Scala High Performance Programming

269 Pages·2016·5.179 MB·English

by  Vincent Theron, Michael Diamant

Scala High Performance
Programming

Leverage Scala and the functional paradigm to build
performant software

Vincent Theron
Michael Diamant

BIRMINGHAM - MUMBAI
Scala High Performance Programming

Copyright © 2016 Packt Publishing

All rights reserved. No part of this book may be reproduced, stored in a retrieval system, or
transmitted in any form or by any means, without the prior written permission of the
publisher, except in the case of brief quotations embedded in critical articles or reviews.
Every effort has been made in the preparation of this book to ensure the accuracy of the
information presented. However, the information contained in this book is sold without
warranty, either express or implied. Neither the authors, nor Packt Publishing, and its
dealers and distributors will be held liable for any damages caused or alleged to be caused
directly or indirectly by this book.
Packt Publishing has endeavored to provide trademark information about all of the
companies and products mentioned in this book by the appropriate use of capitals.
However, Packt Publishing cannot guarantee the accuracy of this information.

First published: May 2016

Production reference: 1250516

Published by Packt Publishing Ltd.
Livery Place
35 Livery Street
Birmingham B3 2PB, UK.
ISBN 978-1-78646-604-4
www.packtpub.com
Credits
Authors Copy Editor

Vincent Theron Priyanka Ravi

Michael Diamant
Reviewer Project Coordinator

Nermin Šerifović Francina Pinto
Commissioning Editor Proofreader

Edward Gordon Safis Editing
Acquisition Editor Indexer

Chaitanya Nair Rekha Nair
Content Development Editor Production Coordinator

Nikhil Borkar Manu Joseph
Technical Editor Cover Work

Madhunikita Sunil Chindarkar Manu Joseph
About the Authors
Vincent Theron is a professional software engineer with 9 years of experience. He
discovered Scala 6 years ago and uses it to build highly scalable and reliable applications.
He designs software to solve business problems in various industries, including online
gambling, financial trading, and, most recently, advertising. He earned a master's degree in
computer science and engineering from Université Paris-Est Marne-la-Vallée. Vincent lives
in the Boston area with his wife, his son, and two furry cats.
To everybody at Packt Publishing, thanks for working so hard to make this book a reality. To
Chaitanya Nair, thanks for reaching out to me with this project. To Nikhil Borkar, thanks for
providing us with guidance along the way. To Michael Diamant, my coauthor, coworker, and friend,
thanks for the knowledge you have brought to this book and for being an inspiration every day. To my
parents, thanks for your love and support and for buying me my first computer. And finally, to my
wife, Julie, thanks for your constant encouragement and for giving me such a wonderful son.

Michael Diamant is a professional software engineer and functional programming
enthusiast. He began his career in 2009 focused on Java and the object-oriented
programming paradigm. After learning about Scala in 2011, he has focused on using Scala
and the functional programming paradigm to build software systems in the financial
trading and advertising domains. Michael is a graduate of Worcester Polytechnic Institute
and lives in the Boston area.
The knowledge I am able to share in this book is the result of a lifetime of support and teaching from
others. I want to recognize my coauthor, Vincent, for pushing me to take on this effort and for all the
hours spent together developing the thoughts contained in our book. All of my current and former
colleagues have helped me sharpen my engineering skills, and without their generosity of sharing
their learning, I would not have been able to write this book. In addition to Vincent, I want to call out
several colleagues that I feel particularly indebted to: Dave Stevens, Gary Malouf, Eugene Kolnick,
and Johnny Everson. Thank you to my parents and my brother for supporting me and shaping me
into the individual I am today. I am deeply appreciative of the support my girlfriend, Anna, gave me
throughout the writing process. And last but not least, thank you to Packt Publishing for helping us
write our first book.
About the Reviewer
Nermin Šerifović has been a Scala enthusiast since 2009, practicing it professionally since
2011. For most of his career, he has focused on building backend platforms using JVM
technologies. Most recently, as VP Engineering at Pingup, he has been leading the
development efforts of a local services booking system.
Nermin is an instructor at Harvard Extension School, where he co-teaches the Concurrent
Programming in Scala course and has also given talks at various conferences.
An active Scala community member, Nermin organized the Boston Area Scala Enthusiasts
user group and was part of the Northeast Scala Symposium founding team. He is a co-
author of the Scala Puzzlers book and co-creator of the Scala Puzzlers website.
Nermin holds an M.Eng in computer science from Cornell University, and his areas of
interest include distributed systems, along with concurrent, reactive, and functional
programming.
www.PacktPub.com
For support files and downloads related to your book, please visit www.PacktPub.com.
eBooks, discount offers, and more
Did you know that Packt offers eBook versions of every book published, with PDF and
ePub files available? You can upgrade to the eBook version at www.PacktPub.com and as a
print book customer, you are entitled to a discount on the eBook copy. Get in touch with us
at [email protected] for more details.
At www.PacktPub.com, you can also read a collection of free technical articles, sign up for a
range of free newsletters and receive exclusive discounts and offers on Packt books and
eBooks.
h t t p s : / / w w w 2 . p a c k t p u b . c o m / b o o k s / s u b s c r i p t i o n / p a c k t l i b
Do you need instant solutions to your IT questions? PacktLib is Packt's online digital book
library. Here, you can search, access, and read Packt's entire library of books.
Why subscribe?
Fully searchable across every book published by Packt
Copy and paste, print, and bookmark content
On demand and accessible via a web browser
Free access for Packt account holders
Get notified! Find out when new books are published by following @PacktEnterprise on
Twitter or the Packt Enterprise Facebook page.
Table of Contents
Preface 1
Chapter 1: The Road to Performance 6
Defining performance 7
Performant software 7
Hardware resources 8
Latency and throughput 9
Bottlenecks 10
Summarizing performance 11
The problem with averages 11
Percentiles to the rescue 15
Collecting measurements 15
Using benchmarks to measure performance 16
Profiling to locate bottlenecks 17
Pairing benchmarks and profiling 17
A case study 18
Tooling 18
Summary 19
Chapter 2: Measuring Performance on the JVM 20
A peek into the financial domain 20
Unexpected volatility crushes profits 23
Reproducing the problem 24
Throughput benchmark 25
Latency benchmark 27
The first latency benchmark 27
The coordinated omission problem 29
The second latency benchmark 30
The final latency benchmark 30
Locating bottlenecks 32
Did I test with the expected set of resources? 34
Was the system environment clean during the profiling? 36
Are the JVM's internal resources performing to expectations? 38
Where are the CPU bottlenecks? 41
What are the memory allocation patterns? 47
Trying to save the day 51
A word of caution 53
A profiling checklist 53
Taking big steps with microbenchmarks 53
Microbenchmarking the order book 55
Summary 61
Chapter 3: Unleashing Scala Performance 62
Value classes 63
Bytecode representation 63
Performance considerations 65
Tagged types – an alternative to value classes 66
Specialization 68
Bytecode representation 69
Performance considerations 71
Tuples 76
Bytecode representation 76
Performance considerations 76
Pattern matching 79
Bytecode representation 79
Performance considerations 80
Tail recursion 86
Bytecode representation 88
Performance considerations 89
The Option data type 92
Bytecode representation 93
Performance considerations 94
Case study – a more performant option 95
Summary 100
Chapter 4: Exploring the Collection API 101
High-throughput systems – improving the order book 102
Understanding historical trade-offs – list implementation 102
List 103
TreeMap 106
Adding limit orders 107
Canceling orders 110
The current order book – queue implementation 111
Queue 111
Improved cancellation performance through lazy evaluation 113
Set 114
Benchmarking LazyCancelOrderBook 119
Lessons learned 123
[ ii ]
Historical data analysis 124
Lagged time series returns 124
Vector 128
Data clean up 129
Handling multiple return series 132
Array 133
Looping with the Spire cfor macro 135
Summary 137
Chapter 5: Lazy Collections and Event Sourcing 138
Improving the client report generation speed 138
Diving into the reporting code 139
Using views to speed up report generation time 142
Constructing a custom view 144
Applying views to improve report generation performance 147
View caveats 150
SeqView extends Seq 150
Views are not memoizers 151
Zipping up report generation 154
Rethinking reporting architecture 155
An overview of Stream 158
Transforming events 161
Building the event sourcing pipeline 166
Streaming Markov chains 170
Stream caveats 174
Streams are memoizers 174
Stream can be infinite 176
Summary 177
Chapter 6: Concurrency in Scala 178
Parallelizing backtesting strategies 178
Exploring Future 180
Future and crazy ideas 184
Future usage considerations 186
Performing side-effects 186
Blocking execution 187
Handling failures 189
Hampering performance through executor submissions 192
Handling blocking calls and callbacks 197
ExecutionContext and blocking calls 197
Asynchronous versus nonblocking 198
Using a dedicated ExecutionContext to block calls 198
[ iii ]