ZNS+: Advanced Zoned Namespace Interface for Supporting In-Storage Zone Compaction
Talkat 15th USENIX Symposium on Operating Systems Design and Implementation (OSDI '21) · 2021-07-14..07-16·
summary
ZNS+ adds internal zone compaction and permits sparse sequential writes so the host offloads data-copy for segment reclamation to the SSD, implements copyback-aware block allocation and a hybrid segment-recycling policy, and runs on a QEMU-based SSD emulator and a Cosmos+ open SSD with F2FS 4.10. Evaluation shows 1.33–2.91× file-system speedup, ~48% less metadata write, over 86% of reclaimed segments handled by threaded logging, and the prototype and code are on GitHub.
AI-written summary · about
chapters (16)
- 0:00Intro
- 0:26Zoned Name Space (ZNS) Storage • The logical address space is divided into fixed-sized zones. • Each zone must be written sequentially and reset explicitly for reuse
- 0:50SSD Architecture 101
- 2:08Why ZNS? Isolation, hot/cold separation
- 2:50Why ZNS? Small L2P Translation Table
- 3:17Why ZNS? GC-less, Predictable
- 4:03New 10 Stack for ZNS Log-Structured File System (LFS) - Append Logging (Sequential Write)
- 4:44F2FS (Flash-Friendly File System) . One of actively maintained Log-structured File Systems (LFS) • Six types of segments: hot, warm and cold segments for each node/data
- 5:58Segment Compaction
- 7:06Robbing Peter to Pay Paul?
- 8:03ZNS+: LFS-Aware ZNS
- 9:28ZNS+-aware LFS • Hybrid segment recycling (HSR): segment cleaning vs. threaded logging
- 10:08Experimental Setup
- 10:42Performance
- 11:23Flash Chip Parallelism
- 12:49Conclusion
related talks
31mFrom Open-Channel SSDs to Zoned Namespaces
24mSSDFS: ZNS SSD ready file system with zero GC overhead
20mZNSwap: un-Block your Swap
20mObject caching on Zoned Storage
03mZoned MC Conference Kick Off
23mImproving data placement for Zoned Linux File systems
25mZonefs: Features Roadmap
14mUnleashing Zoned UFS: Cross-Layer Optimizations for Next-Generation Mobile Storage