Gossip-based clock synchronization for large decentralized systems

Konrad Iwanicki; Maarten van Steen; Spyros Voulgaris

doi:10.1007/11767886_3

Gossip-based clock synchronization for large decentralized systems

Konrad Iwanicki^*, Maarten van Steen, Spyros Voulgaris

^*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Academic › peer-review

33 Citations (Scopus)

Abstract

Numerous large-scale decentralized systems assume loosely synchronized clocks. Existing time protocols have not been designed for deployment in such systems, since they are complex and require manual configuration. We present the Gossiping Time Protocol (GTP), a completely self-managing epidemic time synchronization algorithm for peer-to-peer networks. In GTP, each node synchronizes its time by gossiping with other nodes. The decisions regarding sample evaluation and gossiping frequency are purely local, yet they result in consistent behavior of the whole system. Large-scale experimental evaluation of a 64,500-node network emulated on 65 machines indicates high scalability and reasonable accuracy of GTP.

Original language	English
Title of host publication	Self-Managed Networks, Systems, and Services
Subtitle of host publication	Second IEEE International Workshop, SelfMan 2006, Dublin, Ireland, June 16, 2006. Proceedings
Editors	Alexander Keller, Jean-Philippe Martin-Flatin
Publisher	Springer
Pages	28-42
Number of pages	15
ISBN (Electronic)	978-3-540-34740-8
ISBN (Print)	978-3-540-34739-2
DOIs	https://doi.org/10.1007/11767886_3
Publication status	Published - 1 Jan 2006
Externally published	Yes
Event	2nd IEEE International Workshop on Self-Managed Networks, Systems, and Services, SelfMan 2006 - Dublin, Ireland Duration: 16 Jun 2006 → 16 Jun 2006 Conference number: 2

Publication series

Name	Lecture Notes in Computer Science
Publisher	Springer
Volume	3996
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	2nd IEEE International Workshop on Self-Managed Networks, Systems, and Services, SelfMan 2006
Abbreviated title	SelfMan 2006
Country/Territory	Ireland
City	Dublin
Period	16/06/06 → 16/06/06

Access to Document

10.1007/11767886_3

Cite this

Iwanicki, K., van Steen, M., & Voulgaris, S. (2006). Gossip-based clock synchronization for large decentralized systems. In A. Keller, & J.-P. Martin-Flatin (Eds.), Self-Managed Networks, Systems, and Services: Second IEEE International Workshop, SelfMan 2006, Dublin, Ireland, June 16, 2006. Proceedings (pp. 28-42). (Lecture Notes in Computer Science; Vol. 3996). Springer. https://doi.org/10.1007/11767886_3

Iwanicki, Konrad ; van Steen, Maarten ; Voulgaris, Spyros. / Gossip-based clock synchronization for large decentralized systems. Self-Managed Networks, Systems, and Services: Second IEEE International Workshop, SelfMan 2006, Dublin, Ireland, June 16, 2006. Proceedings. editor / Alexander Keller ; Jean-Philippe Martin-Flatin. Springer, 2006. pp. 28-42 (Lecture Notes in Computer Science).

@inproceedings{5edcbc3fcbb34400b86f2f89487c5c69,

title = "Gossip-based clock synchronization for large decentralized systems",

abstract = "Numerous large-scale decentralized systems assume loosely synchronized clocks. Existing time protocols have not been designed for deployment in such systems, since they are complex and require manual configuration. We present the Gossiping Time Protocol (GTP), a completely self-managing epidemic time synchronization algorithm for peer-to-peer networks. In GTP, each node synchronizes its time by gossiping with other nodes. The decisions regarding sample evaluation and gossiping frequency are purely local, yet they result in consistent behavior of the whole system. Large-scale experimental evaluation of a 64,500-node network emulated on 65 machines indicates high scalability and reasonable accuracy of GTP.",

author = "Konrad Iwanicki and {van Steen}, Maarten and Spyros Voulgaris",

year = "2006",

month = jan,

day = "1",

doi = "10.1007/11767886_3",

language = "English",

isbn = "978-3-540-34739-2",

series = "Lecture Notes in Computer Science",

publisher = "Springer",

pages = "28--42",

editor = "Alexander Keller and Jean-Philippe Martin-Flatin",

booktitle = "Self-Managed Networks, Systems, and Services",

address = "Germany",

note = "2nd IEEE International Workshop on Self-Managed Networks, Systems, and Services, SelfMan 2006, SelfMan 2006 ; Conference date: 16-06-2006 Through 16-06-2006",

}

Iwanicki, K, van Steen, M & Voulgaris, S 2006, Gossip-based clock synchronization for large decentralized systems. in A Keller & J-P Martin-Flatin (eds), Self-Managed Networks, Systems, and Services: Second IEEE International Workshop, SelfMan 2006, Dublin, Ireland, June 16, 2006. Proceedings. Lecture Notes in Computer Science, vol. 3996, Springer, pp. 28-42, 2nd IEEE International Workshop on Self-Managed Networks, Systems, and Services, SelfMan 2006, Dublin, Ireland, 16/06/06. https://doi.org/10.1007/11767886_3

Gossip-based clock synchronization for large decentralized systems. / Iwanicki, Konrad; van Steen, Maarten; Voulgaris, Spyros.
Self-Managed Networks, Systems, and Services: Second IEEE International Workshop, SelfMan 2006, Dublin, Ireland, June 16, 2006. Proceedings. ed. / Alexander Keller; Jean-Philippe Martin-Flatin. Springer, 2006. p. 28-42 (Lecture Notes in Computer Science; Vol. 3996).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Academic › peer-review

TY - GEN

T1 - Gossip-based clock synchronization for large decentralized systems

AU - Iwanicki, Konrad

AU - van Steen, Maarten

AU - Voulgaris, Spyros

N1 - Conference code: 2

PY - 2006/1/1

Y1 - 2006/1/1

N2 - Numerous large-scale decentralized systems assume loosely synchronized clocks. Existing time protocols have not been designed for deployment in such systems, since they are complex and require manual configuration. We present the Gossiping Time Protocol (GTP), a completely self-managing epidemic time synchronization algorithm for peer-to-peer networks. In GTP, each node synchronizes its time by gossiping with other nodes. The decisions regarding sample evaluation and gossiping frequency are purely local, yet they result in consistent behavior of the whole system. Large-scale experimental evaluation of a 64,500-node network emulated on 65 machines indicates high scalability and reasonable accuracy of GTP.

AB - Numerous large-scale decentralized systems assume loosely synchronized clocks. Existing time protocols have not been designed for deployment in such systems, since they are complex and require manual configuration. We present the Gossiping Time Protocol (GTP), a completely self-managing epidemic time synchronization algorithm for peer-to-peer networks. In GTP, each node synchronizes its time by gossiping with other nodes. The decisions regarding sample evaluation and gossiping frequency are purely local, yet they result in consistent behavior of the whole system. Large-scale experimental evaluation of a 64,500-node network emulated on 65 machines indicates high scalability and reasonable accuracy of GTP.

UR - http://www.scopus.com/inward/record.url?scp=33746588203&partnerID=8YFLogxK

U2 - 10.1007/11767886_3

DO - 10.1007/11767886_3

M3 - Conference contribution

AN - SCOPUS:33746588203

SN - 978-3-540-34739-2

T3 - Lecture Notes in Computer Science

SP - 28

EP - 42

BT - Self-Managed Networks, Systems, and Services

A2 - Keller, Alexander

A2 - Martin-Flatin, Jean-Philippe

PB - Springer

T2 - 2nd IEEE International Workshop on Self-Managed Networks, Systems, and Services, SelfMan 2006

Y2 - 16 June 2006 through 16 June 2006

ER -

Iwanicki K, van Steen M, Voulgaris S. Gossip-based clock synchronization for large decentralized systems. In Keller A, Martin-Flatin JP, editors, Self-Managed Networks, Systems, and Services: Second IEEE International Workshop, SelfMan 2006, Dublin, Ireland, June 16, 2006. Proceedings. Springer. 2006. p. 28-42. (Lecture Notes in Computer Science). doi: 10.1007/11767886_3

Gossip-based clock synchronization for large decentralized systems

Abstract

Publication series

Conference

Access to Document

Other files and links

Fingerprint

Cite this