Book Image

Google Cloud for DevOps Engineers

By : Sandeep Madamanchi
Book Image

Google Cloud for DevOps Engineers

By: Sandeep Madamanchi

Overview of this book

DevOps is a set of practices that help remove barriers between developers and system administrators, and is implemented by Google through site reliability engineering (SRE). With the help of this book, you'll explore the evolution of DevOps and SRE, before delving into SRE technical practices such as SLA, SLO, SLI, and error budgets that are critical to building reliable software faster and balance new feature deployment with system reliability. You'll then explore SRE cultural practices such as incident management and being on-call, and learn the building blocks to form SRE teams. The second part of the book focuses on Google Cloud services to implement DevOps via continuous integration and continuous delivery (CI/CD). You'll learn how to add source code via Cloud Source Repositories, build code to create deployment artifacts via Cloud Build, and push it to Container Registry. Moving on, you'll understand the need for container orchestration via Kubernetes, comprehend Kubernetes essentials, apply via Google Kubernetes Engine (GKE), and secure the GKE cluster. Finally, you'll explore Cloud Operations to monitor, alert, debug, trace, and profile deployed applications. By the end of this SRE book, you'll be well-versed with the key concepts necessary for gaining Professional Cloud DevOps Engineer certification with the help of mock tests.
Table of Contents (17 chapters)
1
Section 1: Site Reliability Engineering – A Prescriptive Way to Implement DevOps
6
Section 2: Google Cloud Services to Implement DevOps via CI/CD
Appendix: Getting Ready for Professional Cloud DevOps Engineer Certification

Illustrating the impact of SLAs, SLOs, and error budgets relative to SLI

In this section, we will go through two hands-on scenarios to illustrate how SLO targets are met or missed based on SLI performance over time. SLOs performance will have direct impact on SLAs and error budgets. Changes in the error budget will specifically dictate the priority between the release of new features versus service reliability. For ease of explanation, a 7-day period is taken as the measure of time (ideally, a 28-day period is preferred).

Scenario 1 – New service features introduced; features are reliable; SLO is met

Here are the expectations for this scenario:

  • Expected SLA—95%
  • Expected SLO—98%
  • Measured SLI—Service availability or uptime
  • Measure duration—7 days

Given that the anticipated SLO for service is 98%, here is how the allowed downtime or error budget is calculated (you can use this downtime calculator for reference: https...