CN102404173B - Web服务吞吐率预测方法 - Google Patents

Web服务吞吐率预测方法 Download PDF

Info

Publication number
CN102404173B
CN102404173B CN2011104448475A CN201110444847A CN102404173B CN 102404173 B CN102404173 B CN 102404173B CN 2011104448475 A CN2011104448475 A CN 2011104448475A CN 201110444847 A CN201110444847 A CN 201110444847A CN 102404173 B CN102404173 B CN 102404173B
Authority
CN
China
Prior art keywords
throughput
vector
positive integer
throughput rate
web services
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Expired - Fee Related
Application number
CN2011104448475A
Other languages
English (en)
Other versions
CN102404173A (zh
Inventor
夏云霓
陈鹏
戴刚
罗辛
吴磊
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Chengdu Gkhb Information Technology Co ltd
Chongqing University
Original Assignee
CHENGDU GUOKE HAIBO COMPUTER SYSTEMS Co Ltd
Chongqing University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by CHENGDU GUOKE HAIBO COMPUTER SYSTEMS Co Ltd, Chongqing University filed Critical CHENGDU GUOKE HAIBO COMPUTER SYSTEMS Co Ltd
Priority to CN2011104448475A priority Critical patent/CN102404173B/zh
Publication of CN102404173A publication Critical patent/CN102404173A/zh
Application granted granted Critical
Publication of CN102404173B publication Critical patent/CN102404173B/zh
Expired - Fee Related legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Landscapes

  • Data Exchanges In Wide-Area Networks (AREA)

Abstract

本发明公开了一种web服务吞吐率预测方法,属于软件性能预测领域,按照固定单位时间间隔获取待分析的web服务的吞吐率数据序列,然后根据相邻吞吐率值的变化情况定义多种吞吐率变化的模式,并定义和计算模式之间的转换概率矩阵,最后将当前时刻的吞吐率值加上当前时刻到一个单位时间间隔后时刻的加权平均吞吐率增量,从而获得需要的吞吐率的预测值,该方法避免了传统的Web预测模型和方法中过于依赖SLA规范和假设参数恒定不变的不足,能为WEB服务可信性研究提供模型支撑和分析手段。

Description

Web服务吞吐率预测方法
技术领域
本发明属于软件性能预测领域,特别是涉及一种Web服务吞吐率预测方法。
背景技术
Web服务作为一种新型的分布式构件模型在电子商务、企业应用集成等领域扮演着越来越重要的角色,它被认为是SOA(Service-Oriented Architecture,面向服务的计算构架)中最核心的技术之一,通过Web服务组合技术可以将单个Web服务组合起来成为完整的应用。
为web服务的吞吐率表示系统在单位时间内完成给定任务的次数,Web服务运行在动态变化的网络环境中,其吞吐率值也在频繁变化,这种动态变化的网络环境将会影响Web服务的运行时性能进而导致整个Web服务组合性能的变化,因此web服务性能预测是保证基于服务的应用系统质量的关键技术,也是优化服务组合重要方法。
目前,大部分Web服务的性能预测方法大多以SLA(Service-LevelAgreement,服务等级协议)文档给定的性能参数为准,并假设性能参数恒定不变,忽略了实际运行时性能的波动和变化,由于忽略了实际运行时性能的波动和变化导致这种预测方法精度不高。
发明内容
有鉴于现有技术的上述缺陷,本发明所要解决的技术问题是提供一种精度更高的Web服务吞吐率预测方法。
为实现上述目的,本发明提供了一种web服务吞吐率预测方法,按照固定单位时间间隔获取web服务的吞吐率数据序列;所述吞吐率数据序列为{q(i)},所述i为正整数;
取相邻时刻的吞吐率值q(i)和q(i+1);设定从吞吐率值到变化趋势的映射函
数为map1(i); map 1 ( i ) = - 1 ifq ( i ) - q ( i + 1 ) > d 0 if | q ( i ) - q ( i + 1 ) | ≤ d 1 else ; 所述d为预先给定的阈值;
设定1×p维的趋势模式向量Si;Si=[map1(i),map1(i+1),map1(i+2),...,map1(i+p-1)],1≤i≤i+p-1,p为正整数;
所述趋势模式向量分为k个分类,分别为C1,C2,....,Ck;所述趋势模式向量分类的映射函数为map2();将Si映射为第x个分类,当且仅当 ( x - 1 ) / k &le; ( &Sigma; j = 1 p - 1 2 j + &Sigma; j = 1 p - 1 2 j &times; S i ( j ) ) / &Sigma; i = j p - 1 2 j + 1 < x / k 时,map2(Si)=x,1≤x≤k,k为正整数;
将测试得到的吞吐率序列数据{q(i)}赋给t-p+1个吞吐率向量,t为吞吐率数据序列中数据个数;设定Qi为吞吐率向量,Qi=[q(i),q(i+1)....,q(i+p-1)],1≤i≤t-p+1;设定相邻向量差为Di;Di=Qi+1-Qi
设定不同分类间的转换概率为tra,b所述
M ( l ) = 1 ifmap 2 ( S l ) = a 0 else ; 所述 N ( l ) = 1 ifmap 2 ( S l ) = aandmap 2 ( S l + 1 ) = b 0 else , l,a,b均为正整数;
计算吞吐率平均增量矩阵;设定吞吐率平均增量矩阵为
Figure BDA0000125191670000026
D &OverBar; e , f = tr a , b &times; &Sigma; y = 1 t - p + 1 O ( y ) ;
O ( y ) = D y ifmap 2 ( S y ) = aandmap 2 ( S y + 1 ) = b 0 else ; 所述平均增量矩阵为属于Ca分类的吞吐率向量到属于Cb分类的吞吐率向量之间向量差的平均值;
计算 Q ~ t - p + 2 = Q t - p + 1 + &Sigma; w = 1 k tr map 2 ( S t - p + 1 ) , w &times; D &OverBar; map 2 ( S t - p + 1 ) , w ; w为正整数;所述Qt-p+1向量为t-p+1时刻的吞吐率向量,所述
Figure BDA0000125191670000033
为t-p+2时刻的吞吐率向量。
较佳的,所述获取web服务的吞吐率数据序列由SOAP UI测试平台对web服务的url地址测试得到。
较佳的,所述吞吐率向量为1×p维;第m个向量的第n个分量为q((m-1)*p+n),m、n均为正整数。
本发明的有益效果是:本发明避免了传统的web性能预测模型和方法中过于依赖SLA规范和假设性能参数恒定不变的不足,能为web服务可信性研究提供模型支撑和分析手段,本发明提供了一种精度更高的web服务吞吐率预测方法。
附图说明
图1是本发明的流程示意图。
图2是吞吐率实测数据示意图。
图3是吞吐率实际值与预测值的比较示意图。
具体实施方式
下面结合附图和实施例对本发明作进一步说明:
如图1至图3所示,一种web服务吞吐率预测方法,其特征在于:
按照固定单位时间间隔获取web服务的吞吐率数据序列,所述吞吐率数据序列为{q(i)},所述i为正整数。
取相邻时刻的吞吐率值q(i)和q(i+1);设定从吞吐率值到变化趋势的映射函数为map1(i), map 1 ( i ) = - 1 ifq ( i ) - q ( i + 1 ) > d 0 if | q ( i ) - q ( i + 1 ) | &le; d 1 else , d∈R+,所述d为预先给定的阈值。
设定1×p维的趋势模式向量Si,Si=[map1(i),map1(i+1),map1(i+2),...,map1(i+p-1)],1≤i≤i+p-1,p为正整数;
所述趋势模式向量分为k个分类,分别为C1,C2,....,Ck;所述趋势模式向量分类的映射函数为map2();将Si映射为第x个分类,当且仅当 ( x - 1 ) / k &le; ( &Sigma; j = 1 p - 1 2 j + &Sigma; j = 1 p - 1 2 j &times; S i ( j ) ) / &Sigma; i = j p - 1 2 j + 1 < x / k 时,map2(Si)=x,1≤x≤k,k为正整数;
将测试得到的吞吐率序列数据{q(i)}赋给t-p+1个吞吐率向量,所述吞吐率向量为1×p维,第m个向量的第n个分量为q((m-1)*p+n),m、n均为正整数,所述t为吞吐率数据序列中数据个数,设定Qi为吞吐率向量,Qi=[q(i),q(i+1)....,q(i+p-1)],1≤i≤t-p+1,设定相邻向量差为Di,Di=Qi+1-Qi,所述Di为两个相邻的吞吐率向量的向量差,描述了吞吐率向量在一个单位时间差上的变化值。
设定不同分类间的转换概率为tra,b,所述tra,b代表当前趋势模式向量属于变化模式类Ca的前提下,其相邻的下一个趋势模式向量属于变化模式类Cb的条件概率, tr a , b = &Sigma; l = 1 t - p + 1 N ( l ) &Sigma; l = 1 t - p + 1 M ( l ) , 所述 M ( l ) = 1 ifmap 2 ( S l ) = a 0 else , 所述
N ( l ) = 1 ifmap 2 ( S l ) = aandmap 2 ( S l + 1 ) = b 0 else , 所述l,a,b均为正整数;
计算吞吐率平均增量矩阵;设定吞吐率平均增量矩阵为
Figure BDA0000125191670000046
D &OverBar; e , f = tr a , b &times; &Sigma; y = 1 t - p + 1 O ( y ) , O ( y ) = D y ifmap 2 ( S y ) = aandmap 2 ( S y + 1 ) = b 0 else , 所述平均增量矩阵为属于Ca分类的吞吐率向量到属于Cb分类的吞吐率向量之间向量差的平均值;Qt-p+1向量为吞吐率数据序列中提取出的最后一个吞吐率向量,在其基础上可预测下一个吞吐率向量
Figure BDA0000125191670000053
计算 Q ~ t - p + 2 = Q t - p + 1 + &Sigma; w = 1 k tr map 2 ( S t - p + 1 ) , w &times; D &OverBar; map 2 ( S t - p + 1 ) , w , w为正整数,所述Qt-p+1向量为t-p+1时刻的吞吐率向量,所述为t-p+2时刻的吞吐率向量,最终将向量
Figure BDA0000125191670000056
中最后一个分量值
Figure BDA0000125191670000057
作为对t+1时刻吞吐率的预测值。
为了对方法的正确性和精确性进行验证,在配置为INTEL i5-760处理器、4G内存的PC机上采用SOAPUI测试平台对一个提供天气信息的Web服务的实例的URL地址进行了测试,所述web服务实例的URL地址为http://www.webservicex.net/globalweather.asmx?WSDL,测试的开始时间为2011年10月17日上午10点30分0秒,测试时间间隔为250毫秒,连续获取128个吞吐率测试值,其中前100个数据用于模型建立,剩下28个数据从2011年10月17日上午10点30分25秒开始测得的吞吐率用于模型验证。
模型中的参数设置为:吞吐率向量的维度p=4;阈值d=30;分类数量k=10,按照本专利给出的方法,2011年10月17日上午10点30分25秒以后的吞吐率预测值如线条1所示,真实吞吐率值为线条2所示,采用ARMA时间序列模型预测值的曲线由线条3所示,消除离群点的ARMA模型预测值的曲线由线条4所示。
本发明方法预测值与实际值平均误差率为29%,而ARMA时间序列模型和消除离群点的ARMA模型预测值与实际值平均误差分别为34%和45%。可见,本专利提出的方法,在Web服务性能预测上取得了更好的精度。
以上详细描述了本发明的较佳具体实施例。应当理解,本领域的普通技术人员无需创造性劳动就可以根据本发明的构思作出诸多修改和变化。因此,凡本技术领域中技术人员依本发明的构思在现有技术的基础上通过逻辑分析、推理或者有限的实验可以得到的技术方案,皆应在由权利要求书所确定的保护范围内。

Claims (3)

1.一种web服务吞吐率预测方法,其特征在于:
按照固定单位时间间隔获取web服务的吞吐率数据序列;所述吞吐率数据序列为{q(i)},所述i为正整数;
取相邻时刻的吞吐率值q(i)和q(i+1);设定从吞吐率值到变化趋势的映射函数为map1(i); mapl ( i ) - 1 if q ( i ) - q ( i + 1 ) > d 0 if | q ( i ) - q ( i + 1 ) | &le; d 1 else ; 所述d为预先给定的阈值,d∈R+
设定1×p维的趋势模式向量Si;Si=[map1(i),map1(i+1),map1(i+2),...,map1(i+p-1)],1≤i≤i+p-1,p为正整数;
所述趋势模式向量分为k个分类,分别为C1,C2,....,Ck;所述趋势模式向量分类的映射函数为map2();将Si映射为第x个分类,当且仅当 ( x - 1 ) / k &le; ( &Sigma; j = 1 p - 1 2 j + &Sigma; j = 1 p - 1 2 j &times; S i ( j ) ) / &Sigma; i = j p - 1 2 j + 1 < x / k 时,map2(Si)=x,1≤x≤k,k为正整数;
将测试得到的吞吐率数据序列{q(i)}赋值给t-p+1个吞吐率向量,t为吞吐率数据序列中数据个数;设定Qi为吞吐率向量,Qi=[q(i),q(i+1)....,q(i+p-1)],1≤i≤t-p+1;设定相邻向量差为Di;Di=Qi+1-Qi
设定不同分类间的转换概率为tra,b
Figure FDA00003614605800013
所述 M ( l ) = 1 if map 2 ( S l ) = a 0 else ; 所述 N ( l ) = 1 ifmap 2 ( S 1 ) = aandmap 2 ( S l + 1 ) = b 0 eles , l,a,b均为正整数;
计算吞吐率平均增量矩阵;设定吞吐率平均增量矩阵为
Figure FDA00003614605800016
D &OverBar; e , f = tr a , b &times; &Sigma; y = 1 t - p + 1 O ( y ) ;
O ( y ) = D y ifmap 2 ( S y ) = aandmap 2 ( S y + 1 ) = b 0 else ; 所述平均增量矩阵为属于Ca分类的吞吐率向量到属于Cb分类的吞吐率向量之间向量差的平均值;
计算 Q ~ t - p + 2 = Q t - p + 1 + &Sigma; w = 1 k tr map 2 ( S t - p + 1 ) , w &times; D &OverBar; map 2 ( S t - p + 1 ) , w ; w为正整数;所述Qt-p+1向量为t-p+1时刻的吞吐率向量,所述
Figure FDA00003614605800024
为t-p+2时刻的吞吐率向量。
2.如权利要求1所述的web服务吞吐率预测方法,其特征是:所述获取web服务的吞吐率数据序列由SOAPUI测试平台对web服务的url地址测试得到。
3.如权利要求1所述的web服务吞吐率预测方法,其特征是:所述吞吐率向量为1×p维;第m个向量的第n个分量为q((m-1)*p+n),m、n均为正整数。
CN2011104448475A 2011-12-27 2011-12-27 Web服务吞吐率预测方法 Expired - Fee Related CN102404173B (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN2011104448475A CN102404173B (zh) 2011-12-27 2011-12-27 Web服务吞吐率预测方法

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN2011104448475A CN102404173B (zh) 2011-12-27 2011-12-27 Web服务吞吐率预测方法

Publications (2)

Publication Number Publication Date
CN102404173A CN102404173A (zh) 2012-04-04
CN102404173B true CN102404173B (zh) 2013-11-27

Family

ID=45885986

Family Applications (1)

Application Number Title Priority Date Filing Date
CN2011104448475A Expired - Fee Related CN102404173B (zh) 2011-12-27 2011-12-27 Web服务吞吐率预测方法

Country Status (1)

Country Link
CN (1) CN102404173B (zh)

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102801581B (zh) * 2012-07-16 2015-01-07 重庆大学 一种web服务连接成功率的预测方法
CN107733733B (zh) * 2017-09-04 2021-02-05 南京理工大学 一种确定传输时限下无线网络重传广播机制吞吐率的方法

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101651568A (zh) * 2009-07-01 2010-02-17 青岛农业大学 一种网络流量预测和异常检测方法
CN102170449A (zh) * 2011-04-28 2011-08-31 浙江大学 一种基于协同过滤的Web服务QoS预测方法

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101651568A (zh) * 2009-07-01 2010-02-17 青岛农业大学 一种网络流量预测和异常检测方法
CN102170449A (zh) * 2011-04-28 2011-08-31 浙江大学 一种基于协同过滤的Web服务QoS预测方法

Non-Patent Citations (6)

* Cited by examiner, † Cited by third party
Title
Queuing analysis and performance evaluation of workflow through WFQN;Yunni Xia 等;《Theoretical Aspects of Software Engineering, 2007. TASE "07. First Joint IEEE/IFIP Symposium on》;20070608;第178-187页 *
Stochastic modeling and quality evaluation of component-based software systems;Xia Yunni 等;《Quality Software, 2006. QSIC 2006. Sixth International Conference on》;20061028;第377-384页 *
Xia Yunni 等.Stochastic modeling and quality evaluation of component-based software systems.《Quality Software, 2006. QSIC 2006. Sixth International Conference on》.2006,第377-384页.
Yunni Xia 等.Queuing analysis and performance evaluation of workflow through WFQN.《Theoretical Aspects of Software Engineering, 2007. TASE "07. First Joint IEEE/IFIP Symposium on》.2007,第178-187页.
一种模型驱动的WS-CDL服务组合可靠性预测方法;刘毅 等;《计算机工程与应用》;20110411;第47卷(第11期);第60-66页 *
刘毅 等.一种模型驱动的WS-CDL服务组合可靠性预测方法.《计算机工程与应用》.2011,第47卷(第11期),第60-66页.

Also Published As

Publication number Publication date
CN102404173A (zh) 2012-04-04

Similar Documents

Publication Publication Date Title
Chen et al. Future changes in precipitation extremes over China using the NEX-GDDP high-resolution daily downscaled data-set
Sevlian et al. Detection and statistics of wind power ramps
Bai et al. Passenger demand forecasting with multi-task convolutional recurrent neural networks
Osorio et al. A simulation-based optimization framework for urban transportation problems
Corizzo et al. Anomaly detection and repair for accurate predictions in geo-distributed big data
Ghiassi et al. Reductive bottom-up urban energy computing supported by multivariate cluster analysis
US20140195159A1 (en) Application of artificial intelligence techniques and statistical ensembling to forecast power output of a wind energy facility
CA2996731C (en) Methods and systems for energy use normalization and forecasting
CN105225026A (zh) 一种基于电能服务管理平台的住房空置率评估方法
Brown et al. Disaggregation of household solar energy generation using censored smart meter data
Iftikhar et al. A scalable smart meter data generator using spark
Roth et al. Data-driven, multi-metric, and time-varying (DMT) building energy Benchmarking using smart meter data
CN102739469B (zh) 一种基于时间序列的Web服务响应时间预测方法
CN102404173B (zh) Web服务吞吐率预测方法
Amonkar et al. A k-nearest neighbor space-time simulator with applications to large-scale wind and solar power modeling
Grandón et al. Electricity demand forecasting with hybrid classical statistical and machine learning algorithms: Case study of Ukraine
CN112287503A (zh) 用于交通需求预测的动态空间网络构建方法
Kim et al. Using the National Digital Forecast Database for model-based building controls
Hirata et al. Wind direction modelling using multiple observation points
CN102801581B (zh) 一种web服务连接成功率的预测方法
CN115130788B (zh) 基于大数据分析的用电预测系统
CN102930346B (zh) 一种soa服务平均故障间隔时间预测方法
Zhang et al. A nonstationary and non‐Gaussian moving average model for solar irradiance
CN115688547A (zh) 模拟天气情景和极端天气预测
Lee et al. A cost-effective and heuristic approach for building energy consumption prediction: BES model calibration and forecasting algorithm

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
C14 Grant of patent or utility model
GR01 Patent grant
ASS Succession or assignment of patent right

Owner name: CHENGDU GUOKE HAIBO INFORMATION TECHNOLOGY CO., LT

Free format text: FORMER OWNER: CHONGQING UNIVERSITY

Effective date: 20140504

Free format text: FORMER OWNER: CHENGDU GUOKE HAIBO INFORMATION TECHNOLOGY CO., LTD.

Effective date: 20140504

C41 Transfer of patent application or patent right or utility model
C56 Change in the name or address of the patentee
COR Change of bibliographic data

Free format text: CORRECT: ADDRESS; FROM: 400045 SHAPINGBA, CHONGQING TO: 610041 CHENGDU, SICHUAN PROVINCE

CP01 Change in the name or title of a patent holder

Address after: 400045 Shapingba District, Sha Sha Street, No. 174, Chongqing

Patentee after: Chongqing University

Patentee after: CHENGDU GKHB INFORMATION TECHNOLOGY Co.,Ltd.

Address before: 400045 Shapingba District, Sha Sha Street, No. 174, Chongqing

Patentee before: Chongqing University

Patentee before: CHENGDU GUOKE HAIBO COMPUTER SYSTEMS Co.,Ltd.

TR01 Transfer of patent right

Effective date of registration: 20140504

Address after: 610041, 4 Building 1, ideal center, No. 38 Tianyi street, hi tech Zone, Sichuan, Chengdu

Patentee after: CHENGDU GKHB INFORMATION TECHNOLOGY Co.,Ltd.

Address before: 400045 Shapingba District, Sha Sha Street, No. 174, Chongqing

Patentee before: Chongqing University

Patentee before: CHENGDU GKHB INFORMATION TECHNOLOGY Co.,Ltd.

PE01 Entry into force of the registration of the contract for pledge of patent right

Denomination of invention: Prediction method of Web service throughput rate

Effective date of registration: 20140909

Granted publication date: 20131127

Pledgee: Chengdu high investment financing Company limited by guarantee

Pledgor: CHENGDU GKHB INFORMATION TECHNOLOGY Co.,Ltd.

Registration number: 2014990000743

PLDC Enforcement, change and cancellation of contracts on pledge of patent right or utility model
C53 Correction of patent of invention or patent application
CB03 Change of inventor or designer information

Inventor after: Wu Lei

Inventor after: Chen Peng

Inventor after: Xia Yunni

Inventor after: Dai Gang

Inventor after: Luo Xin

Inventor before: Xia Yunni

Inventor before: Chen Peng

Inventor before: Dai Gang

Inventor before: Luo Xin

Inventor before: Wu Lei

COR Change of bibliographic data

Free format text: CORRECT: INVENTOR; FROM: XIA YUNNI CHEN PENG DAI GANG LUO XIN WU LEI TO: WU LEI CHEN PENG XIA YUNNI DAI GANG LUO XIN

PC01 Cancellation of the registration of the contract for pledge of patent right

Date of cancellation: 20160105

Granted publication date: 20131127

Pledgee: Chengdu high investment financing Company limited by guarantee

Pledgor: CHENGDU GKHB INFORMATION TECHNOLOGY Co.,Ltd.

Registration number: 2014990000743

PLDC Enforcement, change and cancellation of contracts on pledge of patent right or utility model
CF01 Termination of patent right due to non-payment of annual fee

Granted publication date: 20131127

Termination date: 20191227

CF01 Termination of patent right due to non-payment of annual fee