Agent based simulation to evaluate adaptive caching in distributed databases

Research output: Chapter in Book/Report/Conference proceedingChapterpeer-review

1 Citation (Scopus)
208 Downloads (Pure)

Abstract

Caching frequently used data is a common practice to improve query performance in database systems. But traditional algorithms used for cache management prove to be insufficient in distributed environment where groups of users require similar or related data from multiple databases. Repeated data transfers can become a bottleneck leading to long query response time and high resource utilization. Our work focuses on adaptive algorithms to decide on optimal grain of data to be cached and cache refreshment techniques to reduce data transfers. In this paper, we present agent based simulation to investigate and in consequence improve cache management in the distributed database environment. Dynamic grain size and decisions on cache refreshment are made as a result of coordination and interaction between agents. Initial results show better response time and higher data availability compared to traditional caching techniques.

Original languageEnglish
Title of host publicationMulti-Agent Systems and Agreement Technologies
PublisherSpringer International Publishing
Pages455-462
Number of pages8
Volume9571
ISBN (Print)9783319335094
DOIs
Publication statusPublished - 17 Apr 2016
Event13th European Conference on Multi-Agent Systems, EUMAS 2015 and 3rd International Conference on Agreement Technologies, AT 2015 - Athens, Greece
Duration: 17 Dec 201518 Dec 2015

Publication series

NameLecture Notes in Computer Science
PublisherSpringer International Publishing
ISSN (Print)0302-9743

Conference

Conference13th European Conference on Multi-Agent Systems, EUMAS 2015 and 3rd International Conference on Agreement Technologies, AT 2015
Country/TerritoryGreece
CityAthens
Period17/12/201518/12/2015

Keywords

  • Agent based simulation
  • Cache management
  • Distributed databases

Fingerprint

Dive into the research topics of 'Agent based simulation to evaluate adaptive caching in distributed databases'. Together they form a unique fingerprint.

Cite this