DBScholar

Back to papers

Constructing and Analyzing the LSM Compaction Design Space

Summary: Formalizes the LSM-compaction design space via four primitives—trigger, data layout, granularity, and data movement policy—enabling synthesis of existing and novel strategies. Empirically evaluates 10 strategies, reports 12 observations and 7 takeaways to help DB researchers navigate tradeoffs in write/read amplification and space for LSM engines. (summarized by gpt-5-nano on Feb 09 2026)

Paper ID
h7b6b2562ee0ee159
Venue
VLDB
Year
2021
Pagerank
8.0073289e-05
Overall Rank
2,780 | 81.37%
DOI
10.14778/3476249.3476274
PDF
Download (CC BY-NC-ND 4.0)

Incoming Non-self Citations Over Time

Authors

BibTeX Citation

@article{sarkar_vldb21,
        title = {{Constructing and Analyzing the LSM Compaction Design Space}},
        author = {Sarkar, Subhadeep and Staratzis, Dimitris and Zhu, Zichen and Athanassoulis, Manos},
        journal = {PVLDB},
        series = {{VLDB} '21},
        volume = {14},
        number = {11},
        pages = {2216--2229},
        doi = {10.14778/3476249.3476274},
        url = {https://doi.org/10.14778/3476249.3476274},
        year = {2021}
}

Incoming Citations (Sorted by Pagerank)

Showing 27 of 27 citing papers.

Rank Citing Paper Year Venue Pagerank
2,882 Spooky: Granulating LSM-Tree Compactions Correctly 2022 VLDB 7.894356e-05
3,489 Cosine: A Cloud-Cost Optimized Self-Designing Key-Value Storage Engine 2022 VLDB 7.2421958e-05
3,692 SplinterDB and Maplets: Improving the Tradeoffs in Key-Value Store Compaction Policy 2023 SIGMOD 7.0761873e-05
4,771 GRF: A Global Range Filter for LSM-Trees with Shape Encoding 2024 SIGMOD 6.412225e-05
5,065 Dissecting, Designing, and Optimizing LSM-based Data Stores 2022 SIGMOD 6.2789973e-05
5,272 Endure: A Robust Tuning Paradigm for LSM Trees Under Workload Uncertainty 2022 VLDB 6.1887235e-05
5,894 Compactionary: A Dictionary for LSM Compactions 2022 SIGMOD 5.9390155e-05
6,137 Learning to Optimize LSM-trees: Towards A Reinforcement Learning based Key-Value Store for Dynamic Workloads 2023 SIGMOD 5.8576977e-05
7,925 CAMAL: Optimizing LSM-trees via Active Learning 2024 SIGMOD 5.4086685e-05
8,245 ArceKV: Towards Workload-driven LSM-compactions for Key-Value Store Under Dynamic Workloads 2026 VLDB 5.3556365e-05
8,638 Limousine: Blending Learned and Classical Indexes to Self-Design Larger-than-Memory Cloud Storage Engines 2024 SIGMOD 5.2825944e-05
8,666 Mnemosyne: Dynamic Workload-Aware BF Tuning via Accurate Statistics in LSM trees 2025 SIGMOD 5.277223e-05
9,031 ByteCard: Enhancing ByteDance’s Data Warehouse with Learned Cardinality Estimation 2024 SIGMOD 5.219814e-05
9,050 AutoComp: Automated Data Compaction for Log-Structured Tables in Data Lakes 2025 SIGMOD 5.2152045e-05
9,085 BACH: Bridging Adjacency List and CSR Format using LSM-Trees for HGTAP Workloads 2025 VLDB 5.2097768e-05
9,111 MirrorKV: An Efficient Key-Value Store on Hybrid Cloud Storage with Balanced Performance of Compaction and Querying 2023 SIGMOD 5.208561e-05
9,162 Rethinking The Compaction Policies in LSM-trees 2025 SIGMOD 5.2035666e-05
9,725 FluidKV: Seamlessly Bridging the Gap between Indexing Performance and Memory-Footprint on Ultra-Fast Storage 2024 VLDB 5.1210873e-05
9,747 Disco: A Compact Index for LSM-trees 2025 SIGMOD 5.1177772e-05
10,272 MaLT: A Framework for Managing Large Transactions in OceanBase 2025 SIGMOD 5.0356939e-05
10,724 Improving Range Scan Performance in LSM-trees with Group Caching 2026 SIGMOD 4.9605343e-05
10,796 How to Write to SSDs 2026 VLDB 4.9605343e-05
10,834 Swan: Hybrid MVCC Management for Efficient Transaction Processing in LSM-Tree-Based Key-Value Stores 2026 VLDB 4.9605343e-05
10,891 Aker: Density-Aware Approximate Caching for Vector Search 2026 VLDB 4.9605343e-05
11,070 TeXBench: A Unified Benchmarking Suite for Shifting Workloads 2026 VLDB 4.9605343e-05
11,378 Meaningful Data Erasure in the Presence of Dependencies 2025 VLDB 4.9605343e-05
11,640 On Reducing Space Amplification with Multi-Column Compaction in Apache IoTDB 2024 VLDB 4.9605343e-05
Previous Page 1 / 1 Next

Outgoing Citations (Sorted by Pagerank)

Showing 15 of 15 cited papers.

Citations counted here include only citations to other VLDB/SIGMOD/CIDR/PODS papers in this database.

Previous Page 1 / 1 Next

Semantically Similar Papers