What properties should asymmetric quorum systems satisfy?
summary
Shows that asymmetric quorum systems collapse to symmetric ones in the crash-fault model by transforming any asymmetric collection into an equivalent symmetric quorum set, so asymmetric trust gives no benefit there; in the Byzantine model consistency and availability are insufficient, and the existing guild assumption effectively forces symmetry. Introduces a measurable depth for processes, gives a modified reliable-broadcast protocol that reduces the requirement from depth infinity to depth=3, and leaves consensus depth unresolved.
AI-written summary · about
chapters (16)
- 0:00Asymmetric trust and quorum systems
- 2:51Asymmetric generalized trust
- 7:03Classifying correct processes
- 8:46Symmetric properties insufficient
- 10:11Guilts for reliable broadcast
- 12:17Guilts resemble symmetric trust
- 14:25Challenges in symmetric quorums
- 16:30Limitations of asymmetric trust
- 17:12Introducing depth
- 20:49Adapting problems with depth
- 22:36Reliable broadcast with depth
- 25:28Open questions on consensus depth
- 26:55Replacing guilt with depth
- 27:38Future research directions
- 29:02Q&A on quorums
- 33:56Closing remarks
related talks
01h 04mConsensus with Asymmetric Trust
18mAdoB: Bridging Benign and Byzantine Consensus with Atomic Distributed Objects
36mFlexible Byzantine Fault Tolerance
33mSybilQuorum:Bootstrapping Sybil Resistant Quorum Systems from "Trust" Networks
43mWhat We Talk About When We Talk About Distributed Systems
48mDistributed Consensus Algorithms
56mA sledgehammer to crack a nut: why blockchain is not (always) a good idea
22mWhy Are Distributed Systems so Hard?