Sign In Start Free Trial

Book Overview & Buying
Table Of Contents
Feedback & Rating

Distributed Computing with Python

By : Pierfederici

4.3 (3)

Distributed Computing with Python

4.3 (3)

By: Pierfederici

Overview of this book

CPU-intensive data processing tasks have become crucial considering the complexity of the various big data applications that are used today. Reducing the CPU utilization per process is very important to improve the overall speed of applications. This book will teach you how to perform parallel execution of computations by distributing them across multiple processors in a single machine, thus improving the overall performance of a big data processing task. We will cover synchronous and asynchronous models, shared memory and file systems, communication between various processes, synchronization, and more.

Preface

Preface

What this book covers

What you need for this book

Who this book is for

Conventions

Reader feedback

Customer support

Free Chapter

1. An Introduction to Parallel and Distributed Computing

1. An Introduction to Parallel and Distributed Computing

Parallel computing

Distributed computing

Shared memory versus distributed memory

Amdahl's law

The mixed paradigm

Summary

2. Asynchronous Programming

2. Asynchronous Programming

Coroutines

An asynchronous example

Summary

3. Parallelism in Python

3. Parallelism in Python

Multiple threads

Multiple processes

Multiprocess queues

Closing thoughts

Summary

4. Distributed Applications – with Celery

4. Distributed Applications – with Celery

Establishing a multimachine environment

Installing Celery

Testing the installation

A tour of Celery

More complex Celery applications

Celery in production

Celery alternatives – Python-RQ

Celery alternatives – Pyro

Summary

5. Python in the Cloud

5. Python in the Cloud

Cloud computing and AWS

Creating an AWS account

Creating an EC2 instance

Storing data in Amazon S3

Amazon elastic beanstalk

Creating a private cloud

Summary

6. Python on an HPC Cluster

6. Python on an HPC Cluster

Your typical HPC cluster

Job schedulers

Running a Python job using HTCondor

Running a Python job using PBS

Debugging

Summary

7. Testing and Debugging Distributed Applications

7. Testing and Debugging Distributed Applications

The big picture

Common problems – clocks and time

Common problems – software environments

Common problems – permissions and environments

Common problems – the availability of hardware resources

Challenges – the development environment

A useful strategy – logging everything

A useful strategy – simulating components

Summary

8. The Road Ahead

8. The Road Ahead

The first two chapters

The tools

The cloud and the HPC world

Debugging and monitoring

Where to go next

Index

Index

Customer Reviews

4.3 (3)

5 star

66.7%

4 star

0

3 star

33.3%

2 star

0

1 star

0

The tools

In Chapter 3, Parallelism in Python, we looked at a few standard library modules that can be used to introduce (single-node) parallelism in our applications. We experimented with both the threading and multiprocessing modules directly and via the higher-level concurrent.futures module.

We saw how, for non-distributed, parallel applications, Python offers a really robust foundation. The preceding three modules are complete and included in every modern Python distribution. They have no external dependencies, which makes them quite appealing.

We explored a few third-party Python modules for simple distributed computing in Chapter 4, Distributed Applications – with Celery. These included Celery, Python-RQ, and Pyro. We saw how to use them in our code, and mostly, we saw how simple it is to get up and running with each one of them.

They all require some pieces of infrastructure such as message brokers, databases, or name servers, and so they might or might not be of use in all contexts...

Search

Your notes and bookmarks