Condition-based synchronization in data-centric concurrency control

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Data-Centric Concurrency Control (DCCC) shifts the focus of synchronization from control structures to data declaration. It provides a high-level mechanism that abstracts from the low-level concurrency mechanism in use, such as locks, while providing safety properties. When addressing synchronization, the concerns may be split into two categories: atomicity in the accesses to shared resources and predictable execution ordering. Like other high-level concurrency control models, DCCC only deals with the first, precluding its use in many problems that require the second. We tackle this limitation by proposing a novel high-level data-centric condition-based synchronization model, named DCCond, whose expressiveness covers most of the uses of common condition variables, while providing comparable performances.

Original languageEnglish
Title of host publicationSAC '2022
Subtitle of host publicationProceedings of the 37th ACM/SIGAPP Symposium on Applied Computing
Place of PublicationNew York
PublisherAssociation for Computing Machinery
Pages1268-1275
Number of pages8
ISBN (Print)978-145038713-2
DOIs
Publication statusPublished - Apr 2022
Event37th ACM/SIGAPP Symposium on Applied Computing, SAC 2022 - Virtual, Online
Duration: 25 Apr 202229 Apr 2022

Publication series

NameProceedings of the ACM Symposium on Applied Computing
PublisherAssociation for Computing Machinery

Conference

Conference37th ACM/SIGAPP Symposium on Applied Computing, SAC 2022
CityVirtual, Online
Period25/04/2229/04/22

Keywords

  • conditions
  • data-centric concurrency control
  • synchronization

Fingerprint

Dive into the research topics of 'Condition-based synchronization in data-centric concurrency control'. Together they form a unique fingerprint.

Cite this