An approach to fault-tolerant parallel processing on intermittently idle, heterogeneous workstations

Karpjoo Jeong, Dennis Shasha, Surendranath Talla, Peter Wyckoff

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

Abstract

We propose a novel approach to harness the idle cycles of workstations connected by LAN/WANs for long running scientific computations and present performance results for our prototype system called Persistent Linda (PLinda). PLinda offers low runtime overhead and, migration among heterogeneous architectures while retreating quickly when owners return to their workstations. PLinda achieves this by implementing a lightweight transaction model that lacks serializability and durability, but preserves a basic guarantee: if a PLinda execution terminates, it has the same result as some failure-free Linda execution. Further, by storing the state of a PLinda process as a set of core variables as of each transaction commit, a PLinda process can migrate among different architectures. Within the space of lightweight transaction models, we offer three mechanisms that make different tradeoffs between failure-free performance and recovery time. All three mechanisms may be used at the same time in a single application; each process using the mechanism which is best suited for its characteristics. Our experiments illustrate the tradeoffs of the three mechanisms as well as the overall performance of the system on applications from physics and finance.

Original languageEnglish (US)
Title of host publicationDigest of Papers - 27th Annual International Symposium on Fault-Tolerant Computing, FTCS 1997
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages11-20
Number of pages10
ISBN (Electronic)0818678313, 9780818678318
DOIs
StatePublished - Jan 1 1997
Event27th Annual International Symposium on Fault-Tolerant Computing, FTCS 1997 - Seattle, United States
Duration: Jun 24 1997Jun 27 1997

Other

Other27th Annual International Symposium on Fault-Tolerant Computing, FTCS 1997
CountryUnited States
CitySeattle
Period6/24/976/27/97

Fingerprint

Wide area networks
Computer workstations
Finance
Processing
Local area networks
Durability
Physics
Recovery
Experiments

ASJC Scopus subject areas

  • Computer Science Applications
  • Hardware and Architecture
  • Software
  • Safety, Risk, Reliability and Quality

Cite this

Jeong, K., Shasha, D., Talla, S., & Wyckoff, P. (1997). An approach to fault-tolerant parallel processing on intermittently idle, heterogeneous workstations. In Digest of Papers - 27th Annual International Symposium on Fault-Tolerant Computing, FTCS 1997 (pp. 11-20). [614072] Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/FTCS.1997.614072

An approach to fault-tolerant parallel processing on intermittently idle, heterogeneous workstations. / Jeong, Karpjoo; Shasha, Dennis; Talla, Surendranath; Wyckoff, Peter.

Digest of Papers - 27th Annual International Symposium on Fault-Tolerant Computing, FTCS 1997. Institute of Electrical and Electronics Engineers Inc., 1997. p. 11-20 614072.

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

Jeong, K, Shasha, D, Talla, S & Wyckoff, P 1997, An approach to fault-tolerant parallel processing on intermittently idle, heterogeneous workstations. in Digest of Papers - 27th Annual International Symposium on Fault-Tolerant Computing, FTCS 1997., 614072, Institute of Electrical and Electronics Engineers Inc., pp. 11-20, 27th Annual International Symposium on Fault-Tolerant Computing, FTCS 1997, Seattle, United States, 6/24/97. https://doi.org/10.1109/FTCS.1997.614072
Jeong K, Shasha D, Talla S, Wyckoff P. An approach to fault-tolerant parallel processing on intermittently idle, heterogeneous workstations. In Digest of Papers - 27th Annual International Symposium on Fault-Tolerant Computing, FTCS 1997. Institute of Electrical and Electronics Engineers Inc. 1997. p. 11-20. 614072 https://doi.org/10.1109/FTCS.1997.614072
Jeong, Karpjoo ; Shasha, Dennis ; Talla, Surendranath ; Wyckoff, Peter. / An approach to fault-tolerant parallel processing on intermittently idle, heterogeneous workstations. Digest of Papers - 27th Annual International Symposium on Fault-Tolerant Computing, FTCS 1997. Institute of Electrical and Electronics Engineers Inc., 1997. pp. 11-20
@inproceedings{40fae05ca78f40d6ab8195cc6ce00712,
title = "An approach to fault-tolerant parallel processing on intermittently idle, heterogeneous workstations",
abstract = "We propose a novel approach to harness the idle cycles of workstations connected by LAN/WANs for long running scientific computations and present performance results for our prototype system called Persistent Linda (PLinda). PLinda offers low runtime overhead and, migration among heterogeneous architectures while retreating quickly when owners return to their workstations. PLinda achieves this by implementing a lightweight transaction model that lacks serializability and durability, but preserves a basic guarantee: if a PLinda execution terminates, it has the same result as some failure-free Linda execution. Further, by storing the state of a PLinda process as a set of core variables as of each transaction commit, a PLinda process can migrate among different architectures. Within the space of lightweight transaction models, we offer three mechanisms that make different tradeoffs between failure-free performance and recovery time. All three mechanisms may be used at the same time in a single application; each process using the mechanism which is best suited for its characteristics. Our experiments illustrate the tradeoffs of the three mechanisms as well as the overall performance of the system on applications from physics and finance.",
author = "Karpjoo Jeong and Dennis Shasha and Surendranath Talla and Peter Wyckoff",
year = "1997",
month = "1",
day = "1",
doi = "10.1109/FTCS.1997.614072",
language = "English (US)",
pages = "11--20",
booktitle = "Digest of Papers - 27th Annual International Symposium on Fault-Tolerant Computing, FTCS 1997",
publisher = "Institute of Electrical and Electronics Engineers Inc.",

}

TY - GEN

T1 - An approach to fault-tolerant parallel processing on intermittently idle, heterogeneous workstations

AU - Jeong, Karpjoo

AU - Shasha, Dennis

AU - Talla, Surendranath

AU - Wyckoff, Peter

PY - 1997/1/1

Y1 - 1997/1/1

N2 - We propose a novel approach to harness the idle cycles of workstations connected by LAN/WANs for long running scientific computations and present performance results for our prototype system called Persistent Linda (PLinda). PLinda offers low runtime overhead and, migration among heterogeneous architectures while retreating quickly when owners return to their workstations. PLinda achieves this by implementing a lightweight transaction model that lacks serializability and durability, but preserves a basic guarantee: if a PLinda execution terminates, it has the same result as some failure-free Linda execution. Further, by storing the state of a PLinda process as a set of core variables as of each transaction commit, a PLinda process can migrate among different architectures. Within the space of lightweight transaction models, we offer three mechanisms that make different tradeoffs between failure-free performance and recovery time. All three mechanisms may be used at the same time in a single application; each process using the mechanism which is best suited for its characteristics. Our experiments illustrate the tradeoffs of the three mechanisms as well as the overall performance of the system on applications from physics and finance.

AB - We propose a novel approach to harness the idle cycles of workstations connected by LAN/WANs for long running scientific computations and present performance results for our prototype system called Persistent Linda (PLinda). PLinda offers low runtime overhead and, migration among heterogeneous architectures while retreating quickly when owners return to their workstations. PLinda achieves this by implementing a lightweight transaction model that lacks serializability and durability, but preserves a basic guarantee: if a PLinda execution terminates, it has the same result as some failure-free Linda execution. Further, by storing the state of a PLinda process as a set of core variables as of each transaction commit, a PLinda process can migrate among different architectures. Within the space of lightweight transaction models, we offer three mechanisms that make different tradeoffs between failure-free performance and recovery time. All three mechanisms may be used at the same time in a single application; each process using the mechanism which is best suited for its characteristics. Our experiments illustrate the tradeoffs of the three mechanisms as well as the overall performance of the system on applications from physics and finance.

UR - http://www.scopus.com/inward/record.url?scp=26944433892&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=26944433892&partnerID=8YFLogxK

U2 - 10.1109/FTCS.1997.614072

DO - 10.1109/FTCS.1997.614072

M3 - Conference contribution

AN - SCOPUS:26944433892

SP - 11

EP - 20

BT - Digest of Papers - 27th Annual International Symposium on Fault-Tolerant Computing, FTCS 1997

PB - Institute of Electrical and Electronics Engineers Inc.

ER -