FlashAlloc: Dedicating Flash Blocks By Objects
Jonghyeok Park, Soyee Choi, Gihwan Oh, Soojun Im, Moonwook Oh, Sang-Won Lee
摘要
For a write request, today's flash storage cannot distinguish the logical object it comes from ( e.g. , SSTables in RocksDB). In such object-oblivious flash devices, concurrent writes from different objects are simply packed in their arrival order to flash memory blocks; hence data pages from multiple objects with different lifetimes are multiplexed onto the same flash blocks. This multiplexing incurs write amplification, worsening the performance.
Tackling the multiplexing problem, we propose a novel interface for flash storage, FlashAlloc. It is used to pass the logical address ranges of objects to the underlying flash device and thus to enlighten the device to stream writes by objects. The object-aware flash storage can now de-multiplex concurrent writes from multiple objects with distinct deathtimes into per-object dedicated flash blocks. In essence, the interface enables the per-object fine-grained write streaming. Given that popular data stores tend to separate writes by logical objects, we can achieve, compared to the existing solutions, transparent streaming just by calling FlashAlloc upon object creation. Also, FlashAlloc is adaptive to workload changes, and liberates the stream conflicts in the multi-tenant environment.
Our experimental results using an open-source SSD prototype demonstrate that FlashAlloc can reduce the device-level write amplification factor (WAF) under RocksDB, F2FS, and MySQL by 1.5, 2.5, and 0.3, respectively and improve their throughput by 2.7x, 1.8x, and 1.2x, respectively. Also, FlashAlloc can mitigate the WAF interference among tenants: when running RocksDB and MySQL together on the same SSD, FlashAlloc reduced WAF from 2.5 to 1.6 and doubled their throughputs.
问问这篇 Paper
智能体会读完全文。
Lune 把这篇 Paper 索引到了最后一个公式,引用它的顶会 Paper 也一样。你提问,回答直接引用原文。
引用它的顶会 Paper3
- ScaleLFS: A Log-Structured File System with Scalable Garbage Collection for Commodity SSDsJinyong Ha, Sangjin Lee, Hyeonsang Eom, Yongseok SonFAST 2025 · 被引用 7 次
- Moving on From Group Commit: Autonomous Commit Enables High Throughput and Low Latency on NVMe SSDsLam-Duy Nguyen, Adnan Alhomssi, Tobias Ziegler, Viktor LeisSIGMOD 2025 · 被引用 3 次
- How to Write to SSDsBohyun Lee, Tobias Ziegler, Viktor LeisVLDB 2026 · 被引用 2 次
它引用的顶会 Paper5
- ZNS: Avoiding the Block Interface Tax for Flash-based SSDsMatias Bjørling, Abutalib Aghayev, Hans Holmberg, Aravind Ramesh 等USENIX ATC 2021 · 被引用 221 次
- Evolution of Development Priorities in Key-value Stores Serving Large-scale Applications: The RocksDB ExperienceSiying Dong, Andrew Kryczka, Yanqin Jin, Michael StummFAST 2021 · 被引用 110 次
- Toward a Better Understanding and Evaluation of Tree Structures on Flash SSDsDiego Didona, Nikolas Ioannou, Radu Stoica, Kornilios KourtisVLDB 2021 · 被引用 16 次
- Avoiding Read Stalls on Flash StorageMijin An, In-Yeong Song, Yong Ho Song, Sang-Won LeeSIGMOD 2022 · 被引用 10 次
- On Performance Stability in LSM-based Storage SystemsChen Luo, Michael J. CareyVLDB 2020 · 被引用 1 次
相关 Paper
- RLAlloc: A Deep Reinforcement Learning-Assisted Resource Allocation Framework for Enhanced Both I/O Throughput and QoS Performance of Multi-Streamed SSDsMengquan Li, Chao Wu, Congming Gao, Cheng Ji 等DAC 2023 · 被引用 3 次
- Nemo: A Low-Write-Amplification Cache for Tiny Objects on Log-Structured Flash DevicesXufeng Yang, Tingting Tan, Jingxin Hu, Congming Gao 等ASPLOS 2026
- ZMS: Zone Abstraction for Mobile Flash StorageJoo Young Hwang, Seokhwan Kim, Daejun Park, Yong-Gil Song 等USENIX ATC 2024 · 被引用 22 次
- Your Read is Our Priority in Flash StorageMijin An, Soojun Im, Dawoon Jung, Sang Won LeeVLDB 2022 · 被引用 9 次
- LRU-C: Parallelizing Database I/Os for Flash SSDsBo-Hyun Lee, Mijin An, Sang-Won LeeVLDB 2023 · 被引用 14 次
