PostgreSQL: Revealing Secrets Behind YugabyteDB's Industry-Leading Compatibility
summary
YugabyteDB claims Postgres parity by adding read committed isolation (every statement uses a fresh snapshot, no serialization errors) and pessimistic locking (conflicting transactions wait instead of aborting), implemented in v2.15 on a 5‑node cluster. Demo shows no read restart errors under read committed, read restart errors exposed under repeatable read, and deadlock detection via distributed wait‑for graphs and probe messages that identify cycles and abort one waiter so the rest proceed.
AI-written summary · about
chapters (12)
- 0:00PostgreSQL compatibility
- 2:08Why transactional compatibility
- 4:34Transactional compatibility facets
- 6:41Read committed isolation
- 9:32Demo: read committed isolation
- 12:42Read restart errors
- 15:32Demo: read restart errors
- 17:43Roadmap: read committed isolation
- 18:05Pessimistic locking in YugabyteDB
- 19:51Demo: pessimistic locking
- 21:57Deadlock detection in distributed systems
- 25:04Summary and future plans
related talks
25mYugabyteDB Technical Overview
30mTechnical Comparison of YugabyteDB with Oracle and PostgreSQL
26mIntroducing YugabyteDB for PostgreSQL Users
51mEmbrace, Extend, Evolve - Making PostgreSQL Work for a Cloud Native World
27mIntroduction to YugabyteDB – Design and Architecture
26mEfforts Toward Adoption and Uncovering Hidden Features
20mDatabase migration: Insights on Dual Read Write Strategy
51mJump Start Your Developer Journey in 60 Minutes: Go from 0 to Expert