Hotness-Aware Runtime Zone Reset and Reallocation Optimizations for ZNS SSDs

ZNS SSDs reduce address mapping, garbage collection (GC), and over-provisioning costs by organizing storage into sequential-write zones. ZenFS is a file system specifically designed for ZNS SSDs. Runtime zone reset is a key operation in ZenFS, allowing zones to be reclaimed during execution and thereby reducing the need for GC. However, research on how to efficiently use this mechanism remains limited. Specifically, ZenFS uses a fixed reset timing for all zones regardless of their varying hotness, hence failing to optimize write amplification and space utilization. In this work, we propose a hotness-aware optimization method that comprises an adaptive timing control mechanism for runtime zone resets and a dynamic zone reallocation mechanism. The key idea is to adaptively control each zone’s reset timing based on its hotness and the SSD’s available space. Additionally, when space is sufficient, the reallocation mechanism allows zones with fully invalidated data to remain in use, avoiding unnecessary resets. The zone-level timing control optimizes space availability and reduces write amplification in ZNS SSDs. Specifically, hot zones benefit from extended service time before reset, improving space utilization, while cold zones are reset more promptly, minimizing GC-induced erasures and data migration. Moreover, the dynamic zone reallocation mechanism extends the lifetime of zones with invalidated data when space permits, further reducing erasures. Experimental results show that the proposed method significantly reduces the zone reset count by average factors of 1.72 × and 1.57 × in comparison to the traditional and latest approaches, respectively.

Authors

Institutions

Publication Details

Journal
ACM Transactions on Architecture and Code Optimization
Published
2026-06-16
DOI
https://doi.org/10.1145/3819236
Primary Topic
Advanced Data Storage Technologies
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

Hotness-Aware Runtime Zone Reset and Reallocation Optimizations for ZNS SSDs

Masao Okita, Congming Gao, Jingcheng Shen, Fumihiko Ino et al.
ACM Transactions on Architecture and Code Optimization
Advanced Data Storage Technologies
article

Hotness-Aware Runtime Zone Reset and Reallocation Optimizations for ZNS SSDs

Masao Okita, Congming Gao, Jingcheng Shen, Fumihiko Ino, Linbo Long, Shuanglong Huang
article en

Abstract

ZNS SSDs reduce address mapping, garbage collection (GC), and over-provisioning costs by organizing storage into sequential-write zones. ZenFS is a file system specifically designed for ZNS SSDs. Runtime zone reset is a key operation in ZenFS, allowing zones to be reclaimed during execution and thereby reducing the need for GC. However, research on how to efficiently use this mechanism remains limited. Specifically, ZenFS uses a fixed reset timing for all zones regardless of their varying hotness, hence failing to optimize write amplification and space utilization. In this work, we propose a hotness-aware optimization method that comprises an adaptive timing control mechanism for runtime zone resets and a dynamic zone reallocation mechanism. The key idea is to adaptively control each zone’s reset timing based on its hotness and the SSD’s available space. Additionally, when space is sufficient, the reallocation mechanism allows zones with fully invalidated data to remain in use, avoiding unnecessary resets. The zone-level timing control optimizes space availability and reduces write amplification in ZNS SSDs. Specifically, hot zones benefit from extended service time before reset, improving space utilization, while cold zones are reset more promptly, minimizing GC-induced erasures and data migration. Moreover, the dynamic zone reallocation mechanism extends the lifetime of zones with invalidated data when space permits, further reducing erasures. Experimental results show that the proposed method significantly reduces the zone reset count by average factors of 1.72 × and 1.57 × in comparison to the traditional and latest approaches, respectively.

ACM Transactions on Architecture and Code Optimization
Chongqing University of Posts and Telecommunications (CN), Xiamen University (CN), The University of Osaka (JP)
Openalex Percentile: Top 5%
Advanced Data Storage Technologies
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.