(...)Distributed Systems
summary
Defines concrete models (synchronous, asynchronous, semi‑synchronous), states the precise claim of the FLP impossibility (consensus can fail with even one crash under purely asynchronous scheduling), explains which failure detectors and which timing or channel assumptions actually restore consensus, shows how quorum intersection guarantees progress, distinguishes crash vs Byzantine failures, and cites specific books and papers (Nancy Lynch, Paxos/Raft/FLP) as the targeted next reads for implementing fault‑tolerant replication.
AI-written summary · about
chapters (23)
- 0:00Intro
- 0:44Distributed Systems Definition
- 1:27Jargon
- 3:04What to read
- 4:38Books
- 6:16Why
- 8:04Synchronous
- 10:18Asynchronous
- 11:01Interprocess communication
- 11:44Failure modes
- 14:22Safety
- 16:17FLP
- 17:10Synchronous Consensus
- 23:10When do we need Consensus
- 24:08Markos Aguilera
- 25:51Failure Detectors
- 27:09Perfect Failure Detector
- 29:14Accurate Failure Detector
- 30:56quorums
- 32:53disability
- 35:47linearity
- 37:24book recommendation
- 43:31finding paywalled papers
related talks
01h 02mConsistency and Agreements in Distributed Systems
44mThe Road to Akka Cluster, and Beyond
01h 02mAt Least Once
01h 00mThe Secret Life of Distributed Systems
57mScaling the CAP Theorem for Modern Distributed Systems
40mAll In With Determinism for Performance and Testing in Distributed Systems
01h 03mAdventures in Spacetime
43mDistributed Systems