Oh-RAM! one and a half round atomic memory

Theophanis Hadjistasi, Nicolas Nicolaou, Alexander Allister Schwarzmann

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

5 Citations (Scopus)

Abstract

Implementing atomic read/write shared objects in a message-passing system is an important problem in distributed computing. Considering that communication is the most expensive resource, efficiency of read and write operations is assessed primarily in terms of the needed communication and the associated latency. Attiya, Bar-Noy, and Dolev established that two communication round-trip phases involving in total four message exchanges are sufficient to implement atomic operations when a majority of processors are correct. Subsequently Dutta et al. showed that one round involving two communication exchanges is sufficient as long as the system adheres to certain constraints with respect to crashes on the number of readers and writers in the system. It was also observed that three exchanges are sufficient in some settings. This extended abstract presents work that explores algorithms where operations are able to complete in three message exchanges without imposing constraints on the number of participants, i.e., the aim is One and half Round Atomic Memory, hencethenameOh-RAM!Recently Hadjistasi et al. showed that three-exchange implementations are impossible in the MWMR (multi-writer/multi-reader) setting. This paper shows that this is achievable in the SWMR (single-writer/multi-reader) setting, and also achievable for read operations in the MWMR setting by “sacrificing” the performance of write operations. In particular, a SWMR implementation is presented, where reads complete in three and writes complete in two exchanges. Next, a MWMR implementation is given, where reads involve three and writes involve four exchanges. In light of the impossibility result these algorithms are optimal in terms of the number of communication exchanges. Both algorithms are then refined to allow some reads to complete in just two exchanges. These algorithms are evaluated and compared using the NS3 simulator with different topologies and operation loads.

Original languageEnglish (US)
Title of host publicationNetworked Systems - 5th International Conference, NETYS 2017, Proceedings
EditorsAmr El Abbadi, Benoit Garbinato
PublisherSpringer Verlag
Pages117-132
Number of pages16
ISBN (Print)9783319596464
DOIs
StatePublished - Jan 1 2017
Externally publishedYes
Event5th International Conference on Networked Systems, NETYS 2017 - Marrakech, Morocco
Duration: May 17 2017May 19 2017

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume10299 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other5th International Conference on Networked Systems, NETYS 2017
CountryMorocco
CityMarrakech
Period5/17/175/19/17

Fingerprint

Random access storage
Data storage equipment
Communication
Sufficient
Message passing
Distributed computer systems
Computer systems
Simulators
Topology
Crash
Message Passing
Distributed Computing
Latency
Simulator
Resources

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Hadjistasi, T., Nicolaou, N., & Schwarzmann, A. A. (2017). Oh-RAM! one and a half round atomic memory. In A. El Abbadi, & B. Garbinato (Eds.), Networked Systems - 5th International Conference, NETYS 2017, Proceedings (pp. 117-132). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10299 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-319-59647-1_10

Oh-RAM! one and a half round atomic memory. / Hadjistasi, Theophanis; Nicolaou, Nicolas; Schwarzmann, Alexander Allister.

Networked Systems - 5th International Conference, NETYS 2017, Proceedings. ed. / Amr El Abbadi; Benoit Garbinato. Springer Verlag, 2017. p. 117-132 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10299 LNCS).

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

Hadjistasi, T, Nicolaou, N & Schwarzmann, AA 2017, Oh-RAM! one and a half round atomic memory. in A El Abbadi & B Garbinato (eds), Networked Systems - 5th International Conference, NETYS 2017, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 10299 LNCS, Springer Verlag, pp. 117-132, 5th International Conference on Networked Systems, NETYS 2017, Marrakech, Morocco, 5/17/17. https://doi.org/10.1007/978-3-319-59647-1_10
Hadjistasi T, Nicolaou N, Schwarzmann AA. Oh-RAM! one and a half round atomic memory. In El Abbadi A, Garbinato B, editors, Networked Systems - 5th International Conference, NETYS 2017, Proceedings. Springer Verlag. 2017. p. 117-132. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-319-59647-1_10
Hadjistasi, Theophanis ; Nicolaou, Nicolas ; Schwarzmann, Alexander Allister. / Oh-RAM! one and a half round atomic memory. Networked Systems - 5th International Conference, NETYS 2017, Proceedings. editor / Amr El Abbadi ; Benoit Garbinato. Springer Verlag, 2017. pp. 117-132 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{1a5eeb5cfe5646e094d7ba641dc3bb1e,
title = "Oh-RAM! one and a half round atomic memory",
abstract = "Implementing atomic read/write shared objects in a message-passing system is an important problem in distributed computing. Considering that communication is the most expensive resource, efficiency of read and write operations is assessed primarily in terms of the needed communication and the associated latency. Attiya, Bar-Noy, and Dolev established that two communication round-trip phases involving in total four message exchanges are sufficient to implement atomic operations when a majority of processors are correct. Subsequently Dutta et al. showed that one round involving two communication exchanges is sufficient as long as the system adheres to certain constraints with respect to crashes on the number of readers and writers in the system. It was also observed that three exchanges are sufficient in some settings. This extended abstract presents work that explores algorithms where operations are able to complete in three message exchanges without imposing constraints on the number of participants, i.e., the aim is One and half Round Atomic Memory, hencethenameOh-RAM!Recently Hadjistasi et al. showed that three-exchange implementations are impossible in the MWMR (multi-writer/multi-reader) setting. This paper shows that this is achievable in the SWMR (single-writer/multi-reader) setting, and also achievable for read operations in the MWMR setting by “sacrificing” the performance of write operations. In particular, a SWMR implementation is presented, where reads complete in three and writes complete in two exchanges. Next, a MWMR implementation is given, where reads involve three and writes involve four exchanges. In light of the impossibility result these algorithms are optimal in terms of the number of communication exchanges. Both algorithms are then refined to allow some reads to complete in just two exchanges. These algorithms are evaluated and compared using the NS3 simulator with different topologies and operation loads.",
author = "Theophanis Hadjistasi and Nicolas Nicolaou and Schwarzmann, {Alexander Allister}",
year = "2017",
month = "1",
day = "1",
doi = "10.1007/978-3-319-59647-1_10",
language = "English (US)",
isbn = "9783319596464",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "117--132",
editor = "{El Abbadi}, Amr and Benoit Garbinato",
booktitle = "Networked Systems - 5th International Conference, NETYS 2017, Proceedings",

}

TY - GEN

T1 - Oh-RAM! one and a half round atomic memory

AU - Hadjistasi, Theophanis

AU - Nicolaou, Nicolas

AU - Schwarzmann, Alexander Allister

PY - 2017/1/1

Y1 - 2017/1/1

N2 - Implementing atomic read/write shared objects in a message-passing system is an important problem in distributed computing. Considering that communication is the most expensive resource, efficiency of read and write operations is assessed primarily in terms of the needed communication and the associated latency. Attiya, Bar-Noy, and Dolev established that two communication round-trip phases involving in total four message exchanges are sufficient to implement atomic operations when a majority of processors are correct. Subsequently Dutta et al. showed that one round involving two communication exchanges is sufficient as long as the system adheres to certain constraints with respect to crashes on the number of readers and writers in the system. It was also observed that three exchanges are sufficient in some settings. This extended abstract presents work that explores algorithms where operations are able to complete in three message exchanges without imposing constraints on the number of participants, i.e., the aim is One and half Round Atomic Memory, hencethenameOh-RAM!Recently Hadjistasi et al. showed that three-exchange implementations are impossible in the MWMR (multi-writer/multi-reader) setting. This paper shows that this is achievable in the SWMR (single-writer/multi-reader) setting, and also achievable for read operations in the MWMR setting by “sacrificing” the performance of write operations. In particular, a SWMR implementation is presented, where reads complete in three and writes complete in two exchanges. Next, a MWMR implementation is given, where reads involve three and writes involve four exchanges. In light of the impossibility result these algorithms are optimal in terms of the number of communication exchanges. Both algorithms are then refined to allow some reads to complete in just two exchanges. These algorithms are evaluated and compared using the NS3 simulator with different topologies and operation loads.

AB - Implementing atomic read/write shared objects in a message-passing system is an important problem in distributed computing. Considering that communication is the most expensive resource, efficiency of read and write operations is assessed primarily in terms of the needed communication and the associated latency. Attiya, Bar-Noy, and Dolev established that two communication round-trip phases involving in total four message exchanges are sufficient to implement atomic operations when a majority of processors are correct. Subsequently Dutta et al. showed that one round involving two communication exchanges is sufficient as long as the system adheres to certain constraints with respect to crashes on the number of readers and writers in the system. It was also observed that three exchanges are sufficient in some settings. This extended abstract presents work that explores algorithms where operations are able to complete in three message exchanges without imposing constraints on the number of participants, i.e., the aim is One and half Round Atomic Memory, hencethenameOh-RAM!Recently Hadjistasi et al. showed that three-exchange implementations are impossible in the MWMR (multi-writer/multi-reader) setting. This paper shows that this is achievable in the SWMR (single-writer/multi-reader) setting, and also achievable for read operations in the MWMR setting by “sacrificing” the performance of write operations. In particular, a SWMR implementation is presented, where reads complete in three and writes complete in two exchanges. Next, a MWMR implementation is given, where reads involve three and writes involve four exchanges. In light of the impossibility result these algorithms are optimal in terms of the number of communication exchanges. Both algorithms are then refined to allow some reads to complete in just two exchanges. These algorithms are evaluated and compared using the NS3 simulator with different topologies and operation loads.

UR - http://www.scopus.com/inward/record.url?scp=85019771894&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85019771894&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-59647-1_10

DO - 10.1007/978-3-319-59647-1_10

M3 - Conference contribution

SN - 9783319596464

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

SP - 117

EP - 132

BT - Networked Systems - 5th International Conference, NETYS 2017, Proceedings

A2 - El Abbadi, Amr

A2 - Garbinato, Benoit

PB - Springer Verlag

ER -