Database systems: a decade of disruption and innovation
summary
Argues that the cloud breaks two core assumptions — a stable outside world and a monolithic engine — and shows concrete fixes: cloud‑native architectures such as Azure SQL Hyperscale separate compute, storage and log to give ACID while scaling to millions of databases (single DBs into hundreds of terabytes) and deliver four nines availability. Describes controlled overbooking and serverless packing, monthly continuous deployment with automatic rollback, plan‑cache cold starts causing compilation storms (solved with plan‑replay), and warns AI is breaking determinism and testing.
AI-written summary · about
chapters (10)
- 0:00Introduction
- 3:12Database engine innovations
- 9:33Cloud disruption assumptions
- 14:29Cloud DBs vs services
- 21:11Beyond database core
- 25:04Flexible cloud resources
- 35:04Query processing changes
- 40:20AI impact on determinism
- 43:08Databases for AI era
- 49:46Future of cloud benchmarking
related talks
30mThe database trends that are transforming your database infrastructure forever
23mState in a stateless world: data serving in the cloud
28mCloud Native Tech: The Building Blocks of Modern DB Software
28mThe Next Decade in Data Management
32mCloud-Native Data: What is it? - Majors, Dunlap, O’Grady, Richardson, Drewitz
29mModern Database Architecture: From SQL to NoSQL & Distributed SQL
20mThe Future of Databases
11mDatabase Evolution: Partitioning and Indexing