Consistency and isolation for Python programmers
summary
Builds a toy Python telnet server demo that actually runs concurrency tests and shows which anomalies occur and why; implements read uncommitted, read committed, snapshot isolation (with snapshot isolation aborts writes logic), and serializable locking, and explains deadlocks and phantoms. Then models distributed replication, shows causal consistency with timestamps and linearizability tradeoffs. Includes runnable code, points to four or five papers, and explains when operations block, abort, or require consensus so you can judge performance versus correctness.
AI-written summary · about
chapters (7)
- 0:00Start
- 0:53Consistency vs. Isolation
- 4:08READ UNCOMMITTED
- 6:10READ COMMITTED
- 13:16SERIALIZABILITY
- 14:30ISOLATION LEVELS
- 15:41DISTRIBUTED DATABASE
related talks
41mConcurrency, Scalability and Transactions -- Myths and Surprises
30mAn Admin's Cornucopia - Python Is More Than Just A Better Bash
01h 06mLightning Talks & Closing
01h 02mConsistency and Agreements in Distributed Systems
01h 00mCoroutine Concurrency in Python 3 with asyncio
35mDon't Give Up on Serializability Just Yet
37mDistributed SQL Transactions
40mAdvanced asyncio: Solving Real-world Production Problems