Home // CLOUD COMPUTING 2016, The Seventh International Conference on Cloud Computing, GRIDs, and Virtualization // View article

Data Locality via Coordinated Caching for Distributed Processing

Authors:
Max Fischer
Eileen Kuehn

Keywords: Cooperative caching; Coordinated caching; Distributed caching; Batch production systems; Distributed processing

Abstract:
Modern data analysis methods often rely on data locality. Processing applications are executed directly where data is stored. Frameworks enabling this require their own specialised environment and modifications. We propose an alternative ap- proach using coordinated caching integrated into classic batch systems. A custom middleware layer provides relevant data locally on worker nodes. Most importantly no modifications are needed to add data locality to existing workflows. However, considerably more factors must be addressed by distributed caches compared to isolated ones. We investigated our approach with theoretic modelling, simulations, and a prototype implementation. Our evaluations show promising results for both applicability and performance.

Pages: 113 to 118

Copyright: Copyright (c) IARIA, 2016

Publication date: March 20, 2016

Published in: conference

ISSN: 2308-4294

ISBN: 978-1-61208-460-2

Location: Rome, Italy

Dates: from March 20, 2016 to March 24, 2016