Building the Best Cloud Data Architecture: Avoiding Costly Pitfalls
summary
Discusses designing cloud data architectures to avoid wasted capital from upfront on‑prem hardware by using cloud billing and scalability; recommends AWS S3, choosing storage tiers based on access cost, and trading off storage versus compute. Gives examples of failures: a 2014 startup that bought servers and paid to run 80% idle capacity, and a trading project that ingested ~160 TB of tick data across 1,500 symbols then abandoned an expensive data‑warehouse design; notes AWS has about 35% market share.
AI-written summary · about
chapters (10)
- 0:00Introduction
- 3:22Costly cloud architecture
- 9:11Cloud ownership shift
- 14:30Cloud storage benefits
- 22:26Cost-efficient design
- 28:45Common architecture mistakes
- 34:39Cloud vs on-premise
- 43:49Key data system questions
- 48:35Advice for data pros
- 57:23Q&A
related talks
56mBusy Architect’s Guide to the Cloud Offerings
36m3 Pitfalls Everyone Should Avoid with Cloud Data
28mArchitecting in the Cloud for Sustainability
53mHow to build a modern data architecture
30mAnd the CFO Wept: AWS Cost Control
36mArchitecting in the Cloud for Sustainability
17mBoosting Cost Efficiency by Cost Aware Architecture
43mAnalytics capabilities architectural patterns: A journey for data-driven insights