Cross-Backup First: Prioritizing Historical Base Chunks for Delta Compression in Deduplicated Backup Systems
Deduplication-based backup systems often employ delta compression to further reduce storage overhead by encoding similar chunks. However, existing sketch indexing schemes, such as Greedy and SIDC, indiscriminately detect intra-backup similar chunks as base chunks for delta compression, significantly limiting compression effectiveness because these chunks exhibit substantially lower similarity than their cross-backup counterparts. To address this problem, we propose Cross-Backup First (CBF), a strategy that prioritizes cross-backup similar chunks during similarity detection. CBF comprises two implementations: Cross-Backup Preferred Selection (CPS), which distinguishes chunks by their source backup version; and Lazy Sketch Index Update (LSIU), which defers sketch index updates until backup completion to eliminate intra-backup interference. Experimental results based on real-world datasets show that CBF achieves a compression ratio up to 1.87 times that of the corresponding baseline, while incurring negligible impact on backup throughput.
Authors
- Chunguang Li (ORCID: https://orcid.org/0000-0001-6947-267X)
- Yucheng Zhang (ORCID: https://orcid.org/0009-0000-1088-6958)
- Caiyun Xie
- Junhao Liu (ORCID: https://orcid.org/0000-0003-1492-1292)
Institutions
- Nanchang University (CN)
- Nanchang Normal University (CN)
- Chengdu University (CN)
Publication Details
- Journal
- Electronics
- Published
- 2026-10-14
- DOI
- https://doi.org/10.3390/electronics15204712
- Primary Topic
- Advanced Data Storage Technologies
- Type
- article
- Field-Weighted Citation Impact
- 0.00