Cost- and Latency-Efficient Redundant Data Storage in the Cloud

Publikationstyp
Conference Paper
Date Issued
2017-11
Sprache
English
Author(s)
Matt, Johannes  
Waibel, Philipp  
Schulte, Stefan  
TORE-URI
http://hdl.handle.net/11420/11821
Volume
2017-January
Start Page
164
End Page
172
Citation
IEEE 10th International Conference on Service-Oriented Computing and Applications (SOCA 2017)
Contribution to Conference
IEEE 10th International Conference on Service-Oriented Computing and Applications, SOCA 2017  
Publisher DOI
10.1109/SOCA.2017.30
Scopus ID
2-s2.0-85046933122
With the steady increase of offered cloud storage services, they became a popular alternative to local storage systems. Beside several benefits, the usage of cloud storage services can offer, they have also some downsides like potential vendor lock-in or unavailability. Different pricing models, storage technologies and changing storage requirements are further complicating the selection of the best fitting storage solution. In this work, we present a heuristic optimization approach that optimizes the placement of data on cloud-based storage services in a redundant, cost- and latency-efficient way while considering user-defined Quality of Service requirements. The presented approach uses monitored data access patterns to find the best fitting storage solution. Through extensive evaluations, we show that our approach saves up to 30% of the storage cost and reduces the upload and download times by up to 48% and 69% in comparison to a baseline that follows a state-of-the-art approach.
Subjects
Cloud-based storage
Cost-efficient
Data placement optimization
Latency-efficient
Long-term storage