Amphisbaena: A Novel Persistent Buffer Management Strategy to Improve SMR Disk Performance
The explosive growth of massive data makes shingled magnetic recording (SMR) disks a promising candidate for balancing capacity and cost. SMR disks are typically configured with a persistent buffer to reduce the read–modify–write (RMW) overhead introduced by non-sequential writes. Traditional SMR zo...
Main Authors: | , , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2024-01-01
|
Series: | Applied Sciences |
Subjects: | |
Online Access: | https://www.mdpi.com/2076-3417/14/2/630 |
_version_ | 1797340158815633408 |
---|---|
author | Chi Zhang Fangxing Yu Shiqiang Nie Wei Tang Fei Liu Song Liu Weiguo Wu |
author_facet | Chi Zhang Fangxing Yu Shiqiang Nie Wei Tang Fei Liu Song Liu Weiguo Wu |
author_sort | Chi Zhang |
collection | DOAJ |
description | The explosive growth of massive data makes shingled magnetic recording (SMR) disks a promising candidate for balancing capacity and cost. SMR disks are typically configured with a persistent buffer to reduce the read–modify–write (RMW) overhead introduced by non-sequential writes. Traditional SMR zones-based persistent buffers are subject to sequential-write constraints, and frequent cleanups cause disk performance degradation. Conventional magnetic recording (CMR) zones with in-place update capabilities enable less frequent cleanups and are gradually being used to construct persistent buffers in certain SMR disks. However, existing CMR zones-based persistent buffer designs fail to accurately capture hot blocks with long update periods and are limited by an inflexible data layout, resulting in inefficient cleanups. To address the above issues, we propose a strategy called Amphisbaena. First, a two-phase data block classification method is proposed to capture frequently updated blocks. Then, a locality-aware buffer space management scheme is developed to dynamically manage blocks with different update frequencies. Finally, a latency-sensitive garbage collection policy based on the above is designed to mitigate the impact of cleanup on user requests. Experimental results show that Amphisbaena reduces latency by an average of 29.9% and the number of RMWs by an average of 37% compared to current state-of-the-art strategies. |
first_indexed | 2024-03-08T09:58:54Z |
format | Article |
id | doaj.art-896bd4f8c2c74f1bb9828b93ba79f0b0 |
institution | Directory Open Access Journal |
issn | 2076-3417 |
language | English |
last_indexed | 2024-03-08T09:58:54Z |
publishDate | 2024-01-01 |
publisher | MDPI AG |
record_format | Article |
series | Applied Sciences |
spelling | doaj.art-896bd4f8c2c74f1bb9828b93ba79f0b02024-01-29T13:43:15ZengMDPI AGApplied Sciences2076-34172024-01-0114263010.3390/app14020630Amphisbaena: A Novel Persistent Buffer Management Strategy to Improve SMR Disk PerformanceChi Zhang0Fangxing Yu1Shiqiang Nie2Wei Tang3Fei Liu4Song Liu5Weiguo Wu6School of Computer Science and Technology, Xi’an Jiaotong University, Xi’an 710049, ChinaSchool of Computer Science and Technology, Xi’an Jiaotong University, Xi’an 710049, ChinaSchool of Computer Science and Technology, Xi’an Jiaotong University, Xi’an 710049, ChinaByteDance US Infrastructure System Lab, ByteDance Inc., Mountain View, CA 94041, USAByteDance US Infrastructure System Lab, ByteDance Inc., Mountain View, CA 94041, USASchool of Computer Science and Technology, Xi’an Jiaotong University, Xi’an 710049, ChinaSchool of Computer Science and Technology, Xi’an Jiaotong University, Xi’an 710049, ChinaThe explosive growth of massive data makes shingled magnetic recording (SMR) disks a promising candidate for balancing capacity and cost. SMR disks are typically configured with a persistent buffer to reduce the read–modify–write (RMW) overhead introduced by non-sequential writes. Traditional SMR zones-based persistent buffers are subject to sequential-write constraints, and frequent cleanups cause disk performance degradation. Conventional magnetic recording (CMR) zones with in-place update capabilities enable less frequent cleanups and are gradually being used to construct persistent buffers in certain SMR disks. However, existing CMR zones-based persistent buffer designs fail to accurately capture hot blocks with long update periods and are limited by an inflexible data layout, resulting in inefficient cleanups. To address the above issues, we propose a strategy called Amphisbaena. First, a two-phase data block classification method is proposed to capture frequently updated blocks. Then, a locality-aware buffer space management scheme is developed to dynamically manage blocks with different update frequencies. Finally, a latency-sensitive garbage collection policy based on the above is designed to mitigate the impact of cleanup on user requests. Experimental results show that Amphisbaena reduces latency by an average of 29.9% and the number of RMWs by an average of 37% compared to current state-of-the-art strategies.https://www.mdpi.com/2076-3417/14/2/630shingled magnetic recordingRMWin-place updategarbage collectionpersistent buffer |
spellingShingle | Chi Zhang Fangxing Yu Shiqiang Nie Wei Tang Fei Liu Song Liu Weiguo Wu Amphisbaena: A Novel Persistent Buffer Management Strategy to Improve SMR Disk Performance Applied Sciences shingled magnetic recording RMW in-place update garbage collection persistent buffer |
title | Amphisbaena: A Novel Persistent Buffer Management Strategy to Improve SMR Disk Performance |
title_full | Amphisbaena: A Novel Persistent Buffer Management Strategy to Improve SMR Disk Performance |
title_fullStr | Amphisbaena: A Novel Persistent Buffer Management Strategy to Improve SMR Disk Performance |
title_full_unstemmed | Amphisbaena: A Novel Persistent Buffer Management Strategy to Improve SMR Disk Performance |
title_short | Amphisbaena: A Novel Persistent Buffer Management Strategy to Improve SMR Disk Performance |
title_sort | amphisbaena a novel persistent buffer management strategy to improve smr disk performance |
topic | shingled magnetic recording RMW in-place update garbage collection persistent buffer |
url | https://www.mdpi.com/2076-3417/14/2/630 |
work_keys_str_mv | AT chizhang amphisbaenaanovelpersistentbuffermanagementstrategytoimprovesmrdiskperformance AT fangxingyu amphisbaenaanovelpersistentbuffermanagementstrategytoimprovesmrdiskperformance AT shiqiangnie amphisbaenaanovelpersistentbuffermanagementstrategytoimprovesmrdiskperformance AT weitang amphisbaenaanovelpersistentbuffermanagementstrategytoimprovesmrdiskperformance AT feiliu amphisbaenaanovelpersistentbuffermanagementstrategytoimprovesmrdiskperformance AT songliu amphisbaenaanovelpersistentbuffermanagementstrategytoimprovesmrdiskperformance AT weiguowu amphisbaenaanovelpersistentbuffermanagementstrategytoimprovesmrdiskperformance |