A Guardian Angel of File Systems: Monitoring for File System Errors in Da...

on YouTube ↗29m

" summary

Adds a kernel push notification called FAN_FS_ERROR that alerts user-space daemons the instant a filesystem detects corruption, sending fsid, a file handle for the affected inode, an error code and a first-error count. Built by extending fanotify and validated on ext4, merged in Linux 5.16 and used in production for a customer, the tracer demo prints structured events so monitoring systems can act immediately instead of parsing noisy dmesg or polling sysfs; ports to other filesystems and richer metadata are planned.

AI-written summary · about

" chapters (5)

1. 0:00Start
2. 2:18File system errors - Overview
3. 7:04Kernel log - No man's land
4. 18:38Documentation
5. 24:56Future Work

" related talks

• 19mEvaluating File System Reliability on Solid State DrivesBianca Schroeder, Shehbaz Jaffer, Stathis Maneas, Andy Hwang · USENIX ATC '19
• 23mCan Applications Recover from fsync Failures?Ramnatthan Alagappan, Andrea C. Arpaci-Dusseau, Remzi H. Arpaci-Dusseau, Anthony Rebello, Yuvraj Patel · USENIX ATC '20
• 19mConfD: Analyzing Configuration Dependencies of File Systems for Fun and ProfitMai Zheng, Duo Zhang, Om Rameshwar Gatla, Tabassum Mahmud, Carson Love, Ryan Bumann · FAST '23: 21st USENIX Conference on File and Storage Technologies
• 42mCan File Systems Survive in Data-centric World?Viacheslav Dubeyko · Open Source Summit North America 2025
• 08mReal-time monitoring GlusterFS using Events APIsAravinda VK · DevConf 2017
• 42mProduction FS: Adapt or DieClaudia Beresford, Tiago Scolari · Paris Container Day 2017
• 01h 12mRedundancy does not imply fault toleranceAishwarya Ganesan · Hydra 2020
• 19mFlatFS: Flatten Hierarchical File System Namespace on Non-volatile MemoriesMiao Cai, Junru Shen, Baoliu Ye, Hao Huang, Bin Tang · USENIX ATC '22