Reactive Design Patterns

387 Pages·2014·4.68 MB·English

by  Roland Kuhn

Reactive Design Patterns
ROLAND KUHN
WITH BRIAN HANAFEE
AND JAMIE ALLEN
FOREWORD BY JONAS BONÉR
MANNING
SHELTER ISLAND
For online information and ordering of this and other Manning books, please visit
www.manning.com. The publisher offers discounts on this book when ordered in quantity.
For more information, please contact
Special Sales Department
Manning Publications Co.
20 Baldwin Road
PO Box 761
Shelter Island, NY 11964
Email: [email protected]
©2017 by Manning Publications Co. All rights reserved.
No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in
any form or by means electronic, mechanical, photocopying, or otherwise, without prior written
permission of the publisher.
Many of the designations used by manufacturers and sellers to distinguish their products are
claimed as trademarks. Where those designations appear in the book, and Manning
Publications was aware of a trademark claim, the designations have been printed in initial caps
or all caps.
Recognizing the importance of preserving what has been written, it is Manning’s policy to have
the books we publish printed on acid-free paper, and we exert our best efforts to that end.
Recognizing also our responsibility to conserve the resources of our planet, Manning books
are printed on paper that is at least 15 percent recycled and processed without the use of
elemental chlorine.
Manning Publications Co. Development editor: Jennifer Stout
20 Baldwin Road Technical development editor: Brian Hanafee
PO Box 761 Project editors: Tiffany Taylor and Janet Vail
Shelter Island, NY 11964 Line editor: Ben Kovitz
Copyeditor: Tiffany Taylor
Proofreader: Katie Tennant
Technical proofreader: Thomas Lockney
Typesetter: Dottie Marsico
Cover designer: Leslie Haimes
ISBN 9781617291807
Printed in the United States of America
1 2 3 4 5 6 70 8 9 10 – EBM – 22 21 20 19 18 17
To my children
— Roland

brief contents
PART 1 INTRODUCTION................................................................1
1 ■ Why Reactive? 3
2 ■ A walk-through of the Reactive Manifesto 12
3 ■ Tools of the trade 39
PART 2 THE PHILOSOPHY IN A NUTSHELL ...................................65
4 ■ Message passing 67
5 ■ Location transparency 81
6 ■ Divide and conquer 91
7 ■ Principled failure handling 100
8 ■ Delimited consistency 105
9 ■ Nondeterminism by need 113
10 ■ Message flow 120
PART 3 PATTERNS....................................................................125
11 ■ Testing reactive applications 127
12 ■ Fault tolerance and recovery patterns 162
13 ■ Replication patterns 184
14 ■ Resource-management patterns 220
15 ■ Message flow patterns 255
16 ■ Flow control patterns 294
17 ■ State management and persistence patterns 311
v

contents
foreword xv
preface xvii
acknowledgments xix
about this book xxi
about the authors xxiv
PART 1 INTRODUCTION.................................................1
1 Why Reactive? 3
1.1 The anatomy of a Reactive application 4
1.2 Coping with load 5
1.3 Coping with failures 7
1.4 Making the system responsive 8
1.5 Avoiding the ball of mud 10
1.6 Integrating nonreactive components 10
1.7 Summary 11
2 A walk-through of the Reactive Manifesto 12
2.1 Reacting to users 12
Understanding the traditional approach 13 ■ Analyzing
latency with a shared resource 15 ■ Limiting maximum
latency with a queue 16
vii
viii CONTENTS
2.2 Exploiting parallelism 18
Reducing latency via parallelization 18 ■ Improving parallelism
with composable Futures 20 ■ Paying for the serial illusion 21
2.3 The limits of parallel execution 23
Amdahl’s Law 23 ■ Universal Scalability Law 24
2.4 Reacting to failure 26
Compartmentalization and bulkheading 28 ■ Using circuit
breakers 29 ■ Supervision 30
2.5 Losing strong consistency 31
ACID 2.0 33 ■ Accepting updates 34
2.6 The need for Reactive design patterns 35
Managing complexity 36 ■ Bringing programming models
closer to the real world 37
2.7 Summary 38
3 Tools of the trade 39
3.1 Early Reactive solutions 39
3.2 Functionalprogramming 41
Immutability 41 ■ Referential transparency 44 ■ Side
effects 45 ■ Functions as first-class citizens 46
3.3 Responsiveness to users 47
Prioritizing the three performance characteristics 47
3.4 Existing support for Reactive design 48
Green threads 48 ■ Event loops 49 ■ Communicating
Sequential Processes 50 ■ Futures and promises 52
Reactive Extensions 57 ■ The Actor model 58
Summary 62
PART 2 THE PHILOSOPHY IN A NUTSHELL ....................65
4 Message passing 67
4.1 Messages 67
4.2 Vertical scalability 68
4.3 Event-based vs. message-based 69
4.4 Synchronous vs. asynchronous 71
4.5 Flow control 73
CONTENTS ix
4.6 Delivery guarantees 76
4.7 Events as messages 78
4.8 Synchronous message passing 80
4.9 Summary 80
5 Location transparency 81
5.1 What is location transparency? 81
5.2 The fallacy of transparent remoting 82
5.3 Explicit message passing to the rescue 83
5.4 Optimization of local message passing 84
5.5 Messageloss 85
5.6 Horizontal scalability 87
5.7 Location transparency makes testing simpler 87
5.8 Dynamic composition 88
5.9 Summary 90
6 Divide and conquer 91
6.1 Hierarchical problem decomposition 92
Defining the hierarchy 92
6.2 Dependencies vs. descendant modules 94
Avoiding the matrix 95
6.3 Building your own big corporation 96
6.4 Advantages of specification and testing 97
6.5 Horizontal and vertical scalability 98
6.6 Summary 99
7 Principled failure handling 100
7.1 Ownership means commitment 100
7.2 Ownership implies lifecycle control 102
7.3 Resilience on all levels 104
7.4 Summary 104
8 Delimited consistency 105
8.1 Encapsulated modules to the rescue 106
8.2 Grouping data and behavior
according to transaction boundaries 107