DocumentCode :
1680029
Title :
Hybrid caching for cloud storage to support traditional application
Author :
Lin Cao ; Lifu Huang ; Kai Lei ; Zhiming Zhang ; Lian-en Huang
Author_Institution :
Shenzhen Key Lab. for Cloud Comput. Technol. & Applic. (SPCCTA), Peking Univ., Shenzhen, China
fYear :
2012
Firstpage :
11
Lastpage :
15
Abstract :
To take better advantage of cloud storage, it is necessary to migrate traditional application to the cloud. In existing cloud storage platforms, files are divided into chunks and redundantly stored, but in such circumstances, random file write may lead to much delay. It is quite common to access file randomly in traditional applications, therefore the performance of traditional application will be reduced after being migrated to the cloud. In order to solve the problem, we propose the idea of adding client-side caching to augment the cloud storage. In this way, most data access operations are executed on client-side rather than in cloud. File write delay and the cloud load will be reduced consequently. With this point in mind, we designed and implemented LabCloud. Similar with other block/file cloud storages, the system is designed to be chunk layout, but a hybrid caching file system CFS is complemented on the client-side. CFS could be mounted on client-side file system, and is compatible with POSIX file system interface, thus traditional application deployed in local host can be configured and running in it without modification. In this paper we describe the design and implementation of LabCloud and evaluate the system by test suites and experiments. Results demonstrate that it meets the requirement of most traditional applications, and the performance of random file access in cloud is improved.
Keywords :
cache storage; cloud computing; CFS; LabCloud; POSIX file system interface; block-file cloud storages; client-side caching; cloud storage platforms; data access operations; hybrid caching file system; random file access; random file write; POSIX; caching; cloud storage; file system;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Cloud Computing Congress (APCloudCC), 2012 IEEE Asia Pacific
Conference_Location :
Shenzhen
Print_ISBN :
978-1-4673-2903-3
Electronic_ISBN :
978-1-4673-2902-6
Type :
conf
DOI :
10.1109/APCloudCC.2012.6486503
Filename :
6486503
Link To Document :
بازگشت