ctFS - Replacing File Indexing with Hardware Memory Translation through Contiguous File Allocation for Persistent Memory
Offered By: USENIX via YouTube
Course Description
Overview
Explore a 14-minute conference talk from FAST '22 that introduces ctFS, a novel contiguous file system designed to optimize performance for persistent memory. Learn how ctFS eliminates the overhead associated with traditional indexing structures by representing files as contiguous regions of virtual memory, leveraging hardware MMU for efficient address translation. Discover the design challenges, implementation details, and performance benefits of ctFS, including its hierarchy code layout, partition organization, and pswap efficiency. Gain insights into how ctFS outperforms existing file systems like ext4-DAX and SplitFS in real-world workloads such as LevelDB, demonstrating significant performance improvements for persistent memory applications.
Syllabus
Introduction
Precision Memory
Indexing
Design Challenges
Hierarchy Code Layout
Partition Layout
Partition Organization
Pswap
Efficiency
Performance
Benchmarks
Taught by
USENIX
Related Courses
Stanford Seminar - The Quest for Low Storage Latency Changes EverythingStanford University via YouTube Accelerating Apache Spark Shuffle for Data Analytics on Cloud with Remote Persistent Memory Pools
Databricks via YouTube Promise and Pitfalls of Persistent Memory
Strange Loop Conference via YouTube Crimson - A New Ceph OSD for the Age of Persistent Memory and Fast NVMe Storage
USENIX via YouTube Write-Optimized Dynamic Hashing for Persistent Memory
USENIX via YouTube