The following is a transcript from Aval Digital Labs DevOps Engineer and Chaos Engineering Advocate, Yury Niño Roa’s, talk at Chaos Conf 2019, which you can enjoy in the embedded video above. Slides are available here.

Twitter: @yurynino

Okay. Hi, everybody. Thanks for coming. It is an honor to be part of Chaos Conference 2019. So the title of my talk is Hot Recipes for Building Chaos Experiments. Nice to meet you. I am Yury Nino. I work as the DevOps Engineer for Aval Digital Labs, a company in Columbia. I am Chaos Engineering Advocate in my country. I love designing experiments, breaking software applications and solving resilience issues.

So, in my free time, I enjoy cooking and hiking. Recently while I was working in my kitchen, I thought that cooking and Chaos Engineering have a lot of things in common. They are a combination of art and science. For example, when you are practicing chaos experiments, you should be careful choosing fresh ingredients and managing chaos. So with these ideas in my mind, I asked myself, why not write a cookbook about Chaos Engineering? So that is how I started this project, write a cookbook about Chaos Engineering.

So, for this one I check my mom's recipes, my mom's cookbooks. So, I added chaos principles and the chaos maturity model documentary on the first Chaos Engineering book. So, it is a result, I think, I can play it for doing chaos experiment inspired in the recipe. In this case, I have included the ingredients, metal for doing chaos experiments with the steps. And I also included classifications according to the Chaos Maturity Model.

So, if you remember, the Chaos Maturity Model has two metrics, sophistication and adoption. So, in this case our recipe is elemental if the experiments are running manually, and it doesn't run in production. So, when the experiments have metals and process if fully automated, it is classified in advance. Something similar happens with adoption, when there are a few systems correlated with Chaos Engineering, we are talking about in the shadows level. As you progress involving people and sponsors and processes, the recipe reaches a cultural expectation.

So, here we go. For this afternoon, I have prepared three recipes. My first recipe, my first course is for attacking an application. My second recipe is for attacking infrastructure. And finally, I included a sweet desert for doing observe the results.

So, my first recipe is for attacking an application, as I mentioned. And it is classified as simple because it doesn't run in production and in an investment level. So, in this case, this recipe is for applications implemented in Java or Kotlin with the Spring Boot. So, in this case, the ingredients are, one whole Spring Boot application. One theory of Chaos Monkey for Spring Boot dependency. And the steps are, place the application in a bowl. Add Chaos Monkey for a Spring Boot dependency, mix them extraordinarily. Configure the [inaudible 00:03:28] right, like this. So, latency is very spicy for that but you can choose between throw in exceptions or key applications. That is possible with dependency.

So, finally, launch an attack, test and serve. Enjoy seeing the looks and the cost.

So, my second recipe is for attacking the infrastructure. So, in this case, the recipe is classified as sophisticated and in adoption level. So, the ingredients are, whole application running and one Gremlin account, and installation privileges for installation. In this case, I'm going to be using Gremlin, Gremlin account. So, the first step is take your application privileges for adding the Gremlin account. These common works specification you are running during application, for example with MSN Linux. So, the second step is mix them with the Gremlin account in order to have access to a Gremlin account also. And, these common works perfect for this.

Finally, cook and attack, test and serve. Enjoy seeing the results within the Gremlin account.

So, finally, you can complete these recipes with the dessert. In this case, my dessert is inspired in [inaudible 00:04:45]. So, I have chosen Datadog, but there are a lot of flavors in the market. So, [inaudible 00:04:54] configuring the [inaudible 00:04:56] like this. So, that is the result for example.

Finally, the best time to learn about fire is when you're on fire. That is a quote from Jen Hammond, New Relic. Thank you.

See our recap of the entire Chaos Conf 2019 event.

No items found.
Chaos Engineer
Start your free trial

Gremlin's automated reliability platform empowers you to find and fix availability risks before they impact your users. Start finding hidden risks in your systems with a free 30 day trial.