CATER: a policy-based data placement framework for edge storage

Thumbnail Image
Files
CATER__III_.pdf(865.27 KB)
Accepted Version
Date
2024-04-15
Authors
Khalid, Ahmed
Ahearne, Seán
Mehta, Hemant Kumar
Roedig, Utz
Sreenan, Cormac J.
Journal Title
Journal ISSN
Volume Title
Publisher
Institute of Electrical and Electronics Engineers (IEEE)
Published Version
10.1109/PDP62718.2024.00013
Research Projects
Organizational Units
Journal Issue
Abstract
The growing heterogeneity and decentralization in the modern computing paradigm of edge-cloud continuum introduces new constraints on storage systems, such as storage type, associated processors, privacy, scarce resources, compliance, GDPR and geographical restrictions. While existing distributed data and object stores can ensure data availability and fault-tolerance, they are not flexible or dynamic enough to address these diverse set of constraints. In this paper, we introduce a modular policy-driven data placement framework, CATER, designed to seamlessly integrate with existing storage systems and overcome the aforementioned limitations. CATER formulates the data placement problem as an optimization model, incorporating data collocation and hardware constraints. We integrated a pro-totype of CATER with Apache Ozone and conducted experiments and simulations. Results show a 23% improvement in data placement while respecting 100 % of the constraints.
Description
Keywords
Data placement , Distributed storage system , Apache Ozone , GDPR , Edge computing
Citation
Khalid, A., Ahearne, S., Mehta, H.K., Roedig, U. and Sreenan, C.J. (2024) ‘Cater: a policy-based data placement framework for edge storage’, 2024 32nd Euromicro International Conference on Parallel, Distributed and Network-Based Processing (PDP). Dublin, Ireland, 20-22 March, pp. 26–34. https://doi.org/10.1109/PDP62718.2024.00013
Link to publisher’s version
https://ieeexplore.ieee.org/xpl/conhome/10495317/proceeding
URI
https://hdl.handle.net/10468/17685
Collections
Computer Science - Conference Items
Copyright
© 2024, IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.