Transactions are one of the most misunderstood concepts in software engineering.
Every developer knows the ACID acronym — but do you know what isolation actually
guarantees? Do you know what write skew is, and why it can happen even with
"correct" transactions under snapshot isolation?
In this video, we go deep on Chapter 7 of Designing Data-Intensive Applications
by Martin Kleppmann — the definitive guide to transactions in modern data systems.
What you'll learn:
✅ The REAL meaning of ACID — and why C doesn't belong in the acronym
✅ The difference between Read Committed, Snapshot Isolation, and Serializable
✅ Every race condition catalogued: dirty reads, lost updates, write skew, phantoms
✅ How MVCC (Multi-Version Concurrency Control) works internally in PostgreSQL
✅ Two-Phase Locking vs Serializable Snapshot Isolation — performance tradeoffs
✅ How to choose the right isolation level for YOUR system
Whether you're preparing for system design interviews, building high-scale backends,
or studying for a staff engineering role, this is the transaction deep-dive you need.
