Enhancing Data Recovery in Deduplication Backup Systems: A Novel Rewriting Algorithm to Minimize Fragmentation Effects

Abstract

Deduplication technology detects and eliminates redundant data, significantly conserving disk space, thus finding extensive applications across various domains. Nevertheless, the removal of numerous duplicate data blocks leads to data fragmentation, adversely affecting data recovery performance. To address fragmentation issues, rewriting algorithms have been introduced. However, current rewriting algorithms fail to precisely identify fragmented blocks. To resolve this issue, we propose a new rewriting algorithm called MERW. MERW takes into account fragmentation information on a broader scale and selects the most optimal fragmented blocks. Consequently, MERW effectively mitigates data fragmentation by accurately rewriting fragmented blocks. Experimental results demonstrate that MERW enhances recovery performance by 48% and improves the deduplication ratio by 6%.