From: "Oleg Pashko" Date: Thursday, May 23, 2002 3:00 AM Fundamentals of Fault-Tolerant Distributed Computing in Asynchronous Environments 1. Comprehensive summary of the literature survey in the field of fault tolerance in distributed computing. Some fundamental concepts (like fault, fault tolerance, redundancy, etc.) are formally defined. Terminology formalization and basic model outline are presented. 2 main phases to achieve 4forms of fault tolerance are also described. 2a.2 Survey of the fundamental methodologies was useful, but a lot of formalism seems redundant at times. 2b.Not applicable - no simulations or experiments. 2c.The style of the paper is close to those in math logic. And as it usually happens with newly introduced formalism, its usefulness remains to be seen, particularly in such a dynamic & fast changing area as distributed systems. 3.The attempt to introduce formal and rigorous definitions of widely (and loosely) used concepts and methods is very interesting. 4.same as 2c.Practical applications of the paper are not clear. 5-7.Can that formalism be extended to another areas of IT? Distributing Trust on the Internet 1.An architecture for secure and fault-tolerant service replication in an asynchronous network is described. The outlined model relies on protocols for Bythantine agreement and atomic broadcast.The trusted services are distributed among a number of servers that guarantees proper functioning some under malicious attack. General failure patterns & protocols are introduced. 2a.4. Interesting research on an important subject of how to establish the fault tolerance in asynchronous networks. 2.b.Not applicable - no simulations or experiments. 2.c.Lack of simulations or real experiments. 3.I was pleased to learn about threshold public-key cryptography and atomic broadcast protocol. 4. same as 2c.Not sure if it'll work OK in the real Internet. 5-7 .Can these protocol be extended to dynamic, p2p systems?