Enabling High-Performance and Secure Userspace NVM File Systems with the Trio Architecture
Diyu Zhou, Vojtech Aschenbrenner, Tao Lyu, Jian Zhang, Sudarsun Kannan, Sanidhya Kashyap
摘要
Userspace library file systems (LibFSes) promise to unleash the performance potential of non-volatile memory (NVM) by directly accessing it and enabling unprivileged applications to customize their LibFSes to their workloads. Unfortunately, such benefits pose a significant challenge to ensuring metadata integrity. Existing works either underutilize NVM's performance or forgo critical file system security guarantees.
We present Trio, a userspace NVM file system architecture that resolves this inherent tension with a clean decoupling among file system design, access control, and metadata integrity enforcement. Our key insight is that other state (i.e., auxiliary state) in a file system can be regenerated from its "ground truth" state (i.e., core state). Thus, Trio explicitly defines the data structure of a single core state and shares it as common knowledge among its LibFSes and the trusted entity. Enabled by this, a LibFS can directly access NVM without involving the trusted entity and can be customized with its private auxiliary state. The trusted entity enforces metadata integrity by verifying the core state of a file when its write access is transferred from one LibFS to another. We design a generic POSIX-like file system called ArckFS and two customized file systems based on the Trio architecture. Our evaluation shows that ArckFS outperforms existing NVM file systems by 3.1× to 17× on LevelDB while the customized file systems further outperform ArckFS by 1.3×.
问问这篇 Paper
智能体会读完全文。
Lune 把这篇 Paper 索引到了每一个公式,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 Paper12
- ByteFS: System Support for (CXL-based) Memory-Semantic Solid-State DrivesShaobo Li, Yirui Eric Zhou, Hao Ren, Jian HuangASPLOS 2025 · 被引用 13 次
- Optimizing File Systems on Heterogeneous Memory by Integrating DRAM Cache with Virtual Memory ManagementYubo Liu, Yuxin Ren, Mingrui Liu, Hongbo Li 等FAST 2024 · 被引用 13 次
- Ethane: An Asymmetric File System for Disaggregated Persistent MemoryMiao Cai, Junru Shen, Baoliu YeUSENIX ATC 2024 · 被引用 9 次
- DJFS : Directory-Granularity Filesystem Journaling for CMM-H SSDsSeung Won Yoo, Joontaek Oh, Myeongin Cheon, Bonmoo Koo 等FAST 2025 · 被引用 8 次
- SquirrelFS: using the Rust compiler to check file-system crash consistencyHayley LeBlanc, Nathan Taylor, James Bornholt, Vijay ChidambaramOSDI 2024 · 被引用 7 次
它引用的顶会 Paper13
- An Empirical Guide to the Behavior and Use of Scalable Persistent MemoryJian Yang, Juno Kim, Morteza Hoseinzadeh, Joseph Izraelevitz 等FAST 2020 · 被引用 470 次
- FlatStore: An Efficient Log-Structured Key-Value Storage Engine for Persistent MemoryYoumin Chen, Youyou Lu, Fan Yang, Qing Wang 等ASPLOS 2020 · 被引用 166 次
- Characterizing and Modeling Non-Volatile Memory SystemsZixuan Wang, Xiao Liu, Jian Yang, Theodore Michailidis 等MICRO 2020 · 被引用 88 次
- ctFS: Replacing File Indexing with Hardware Memory Translation through Contiguous File Allocation for Persistent MemoryRuibin Li, Xiang Ren, Xu Zhao, Siwei He 等FAST 2022 · 被引用 43 次
- Maximizing Persistent Memory Bandwidth Utilization for OLAP WorkloadsBjörn Daase, Lars Jonas Bollmeier, Lawrence Benson, Tilmann RablSIGMOD 2021 · 被引用 38 次
相关 Paper
- Analyzing and Enhancing ArckFS: An Anecdotal Example of Benefits of Artifact EvaluationJonguk Jeon, Subeen Park, Sanidhya Kashyap, Sudarsun Kannan 等SOSP 2025
- Simurgh: a fully decentralized and secure NVMM user space file systemNafiseh Moti, Frederic Schimmelpfennig, Reza Salkhordeh, David Klopp 等SC 2021 · 被引用 11 次
- FlatFS: Flatten Hierarchical File System Namespace on Non-volatile MemoriesMiao Cai, Junru Shen, Bin Tang, Hao Huang 等USENIX ATC 2022 · 被引用 8 次
- MadFS: Per-File Virtualization for Userspace Persistent Memory FilesystemsShawn Zhong, Chenhao Ye, Guanzhou Hu, Suyan Qu 等FAST 2023 · 被引用 27 次
- Libnvmmio: Reconstructing Software IO Path with Failure-Atomic Memory-Mapped InterfaceJungsik Choi, Jaewan Hong, Youngjin Kwon, Hwansoo HanUSENIX ATC 2020 · 被引用 36 次
