architect in charge of a removable media management system developed at the IBM
Software Development Open Systems Lab in Mainz, Germany. He provides in-depth
knowledge on tape and tape management.
Our intention has been to take off our ‘IBM hats’ and write this book from an unbiased
viewpoint. Of course, as employees of IBM, experience and opinions have flowed into
this book that have been formed in our daily work. Despite this, the book is our personal
work and has no connection with IBM apart from our employee relationship. In particular,
this book does not present any official opinions of IBM.

Acknowledgements
To conclude the preface we would like to thank a few people who have made a significant
contribution to this book. From a chronological point of view we should first mention the
editors of iX magazine and the readers’ department of dpunkt.verlag, who set the whole
thing in motion in March 2001 with the question ‘Could you see yourselves writing a

book on the subject of storage in the network?’
Regarding content, our colleagues from the IBM Mainz storage community, especially
the former SAN Lab and the current TotalStorage Interoperability Centre deserve mention:
without the collaboration on storage hardware and software with customers and employees
of partner companies, business partners and IBM, and without the associated knowledge
exchange, we would lack the experience and knowledge that we have been able to put into