SIMILARITY AND LOCATION AWARE SCALABLE DEDUPLICATION SYSTEM FOR STORAGE SYSTEMS

Abstract

Big data is extensively considered as potentially the coming dominant technology in IT assiduity. It offers simplified system conservation and scalable resource operation with storehouse systems. As a abecedarian technology of cloud computing, storehouse has been a hot exploration content in recent times. The high outflow of virtualization has been well addressed by tackle advancement in CPU assiduity, and by software perpetration enhancement in hypervisors themselves. still, the high demand on storehouse image storehouse remains a grueling problem. Being systems have made sweats to reduce storehouse image storehouse consumption by means of deduplication within a storehouse area network system. nonetheless, storehouse area network can not satisfy the adding demand of large- scale storehouse hosting for cloud computing because of its cost limitation. In this design, we propose SILO, a scalable deduplication train system that has been particularly designed for large- scale storehouse deployment. Its design provides fast storehouse deployment with similarity and position grounded point indicator for data transfer and low storehouse consumption by means of deduplication on storehouse images. It also provides a comprehensive set of storehouse features including instant cloning for storehouse images, on- demand costing through a network, and caching with original disks by dupe- on- read ways. Trials show that SILO features perform well and introduce minor performance outflow.