Book Image

High-Performance Programming in C# and .NET

By : Jason Alls
Book Image

High-Performance Programming in C# and .NET

By: Jason Alls

Overview of this book

Writing high-performance code while building an application is crucial, and over the years, Microsoft has focused on delivering various performance-related improvements within the .NET ecosystem. This book will help you understand the aspects involved in designing responsive, resilient, and high-performance applications with the new version of C# and .NET. You will start by understanding the foundation of high-performance code and the latest performance-related improvements in C# 10.0 and .NET 6. Next, you’ll learn how to use tracing and diagnostics to track down performance issues and the cause of memory leaks. The chapters that follow then show you how to enhance the performance of your networked applications and various ways to improve directory tasks, file tasks, and more. Later, you’ll go on to improve data querying performance and write responsive user interfaces. You’ll also discover how you can use cloud providers such as Microsoft Azure to build scalable distributed solutions. Finally, you’ll explore various ways to process code synchronously, asynchronously, and in parallel to reduce the time it takes to process a series of tasks. By the end of this C# programming book, you’ll have the confidence you need to build highly resilient, high-performance applications that meet your customer's demands.
Table of Contents (22 chapters)
1
Part 1: High-Performance Code Foundation
7
Part 2: Writing High-Performance Code
16
Part 3: Threading and Concurrency

Summary

In this chapter, we began with a high-level overview of the task-based asynchronous pattern. Things we covered were naming, parameters, return types, initializing asynchronous operations, exceptions, and optionally providing ways to report progress updates and cancel operations. We saw that we can have asynchronous operations that allow cancellation, and those that don't allow cancellation. Plus, we learned that when a cancellation has been requested, the cancellation will either go ahead or be ignored. Completed tasks can have a completed state of Canceled, RanToCompletion, or Faulted.

We then benchmarked three different ways of calling a method synchronously, using Task.Run, and asynchronously. Using Task.Run took the longest time, followed by running the method synchronously, and running the method asynchronously was by far the quickest way to run the method.

Then we benchmarked GetAwaiter.GetResult(), Result, and Wait for both Task and TaskValue. We saw that...