ReStore: A Reinforcement Learning Approach for Data Migration in Multi-Tiered Storage
Summary: ReStore uses per-tier reinforcement-learning agents for adaptive, page-level migration, jointly modeling access patterns and SSD read/write asymmetry and parallelism. It delivers up to 6× lower runtime and 48× fewer migrations across benchmarks and traces. (summarized by gpt-5.6-luna on Jul 26 2026)
Incoming Non-self Citations Over Time
No non-self incoming citations found for this paper in this database.
Authors
- 1. Tianru Zhang (Uppsala University)
- 2. Tarikul Islam Papon (University of Massachusetts Boston)
- 3. Teona Bagashvili (Boston University)
- 4. Salman Toor (Uppsala University)
- 5. Manos Athanassoulis (Boston University)
BibTeX Citation
@inproceedings{zhang_sigmod26,
title = {{ReStore: A Reinforcement Learning Approach for Data Migration in Multi-Tiered Storage}},
author = {Zhang, Tianru and Papon, Tarikul Islam and Bagashvili, Teona and Toor, Salman and Athanassoulis, Manos},
series = {{SIGMOD} '26},
booktitle = {Proceedings of the {ACM} {SIGMOD} International Conference on Management of Data},
publisher = {Association for Computing Machinery},
doi = {10.1145/3802104},
url = {https://dl.acm.org/doi/10.1145/3802104},
year = {2026}
}
Incoming Citations (Sorted by Pagerank)
Showing 1 of 1 citing papers.
| Rank | Citing Paper | Year | Venue | Pagerank |
|---|---|---|---|---|
| 10,994 | ReStore-in-Action: Adaptive Multi-Tier Storage Management via Intelligent Data Migration Policy | 2026 | VLDB | 4.9793485e-05 |
Previous
Page 1 / 1
Next
Outgoing Citations (Sorted by Pagerank)
Showing 10 of 10 cited papers.
Citations counted here include only citations to other VLDB/SIGMOD/CIDR/PODS papers in this database.
| Rank | Cited Paper | Year | Venue | Pagerank |
|---|---|---|---|---|
| 212 | The LRU-K Page Replacement Algorithm For Database Disk Buffering | 1993 | SIGMOD | 0.00024767296 |
| 340 | THE 5 MINUTE RULE FOR TRADING MEMORY FOR DISC ACCESSES and THE 10 BYTE RULE FOR TRADING MEMORY FOR CPU TIME | 1987 | SIGMOD | 0.00020539958 |
| 580 | Efficient Pattern Matching over Event Streams | 2008 | SIGMOD | 0.00016085679 |
| 593 | 2Q: A Low Overhead High Performance Buffer Management Replacement Algorithm | 1994 | VLDB | 0.00015836342 |
| 1,460 | Online Maintenance of Very Large Random Samples on Flash Storage | 2008 | VLDB | 0.0001058438 |
| 1,524 | Trekking Through Siberia: Managing Cold Data in a Memory-Optimized Database | 2014 | VLDB | 0.00010356318 |
| 4,008 | Automating Distributed Tiered Storage Management in Cluster Computing | 2020 | VLDB | 6.8591707e-05 |
| 6,617 | CAVE: Concurrency-Aware Graph Processing on SSDs | 2024 | SIGMOD | 5.7327861e-05 |
| 10,243 | The Need for a New I/O Model | 2021 | CIDR | 5.0525742e-05 |
| 11,137 | ACE-in-Action: A Smart DBMS Bufferpool for SSDs | 2025 | SIGMOD | 4.9793485e-05 |
Previous
Page 1 / 1
Next
Semantically Similar Papers
| # | Overall Rank | Paper | Year | Venue |
|---|---|---|---|---|
| 1 | 6,557 | Optimal Splitters for Temporal and Multi-version Databases | 2013 | SIGMOD |
| 2 | 10,931 | A Unified Bandwidth Orchestration Framework for Hierarchical Data Storage Systems | 2026 | VLDB |
| 3 | 12,689 | Resiliency-Aware Data Management | 2011 | VLDB |
| 4 | 4,987 | Optimizing Databases by Learning Hidden Parameters of Solid State Drives | 2020 | VLDB |
| 5 | 2,411 | Towards Elastic Transactional Cloud Storage with Range Query Support | 2010 | VLDB |
| 6 | 10,856 | How Much Can RocksDB Chew? Achieving Near-Zero Write Stalls with Sustainable RocksDB | 2026 | VLDB |
| 7 | 5,360 | LeanStore: A High-Performance Storage Engine for NVMe SSDs | 2024 | VLDB |
| 8 | 6,124 | Learning to Optimize LSM-trees: Towards A Reinforcement Learning based Key-Value Store for Dynamic Workloads | 2023 | SIGMOD |
| 9 | 4,008 | Automating Distributed Tiered Storage Management in Cluster Computing | 2020 | VLDB |
| 10 | 10,994 | ReStore-in-Action: Adaptive Multi-Tier Storage Management via Intelligent Data Migration Policy | 2026 | VLDB |