FAST '22
as the organisers put it
20th USENIX Conference on File and Storage Technologies (FAST '22), February 22–24, 2022, Hyatt Regency Santa Clara, Santa Clara, CA, USA. FAST brings together storage-system researchers and practitioners to explore new directions in the design, implementation, evaluation, and deployment of storage systems.
| title | author | length |
|---|---|---|
| 25 Years of Storage Research and Education: A Retrospective | Remzi H. Arpaci-Dusseau | 56m |
A career-long claim is that specific measurements plus simple system changes yield huge wins: a cluster-built sorter (NOWsort) hits ~6 GB/min versus the prior… | ||
| NyxCache: Flexible and Efficient Multi-tenant Persistent Memory Caching | Kan Wu, Kaiwei Tu, Yuvraj Patel, Rathijit Sen, Kwanghyun Park, Andrea C. Arpaci-Dusseau, Remzi H. Arpaci-Dusseau | 16m |
NyxCache is an access-regulation framework for multi-tenant persistent-memory (PM) caches that implements runtime microcontrolled experiments to identify which… | ||
| HTMFS: Strong Consistency Comes for Free with Hardware Transactional Memory... | Jifei Yi, Mingkai Dong, Fangnuo Wu, Haibo Chen | 15m |
Claims that hardware transactional memory plus a small runtime gives strong per-request crash atomicity while staying fast. The system implements HOP: split… | ||
| ctFS: Replacing File Indexing with Hardware Memory Translation through Contiguous File... | Ruibin Li, Xiang (Jenny) Ren, Xu Zhao, Siwei He, Michael Stumm, Ding Yuan | 14m |
ctFS replaces software extent trees by allocating every file as a contiguous virtual region, storing only the start virtual address in the inode and relying on… | ||
| FORD: Fast One-sided RDMA-based Distributed Transactions for Disaggregated Persistent... | Ming Zhang, Yu Hua, Pengfei Zuo, Lurong Liu | 16m |
FORD uses one-sided RDMA to bypass remote CPUs and makes three concrete protocol changes: hitchhiker locking that batches read+lock into one RDMA to cut… | ||
| Closing the B+-tree vs. LSM-tree Write Amplification Gap on Modern Storage Hardware... | Yifan Qiao, Xubin Chen, Ning Zheng, Jiangpeng Li, Yang Liu, Tong Zhang | 13m |
Shows three concrete changes — deterministic page shadowing, localized modification logging, and sparse redo logging — applied to a B+-tree implementation and… | ||
| TVStore: Automatically Bounding Time Series Storage via Time-Varying Compression | Yanzhe An, Yue Su, Yuqing Zhu, Jianmin Wang | 16m |
FAST '22 - TVStore: Automatically Bounding Time Series Storage via Time-Varying Compression
Yanzhe An, Tsinghua University; Yue Su, Huawei Technologies Co… | ||
| Removing Double-Logging with Passive Data Persistence in LSM-tree based Relational... | Kecheng Huang, Zhaoyan Shen, Zhiping Jia, Zili Shao, Feng Chen | 15m |
Proposes a concrete redesign that removes WAL from storage engine in LSM-tree RDBs, implemented as a MyRocks prototype and evaluated with Linkbench and TPC-C… | ||
| Improving the Reliability of Next Generation SSDs using WOM-v Codes | Shehbaz Jaffer, Kaveh Mahdaviani, Bianca Schroeder | 14m |
Implements non-binary WOM-v codes in the Linux lightNVM/FEMU stack (700 LOC) and adds a QLC extension; evaluates on 800+ real traces and a FEMU-based testbed… | ||
| GuardedErase: Extending SSD Lifetimes by Protecting Weak Wordlines | Duwon Hong, Myungsuk Kim, Geonhee Cho, Dusol Lee, Jihong Kim | 14m |
Proposes GuardedErase, a per‑wordline erase‑voltage control that slows wear on weak wordlines so blocks reach their true endurance instead of failing early… | ||
| Hardware/Software Co-Programmable Framework for Computational SSDs to Accelerate Deep... | Miryeong Kwon, Donghyun Gouk, Sangwon Lee, Myoungsoo Jung | 13m |
HolisticGNN implements a FPGA-based computational SSD that stores graphs as neighbor-oriented adjacency lists, exposes RPC over PCIe and runs programmable GNN… | ||
| Operational Characteristics of SSDs in Enterprise Storage Systems: A Large-Scale Field... | Stathis Maneas, Kaveh Mahdaviani, Tim Emami, Bianca Schroeder | 14m |
Study uses NetApp field telemetry covering almost 2 million SSDs over a decade and shows most drives consume <1% of their PE cycles per year so 86% survive… | ||
| Hydra : Resilient and Highly Available Remote Memory | Youngmoon Lee, Hasan Al Maruf, Mosharaf Chowdhury, Asaf Cidon, Kang G. Shin | 15m |
Hydra is a remote-memory system that encodes 4 KB pages on-the-fly and serves reads/writes with single-digit μs latency by asynchronous encode/write, k+r… | ||
| MT^2: Memory Bandwidth Regulation on Hybrid NVM/DRAM Platforms | Jifei Yi, Benchao Dong, Mingkai Dong, Ruizhe Tong, Haibo Chen | 15m |
MT^2 detects noisy neighbors on hybrid NVM/DRAM servers by measures four bandwidth types via PMU plus application reports, uses memory-access uses latency to… | ||
| Aurogon: Taming Aborts in All Phases for Distributed In-Memory Transactions | Tianyang Jiang, Guangyan Zhang, Zhiyue Li, Weimin Zheng | 15m |
Distributed in-memory system Aurogon implements 2LClock (two-layer RDMA-aware clock sync), adaptive request deferral for hotspot reads, and pre-attached writes… | ||
| DedupSearch: Two-Phase Deduplication Aware Keyword Search | Nadav Elias, Philip Shilane, Sarai Sheinvald, Gala Yadgar | 13m |
DedupSearch scans each unique chunk once in a two-phase: physical then logical pipeline, recording exact matches keyed by chunk fingerprint and saving boundary… | ||
| DeepSketch: A New Machine Learning-Based Reference Search Technique... | Jisung Park, Jeonggyun Kim, Yeseong Kim, Sungjin Lee, Onur Mutlu | 15m |
A machine-learned reference search called DeepSketch extracts binary sketches via a small DNN (three conv + two FC layers) and uses clustering plus transfer… | ||
| The what, The from, and The to: The Migration Games in Deduplicated Systems | Roei Kisous, Ariel Kolikant, Abhinav Duggal, Sarai Sheinvald, Gala Yadgar | 13m |
Frames deduplicated migration as an optimization to minimize total stored blocks subject to per-volume load balance and a traffic budget, and evaluates three… | ||
| DUPEFS: Leaking Data Over the Network With Filesystem Deduplication Side Channels | Andrei Bacs, Saidgani Musaev, Kaveh Razavi, Cristiano Giuffrida, Herbert Bos | 15m |
Shows a practical remote channel against ZFS inline deduplication that leaks OAuth access tokens from an nginx access log by exploiting timing differences… | ||
| Test of Time Award: Analysis of Data Corruption in the Storage Stack | Lakshmi N. Bairavasundaram, Garth Goodson, Bianca Schroeder, Andrea C. Arpaci-Dusseau, Remzi H. Arpaci-Dusseau | 09m |
Analysis of telemetry covers 1.5 million drives and records 400,000+ corruptions detected as checksum mismatches. Nearline drives show an order-of-magnitude… | ||
| FusionFS: Fusing I/O Operations using CISCOps in Firmware File Systems | Jian Zhang, Yujie Ren, Sudarsun Kannan | 17m |
FusionFS is a firmware-level in-storage filesystem that fuses multiple I/O and data-processing steps into single offloaded operations called CISCOps, reducing… | ||
| InfiniFS: An Efficient Metadata Service for Large-Scale Distributed Filesystems | Wenhao Lv, Youyou Lu, Yiming Zhang, Peile Duan, Jiwu Shu | 15m |
InfiniFS implements access-content decoupling to partition directory metadata for both locality and load balance, a speculative path resolution that predicts… | ||
| ScaleXFS: Getting scalability of XFS back on the ring | Dohyun Kim, Kwangwon Min, Joontaek Oh, Youjip Won | 14m |
ScaleXFS is a modified XFS that removes journaling bottlenecks by adding Double Committed Item List, Per-core In-memory Logging, and Strided Space Counting… | ||
| exF2FS: Transaction Support in Log-Structured Filesystem | Joontaek Oh, Sion Ji, Yongjin Kim, Youjip Won | 16m |
exF2FS is a log‑structured F2FS variant that implements file‑spanning atomic updates by exposing Membership-Oriented Transaction, allowing the app to list… | ||
| Generating Realistic Wear Distributions for SSDs (Work-in-Progress Report) | Ziyang Jiao, Bryan S. Kim | 05m |
A machine-learning aging framework samples real workloads in a simulator, builds lightweight per-block regression models from observed internal SSD activity… | ||
| O-AFA:Order Preserving All Flash Array for Journaling Filesystem (Work-in-Progress Report) | Seung Won Yoo, Joontaek Oh, Youjip Won | 05m |
Prototype order-preserving all-flash array firmware that tracks barrier flags per-LBA and groups writes into hash barrier stripes and cache barrier stripes… | ||
| Exploring Data Deduplication in LSM Tree-based Key-Value Stores (Work-in-Progress Report) | Safdar Jamil, Awais Khan, Youngjae Kim | 05m |
Explores adding value-based deduplication to LSM trees, showing inline duplication (fingerprinting on writes with a chunk information table) causes severe… | ||
| A Combinatorial Writing Scheme for High Throughput DNA-based... (Work-in-Progress Report) | Swapnil Bhatia, Kevin Gildea | 08m |
A combinatorial write method maps a lexicographically ordered tree of DNA components to bit positions and encodes ones by assembling the corresponding… | ||
| Overcoming Capacitance Constraints in SSD (Work-in-Progress Report) | Sanghyun Nam, Seungmin Shin, Sungjin Lee, Bryan S. Kim, Eunji Lee | 06m |
Proposes a divided-architecture SSD that avoids full capacitor-backed PRP by protecting only 1% of mapping entries and forcing cold mapping pages to land in… | ||
| Understanding Configuration Issues in Storage Systems (Work-in-Progress Report) | Tabassum Mahmud, Mai Zheng | 06m |
Finds that many storage bugs require satisfying multi-level configuration dependencies across components and backs that claim with 70 bug patches collected… | ||
| Early experience of Utilizing Persistent Memory for Database... (Work-in-Progress Report) | Tsuyoshi Ozawa, Ryoji Kawamichi, Yuto Hayamizu, Kazuo Goda, Masaru Kitsuregawa | 06m |
Compares using persistent memory vs DRAM for bulk loading TPCH lineitem on a homegrown database engine, measuring four configs (DRAM/PMem as sort buffer and… | ||
| A Log-Structured Merge Tree-aware Message Authentication Scheme for Persistent... | Igjae Kim, J. Hyun Kim, Minu Chung, Hyungon Moon, Sam H. Noh | 13m |
TWEEZER extends RocksDB and implements per-SST authentication keys plus a hash chain to avoid slow hardware counters, reproduces Speicher for comparison, and… | ||
| Practicably Boosting the Processing Performance of BFS-like Algorithms... | Tsun-Yu Yang, Yuhong Liang, Ming-Chang Yang | ? |
Builds an ordering called IOOrder that combines a BFS-from-high-degree reordering (BFDS ordering), an exponential-size binning layout (auto-beam) and a cheap… | ||
| DEPART: Replica Decoupling for Distributed Key-Value Storage | Qiang Zhang, Yongkun Li, Patrick P. C. Lee, Yinlong Xu, Si Wu | 16m |
DEPART implements replica decoupling by storing primaries in an LSM and redundant copies in a separate two-layer log with tunable ordering, implemented as a… | ||
| PAIO: General, Portable I/O Optimizations With Minor Application Modifications | Ricardo Macedo, Yusuke Tanimura, Jason Haga, Vijay Chidambaram, Jose Pereira, João Paulo | 16m |
PAIO is a user‑level library that inserts data plane stages between layers and a centralized control plane to enforce storage policies by propagating… | ||
| Separating Data via Block Invalidation Time Inference for Write Amplification Reduction... | Qiuping Wang, Jinhong Li, Patrick P. C. Lee, Tao Ouyang, Chao Shi, Lilong Huang | 15m |
Claims an algorithm, SepBIT, that infers each block’s block-invalidation time (BIT) from workload traces and groups blocks with similar estimated BITs to cut… | ||
| CacheSifter: Sifting Cache Files for Boosted Mobile Performance and Lifetime | Yu Liang, Riwei Pan, Tianyu Ren, Yufei Cui, Rachata Ausavarungnirun, Xianzhang Chen, Changlong Li, Tei-Wei Kuo | 16m |
FAST '22 - CacheSifter: Sifting Cache Files for Boosted Mobile Performance and Lifetime
Yu Liang, Department of Computer Science, City University of Hong Kong… | ||
| nothing matches | ||