Book Image

Mastering Python Networking

Book Image

Mastering Python Networking

Overview of this book

This book begins with a review of the TCP/ IP protocol suite and a refresher of the core elements of the Python language. Next, you will start using Python and supported libraries to automate network tasks from the current major network vendors. We will look at automating traditional network devices based on the command-line interface, as well as newer devices with API support, with hands-on labs. We will then learn the concepts and practical use cases of the Ansible framework in order to achieve your network goals. We will then move on to using Python for DevOps, starting with using open source tools to test, secure, and analyze your network. Then, we will focus on network monitoring and visualization. We will learn how to retrieve network information using a polling mechanism, ?ow-based monitoring, and visualizing the data programmatically. Next, we will learn how to use the Python framework to build your own customized network web services. In the last module, you will use Python for SDN, where you will use a Python-based controller with OpenFlow in a hands-on lab to learn its concepts and applications. We will compare and contrast OpenFlow, OpenStack, OpenDaylight, and NFV. Finally, you will use everything you’ve learned in the book to construct a migration plan to go from a legacy to a scalable SDN-based network.
Table of Contents (22 chapters)
Title
Humble Bundle
Credits
Foreword
About the Author
About the Reviewer
www.PacktPub.com
Customer Feedback
Preface
12
OpenStack, OpenDaylight, and NFV

Greenfield deployment


A greenfield deployment, or building a network from scratch, is perhaps the easiest in terms of co-exist an SDN OpenFlow network with traditional network. The only interaction between your SDN network to traditional network equipment will almost only be at the edge, which in today's world, almost always uses BGP as the exchange protocol. Your external facing node will be using BGP, such as what we have seen with the Ryu BGP library, while internally you are free to use any type of connectivity and topology you wish. If you already have a way to isolate your datacenter network with this approach, in the case of a greenfield deployment you can use the same method. If you use a bottom up approach, you can start by calculating the amount of compute resource you need, tally it up to the number of racks, and determine the number of top of rack switches as a starting point. Then calculate the number of over subscription, if any, at each of the aggregation layers, leaves, spines...