Author:       2014-05-27

Currently, large-scale data centers need to offer high data reliability via data backups. However, the practical applications make use of remote backups, which incur expensive network bandwidth costs. Therefore, we need to study the cost-effective and fast remote backups via exploiting data properties of network data, thus improving remote transmission.

Yu Hua from the Information Storage and Optical Display Division, Wuhan National Laboratory for Optoelectronics proposes a new network transmission scheme, called Neptune, which exploits the data redundancy. This scheme leverages fast data deduplication and efficient delta compression to improve network transmission quality. The semantic-aware grouping can narrow the searching scope, while short-cut delta chains can further improve the quality of network transmission. Extensive experimental results demonstrate that Neptune can support remote backup services for large-scale data centers.

This work is published in INFOCOM 2014 and supported in part by the NSFC under grant 61173043, National Basic Research 973 Program of China under Grant 2011CB302301, and NSFC 61025008.