Book Image

Python Microservices Development

Book Image

Python Microservices Development

Overview of this book

We often deploy our web applications into the cloud, and our code needs to interact with many third-party services. An efficient way to build applications to do this is through microservices architecture. But, in practice, it's hard to get this right due to the complexity of all the pieces interacting with each other. This book will teach you how to overcome these issues and craft applications that are built as small standard units, using all the proven best practices and avoiding the usual traps. It's a practical book: you’ll build everything using Python 3 and its amazing tooling ecosystem. You will understand the principles of TDD and apply them. You will use Flask, Tox, and other tools to build your services using best practices. You will learn how to secure connections between services, and how to script Nginx using Lua to build web application firewall features such as rate limiting. You will also familiarize yourself with Docker’s role in microservices, and use Docker containers, CoreOS, and Amazon Web Services to deploy your services. This book will take you on a journey, ending with the creation of a complete Python application based on microservices. By the end of the book, you will be well versed with the fundamentals of building, designing, testing, and deploying your Python microservices.
Table of Contents (20 chapters)
Title Page
Credits
About the Author
About the Reviewer
www.PacktPub.com
Customer Feedback
Preface
Introduction

Summary


In this chapter, we've seen how to add some instrumentation in our microservices and at the web server level. We've also learned how to set up Graylog to centralize and use all the generated logs and performance metrics.

Graylog uses Elasticsearch to store all the data, and that choice offers fantastic search features that will make your life easier to look for what's going on. The ability to add alerts is also useful for being notified when something's wrong. But deploying Graylog should be considered carefully. An Elastic Search cluster is heavy to run and maintain once it has a lot of data.

For your metrics, time-series based systems such as ;InfluxDB (open source) from InfluxData (https://www.influxdata.com/) is a faster and lightweight alternative. But it's not meant to store raw logs and exceptions.

So if you just care about performance metrics and exceptions, maybe a good solution would be to use a combination of tools: Sentry for your exceptions and InfluxDB for tracking performances...