Super large, ultra fast, mega scaling cloud computing
summary
A large Norwegian oil operator collects billions of time-series points from thousands of platform sensors and ships it into Azure, then runs two ingestion paths (15-minute batch and low-latency real time), stores petabytes in a Data Lake + Databricks stack, uses Event Hubs/IoT Hub, Stream Analytics/Time Series Insights and SQL DW for curated views, and deploys analytics to edge modules for local filtering and ML. The first VM/SQL lift‑and‑shift failed at scale; rearchitecture to cloud-native services and edge worked and reduced costs and latency.
AI-written summary · about
chapters (18)
- 0:00Introduction
- 4:05Evil of Linux
- 5:08SP 500
- 8:37IoT
- 11:10Power plants
- 16:06Oil companies
- 21:16Brilliant idea
- 21:33Digital roadmap
- 22:42Data platform
- 26:15Hololens
- 28:08Supercenter
- 29:27Data ingestion
- 37:38Sharing data
- 39:55Scalability
- 41:28Do things differently
- 43:06IOT Central
- 48:50Example
- 49:44Thank you
related talks
36mThe MODERN Modern Data Stack: Building an Open Distributed Data Warehouse Beyond...
32mDiscover the Modern Way to Build Real-Time Analytics Solutions at Any Scale
43mLeveraging Azure PaaS for Real-time Social Media Analysis
58mBuilding a green supercomputer in the cloud
43mDatabase capabilities you need for your next awesome app Q&A
28mBig Data & Cloud Forum – Djalma G. dos Santos
08mReal-Time Data Systems: Speed, Scale, and Resilience
42mBuilding a green supercomputer in the cloud