Book Image

ASP.NET Core 1.0 High Performance

By : James Singleton, Pawan Awasthi
Book Image

ASP.NET Core 1.0 High Performance

By: James Singleton, Pawan Awasthi

Overview of this book

ASP.NET Core is the new, open source, and cross-platform, web-application framework from Microsoft. It's a stripped down version of ASP.NET that's lightweight and fast. This book will show you how to make your web apps deliver high performance when using it. We'll address many performance improvement techniques from both a general web standpoint and from a C#, ASP.NET Core, and .NET Core perspective. This includes delving into the latest frameworks and demonstrating software design patterns that improve performance. We will highlight common performance pitfalls, which can often occur unnoticed on developer workstations, along with strategies to detect and resolve these issues early. By understanding and addressing challenges upfront, you can avoid nasty surprises when it comes to deployment time. We will introduce performance improvements along with the trade-offs that they entail. We will strike a balance between premature optimization and inefficient code by taking a scientific- and evidence-based approach. We'll remain pragmatic by focusing on the big problems. By reading this book, you'll learn what problems can occur when web applications are deployed at scale and know how to avoid or mitigate these issues. You'll gain experience of how to write high-performance applications without having to learn about issues the hard way. You'll see what's new in ASP.NET Core, why it's been rebuilt from the ground up, and what this means for performance. You will understand how you can now develop on and deploy to Windows, Mac OS X, and Linux using cross-platform tools, such as Visual Studio Code.
Table of Contents (18 chapters)
ASP.NET Core 1.0 High Performance
Credits
Foreword
About the Author
Acknowledgments
About the Reviewer
www.PacktPub.com
Preface
2
Measuring Performance Bottlenecks

Staying alert


Whenever you perform testing, particularly UI or performance testing, you will get noisy results. Reliability is not perfect, and there will always be failures that are not due to bugs in the code. You shouldn't let these false positives cause you to ignore failing tests, and although the easiest course of action may be disabling them, the correct thing to do to make them more reliable.

Tip

The scientifically minded know that there is no such thing as a perfect filter in binary classification, and always look at the precision and recall of a system. Knowing the rate of false positives and negatives is important to get a good idea of the accuracy and tradeoffs involved.

To avoid testing fatigue, it can be helpful to engage developers and instill a responsibility to fix failing tests. You don't want everyone thinking that it's somebody else's problem. It should be pretty easy to see who broke a test by the commit in version control, and it's then their job to fix the broken test...