Kubernetes and Selenium Grid for Highly Scalable Browser and Device Farm

Explore a comprehensive solution for highly scalable browser and device farms using Kubernetes and Selenium Grid in this 50-minute conference talk. Dive into the challenges posed by cloud computing and microservices to Selenium Grid-based solutions for concurrent UI automation test execution. Learn how to optimize cloud costs, address the needs of micro-frontends, and improve feedback cycle speed in pre-merge pipelines. Discover the benefits of using Kubernetes with Docker, Helm, and Traefik to create on-demand unique grids, achieve faster creation and deletion of hubs/nodes, and optimize resource allocation. Gain valuable insights on fine-tuning CPU and memory for browser pods, configuring resources for different scale hubs, and implementing a long-term solution for companies transitioning to the cloud. Cover topics such as continuous testing pipelines, micro-frontends, multi-cloud environments, autoscaling worker nodes, EKS on Fargate, resource QoS, and cost optimization strategies.

Intro
Key Tenets of Continuous Testing Pipeline
Data Centre/Jenkins
Micro-frontends
Multiple Cloud A/C
Solution
Big Picture
End to End flow
Limiting access to NLB
Creating - Scaling
Autoscaling of Worker Nodes
EKS on Fargate - Serverless
Warm Worker Node Pool
Https based Hub
Resource Qos
One browser per Pod
Hub and Node in different subnet
Selenium Changes and Cost
Prevent IP Address bleeding - EKS
CLOUD PROVIDERS VS SELENIUM GRID
SPOT INSTANCE IN EKS
SCALE THE TEST INFRASTRUCTURE
SCALE THE TEST RUNNERS VERTICALLY
RETURN ON INVESTMENT