Part 3: practice what you preach when deploying Selenium 4.x
summary
Deploys a Selenium Grid on AWS using Terraform + AWS with Packer-built images, Puppet provisioning, a Bastion host inside a public VPC and hub/nodes in a private subnet, and proves autoscaling with autoscaling spot instances and scale-to-zero to control cost. Demonstrates MFA for root, CloudTrail logging, billing alerts, Secrets Manager, and a scriptable proxy for third‑party traffic. Live demo hits MFA/console restrictions but the environment is created, tests register to the hub and the stack is torn down. Notes a gap: no HTTPS between hub and nodes and need for node authorization.
AI-written summary · about
chapters (16)
- 0:00Intro
- 0:48The unofficial speaking school
- 1:13Introduction
- 1:47What is burnout
- 2:16Mental health resources
- 2:37Breaking rule 2
- 12:56Networking
- 14:27Getting into instances
- 18:21Cattle not pets
- 21:45Logging into the hub
- 24:12CloudTrail
- 26:13Billing Alerts
- 28:01Autoscaling Groups
- 30:24Lifecycle Events
- 31:42Scriptable Proxy
- 33:18Cloud Formation
related talks
47mBuilding Highly Scalable Appium Infra with Terraform &Kubernetes
34mGood, Cheap, and Fast, Scaling Your Selenium Grid in the Cloud
45mAWS: Plumbing Is Important
55mBest practices
19mSynchronization of Infrastructure Provisioning & Application Deployments
29mBuilding for massive scale and security with the HashiCorp Cloud Platform
01h 32mUsing Terraform and AWS to create environments on the fly
26mOn-demand Test Infrastructure on ECS as code