DobLIX: A Dual-Objective Learned Index for Log-Structured Merge Trees

Summary: DobLIX: a dual-objective learned index for LSM trees that trains models to minimize both index lookup error and downstream on-disk data-access cost, addressing I/O-heavy lookups when index parts live on disk. Includes an RL tuner for online adaptation, yielding 1.19×–2.21× throughput gains in RocksDB while preserving write efficiency. (summarized by gpt-5-mini on Feb 09 2026)

Paper ID: h0d0e6d7f0502a179
Venue: VLDB
Year: 2025
Pagerank: 5.0392037e-05
Overall Rank: 10,307 | 30.73%
DOI: 10.14778/3749646.3749667
PDF: Download (CC BY-NC-ND 4.0)

Incoming Non-self Citations Over Time

Authors

1. Alireza Heidari (Huawei)
2. Amirhossein Ahmadi (Huawei)
3. Wei Zhang (Huawei)

BibTeX Citation

@article{heidari_vldb25,
        title = {{DobLIX: A Dual-Objective Learned Index for Log-Structured Merge Trees}},
        author = {Heidari, Alireza and Ahmadi, Amirhossein and Zhang, Wei},
        journal = {PVLDB},
        series = {{VLDB} '25},
        volume = {18},
        number = {11},
        pages = {3965--3978},
        doi = {10.14778/3749646.3749667},
        url = {https://doi.org/10.14778/3749646.3749667},
        year = {2025}
}

Incoming Citations (Sorted by Pagerank)

Showing 1 of 1 citing papers.

Rank	Citing Paper	Year	Venue	Pagerank
10,350	Replace or Augment: Rethinking the Role of Learned Indexes in Relational Databases	2027	VLDB	4.9769913e-05

Outgoing Citations (Sorted by Pagerank)

Showing 29 of 29 cited papers.

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

Rank	Cited Paper	Year	Venue	Pagerank
40	The Case for Learned Index Structures	2018	SIGMOD	0.00046363107
422	ALEX: An Updatable Adaptive Learned Index	2020	SIGMOD	0.00018488849
458	The PGM-index: a fully-dynamic compressed learned index with provable worst-case bounds	2020	VLDB	0.00017880664
753	Dostoevsky: Better Space-Time Trade-Offs for LSM-Tree Based Key-Value Stores via Adaptive Removal of Superfluous Merging	2018	SIGMOD	0.00014232329
768	FITing-Tree: A Data-aware Index Structure	2019	SIGMOD	0.00014107655
835	Benchmarking Learned Indexes	2021	VLDB	0.00013575971
884	HoloDetect: Few-Shot Learning for Error Detection	2019	SIGMOD	0.00013263269
911	Finding Frequent Items in Data Streams	2008	VLDB	0.0001312057
1,440	LISA: A Learned Index Structure for Spatial Data	2020	SIGMOD	0.00010638444
2,445	Approximate Denial Constraints	2020	VLDB	8.4546614e-05
2,582	Are Updatable Learned Indexes Ready?	2022	VLDB	8.2641447e-05
2,861	APEX: A High-Performance Learned Index on Persistent Memory	2022	VLDB	7.9295281e-05
3,621	Learned Index: A Comprehensive Experimental Evaluation	2023	VLDB	7.1510804e-05
3,710	CARMI: A Cache-Aware Learned Index with a Cost-based Construction Algorithm	2022	VLDB	7.0775695e-05
4,366	DILI: A Distribution-Driven Learned Index	2023	VLDB	6.6330579e-05
4,635	PLIN: A Persistent Learned Index for Non-Volatile Memory with High Performance and Instant Recovery	2023	VLDB	6.4904555e-05
4,753	GRF: A Global Range Filter for LSM-Trees with Shape Encoding	2024	SIGMOD	6.4334982e-05
5,041	Dissecting, Designing, and Optimizing LSM-based Data Stores	2022	SIGMOD	6.2998284e-05
5,250	Endure: A Robust Tuning Paradigm for LSM Trees Under Workload Uncertainty	2022	VLDB	6.2092552e-05
5,310	CliffGuard: A Principled Framework for Finding Robust Database Designs	2015	SIGMOD	6.184026e-05
5,378	Updatable Learned Indexes Meet Disk-Resident DBMS - From Evaluations to Design Choices	2023	SIGMOD	6.1551053e-05
6,123	Learning to Optimize LSM-trees: Towards A Reinforcement Learning based Key-Value Store for Dynamic Workloads	2023	SIGMOD	5.8771312e-05
6,613	Making In-Memory Learned Indexes Efficient on Disk	2024	SIGMOD	5.7326443e-05
7,082	LITS: An Optimized Learned Index for Strings	2024	VLDB	5.6010119e-05
7,239	CaaS-LSM: Compaction-as-a-Service for LSM-based Key-Value Stores in Storage Disaggregated Infrastructure	2024	SIGMOD	5.5761767e-05
8,326	How to Grow an LSM-tree? Towards Bridging the Gap Between Theory and Practice	2025	SIGMOD	5.3522236e-05
8,402	The Case for Learned In-Memory Joins	2023	VLDB	5.3375308e-05
8,619	Limousine: Blending Learned and Classical Indexes to Self-Design Larger-than-Memory Cloud Storage Engines	2024	SIGMOD	5.3001199e-05
10,182	DumpKV: Learning based lifetime aware garbage collection for key value separation in LSM-tree	2025	VLDB	5.0638741e-05

Semantically Similar Papers

#	Overall Rank	Paper	Year	Venue
1	10,188	SWIX: A Memory-efficient Sliding Window Learned Index	2024	SIGMOD
2	4,366	DILI: A Distribution-Driven Learned Index	2023	VLDB
3	422	ALEX: An Updatable Adaptive Learned Index	2020	SIGMOD
4	6,123	Learning to Optimize LSM-trees: Towards A Reinforcement Learning based Key-Value Store for Dynamic Workloads	2023	SIGMOD
5	817	SlimDB: A Space-Efficient Key-Value Storage Engine For Semi-Sorted Data	2017	VLDB
6	7,593	Efficient Data Ingestion and Query Processing for LSM-Based Storage Systems	2019	VLDB
7	10,586	High Performance or Low Memory? An Updatable Learned Index Framework for Time-Space Tradeoff	2026	SIGMOD
8	11,073	LiBox: A Learned Index as an Array to Minimize Last-Mile Search	2026	VLDB
9	10,656	From Learning to Recycling: A Log-Structured Learned-Less Index	2026	SIGMOD
10	10,500	On Self-Designing Learned Indexes	2026	SIGMOD