Extended Summaries
IEEJ Transactions on Electronics Information and Systems · 2014
In production use backup system, to reduce the amount of stored data in backup storage and to shorten the backup process time are the top two priority requirements.Recently, "Deduplication" backup technology is widely used which removes redundant portion from the data to reduce the amount of backup storage.In the system, the data is divided into plural small segments which is called "chunks", then reduce duplicate chunks from the data.However, any conventional algorithm has a heavy trade-off caused by the size of chunk, that is, when smaller chunk size is implemented, it can provide more precise deduplication, therefore higher deduplication rate, but on the other hand requires big performance degradation.In case of bigger size, the situation is opposite.This paper proposes a new approach which can reduce the tradeoff, then evaluate it can provide higher deduplication rate with less performance degradation to conventional way.