Modern Database Architecture: From SQL to NoSQL & Distributed SQL
summary
A clear claim: distributed SQL preserves full transactional semantics while scaling across dozens or hundreds of nodes, using automatic sharding and consensus (Paxos/Raft) under the hood. The presentation gives concrete trade-offs: replication gives near‑unlimited read capacity but allows milliseconds‑to‑seconds staleness; two‑phase commit and distributed transactions work but incur blocking and latency. A team building ML infrastructure behind Mo and Chhat reports over three million months, and examples include cache+replica setups and Redis‑style in‑memory clusters that demonstrably trade consistency for availability.
AI-written summary · about
chapters (9)
- 0:00Introduction and Speaker Background
- 0:26Early Database Systems
- 1:00The Relational Model Revolution
- 2:45Challenges of the Internet Era
- 3:35Scaling Databases: Vertical and Horizontal Approaches
- 7:55Distributed Systems and the CAP Theorem
- 14:33Modern NoSQL Databases
- 20:33Consistency Models in Distributed Databases
- 28:34Conclusion and Final Thoughts
related talks
26mNewSQL: concepts, realities and when you should embrace it
31mSQL to NoSQL to NewSQL and the rise of polyglot persistence Paul Dix
01h 01mNoSQL: An Analysis
27mScaling for Intelligence: Distributed SQL Databases in the Age of AI
43mThe State of NoSQL in 2021
45mSay YES! to NoSQL: A Guide on When to Ditch Relational Databases
18mFrom SQL to NOSQL - a gentle intro for Devs
45mNoSQL: Death to Relational Databases