Microsoft
®
Exchange
Server 2007:
Tony Redmond’s Guide to
Successful Implementation
This page intentionally left blank
Microsoft
®
Exchange
Server 2007:
Tony Redmond’s Guide to
Successful Implementation
Tony Redmond
Amsterdam • Boston • Heidelberg • London • New York • Oxford
Paris • San Diego• San Francisco • Singapore • Sydney • Tokyo

Digital Press is an imprint of Elsevier
Digital Press is an imprint of Elsevier
30 Corporate Drive, Suite 400, Burlington, MA 01803, USA
Linacre House, Jordan Hill, Oxford OX2 8DP, UK
Copyright © 2007, Hewlett-Packard Development Company, L.P. Published by
Elsevier. 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, or otherwise, without the prior written permission of the publisher.
Permissions may be sought directly from Elsevier’s Science & Technology Rights
Department in Oxford, UK: phone: (+44) 1865 843830, fax: (+44) 1865 853333,
E-mail: You may also complete your request online
via the Elsevier homepage (), by selecting “Support & Contact”

1.4.1 Modifying and removing servers 27
1.4.2 Validating the installation 27
1.4.3 Third-party software 28
1.5 Server roles 28
1.5.1 Services 32
1.6 Licensing 36
1.6.1 Version numbers 40
1.6.2 32-bit Exchange 2007? 41
1.7 Support 42
1.8 Challenges for Exchange 2007 42
1.9 Into the future 45
vi Contents
2 Exchange, Windows, and the Active Directory 47
2.1 Active Directory and Exchange 47
2.1.1 Domain Designs 48
2.2 Active Directory replication 50
2.2.1 Replication basics 51
2.2.2 When Active Directory replication happens 53
2.2.3 Active Directory naming contexts 55
2.2.4 Transforming Domain controllers into
Global Catalogs 58
2.2.5 USNs and replication 60
2.2.6 Urgent replication 64
2.2.7 Intrasite and Intersite replication 65
2.2.8 High-watermark vector and up-to-date vector tables 68
2.2.9 Changes in Active Directory replication in Windows 2003 70
2.3 Exchange’s Active Directory Topology service 71
2.3.1 DSAccess (or ADAccess) 72
2.3.2 How many Global Catalog servers do I need? 75
2.3.3 Where are my Global Catalogs? 76

3.5 Moving users 133
3.5.1 Moving mailboxes 134
3.5.2 Logging mailbox moves 138
3.6 Using distribution groups 140
3.6.1 Forming groups 142
3.6.2 Group changes in Exchange 2007 145
3.6.3 Expanding distribution lists 147
3.6.4 How many objects can I have in a group? 148
3.6.5 Managing group membership 149
3.6.6 Protected groups (and users) 152
3.7 Using groups for permissions 154
3.7.1 Managing distribution groups from Outlook 154
3.8 Dynamic distribution groups 156
3.8.1 Changing filters and conditions for dynamic
distribution groups 157
3.8.2 A note on OPATH 159
3.8.3 A new UI for dynamic groups 160
3.8.4 Creating New dynamic groups 162
3.8.5 Using dynamic Distribution groups 167
3.9 Mailbox quotas 168
3.9.1 Setting mailbox quotas 170
3.10 Email address policies 173
3.10.1 Mailbox moves and email address policies 178
3.10.2 Queries that drive email address policies 178
3.11 Address lists 183
3.11.1 Upgrading Address Lists to Exchange 2007 format 187
3.12 User naming conventions 188
3.13 Server naming conventions 192
3.14 Moving from the basics 194
4 The Exchange Management Shell 195

4.8 Reporting mailbox data 275
4.8.1 Special properties 282
4.9 Using the shell for other management tasks 284
4.10 Command validation 287
4.11 Working with remote servers 290
4.12 Working with non-Exchange 2007 servers 291
4.13 Testing Exchange 2007 292
4.13.1 Client connections 294
4.13.2 Mail Flow 295
4.13.3 Miscellaneous test commands 297
4.14 PowerShell for Exchange administrators 297
5 The Store 301
5.1 Introducing the Store 301
5.2 Differences in the Exchange 2007 Store 306
Contents ix
Contents
5.2.1 Are 64 bits that important? 307
5.2.2 Trading memory for I/O 312
5.2.3 The decrease in storage costs 317
5.3 No more streaming database 318
5.4 Tables and items 320
5.5 Storage groups 323
5.5.1 Creating a new storage group and database 327
5.5.2 Working with storage groups and databases 329
5.6 Transaction logs 331
5.6.1 Circular logging 335
5.6.2 Creating new transaction logs 337
5.6.3 Reserved logs 338
5.6.4 Transactions, buffers, and commitment 339
5.6.5 Transaction log I/O 341

5.15.4 Backups and storage groups 415
5.15.5 Checkpoint file 421
5.15.6 The future of streaming backups 426
5.16 Moving from the Store 427
6 Exchange Transport and Routing 429
6.1 The evolution of routing 429
6.2 Change through experience 430
6.2.1 Hidden administrative and routing groups 433
6.3 Exchange 2007 transport architecture 435
6.3.1 The critical role of hub transport servers 438
6.3.2 Receive connectors 440
6.3.3 Send connectors 447
6.3.4 Linking Exchange 2003 and Exchange 2007 453
6.3.5 Multiple routes into Exchange 2003 458
6.3.6 Decommissioning Exchange 2003 routing groups 458
6.3.7 Handling Exchange 2003 link state updates
during migration 458
6.3.8 Foreign connectors 459
6.3.9 Authorization 460
6.3.10 Accepted domains 460
6.3.11 Transport storage 461
6.4 Routing ABC 464
6.4.1 Resolving multiple paths 467
6.4.2 Most specific connector 467
6.4.3 Connector cost 469
6.4.4 Closest proximity 469
6.4.5 The role of hub routing sites 470
6.4.6 Site link costs versus routing costs 471
6.4.7 Instructing mailbox servers 472
6.4.8 Bypassing some connections 472

6.11.11 Connection filtering 536
6.11.12 Sender filtering 538
6.11.13 Address Rewrite agent 539
6.11.14 Sender ID agent 541
6.11.15 Content filtering 547
6.11.16 Content Filter updates 550
6.11.17 Per-user SCL processing 553
6.11.18 Safelist Aggregation 554
6.11.19 Sender reputation 557
6.11.20 Recipient filtering 559
6.11.21 Blocking file attachments 560
6.11.22 Attachment filtering 562
6.11.23 Edge transport rules 563
6.11.24 Available Edge 565
6.12 Client-side spam suppression 567
6.12.1 Outlook’s Junk Mail Filter 568
6.12.2 Postmarks 573
6.12.3 Restricting OOF and other notifications 574
6.13 Routing onwards 580
xii Contents
7 Clients 581
7.1 Outlook 583
7.1.1 Outlook web services 585
7.1.2 Understanding Outlook’s relationship with Exchange 591
7.1.3 Deploying cached Exchange mode 596
7.1.4 Address caching 599
7.1.5 MAPI compression and buffers 600
7.1.6 Conflict resolution 602
7.1.7 Preventing MAPI clients from connecting 603
7.1.8 Outlook 2007 and Exchange 5.5 607

7.7.11 Client settings 684
Contents xiii
Contents
7.8 Internet client access protocols 684
7.8.1 IMAP4 685
7.8.2 The Exchange 2007 IMAP server 689
7.9 Mobile clients 694
7.9.1 Selecting mobile devices 696
7.9.2 Server-based ActiveSync 698
7.10 Windows Mobile 6.0 and Exchange 2007 702
7.10.1 ActiveSync policies 706
7.10.2 Managing mobile devices through EMC 711
7.10.3 Moving mailboxes to Exchange 2007 and ActiveSync 713
7.10.4 Estimating network traffic for mobile devices 715
7.10.5 Analyzing ActiveSync logs 717
7.10.6 Wiping mobile devices 719
7.10.7 Debugging synchronization 721
7.11 Comparing Windows Mobile and BlackBerry 723
7.11.1 Processing the mail 725
7.11.2 Other messaging options for Windows Mobile 730
7.11.3 Power management 731
7.11.4 Input flexibility 732
7.12 Unified Communications 735
7.13 Unified Messaging 737
7.13.1 Client Access to voicemail 741
7.13.2 Dealing with voicemail 745
7.13.3 Voice synthesis 747
7.13.4 Pure voicemail 748
7.13.5 The magic of SIP 749
7.13.6 Speech Grammars 752

8.5.5 The Managed Folder Assistant 829
8.5.6 Logging Managed Folder activity 831
8.5.7 Using Managed Folders 833
8.5.8 Harvesting information from managed folders 835
8.6 Message classifications 837
8.6.1 Adding intelligence to classification through rules 844
8.7 Copying user mailboxes 848
8.7.1 Auditing 853
8.8 Free and busy 853
8.8.1 Looking at free and busy data 855
8.8.2 Free and busy in Exchange 2007 861
8.8.3 Changes in Outlook 2007 863
8.8.4 Cross-forest free and busy 866
9 Hardware and Performance 867
9.1 Moving toward 64-bit Exchange 867
9.2 Buying servers for Exchange 2007 870
9.3 The storage question 876
9.4 RPC pop-ups 881
9.5 Clusters and Exchange 882
9.6 Continuous replication and Exchange 2007 888
9.6.1 Concepts 889
9.7 Deploying Local Continuous Replication (LCR) 892
9.7.1 How LCR works 897
9.7.2 LCR operations 900
Contents xv
Contents
9.7.3 LCR restrictions 903
9.7.4 LCR database transition 904
9.8 Deploying Cluster Continuous Replication (CCR) 906
9.8.1 Comparing CCR and traditional clusters 910

10.5.5 PFDavAdmin 971
10.5.6 LogParser 973
10.5.7 Outlook Spy 978
10.6 Bits and pieces 978
10.6.1 Where the Exchange team hangs out 978
10.6.2 Online Forums 979
xvi Contents
10.7 Conferences 979
10.7.1 Magazines 980
10.7.2 How Exchange uses registry keys 980
10.8 Good reference books 981
A Appendix 983
A.1 Message Tracking Log Format 983
A.2 Events noted in Message Tracking Logs 985
B Important Exchange PowerShell commands 987
B.1 Recipient management commands 987
B.2 Exchange server administrative Commands 990
B.3 Databases and Storage Groups 993
B.4 Address Lists and Email Policies 995
B.5 Queues and Messages 995
B.6 Edge Synchronization 996
B.7 Routing 997
B.8 ActiveSync 998
B.9 Public folders 999
B.10 Transport and journal rules 1000
B.11 IMAP and POP 1001
B.12 Active Directory commands 1002
B.13 Testing Exchange 2007 1003
B.14 Basic PowerShell 1004
B.15 PowerShell control commands 1005

from a predominantly GUI-centric approach to server management to the
use of the PowerShell scripting language as the basis of many management
operations. The need to understand and appreciate the change has to occur
xviii Preface
before you can adequately document and describe the benefits and this
increases the effort required to write the book. I must admit that it took me
time to realize the full benefit of interacting with Exchange through the shell,
but now I am at the point where I wonder why Microsoft never provided
such a powerful interface in the past!
The degree of change that exists in Exchange 2007 means that it is diffi-
cult to cover everything in one book. I have therefore elected to cover the
parts of Exchange that I think are of most interest to the majority of admin-
istrators and have left other components for you to discover through the
material that Microsoft publishes or perhaps another book, written by me or
someone else. Please accept my apology if I have not covered something that
you think is important and treat this as a challenge and opportunity for you
to write about the topic yourself. There are many magazines, blogs, and other
ways of spreading information about Exchange.
From time to time, I wander back down the path to consider some
aspect of Exchange 2003. While this book is firmly focused on Exchange
2007, the vast majority of companies that will deploy Exchange 2007 will do
so by migrating from Exchange 2003 and will therefore run both products
alongside each other for some period. For large organizations, the period
might extend to a year or more as it is unlikely that few will complete their
migration to a pure Exchange 2007 environment quickly. With this in mind,
it is fair and reasonable to document how things work with Exchange 2003,
especially when these servers operate with Exchange 2007.
So what is in the book? To set the context, Chapter 1 starts with an over-
view of the development of Exchange from 4.0 to 2007 and then describes the
themes that Microsoft employed to focus the development priorities for

One of my reviewers noticed this point, and complained that I did not think
that POP3 was an important protocol. Using Exchange 2007 as a hosting
platform is a pretty specialized business and I apologize in advance if I offend
anyone by my concentration on how to deploy Exchange 2007 most effec-
tively for medium to large enterprises.
All errors and omissions are mine, especially in the code samples selected
to illustrate the power of the Exchange Management Shell. PowerShell sam-
ples are indicated in the courier typeface like so:
Get-Mailbox –id Redmond | Select DisplayName
Any output from the commands is shown as follows:
DisplayName: Tony Redmond
While all the code worked on one or more test systems, experience tells
me that errors can creep in the process required to take code from a system
through editing and publishing to the final content in a book. This is espe-
cially so when the underlying code changes from build to build as the engi-
neers push to finish the product and generate a knock-on effect of changes to
commands and individual parameters. This book does not pretend to be a
comprehensive guide to PowerShell programming or to the Exchange Man-
agement Shell and the examples are there to give you a taste of what you can
now do to automate management operations, so any errors that do creep in
should be pretty obvious and easily solved—I hope!
Books do not happen overnight and they represent a lot of work. I have
gained enormously from being able to work alongside some tremendous
experts in enterprise messaging, both inside and outside HP. I acknowledge
the contribution of groups such as my own team, who humored me when I
was writing. The Exchange 2007 academy tutors allowed me to ask many
xx Preface
questions as I probed the content that they generated to train HP consultants
and customers. I must also acknowledge the huge contribution made by the
enterprise messaging team at HP including Kathy Pollert, Mike Ireland, and

I hope you will find with your deployment of Exchange 2007 that
we’ve delighted all three. Since 2005, we’ve been testing Exchange 2007
with more organizations and more end users than any previous release of
Exchange. The end result is a product that we are very proud of here in
Redmond, Washington. We look forward to receiving your feedback about
Exchange 2007 over the coming years.
On behalf of the entire Exchange team, thank you for choosing
Microsoft Exchange!
Terry Myerson ()
General Manager, Exchange Server
Microsoft Corporation
This page intentionally left blank
1
1
Introduction
1.1 A decade and counting of Exchange deployments
Microsoft shipped Exchange 4.0 in March 1996 after a gestation period of
some four years. The new messaging server went through many different
design phases. Microsoft grappled with the challenge of enterprises and small
companies, figured out what they had to do to be competitive, understood
how best to migrate users from other platforms (including their own), and
achieved the necessary performance and scalability levels—albeit limited by
the capabilities of Windows NT 3.51 and the available hardware.
Exchange replaced Microsoft Mail and went into immediate competi-
tion with other messaging systems such as those favored by large corpora-
tions (IBM PROFS, Digital Equipment Corporation’s ALL-IN-1 and
MailWorks, and HP OpenMail) and the PC LAN-based systems such as
Lotus cc:Mail, Banyan Vines, Novell GroupWise, and Lotus Notes.
Exchange 4.0 was the first version that implemented the initial Exchange
architecture and this generation subsequently spanned Exchange 5.0 and 5.5,

Government and telecommunications bodies led the charge toward a
common standard for directories that eventually resulted in the X.500 stan-
dard. While X.500 offered the potential that it could eventually result in a
global directory standard that everyone used to connect directories to, direc-
tory synchronization was another black art in 1996. It was common to have
weekly or monthly synchronization runs to merge directory data to provide a
common view of users across multiple systems. Email addresses were more
convoluted (mine was then ) than today
as most organizations now use the standard SMTP convention of first-
name.last-name@domain. Of course, X.500 has long since faded into the
background and LDAP is now the most widely used standard for directory
access and interoperability. We can still see the influence of X.500 in some
enterprise directories and in the design principles that Microsoft followed to
build the original Exchange Directory Store and then the Active Directory,
but few Exchange administrators bother about X.500 now.
The ease of connectivity established by SMTP, its extensions (ESMTP),
and the easy access that we now enjoy to the Internet has revolutionized
email. This is true for corporate users and personal users. Ten years ago it
would have been difficult to predict the success and ease of access that peo-
ple around the world enjoy to email systems such as Hotmail, Gmail, and
Yahoo mail.
1.1.2 The protocol wars
MAPI is the great survivor of the protocol wars. MAPI is actually an API, but
many people refer to MAPI as a protocol, in the same way as they refer to