Bounded wait-free f-resilient atomic byzantine data storage systems for an unbounded number of clients

Rida Bazzi, Yin Ding

Research output: Chapter in Book/Report/Conference proceedingConference contribution

2 Citations (Scopus)

Abstract

We present the first direct bounded wait-free implementation of a replicated register with atomic semantics in a system with an unbounded number of clients and in which up to / servers are subject to Byzantine failures. In a system with n ≥ 4f + i + 1 servers, and in the presence of a single writer, our implementation requires 5 messages from the reader and at most 6 + 2(f - i) messages per correct server per read operation and 2 request and 2 reply messages per server for each write operation. Unlike previous solutions, the number of messages is independent of the number of write operations that are concurrent with a read operation. For the case of multiple writers, a read operation requires 5 messages for the reader and no more than 6 + 2c(f - i) reply messages per correct server, where c is the number of writers that execute concurrently with the read operations, and a write operation requires 4 request and 4 reply messages per server. The message requirements of our wait-free implementations are considerably better in the worst case than those of the best known non wait-free implementations. If there is a bound on the number of writers, the total number of messages sent by a server is linear in the number of read operations, so faulty servers that send too many messages will be detected as faulty. This implementation does not rule out the possibility that a reader receives and discards many delayed messages in a read operation, so it is bounded only in an amortized sense. We describe a bounded solution in which a read operation will not receive more than a constant number of messages from a server without detecting the failure of the server. No other solution is bounded - in an amortized sense or otherwise.

Original languageEnglish (US)
Title of host publicationLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Pages299-313
Number of pages15
Volume4167 LNCS
StatePublished - 2006
Event20th International Symposium on Distributed Computing, DISC 2006 - Stockholm, Sweden
Duration: Sep 18 2006Sep 20 2006

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume4167 LNCS
ISSN (Print)03029743
ISSN (Electronic)16113349

Other

Other20th International Symposium on Distributed Computing, DISC 2006
CountrySweden
CityStockholm
Period9/18/069/20/06

Fingerprint

Wait-free
Information Storage and Retrieval
Data Storage
Storage System
Information Systems
Servers
Server
Data storage equipment
Semantics
Bounded Solutions
Concurrent
Computer systems

Keywords

  • Atomic
  • Byzantine
  • Fault tolerance
  • Replication
  • Timestamps
  • Wait-free

ASJC Scopus subject areas

  • Computer Science(all)
  • Biochemistry, Genetics and Molecular Biology(all)
  • Theoretical Computer Science

Cite this

Bazzi, R., & Ding, Y. (2006). Bounded wait-free f-resilient atomic byzantine data storage systems for an unbounded number of clients. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 4167 LNCS, pp. 299-313). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 4167 LNCS).

Bounded wait-free f-resilient atomic byzantine data storage systems for an unbounded number of clients. / Bazzi, Rida; Ding, Yin.

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 4167 LNCS 2006. p. 299-313 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 4167 LNCS).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Bazzi, R & Ding, Y 2006, Bounded wait-free f-resilient atomic byzantine data storage systems for an unbounded number of clients. in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). vol. 4167 LNCS, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 4167 LNCS, pp. 299-313, 20th International Symposium on Distributed Computing, DISC 2006, Stockholm, Sweden, 9/18/06.
Bazzi R, Ding Y. Bounded wait-free f-resilient atomic byzantine data storage systems for an unbounded number of clients. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 4167 LNCS. 2006. p. 299-313. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
Bazzi, Rida ; Ding, Yin. / Bounded wait-free f-resilient atomic byzantine data storage systems for an unbounded number of clients. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 4167 LNCS 2006. pp. 299-313 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{0fc84f4879c640448e5f84ec9bc8196a,
title = "Bounded wait-free f-resilient atomic byzantine data storage systems for an unbounded number of clients",
abstract = "We present the first direct bounded wait-free implementation of a replicated register with atomic semantics in a system with an unbounded number of clients and in which up to / servers are subject to Byzantine failures. In a system with n ≥ 4f + i + 1 servers, and in the presence of a single writer, our implementation requires 5 messages from the reader and at most 6 + 2(f - i) messages per correct server per read operation and 2 request and 2 reply messages per server for each write operation. Unlike previous solutions, the number of messages is independent of the number of write operations that are concurrent with a read operation. For the case of multiple writers, a read operation requires 5 messages for the reader and no more than 6 + 2c(f - i) reply messages per correct server, where c is the number of writers that execute concurrently with the read operations, and a write operation requires 4 request and 4 reply messages per server. The message requirements of our wait-free implementations are considerably better in the worst case than those of the best known non wait-free implementations. If there is a bound on the number of writers, the total number of messages sent by a server is linear in the number of read operations, so faulty servers that send too many messages will be detected as faulty. This implementation does not rule out the possibility that a reader receives and discards many delayed messages in a read operation, so it is bounded only in an amortized sense. We describe a bounded solution in which a read operation will not receive more than a constant number of messages from a server without detecting the failure of the server. No other solution is bounded - in an amortized sense or otherwise.",
keywords = "Atomic, Byzantine, Fault tolerance, Replication, Timestamps, Wait-free",
author = "Rida Bazzi and Yin Ding",
year = "2006",
language = "English (US)",
isbn = "3540446249",
volume = "4167 LNCS",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
pages = "299--313",
booktitle = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

}

TY - GEN

T1 - Bounded wait-free f-resilient atomic byzantine data storage systems for an unbounded number of clients

AU - Bazzi, Rida

AU - Ding, Yin

PY - 2006

Y1 - 2006

N2 - We present the first direct bounded wait-free implementation of a replicated register with atomic semantics in a system with an unbounded number of clients and in which up to / servers are subject to Byzantine failures. In a system with n ≥ 4f + i + 1 servers, and in the presence of a single writer, our implementation requires 5 messages from the reader and at most 6 + 2(f - i) messages per correct server per read operation and 2 request and 2 reply messages per server for each write operation. Unlike previous solutions, the number of messages is independent of the number of write operations that are concurrent with a read operation. For the case of multiple writers, a read operation requires 5 messages for the reader and no more than 6 + 2c(f - i) reply messages per correct server, where c is the number of writers that execute concurrently with the read operations, and a write operation requires 4 request and 4 reply messages per server. The message requirements of our wait-free implementations are considerably better in the worst case than those of the best known non wait-free implementations. If there is a bound on the number of writers, the total number of messages sent by a server is linear in the number of read operations, so faulty servers that send too many messages will be detected as faulty. This implementation does not rule out the possibility that a reader receives and discards many delayed messages in a read operation, so it is bounded only in an amortized sense. We describe a bounded solution in which a read operation will not receive more than a constant number of messages from a server without detecting the failure of the server. No other solution is bounded - in an amortized sense or otherwise.

AB - We present the first direct bounded wait-free implementation of a replicated register with atomic semantics in a system with an unbounded number of clients and in which up to / servers are subject to Byzantine failures. In a system with n ≥ 4f + i + 1 servers, and in the presence of a single writer, our implementation requires 5 messages from the reader and at most 6 + 2(f - i) messages per correct server per read operation and 2 request and 2 reply messages per server for each write operation. Unlike previous solutions, the number of messages is independent of the number of write operations that are concurrent with a read operation. For the case of multiple writers, a read operation requires 5 messages for the reader and no more than 6 + 2c(f - i) reply messages per correct server, where c is the number of writers that execute concurrently with the read operations, and a write operation requires 4 request and 4 reply messages per server. The message requirements of our wait-free implementations are considerably better in the worst case than those of the best known non wait-free implementations. If there is a bound on the number of writers, the total number of messages sent by a server is linear in the number of read operations, so faulty servers that send too many messages will be detected as faulty. This implementation does not rule out the possibility that a reader receives and discards many delayed messages in a read operation, so it is bounded only in an amortized sense. We describe a bounded solution in which a read operation will not receive more than a constant number of messages from a server without detecting the failure of the server. No other solution is bounded - in an amortized sense or otherwise.

KW - Atomic

KW - Byzantine

KW - Fault tolerance

KW - Replication

KW - Timestamps

KW - Wait-free

UR - http://www.scopus.com/inward/record.url?scp=33845203330&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33845203330&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:33845203330

SN - 3540446249

SN - 9783540446248

VL - 4167 LNCS

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 299

EP - 313

BT - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

ER -