DISTRIBUTED MULTI-VERSION OPTIMISTIC CONCURRENCY CONTROL FOR RELATIONAL DATABASES.

Divyakant Agrawal, Arthur J. Bernstein, Pankaj Gupta, Soumitra Sengupta

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

5 Citations (Scopus)

Abstract

Concurrency control algorithms have traditionally been based on locking and timestamp ordering mechanisms. Recently, optimistic schemes have been proposed. Here a distributed, multiversion, optimistic concurrency control scheme is described which is particularly advantageous in a query-dominant environment. The drawbacks of the original optimistic concurrency control scheme, namely, that inconsistent views may be seen by transactions (potentially causing unpredictable behavior) and that read-only transactions must be validated, have been eliminated in the proposed algorithm. Read-only transactions execute in a completely asynchronous fashion and are therefore processed with very little overhead. As opposed to timestamp algorithms, read-only transactions cannot cause aborts of read-write transactions. The effects of global transactions on local transaction processing are minimized. The algorithm is also free from deadlock and cascading rollback problems.

Original languageEnglish
Title of host publicationProceedings - IEEE Computer Society International Conference
EditorsAlan G. Bell
Place of PublicationNew York, NY, USA
PublisherIEEE
Pages416-421
Number of pages6
ISBN (Print)0818606924
Publication statusPublished - 1 Jan 1986
Externally publishedYes

Fingerprint

Concurrency control
Processing

ASJC Scopus subject areas

  • Engineering(all)

Cite this

Agrawal, D., Bernstein, A. J., Gupta, P., & Sengupta, S. (1986). DISTRIBUTED MULTI-VERSION OPTIMISTIC CONCURRENCY CONTROL FOR RELATIONAL DATABASES. In A. G. Bell (Ed.), Proceedings - IEEE Computer Society International Conference (pp. 416-421). New York, NY, USA: IEEE.

DISTRIBUTED MULTI-VERSION OPTIMISTIC CONCURRENCY CONTROL FOR RELATIONAL DATABASES. / Agrawal, Divyakant; Bernstein, Arthur J.; Gupta, Pankaj; Sengupta, Soumitra.

Proceedings - IEEE Computer Society International Conference. ed. / Alan G. Bell. New York, NY, USA : IEEE, 1986. p. 416-421.

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

Agrawal, D, Bernstein, AJ, Gupta, P & Sengupta, S 1986, DISTRIBUTED MULTI-VERSION OPTIMISTIC CONCURRENCY CONTROL FOR RELATIONAL DATABASES. in AG Bell (ed.), Proceedings - IEEE Computer Society International Conference. IEEE, New York, NY, USA, pp. 416-421.
Agrawal D, Bernstein AJ, Gupta P, Sengupta S. DISTRIBUTED MULTI-VERSION OPTIMISTIC CONCURRENCY CONTROL FOR RELATIONAL DATABASES. In Bell AG, editor, Proceedings - IEEE Computer Society International Conference. New York, NY, USA: IEEE. 1986. p. 416-421
Agrawal, Divyakant ; Bernstein, Arthur J. ; Gupta, Pankaj ; Sengupta, Soumitra. / DISTRIBUTED MULTI-VERSION OPTIMISTIC CONCURRENCY CONTROL FOR RELATIONAL DATABASES. Proceedings - IEEE Computer Society International Conference. editor / Alan G. Bell. New York, NY, USA : IEEE, 1986. pp. 416-421
@inproceedings{bfc9966123964422a815ac779843ca89,
title = "DISTRIBUTED MULTI-VERSION OPTIMISTIC CONCURRENCY CONTROL FOR RELATIONAL DATABASES.",
abstract = "Concurrency control algorithms have traditionally been based on locking and timestamp ordering mechanisms. Recently, optimistic schemes have been proposed. Here a distributed, multiversion, optimistic concurrency control scheme is described which is particularly advantageous in a query-dominant environment. The drawbacks of the original optimistic concurrency control scheme, namely, that inconsistent views may be seen by transactions (potentially causing unpredictable behavior) and that read-only transactions must be validated, have been eliminated in the proposed algorithm. Read-only transactions execute in a completely asynchronous fashion and are therefore processed with very little overhead. As opposed to timestamp algorithms, read-only transactions cannot cause aborts of read-write transactions. The effects of global transactions on local transaction processing are minimized. The algorithm is also free from deadlock and cascading rollback problems.",
author = "Divyakant Agrawal and Bernstein, {Arthur J.} and Pankaj Gupta and Soumitra Sengupta",
year = "1986",
month = "1",
day = "1",
language = "English",
isbn = "0818606924",
pages = "416--421",
editor = "Bell, {Alan G.}",
booktitle = "Proceedings - IEEE Computer Society International Conference",
publisher = "IEEE",

}

TY - GEN

T1 - DISTRIBUTED MULTI-VERSION OPTIMISTIC CONCURRENCY CONTROL FOR RELATIONAL DATABASES.

AU - Agrawal, Divyakant

AU - Bernstein, Arthur J.

AU - Gupta, Pankaj

AU - Sengupta, Soumitra

PY - 1986/1/1

Y1 - 1986/1/1

N2 - Concurrency control algorithms have traditionally been based on locking and timestamp ordering mechanisms. Recently, optimistic schemes have been proposed. Here a distributed, multiversion, optimistic concurrency control scheme is described which is particularly advantageous in a query-dominant environment. The drawbacks of the original optimistic concurrency control scheme, namely, that inconsistent views may be seen by transactions (potentially causing unpredictable behavior) and that read-only transactions must be validated, have been eliminated in the proposed algorithm. Read-only transactions execute in a completely asynchronous fashion and are therefore processed with very little overhead. As opposed to timestamp algorithms, read-only transactions cannot cause aborts of read-write transactions. The effects of global transactions on local transaction processing are minimized. The algorithm is also free from deadlock and cascading rollback problems.

AB - Concurrency control algorithms have traditionally been based on locking and timestamp ordering mechanisms. Recently, optimistic schemes have been proposed. Here a distributed, multiversion, optimistic concurrency control scheme is described which is particularly advantageous in a query-dominant environment. The drawbacks of the original optimistic concurrency control scheme, namely, that inconsistent views may be seen by transactions (potentially causing unpredictable behavior) and that read-only transactions must be validated, have been eliminated in the proposed algorithm. Read-only transactions execute in a completely asynchronous fashion and are therefore processed with very little overhead. As opposed to timestamp algorithms, read-only transactions cannot cause aborts of read-write transactions. The effects of global transactions on local transaction processing are minimized. The algorithm is also free from deadlock and cascading rollback problems.

UR - http://www.scopus.com/inward/record.url?scp=0022564909&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0022564909&partnerID=8YFLogxK

M3 - Conference contribution

SN - 0818606924

SP - 416

EP - 421

BT - Proceedings - IEEE Computer Society International Conference

A2 - Bell, Alan G.

PB - IEEE

CY - New York, NY, USA

ER -