100K-User PWA Load Simulation

Our live-event Progressive Web App must hold steady under 100,000 concurrent visitors. I’m looking for a senior performance engineer who can design and run a realistic, end-to-end load campaign that goes far beyond basic QA scripting. The work starts with an architecture blueprint: distributed k6 and/or Locust generators spun up on AWS (EKS, Fargate or a similar elastic setup) so we can dial traffic up smoothly while keeping costs under control. You’ll script full user journeys—browse, add-to-cart, checkout and payment—but the cart and checkout path will carry the heaviest weight because it represents our revenue peak during a live drop. API-level hits and lightweight browser-level probes should be balanced to reflect actual use; I’ll listen to your reasoning on the exact split. Once the rig is ready, you will: • Build parameterised test scripts covering geo mix, think-time and ramp patterns • Orchestrate step and spike tests up to 100 K active sessions • Capture system-wide metrics and trace bottlenecks (autoscaling, DB locks, queue back-pressure, etc.) • Deliver a concise report with findings, graphs and remediation suggestions Acceptance criteria: reproducible test scripts in Git, Terraform/Kubernetes manifests for the load cluster, run book, results dashboard link and a PDF summary highlighting every threshold breach. When you reply, include past work that proves you’ve handled similar-scale traffic, outline the stack you’d choose, explain how you separate API from browser testing, cite the largest load you’ve executed and describe how you validate payments safely without touching real money.

Project ID: 40377051