TY - GEN
T1 - A novel block-level continuous data protection system
AU - Deng, Yibing
AU - Chen, Shanguang
AU - Hu, Wei
AU - Gao, Feng
AU - Liu, Chuanyi
PY - 2010
Y1 - 2010
N2 - Continuous Data Protection (CDP) techniques have received great interest both from academia and industry in recent years. Aiming at the challenges to design and implement a CDP system, this paper proposes GSP-BCDP: a novel continuous data protection architecture at block level. Compared with other approaches, it uses a global shared storage pool to actually store all the versions of data contents, and each version of each logical volume bonded with a restore index table. The recovery of GSP-BCDP is per volume basis, rather than per block basis in other schemes, so GSP-BCDP can provide a much faster version restore than others. Furthermore, through global data de-duplication techniques, GSP-BCDP can not only reduce the duplications along the history of the same data stream, but also de-duplicate the redundancies among data streams or identities between different users. Experimental results demonstrate that GSP-BCDP can improve the recovery time up to more than 10 times, and with storage space compression ratio of 1.2 times in average.
AB - Continuous Data Protection (CDP) techniques have received great interest both from academia and industry in recent years. Aiming at the challenges to design and implement a CDP system, this paper proposes GSP-BCDP: a novel continuous data protection architecture at block level. Compared with other approaches, it uses a global shared storage pool to actually store all the versions of data contents, and each version of each logical volume bonded with a restore index table. The recovery of GSP-BCDP is per volume basis, rather than per block basis in other schemes, so GSP-BCDP can provide a much faster version restore than others. Furthermore, through global data de-duplication techniques, GSP-BCDP can not only reduce the duplications along the history of the same data stream, but also de-duplicate the redundancies among data streams or identities between different users. Experimental results demonstrate that GSP-BCDP can improve the recovery time up to more than 10 times, and with storage space compression ratio of 1.2 times in average.
UR - https://www.scopus.com/pages/publications/77957787540
M3 - 会议稿件
AN - SCOPUS:77957787540
SN - 9788988678183
T3 - NISS2010 - 4th International Conference on New Trends in Information Science and Service Science
SP - 242
EP - 247
BT - NISS2010 - 4th International Conference on New Trends in Information Science and Service Science
T2 - 4th International Conference on New Trends in Information Science and Service Science, NISS2010
Y2 - 11 May 2010 through 13 May 2010
ER -