How I Pwned My Cloud Costs
summary
A public breach-check service handling billions of records stays on a coffee-budget by replacing autoscaling app-service instances with Azure Functions (67k requests cost $0.14 in one example), fronting with CloudFlare (99.62% cache hit, 32.4M requests/week) and moving index data from Table Storage to Blob Storage (latency cut from ~122ms to ~54ms). The result: huge traffic spikes handled cheaply and with practical rollovers using CloudFlare Workers.
AI-written summary · about
chapters (17)
- 0:00Intro
- 3:00Autoscale
- 6:00Example
- 11:30The Martin Lewis Money Show
- 14:00The Goldilocks Principle
- 14:45Pwned Password Demo
- 18:20Function Execution Counts
- 20:45Function Pricing
- 24:05Data Storage
- 24:55Table Storage
- 26:55Blob Storage
- 29:30Cloudflare
- 31:40Cloud Data
- 36:50Rolling Over
- 38:30CloudFlare Worker
- 39:55Concerns
- 42:15Questions
related talks
37mHave I Been Pwned: Serving billions of requests and terabytes of data without going broke!
45mCloud Computing: Cloud in a cost-effective way
46mAzure on the cheap
53mAzure for Less Than One Dollar a Day
44mFinOps for Engineers: Why Your Architecture Has a Price Tag
37mA startup's infrastructure - how to make scaling a sales, not a tech issue
20mHow Do I Reduce My Cloud Bill?
46m10 Strategies to keep your Cloud costs under control