Architecting for Scale: High Availability for Your Growing Applications

by Lee Atchison

Every day, companies struggle to scale critical applications. As traffic volume and data demands increase, these applications become more complicated and brittle, exposing risks and compromising availability. This practical guide shows IT, devops, and system reliability managers how to prevent an application from becoming slow, inconsistent, or downright unavailable as it grows.

Scaling isn't just about handling more users; it's also about managing risk and ensuring availability. Author Lee Atchison provides basic techniques for building applications that can handle huge quantities of traffic, data, and demand without affecting the quality your customers expect.

In five parts, this book explores:



Availability: learn techniques for building highly available applications, and for tracking and improving availability going forward

Risk management: identify, mitigate, and manage risks in your application, test your recovery/disaster plans, and build out systems that contain fewer risks

Services and microservices: understand the value of services for building complicated applications that need to operate at higher scale

Scaling applications: assign services to specific teams, label the criticalness of each service, and devise failure scenarios and recovery plans

Cloud services: understand the structure of cloud-based services, resource allocation, and service distribution

Genres: Technology, Architecture, Software, Programming, Engineering, Computer Science, Technical

228 pages, Paperback

First published August 6, 2016

Reviews

[Emre Sevinç · Rating 4 out of 5]

The book provides a very high-level overview of the fundamental aspects of building highly-available service-based information systems, for cloud-based or on-premises systems. The overview is pretty good, but don't expect a lot of details. The author's experience based on hard lessons learned from his years working for AWS and New Relic definitely shows, and some chapters get much more attention than the others.

I've particularly liked chapters on risk management and mitigation, especially the discussion about creating and managing a risk matrix, as well as the chapter dedicated to clarifying concepts of likelihood and severity. You don't see this topic taken this seriously in many texts discussing the building of highly available and scalable systems, but without proper risk management mindset, you are almost always one step away from a big mess waiting to happen, without you having planned for it.

The sections on microservices are rather light, and you'll definitely need another book on this important architectural pattern. The chapters and discussion on establishing and clarifying service ownership, service tiers, and service-level agreements again provide a good overview, and draws attention to the importance of percentiles when it comes to establishing baselines for performance and checking if things are within expected bounds. To some experienced readers these will sound pretty basic, but nevertheless I think these discussions have educational value for many readers.

The last part of the book is looking at things from cloud computing, and even though the chapters are short, the author's guidance provide valuable experience summary for different types of use cases, as well as drawing attention to pitfalls in this complex world of computing.

Conclusion: If you want to a good overview of how current scalable, highly available, microservices-based software intensive applications are built, operated, risk-managed using cloud computing, this book accomplishes this objective. But if you want more details, you'll have to dive into other books.

[Gabriel Santos · Rating 3 out of 5]

It is a good book. In my opinion, it might benefit inexperienced architects not used to deal with large scale applications. Don’t get me wrong, the concepts described by the author are rather useful. However these same concepts are already widespread among other books.

The book talks about stosa, which is a rather simple form of organization, similar to other topologies presented in other works, lacking mentioning Conway’s Law properly and relating everything with domains.

Finally, he talks about the cloud and its adoption. Once again, the concepts presented are pretty simple and shallow. Hence, it is a good start for newcomers architects to get used with some really important concepts when it comes to architecting for scalability

[Ivan Chernov · Rating 4 out of 5]

Книга понравилась своей лаконичностью. Написана она по принципу: если у вас ситуация А, то делайте Б, потому что В. Это круто, потому что много технический писателей грешат описать как можно более подробно все краевые случаи. Слог простой из-за чего на прочтение всей книги можно уделить всего пару вечеров.
Минусом данной книги назвал секцию про облака, потому что описаны конкретные реализации вещей на примере AWS, но не думаю что они дают такой же фундамент, как все остальные главы.

Итого: рекомендую прочесть :)

[Peter Sichevsky · Rating 4 out of 5]

Definitely worth a read. You will not find any new ideas in this book.

[Vinayak Hegde · Rating 4 out of 5]

A good overview of how to scale system. Talks more about the (thought) process rather than architectural patterns (and specific software) but that is fine. I liked the risk mitigation matrix. I had used something similar in a couple of earlier jobs. The overall section on risk assessment and mitigation was quite good. Much hard fought advice here. The idea on game days to do testing on production was also good. I know lots of folks dither when testing in production but sometimes there is a choice between doing this and failure at prime time. It broadens the concept of chaos monkey which Netflix pioneered.

The sections on microservices especially advice on how to think, layer and build microservices (without overdoing it) is quite good. Goes into depth abpout the effect on team structure as well as SLAs and contracts between services. I wish the author had gone on more deeper into service meshes. An good overview of scaling strategies and cloud follows in the later sections. Overall a good high-level overview of building to scale.

[Natallia · Rating 5 out of 5]

I just love this book but it’s definitely not for everyone. It’s an overview how to deal with heavy systems and common pitfalls overcome approaches. The narrative style is strict and mostly contains only necessary information organized in a logical way. I would guess that the author is an engineer from the style of writing. I would recommend this book to someone who is interested in general overview, novice team leaders or experts interested in solution architecture as the first step.

[Leonid Panich · Rating 3 out of 5]

It is good as a starting point about architecture, but is not deep enough and repetitive. Some statements are provided without argumentation.

[Gleb Radchenko · Rating 3 out of 5]

This book provides a brief high-level review of modern approaches and principles of web-applications design. The author describes the basics of such terms as availability, scalability, risk assessment and basic principles of microservices. Unfortunately, the text of the book is too verbose. The descriptions are prolix and often very repetitive. It looks like the main content of the book could be explained in a much more compact way.

[Dmitry · Rating 2 out of 5]

Что понравилось: введение в продукты облака Amazon(никакой связи с тем фактом, что автор проработал там 7 лет), упоминания PCI, SOX, HIPAA.
Что не понравилось: книга слишком уж обзорная, я бы рекомендовал читать такое на втором курсе университета, и уж точно не рассчитывать на пользу для уже сформировавшегося специалиста. Список паттернов распределенной архитектуры на msdn полезнее.

[Mahdi Taghizadeh · Rating 4 out of 5]

Expect a lot about AWS and cloud best practices but you will have a good big picture of what a large scale application looks like and what elements are involved. If you don’t have time to read the whole book, “Chapter 26. Putting It All Together” is a good summary of everything discussed.

[Venkatesh-Prasad · Rating 5 out of 5]

The book is about architecture for scale with specific focus on availability. It is not about scalability design patterns.

That said, I liked the book was brisk, short, and precise. While the book dives deep into availability (definitions and different metrics), it also touches on topics that other books that I have read seldom cover, e.g., risk management, SLA. Dedicating a part for each of these topics, the book provides a reasonably deep treatment of the topics while nicely connecting them to each other. This connection explicates why each of the topics should be considered to architect systems that can scale successfully.

While I would not recommend the book for folks interested learning about scalability patterns, I would highly recommend the book for folks interested in learning about considering scale beyond design and implementation, i.e., in architecture, during planning, for maintenance.

[Liam Hazell · Rating 2 out of 5]

Well written, but I definitely was not the intended audience of the book. I think there’s a small niche of people who this is intended for and its upper management who want to broadly understand the issues going on with scaling an application. I was much more hoping for a deeper technical dive on how to actually scale systems.

If you fit the mentioned audience, then I think this book would be fantastic for you.

[Anup · Rating 4 out of 5]

A quick and light reading book which deals with how to make web applications scalable. Author has worked at Amazon and thus draws on knowledge from company. It was informative in making decisions on what are the considerations on choosing a cloud solution and cost impacts. Also learnt some new concepts such as Game Day, AWS Lambda.

[Mansur Mustaquim · Rating 4 out of 5]

This book provides a good quick overview of the elements that go into making a modern scalable application. Good for a breadth-wise introduction for technical managers and architects. Deals with not just technical aspects but also organizational aspects that would be required in maintaining high-availability software.

[Priyank · Rating 2 out of 5]

The book gives a very high-level view of how applications are architected and run at scale. It also covers the organizational aspect of running applications at scale.
Though I felt this to be more of a light read for someone already familiar with these topics rather than being an informative book for someone starting with these concepts.

[Christoph Kappel · Rating 5 out of 5]

This was a quick read and a good round-up about the usual topics related to scaling. I especially like the fact that organisational scaling is also part of it and about many of the risks related to cloud.

Bonus: I think I finally understand what Edge computing is all about and I found the part about the different adoption rate of cloud-based products based on geography quite interesting.

[Rodrigo Ramos · Rating 5 out of 5]

A greag book for think in HA applications and all it concerns. I really figure out who and how acomplish with supports responsabilities. Its important hightlight that the autor does not mention any particular technology.

[Jacek · Rating 4 out of 5]

A lot of helpful ideas that can give you a theoretical vision of scalable, distributed systems.
However if you already had a knowledge about 12 factor idea, docker, orchestration or something similar this book will not give you nothing new

[Nikolay · Rating 2 out of 5]

Такое чувство, что книга написана ради книги. Опыт автора чувствуется, мысли изложены полезные и книга безусловно заставляет задуматься, но мне кажется, что материал можно было бы ужать до 2-3 статей на хабр без ущерба для понимания.

[Shenyu · Rating 5 out of 5]

I enjoyed reading this book due to the proper usage of real-world examples, which made a lot of sense to me. Though the last part (Cloud Services) seems to be a little outdated, the first four parts are all very practical and informative.

[Paul · Rating 3 out of 5]

Где-то в отзывах увидел фразу, что книга "обо всем и ни о чем": полностью согласен. В ней описаны базовые принципы построения масштабируемых приложений, при этом, почти все это описано довольно абстрактно и без конкретики. Перевод местами ужасен и читается очень тяжело.

[Nilendu Misra · Rating 3 out of 5]

This review is for the second edition of the book. Sufficiently high level walkthrough of key concepts with many new ideas introduced -
- failure loops
- flying 2-mistakes high
- keep a “5th computer” like space shuttle

[Francois D’Agostini · Rating 4 out of 5]

Good overview books with ideas already mentioned a lot in other books.

However, the content itself is easy to read and makes sense.
I liked the content around Edge computing, STOSA and risk management

Quick read, and mostly a refresher, but not unpleasant

[Alex · Rating 3 out of 5]

Very generic overview of scalability.

[Deiwin Sarjas · Rating 2 out of 5]

Nothing new or original. Advice is mostly too high-level to be actionable. And then at other times it spent pages explaining trivial calculations.

[Paul · Rating 3 out of 5]

Very brief introduction to highly reliable services design. May be helpful for junior system administrators, but you should not expect any revelations there.

[Mehul · Rating 4 out of 5]

Good quick read

Good quick read
High level and useful tips and suggestions.
Useful for DevOps.
You can use it to build frameworks for managing you applications

[Geoffrey Kerr]

Good overview

Does not go very deep, but provides a reasonable overview of scale. Good for managers or for devs looking for a quick refresher.

[Guillaume Humbert · Rating 3 out of 5]

Good book that provides condensed information on the topic, however I would have loved that the author went in much more details.

[Łukasz Słonina · Rating 4 out of 5]

General concepts of applications scaling. High level only, but definitely worth reading. Good chapters about risk.