Already registered? Log in now to personalize your experience!
Test the performance of any large-scale web applications by simulating http loads from several cloud platforms using the power of Kubernetes job parallelization and autoscaling, Flux2 for multi-cloud clusters setup and jobs executions, Linkerd to collect web performances and Prometheus for storing performances metrics to see what happen in real time on a single dashboard.
The purpose of this talk is to demonstrate how to use the parallelization power of Kubernetes Jobs combined with observability and monitoring tools in order to visualize events and performances. All managed in a fast, simple and industrialized way by Flux 2 as captain to easily run this from many Kubernetes clusters hosted in several different cloud platforms.
I developed this approach when I was working on very high volume APIs for mobile applications to be able to simulate billions of connections on a planetary scale in the most realistic conditions.
Thanks to the combination of all these CNCF tools and the orchestration of Kubernetes - after a lot of testing and experimentation - a simple Yaml configuration file allows us to overcome this challenge super easily.
Multi Cloud http load testing with Kubernetes, Flux2, Linkerd & Prometheus