chapter two

2 First cup of chaos & blast radius

 

This chapter covers

  • Setting up the VM to run through accompanying code
  • Basics of Linux forensics - why did my process die?
  • Your first chaos experiment with a simple bash script
  • Blast radius

In the previous chapter you covered what chaos engineering is and what a chaos experiment template looks like. It is now time to get your hands dirty and implement some from scratch! I’m going to take you step by step through building your first chaos experiment, using nothing more than a few lines of bash. I’ll also use the occasion to introduce and illustrate new concepts like blast radius.

Just one last pit-stop before we’re off to our journey - let’s set up the workspace.

DEFINITIONS

I’ll bet you’re wondering what a blast radius is. Let me explain. Much like an explosive, a software component can go wrong and break other things it connects to. We often speak of a blast radius to describe the maximum of things which can be affected by something going wrong. I’ll be teaching you more about it as you read this chapter.

2.1   Setup - working with the code in this book

2.2   Scenario

2.3   Linux forensics 101

2.3.1   Exit codes

2.3.2   Killing processes

2.3.3   Out Of Memory (OOM) Killer

2.4   The first chaos experiment

2.4.1   Visibility

2.4.2   Steady state

2.4.3   Hypothesis

2.4.4   Run the experiment

2.5   Blast radius

2.6   Digging deeper

2.6.1   Saving the world

2.7   Summary