#ATAGTR2023 Speaker

Implementation of Chaos Experiments – a pocket friendly and simple way

In the modern world of Cloud computing and containerization, the necessity of serving the customers/end users should be fast and accurate without any loss of confidence and revenue.  To improve confidence and better shaping of infrastructure, Chaos Engineering plays a major role where we conduct all the possible experiments ahead of releasing the grand application for customers/end users.

ChaosBlade is one such tool to perform controlled experiments in the infrastructure before Go-Live to avoid disasters and to provide a smooth experience for end users.  ChaosBlade is an open-source tool developed by Alibaba.  Initially, the tool has been developed to test the Alibaba Cloud, due to its capability the tool has been made open source which is pocket friendly, mainly for start-ups. 

ChaosBlade has the ability to carry out experiments on Host (OnPrem or Cloud), Containers, and Applications.  The various types of chaos experiments that are available on CPU, Memory, Network, Disk, etc for Host Level.  Also, experiments are available for Pod, Container & nodes, and experiments can be carried out for the Java application.

ChaosBlade experiments have the capability to carry out the experiments through the command line, APIs, and ChaosBlade Dashboard.  ChaosBlade can be integrated with the CI/CD process where the experiments can be triggered through API calls in intervals or through a scheduler.  It’s really fun to carry out the experiments along with the performance testing of the applications, where the experiments through API calls can be integrated through the performance scenarios and control the experiments through wait times which gives more insights into the behavior of infrastructure and can be monitored through the various monitoring tools.

The main difference between the other open-source Chaos engineering tools and ChaosBlade is that the tool has the capability to perform fault drills through APIs, the command line, and through ChaosBlade dashboard.  Implementation of the ChaosBlade tool is simple compared to the other Chaos engineering tools.  Overall, ChaosBlade is a package of various fault drills, open-source, and easy-to-use tools.

Table Contents:

Lab Requirements:

1. Linux based Cloud machines – 1 (Chaos Blade Tool Installation)
2. Java based Application hosted on cloud
3. Containerized application

Pre-Requisites:

1. Basic understanding of Chaos Engineering
2. Basic understanding of Cloud concepts