When the biggest companies like Amazon or Netflix want to test their systems’ strength, they use chaos engineering to simulate the worst-case scenarios and find potential issues before they happen.
At the AWS re:Invent summit on Tuesday, Amazon CTO, Werner Vogels, showcased new Chaos Engineering as a service, called AWS Fault Injection Simulator.
The name is not all showy or even indicative of the technology’s awesomeness, but Vogels says that it will help all companies have the capabilities to test their resilience.
He says that Amazon believes every company should have access to the service, and not just the industry’s giants.
Making Chaos Engineering widely available
As he explained it, the goal of chaos engineering is to give insight into how an app responds to issues by injecting failure into the system. The typical way to conduct the test involves running it against production systems.
AWS Fault Injection Simulator gives users a fully managed service to run the experiments on apps running on AWS hardware.
FIS will make it easy to run experiments like these safely. Vogel says that they built the system to follow the typical chaos experimental workflow where one can understand the steady-state, set a hypothesis, and inject faults into the app.
You’ll have to wait
When the experiment is done, the ones running can then collect the data and know if the hypothesis was confirmed. The data collected by CloudWatch will then show the app developers where they need to make changes to avoid the scenario that they just simulated.
The service may have been announced this week but will not be available until some time in 2021.
There are similar services offered in the same vein, like Gremlin, which provides broad Chaos Engineering Services as a service.