Analytical Model for Service Response Time in Distributed Systems

Abdulhalim Dandoush, M. Elgamel
{"title":"Analytical Model for Service Response Time in Distributed Systems","authors":"Abdulhalim Dandoush, M. Elgamel","doi":"10.1109/ICCTA35431.2014.9521610","DOIUrl":null,"url":null,"abstract":"Peer-to-peer (P2P) systems have emerged as a cheap, scalable and self-repairing file sharing, backup/storage and video on-demand (VoD) solutions. Such distributed systems rely on data fragmentation and distributed storage of data over multiple peers. Files are usually partitioned into fixed-size blocks that are themselves partitioned into fragments. In such systems, some peers (clients) exploit the power of parallelism to download blocks/fragments of data stored in a distributed way over some other peers (servers). This idea is deployed even over a network of servers (mirrors and cache) to reduce download time. Determining response times in parallel downloading with capacity constraints on both the client downloads and server uploads is then an important measure of the quality of the service given by these systems. This download time depends on the throughput of the different connections created to obtain the fragments of the given file. This paper enriches the literature with a simple stochastic model of the download process in such large and complex systems. In particular, it characterizes the response time performance metric and assesses its expectation through Markovian model and validates the study through the packet level Network Simulator NS-2. The relative error between the numerical results of the proposed model and the result of the simulation is reported. The comparison shows that the Model characterizes with high accuracy the response time under realistic assumptions.","PeriodicalId":162170,"journal":{"name":"2014 24th International Conference on Computer Theory and Applications (ICCTA)","volume":"23 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2014-10-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2014 24th International Conference on Computer Theory and Applications (ICCTA)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ICCTA35431.2014.9521610","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0

Abstract

Peer-to-peer (P2P) systems have emerged as a cheap, scalable and self-repairing file sharing, backup/storage and video on-demand (VoD) solutions. Such distributed systems rely on data fragmentation and distributed storage of data over multiple peers. Files are usually partitioned into fixed-size blocks that are themselves partitioned into fragments. In such systems, some peers (clients) exploit the power of parallelism to download blocks/fragments of data stored in a distributed way over some other peers (servers). This idea is deployed even over a network of servers (mirrors and cache) to reduce download time. Determining response times in parallel downloading with capacity constraints on both the client downloads and server uploads is then an important measure of the quality of the service given by these systems. This download time depends on the throughput of the different connections created to obtain the fragments of the given file. This paper enriches the literature with a simple stochastic model of the download process in such large and complex systems. In particular, it characterizes the response time performance metric and assesses its expectation through Markovian model and validates the study through the packet level Network Simulator NS-2. The relative error between the numerical results of the proposed model and the result of the simulation is reported. The comparison shows that the Model characterizes with high accuracy the response time under realistic assumptions.
分布式系统服务响应时间分析模型
点对点(P2P)系统作为一种廉价、可扩展和自修复的文件共享、备份/存储和视频点播(VoD)解决方案而出现。这种分布式系统依赖于数据碎片和数据在多个对等点上的分布式存储。文件通常被划分为固定大小的块,这些块本身又被划分为片段。在这样的系统中,一些对等体(客户端)利用并行性的力量下载以分布式方式存储在其他对等体(服务器)上的数据块/片段。这个想法甚至部署在服务器网络(镜像和缓存)上,以减少下载时间。在客户端下载和服务器上传都有容量限制的情况下,确定并行下载的响应时间是衡量这些系统提供的服务质量的重要指标。这个下载时间取决于为获取给定文件的片段而创建的不同连接的吞吐量。本文用一个简单的随机模型来丰富这类大型复杂系统的下载过程。特别是,通过马尔可夫模型表征响应时间性能指标并评估其期望,并通过数据包级网络模拟器NS-2验证研究结果。给出了模型数值结果与仿真结果的相对误差。仿真结果表明,该模型能较准确地描述实际假设下的响应时间。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
求助全文
约1分钟内获得全文 求助全文
来源期刊
自引率
0.00%
发文量
0
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
copy
已复制链接
快去分享给好友吧!
我知道了
右上角分享
点击右上角分享
0
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:604180095
Book学术官方微信