Mastering Concurrency in Python

Description

Immerse yourself in the world of Python concurrency and tackle the most complex concurrent programming problems Key Features Explore the core syntaxes, language features and modern patterns of concurrency in Python Understand how to use concurrency to keep data consistent and applications responsive Utilize application scaffolding to design highly-scalable programs Book Description Python is one of the most popular programming languages, with numerous libraries and frameworks that facilitate high-performance computing. Concurrency and parallelism in Python are essential when it comes to multiprocessing and multithreading; they behave differently, but their common aim is to reduce the execution time. This book serves as a comprehensive introduction to various advanced concepts in concurrent engineering and programming. Mastering Concurrency in Python starts by introducing the concepts and principles in concurrency, right from Amdahl's Law to multithreading programming, followed by elucidating multiprocessing programming, web scraping, and asynchronous I/O, together with common problems that engineers and programmers face in concurrent programming. Next, the book covers a number of advanced concepts in Python concurrency and how they interact with the Python ecosystem, including the Global Interpreter Lock (GIL). Finally, you'll learn how to solve real-world concurrency problems through examples. By the end of the book, you will have gained extensive theoretical knowledge of concurrency and the ways in which concurrency is supported by the Python language What you will learn Explore the concepts of concurrency in programming Explore the core syntax and features that enable concurrency in Python Understand the correct way to implement concurrency Abstract methods to keep the data consistent in your program Analyze problems commonly faced in concurrent programming Use application scaffolding to design highly-scalable programs Who this book is for This book is for developers who wish to build high-performance applications and learn about signle-core, multicore programming or distributed concurrency. Some experience with Python programming language is assumed. Cover......Page 1 Title Page......Page 2 Copyright and Credits......Page 3 Dedication......Page 4 About Packt......Page 5 Contributors......Page 6 Table of Contents......Page 8 Preface......Page 17 Chapter 1: Advanced Introduction to Concurrent and Parallel Programming......Page 24 Concurrent versus sequential......Page 25 Example 1 – checking whether a non-negative number is prime......Page 26 Concurrent versus parallel......Page 29 Not everything should be made concurrent......Page 30 Inherently sequential......Page 31 Example 2 – inherently sequential tasks......Page 32 The history, present, and future of concurrency......Page 34 The history of concurrency......Page 35 The present......Page 36 The future......Page 38 A brief overview of mastering concurrency in Python......Page 40 Why Python?......Page 41 General setup......Page 43 Downloading example code......Page 44 Summary......Page 45 Further reading......Page 46 Technical requirements......Page 47 Terminology......Page 48 The formula for Amdahl's Law......Page 49 Implications......Page 50 Amdahl's Law's relationship to the law of diminishing returns......Page 51 How to simulate in Python......Page 52 Practical applications of Amdahl's Law......Page 56 Summary......Page 57 Further reading......Page 58 Chapter 3: Working with Threads in Python......Page 59 Threads versus processes......Page 60 Multithreading......Page 61 An example in Python......Page 63 The threading module in Python 3......Page 67 Creating a new thread in Python......Page 68 Starting a thread with the thread module......Page 69 Starting a thread with the threading module......Page 71 The concept of thread synchronization......Page 74 An example in Python......Page 75 A connection between real-life and programmatic queues......Page 77 The queue module......Page 78 Queuing in concurrent programming......Page 79 Multithreaded priority queue......Page 83 Summary......Page 84 Further reading......Page 85 Technical requirements......Page 86 Starting from managing files......Page 87 The with statement as a context manager......Page 88 The with statement in concurrent programming......Page 90 Example of deadlock handling......Page 91 Questions......Page 93 Further reading......Page 94 Technical requirements......Page 95 HTML......Page 96 HTTP requests......Page 98 HTTP status code......Page 99 The requests module......Page 100 Making a request in Python......Page 101 Running a ping test......Page 103 Concurrent web requests......Page 104 Spawning multiple threads......Page 105 Refactoring request logic......Page 107 Support from httpstat.us and simulation in Python......Page 109 Timeout specifications......Page 110 Error handling......Page 114 Avoid making a large number of requests......Page 115 Further reading......Page 117 Chapter 6: Working with Processes in Python......Page 118 The concept of a process......Page 119 Processes versus threads......Page 121 Multiprocessing......Page 122 Introductory example in Python......Page 124 The process class......Page 126 The Pool class......Page 127 Determining the current process......Page 128 Waiting for processes......Page 131 Interprocess communication......Page 134 Message passing for a single worker......Page 135 Message passing between several workers......Page 137 Summary......Page 143 Further reading......Page 144 Technical requirements......Page 145 Properties of a reduction operator......Page 146 Examples and non-examples......Page 147 Example implementation in Python......Page 149 Summary......Page 154 Further reading......Page 155 Technical requirements......Page 156 Python as an image processing tool......Page 157 Installing OpenCV and NumPy......Page 158 RGB values......Page 159 Coordinates inside an image......Page 160 OpenCV API......Page 161 Image processing techniques......Page 163 Grayscaling......Page 164 Thresholding......Page 166 Applying concurrency to image processing......Page 171 Choosing the correct way (out of many)......Page 175 Summary......Page 178 Further reading......Page 179 Technical requirements......Page 180 A quick analogy......Page 181 Asynchronous versus other programming models......Page 182 Asynchronous versus synchronous programming......Page 183 Asynchronous versus threading and multiprocessing......Page 184 An example in Python......Page 185 Questions......Page 188 Further reading......Page 189 Technical requirements......Page 190 Coroutines, event loops, and futures......Page 191 Asyncio API......Page 193 The asyncio framework in action......Page 194 Asynchronously counting down......Page 195 A note about blocking functions......Page 199 Asynchronous prime-checking......Page 200 Improvements from Python 3.7......Page 204 Inherently blocking tasks......Page 205 concurrent.futures as a solution for blocking tasks......Page 206 Examples in Python......Page 207 Summary......Page 211 Questions......Page 212 Further reading......Page 213 Chapter 11: Building Communication Channels with asyncio......Page 214 Communication protocol layers......Page 215 Asynchronous programming for communication channels......Page 217 Transports and protocols in asyncio......Page 218 The big picture of asyncio's server client......Page 220 Starting a server......Page 221 Installing Telnet......Page 223 Simulating a connection channel......Page 224 Sending messages back to clients......Page 225 Closing the transports......Page 226 Client-side communication with aiohttp......Page 228 Fetching a website's HTML code......Page 229 Writing files asynchronously......Page 231 Summary......Page 233 Further reading......Page 234 Technical requirements......Page 235 The Dining Philosophers problem......Page 236 Deadlock in a concurrent system......Page 239 Python simulation......Page 240 Implementing ranking among resources......Page 244 Ignoring locks and sharing resources......Page 250 An additional note about locks......Page 252 The concept of livelock......Page 253 Further reading......Page 256 Technical requirements......Page 257 What is starvation?......Page 258 Scheduling......Page 259 Causes of starvation......Page 260 Starvation's relationship to deadlock......Page 261 Problem statement......Page 262 The first readers-writers problem......Page 263 The second readers-writers problem......Page 267 The third readers-writers problem......Page 270 Solutions to starvation......Page 272 Summary......Page 273 Further reading......Page 274 Technical requirements......Page 275 Critical sections......Page 276 How race conditions occur......Page 277 Simulating race conditions in Python......Page 279 The effectiveness of locks......Page 281 Implementation in Python......Page 283 The downside of locks......Page 284 Turning a concurrent program sequential......Page 285 Locks do not lock anything......Page 287 Security......Page 288 Operating systems......Page 289 Networking......Page 290 Questions......Page 291 Further reading......Page 292 Technical requirements......Page 293 An analysis of memory management in Python......Page 294 The problem that the GIL addresses......Page 297 Problems raised by the GIL......Page 298 How to work with the GIL......Page 300 Implementing multiprocessing, rather than multithreading ......Page 301 Summary......Page 303 Further reading......Page 304 Chapter 16: Designing Lock-Based and Mutex-Free Concurrent Data Structures......Page 305 LocklessCounter and race conditions......Page 306 Embedding locks in the data structure of the counter......Page 309 The concept of scalability......Page 311 Analysis of the scalability of the counter data structure......Page 313 The idea behind approximate counters......Page 316 Implementing approximate counters in Python......Page 318 Mutex-free concurrent data structures in Python......Page 324 The impossibility of being lock-free in Python......Page 325 Introduction to the network data structure......Page 326 Implementing a simple network data structure in Python and race conditions......Page 327 RCU as a solution......Page 331 Building on simple data structures......Page 334 Questions......Page 335 Further reading......Page 336 Technical requirements......Page 337 The components of Python memory manager......Page 338 Memory model as a labeled directed graph......Page 339 In the context of concurrency......Page 341 What does it mean to be atomic?......Page 342 The GIL reconsidered......Page 343 Atomic versus nonatomic......Page 344 Simulation in Python......Page 345 Questions......Page 348 Further reading......Page 349 Technical requirements......Page 350 The theory of server-side communication......Page 351 The API of the socket module......Page 353 Building a simple echo server......Page 354 The underlying calculation logic......Page 358 Implementing the calculator server......Page 359 Analyzing the concurrency of the server......Page 362 Generators in Python......Page 366 Asynchronous generators and the send method......Page 368 Making the server non-blocking......Page 370 Questions......Page 376 Further reading......Page 377 Technical requirements......Page 378 Installing APScheduler......Page 379 APScheduler functionalities......Page 380 Executor classes......Page 382 Common scheduler methods......Page 383 Blocking scheduler......Page 384 Background scheduler......Page 385 Executor pool......Page 386 Running on the cloud......Page 388 Testing and concurrency in Python......Page 389 Unit testing......Page 390 Testing programs concurrently......Page 392 Debugging tools and techniques......Page 396 Debugging and concurrency......Page 397 Summary......Page 398 Questions......Page 399 Further reading......Page 400 Assessments......Page 401 Other Books You May Enjoy......Page 426 Index......Page 429