Initial page

I’ve been working with the ‘big three’ cloud providers (AWS, Azure and Google) for many years; I’ve had a VPC and an EC2 instance up in AWS for at least 6, and worked on lots of projects in Azure. But I’ve never taken the time to really do some structured study on any of these. So it’s time to remedy this.

But why learn each cloud provider individually? There’s so much overlap in functionality between the big three that I think it makes more sense to learn each of them in parallel. You also get to contrast them against each other, and find the situations in which one may do something in a better way than the other.

I’m also a big advocate of automation (where it makes sense), and the thought of trying to learn by pointing and clicking through a portal immediately makes me want to turn off my laptop. So all of the scenarios will spun up (and down) using infrastructure-as-code, preferably using Terraform, and where that’s not possible using the providers’ CLI tools.

How It’s Structured

Each chapter on this book will focus on high’ish level concept (basic networking, DNS, serverless). Within each chapter we will create a number of different scenarios that apply to this concept. I’ll link to the Terraform code that creates the scenario, and we will also run through some tests to prove out the operational status. For example we might ping outbound from an EC2 instance, make an HTTP request to a load balancer, or a DNS request to a DNS server.

How It’s Generated

Each page is generated from an Rmarkdown document, with inline code being ‘knitted’ into the document using the knitr package. This is similar to Jupyter notebooks, however as an advocate of R I would say it’s more powerful.

The generated markdown, along with the Terraform files for each scenario, and any other secondary files, and commited and pushed up to a github repository. This repository is linked with a gitbooks site called bigthree.study.foletta.org. Whenever a new commit is pushed, this gitbooks site is auto-generated from the markdown files.