LaDy: Enabling Locality-aware Deduplication Technology on Shingled Magnetic Recording Drives

以作者查詢圖書館館藏

、以作者查詢臺灣博碩士

、以作者查詢全國書目

、勘誤回報

、線上人數：18

、訪客IP：18.217.182.45

姓名

張榮綉(Jung-Hsiu Chang) 查詢紙本館藏

畢業系所

資訊工程學系

論文名稱

(LaDy: Enabling Locality-aware Deduplication Technology on Shingled Magnetic Recording Drives)

相關論文

★ 重新思考虛擬記憶體管理的方式以開放通道式固態硬碟最大限度地減少深度學習推薦系統演算法的讀寫流量	★ 開啟製程相似檢查方法在組裝超級塊上以最小化額外的寫入延遲
★ On Minimizing Writing Overhead to Establish a Low-latency LSM-tree on Skyrmion-based Racetrack Memory	★ WABE: Rethinking B-epsilon-tree to Minimize Write-amplification on NAND Flash Memory
★ Rethinking Bϵ tree Indexing Structure over NVM with the Support of Multi-write Modes	★ Prophet’s Insight: Unleashing Deduplication System Performance in Multi-tier Storage Systems
★ Freeing the Power of High Parallelism: Accelerating the Bϵ-Tree Indexing Scheme Performance on Open-Channel SSD	★ Applying Content-Defined Chunking to OCSSD-based Deduplication Systems
★ GraLoc: Preserving Graph Locality to Minimize Read and Write Amplification on NAND Flash Memory	★ On Minimizing Writing Overhead to Establish a Low-latency LSM-tree on Skyrmion-based Racetrack Memory

檔案

[Endnote RIS 格式]

[Bibtex 格式]

[相關文章]

[文章引用]

[完整記錄]

[館藏目錄]

至系統瀏覽論文 (2025-6-30以後開放)

摘要(中)

硬碟驅動器在過去幾年中最普遍使用的技術是傳統磁記錄技術。傳統磁記錄技術是通過寫入彼此平行而不重疊的磁軌來記錄資料，磁道並排寫入，磁道之間不重疊。但是傳統磁記錄技術已經無法滿足日益增長的存儲容量需求。為了增加存儲容量，疊瓦磁記錄技術通過重疊磁道來增加硬盤的面密度。在本文中，我們將介紹重複數據刪除的技術，這是常被用在固態硬碟的方法，透過這個方法可以提升存儲量。若將重複數據刪除的技術應用在疊瓦磁記錄硬碟上則可以更進一步增加資料儲存量，但是這會失去數據局部性也就會增加讀取資料的負擔，因為重複的資料會分散在整個硬碟上。所以為了增加存儲容量並減少數據局部性降低產生的影響，我們將導入我們的方法，位置感知重複數據刪除技術。我們提出的方法與基線 CAFTL相比，在 Systor ′17 Traces 中，總讀取反應時間上減少了87.3%，總存儲時間上減少了 76.75%。

摘要(英)

Conventional Magnetic Recording (CMR) technique used in hard drive is common in the past few years. It works by aligning the poles of the magnetic elements, which represent bits of data, perpendicularly to the surface of the disk. Magnetic tracks are written side by side and it is not overlapping between tracks. However, CMR technique is not enough for the growing need of the storage capacity. To increase the storage capacity, Shingled Magnetic Recording (SMR) cite{smr01} technique increases the areal density of hard disks by overlapping magnetic tracks. In this paper we will introduce the deduplication technique, which is normally used in the Solid State Disk to further increase the storage capacity. With the deduplication technology on SMR hard disks, it would increase the storage capacity but, for the trade off, it would lose the data locality, which leads to the overhead of the read request because the duplicate data is scattered all over the hard disk. To increase the storage capacity and also reduce the influence of losing data locality, we introduce our method, Locality-aware Deduplication Technology (Lady). The performance of the proposed approach, i.e Lady, in the Systor ‘17 Traces, reduces 87.3% on total read response time, and 76.75% for the total access time compared to the baseline, CAFTL on SMR.

關鍵字(中)

★ 除複
★ 疊瓦磁記錄
★ 儲存裝置

關鍵字(英)

★ Deduplication
★ SMR drives
★ Storage system

論文目次

Contents
1 Introduction 1
2 Background 3
2.1 SMR hard disk introduction . . . . . . . . . . . . . . . . . . . 3
2.2 Deduplication technology . . . . . . . . . . . . . . . . . . . 5
3 Motivation 7
4 Locality-aware deduplication technology on SMR 9
4.1 Deduplication range definition . . . . . . . . . . . . . . . . . 9
4.2 Deduplication method . . . . . . . . . . . . . . . . . . . . . . 10
4.2.1 Data structure of storing sample and fingerprint . . . . 10
4.2.2 Example of writing a duplicate chunk . . . . . . . . . 12
4.2.3 Example of writing a unique chunk . . . . . . . . . . 13
4.3 Mapping table of the Lady . . . . . . . . . . . . . . . . . . . 13
4.4 Garbage collection of the Lady . . . . . . . . . . . . . . . .15
4.4.1 Choose victim zones . . . . . . . . . . . . . . . . . . 15
4.4.2 Read cold chunk flow . . . . . . . . . . . . . . . . . 16
4.4.3 Read hot chunk flow . . . . . . . . . . . . . . . . . . 19
5 Evaluation 20
5.1 Baseline Introduction . . . . . . . . . . . . . . . . . . . . . . 20
5.2 Implementations . . . . . . . . . . . . . . . . . . . . . . . . 20
5.3 Trace Information . . . . . . . . . . . . . . . . . . . . . . . . 21
5.4 Experiment Design . . . . . . . . . . . . . . . . . . . . . . . 21
5.4.1 Different deduplication ranges performance . . . . . . 21
5.4.2 Different duplicate input data rate performance . . . . 23
5.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

參考文獻

References
[1] Tim Feldman and Garth Gibson. Shingled magnetic recording areal density increase requires new data management. volume 38, pages 22–30, june 2013.
[2] Feng Chen, Tian Luo, and Xiaodong Zhang. {CAFTL}: A {Content-Aware} flash translation layer enhancing the lifespan of flash memory based solid state drives. In 9th USENIX Conference on File and Storage Technologies (FAST 11), 2011.
[3] Chun-Feng Wu, Ming-Chang Yang, and Yuan-Hao Chang. Improving runtime performance of deduplication system with host-managed smr storage drives. In 2018 55th ACM/ESDA/IEEE Design Automation Conference (DAC), pages 1–6. IEEE, 2018.
[4] Ycsb rocksdb ssd block i /o traces. URL http://iotta.snia.org/traces/block-io.
[5] Msr cambridge block i /o traces. URL http://iotta.snia.org/traces/388.
[6] Systor ’17 paper (omitted files) block i /o traces. URL http://iotta.snia.org/traces/block-io/4964.
[7] Fenggang Wu, Baoquan Zhang, Zhichao Cao, Hao Wen, Bingzhe Li, Jim Diehl, Guohua Wang, and David HC Du. Data management design for interlaced magnetic recording. In 10th USENIX Workshop on Hot Topics in Storage and File Systems (HotStorage 18), 2018.
[8] Yong-Ching Lin, Shuo-Han Chen, Tseng-Yi Chen, Yuan-Hao Chang, Ming-Chang Yang, Hsin-Wen Wei, and Wei-Kuan Shih. Management designs for emerging shingled magnetic recording drives. In Symposium on Digital Life Technologies (DLT) (Chiayi, Taiwan), 2016. URL http://ir.sinica.edu.tw/handle/201000000A/39427.
[9] Garth Gibson and Greg Ganger. Principles of operation for shingled disk devices. Canregie Mellon Parallel Data Laboratory, CMU-PDL-11-107, 2011.
[10] Garth Gibson and Milo Polte. Directions for shingled-write and twodi-mensional magnetic recording system architectures: Synergies with solid-state disks. Parallel Data Lab, Carnegie Mellon Univ., Pittsburgh,PA, Tech. Rep. CMU-PDL-09-014, 2009.
[11] Fips 180-1, secure hash standard. April 1995.
[12] Disksim. URL http://www.pdl.cmu.edu/DiskSim/.
[13] Database of validated disk parameters. URL http://www.pdl.cmu.edu/DiskSim/diskspecs.shtml.

指導教授

陳增益(Tseng-Yi Chen)

審核日期

2022-7-5

推文