DocumentCode :
2627774
Title :
iOSDC : A Novel Autonomous Intelligent OSD Cluster
Author :
Yue, Yinliang ; Wang, Fang ; Feng, Dan
Author_Institution :
Huazhong Univ. of Sci. & Technol., Wuhan
fYear :
2007
fDate :
21-23 Nov. 2007
Firstpage :
1547
Lastpage :
1552
Abstract :
As the requirement for Peta-scale massive storage system, Object-based Storage Device (OSD) protocol has already been the mainstream in the next generation storage technology. However, in recent research, most actions in OSD remain the passive way, because of the request-driven model. As partial storage management functions are offloaded to OSD, the construction of an intelligent OSD has caught the fancy of most researchers. In this paper, we propose a novel autonomous intelligent OSD Cluster (iOSDC). In iOSDC, every OSD is intelligent, called intelligent OSD (iOSD) and all the objects are intelligent, called intelligent OBJECT (iOBJECT). Methods, rules, trigger and feed back mechanism are introduced into object defined in OSD-2 draft, so an object is comprised of data, attributes, methods, rules, trigger and feedback mechanism. After the relationship among these six components is built up, the object is no longer passive, but active. Furthermore, some specific objects are introduced into OSDs, these specific objects can enhance the intelligence of OSD. Collaboration of all the iOSDs gives birth to the intelligent OSD Cluster. Based on the requirements of iOSDC, some enhancements have been performed onto the T10 OSD- 2 draft.
Keywords :
knowledge based systems; object-oriented programming; storage management; Peta-scale massive storage system; autonomous intelligent OSD cluster; iOSDC; intelligent OBJECT; object-based storage device protocol; partial storage management; request-driven model; Computer science; Computer science education; Data storage systems; Educational technology; Feedback; Feeds; Humans; Information technology; Laboratories; Protocols;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Convergence Information Technology, 2007. International Conference on
Conference_Location :
Gyeongju
Print_ISBN :
0-7695-3038-9
Type :
conf
DOI :
10.1109/ICCIT.2007.115
Filename :
4420474
Link To Document :
بازگشت