μScope：评估存储堆栈对SSD延迟变化的鲁棒性

IF 3.7 2区计算机科学 Q1 COMPUTER SCIENCE, HARDWARE & ARCHITECTURE

Journal of Systems Architecture Pub Date : 2025-04-24 DOI:10.1016/j.sysarc.2025.103405

Linxiao Bai , Shanshan Li , Zhouyang Jia , Yu Jiang , Yuanliang Zhang , Zichen Xu , Bin Lin , Si Zheng , Xiangke Liao

{"title":"μScope：评估存储堆栈对SSD延迟变化的鲁棒性","authors":"Linxiao Bai , Shanshan Li , Zhouyang Jia , Yu Jiang , Yuanliang Zhang , Zichen Xu , Bin Lin , Si Zheng , Xiangke Liao","doi":"10.1016/j.sysarc.2025.103405","DOIUrl":null,"url":null,"abstract":"<div><div>The rapid development of Solid State Disks (SSDs) drastically reduces device latency from <span><math><mrow><mn>100</mn><mspace></mspace><mi>μ</mi><mi>s</mi></mrow></math></span> to around <span><math><mrow><mn>10</mn><mspace></mspace><mi>μ</mi><mi>s</mi></mrow></math></span>. However, performance advertised is not always performance delivered. Background operations (e.g., garbage collection and wear leveling) inside the SSDs now may severely influence the performance. In addition, SSDs are also susceptible to fail-slow failures. Traditionally, studying SSD-based stack focuses on understanding the SSD internal behaviors or discussing the impacts of software stack on throughput.</div><div>In this paper, we conduct an extensive study on software stack atop the low-latency SSDs, especially under device latency variations. We build <span><math><mi>μ</mi></math></span>Scope to overcome two major challenges, including achieving fine-grained latency injection and low-overhead monitoring, in profiling. Via <span><math><mi>μ</mi></math></span>Scope, we manage to obtain three major lessons in access patterns, consistency trade-offs and consecutive performance variations which shall benefit developers for further optimizations.</div></div>","PeriodicalId":50027,"journal":{"name":"Journal of Systems Architecture","volume":"164 ","pages":"Article 103405"},"PeriodicalIF":3.7000,"publicationDate":"2025-04-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"μScope: Evaluating storage stack robustness against SSD’s latency variation\",\"authors\":\"Linxiao Bai , Shanshan Li , Zhouyang Jia , Yu Jiang , Yuanliang Zhang , Zichen Xu , Bin Lin , Si Zheng , Xiangke Liao\",\"doi\":\"10.1016/j.sysarc.2025.103405\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<div><div>The rapid development of Solid State Disks (SSDs) drastically reduces device latency from <span><math><mrow><mn>100</mn><mspace></mspace><mi>μ</mi><mi>s</mi></mrow></math></span> to around <span><math><mrow><mn>10</mn><mspace></mspace><mi>μ</mi><mi>s</mi></mrow></math></span>. However, performance advertised is not always performance delivered. Background operations (e.g., garbage collection and wear leveling) inside the SSDs now may severely influence the performance. In addition, SSDs are also susceptible to fail-slow failures. Traditionally, studying SSD-based stack focuses on understanding the SSD internal behaviors or discussing the impacts of software stack on throughput.</div><div>In this paper, we conduct an extensive study on software stack atop the low-latency SSDs, especially under device latency variations. We build <span><math><mi>μ</mi></math></span>Scope to overcome two major challenges, including achieving fine-grained latency injection and low-overhead monitoring, in profiling. Via <span><math><mi>μ</mi></math></span>Scope, we manage to obtain three major lessons in access patterns, consistency trade-offs and consecutive performance variations which shall benefit developers for further optimizations.</div></div>\",\"PeriodicalId\":50027,\"journal\":{\"name\":\"Journal of Systems Architecture\",\"volume\":\"164 \",\"pages\":\"Article 103405\"},\"PeriodicalIF\":3.7000,\"publicationDate\":\"2025-04-24\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Journal of Systems Architecture\",\"FirstCategoryId\":\"94\",\"ListUrlMain\":\"https://www.sciencedirect.com/science/article/pii/S1383762125000773\",\"RegionNum\":2,\"RegionCategory\":\"计算机科学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q1\",\"JCRName\":\"COMPUTER SCIENCE, HARDWARE & ARCHITECTURE\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Journal of Systems Architecture","FirstCategoryId":"94","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S1383762125000773","RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"COMPUTER SCIENCE, HARDWARE & ARCHITECTURE","Score":null,"Total":0}

引用次数: 0

摘要

随着固态硬盘（ssd）技术的快速发展，设备延迟从100μs大幅降低到10μs左右。然而，宣传的性能并不总是交付的性能。ssd内部的后台操作（例如，垃圾收集和损耗均衡）现在可能会严重影响性能。此外，ssd还容易出现慢速故障。传统上，对基于SSD的堆栈的研究主要集中在了解SSD内部行为或讨论软件堆栈对吞吐量的影响。在本文中，我们对低延迟ssd上的软件堆栈进行了广泛的研究，特别是在设备延迟变化的情况下。我们构建μScope来克服分析中的两个主要挑战，包括实现细粒度延迟注入和低开销监控。通过μScope，我们在访问模式、一致性权衡和连续的性能变化方面获得了三个主要的经验教训，这将有利于开发人员进一步优化。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

查看原文本刊更多论文

μScope: Evaluating storage stack robustness against SSD’s latency variation

The rapid development of Solid State Disks (SSDs) drastically reduces device latency from

100 μ s

to around

10 μ s

. However, performance advertised is not always performance delivered. Background operations (e.g., garbage collection and wear leveling) inside the SSDs now may severely influence the performance. In addition, SSDs are also susceptible to fail-slow failures. Traditionally, studying SSD-based stack focuses on understanding the SSD internal behaviors or discussing the impacts of software stack on throughput.

In this paper, we conduct an extensive study on software stack atop the low-latency SSDs, especially under device latency variations. We build

μ

Scope to overcome two major challenges, including achieving fine-grained latency injection and low-overhead monitoring, in profiling. Via

μ

Scope, we manage to obtain three major lessons in access patterns, consistency trade-offs and consecutive performance variations which shall benefit developers for further optimizations.

求助全文

通过发布文献求助，成功后即可免费获取论文全文。去求助

来源期刊

Journal of Systems Architecture 工程技术-计算机：硬件

CiteScore

8.70

自引率

15.60%

发文量

226

审稿时长

46 days

期刊介绍： The Journal of Systems Architecture: Embedded Software Design (JSA) is a journal covering all design and architectural aspects related to embedded systems and software. It ranges from the microarchitecture level via the system software level up to the application-specific architecture level. Aspects such as real-time systems, operating systems, FPGA programming, programming languages, communications (limited to analysis and the software stack), mobile systems, parallel and distributed architectures as well as additional subjects in the computer and system architecture area will fall within the scope of this journal. Technology will not be a main focus, but its use and relevance to particular designs will be. Case studies are welcome but must contribute more than just a design for a particular piece of software. Design automation of such systems including methodologies, techniques and tools for their design as well as novel designs of software components fall within the scope of this journal. Novel applications that use embedded systems are also central in this journal. While hardware is not a part of this journal hardware/software co-design methods that consider interplay between software and hardware components with and emphasis on software are also relevant here.