The Optimal ChunkSize Pair Choosing Method for Dual Layered Deduplication Backup System

Abstract

This paper proposes a multiple layers deduplication system for backup process in IT environment. The proposed system eliminates the duplication in target data by applying a series of plural ChunkSizes. The effectiveness of proposed system is evaluated by a simulator and a prototype machine on a linux server, comparing to a conventional single chunk system.The system realizes better deduplication capability with much less number of duplicated chunk reduction. This paper then proposes a choosing method to choose the optimal ChunkSizes to be set in layers. It maximizes the deduplication capability and minimizes the performance degradation caused by chunk reductions. © Springer International Publishing Switzerland 2014.