本文選題：云計(jì)算 + 平臺(tái)即服務(wù)��； 參考：《哈爾濱工業(yè)大學(xué)》2017年碩士論文

【摘要】：隨著云計(jì)算時(shí)代的到來,各大服務(wù)運(yùn)營商紛紛推出了基于容器技術(shù)的輕量級(jí)Paa S(平臺(tái)即服務(wù))平臺(tái),為人們提供高效便捷的業(yè)務(wù)應(yīng)用部署服務(wù)。在現(xiàn)有的云計(jì)算服務(wù)模式下,計(jì)量計(jì)費(fèi)系統(tǒng)是實(shí)現(xiàn)云端資源經(jīng)濟(jì)效益的必備手段,此外如何在云環(huán)境下保證計(jì)量計(jì)費(fèi)系統(tǒng)運(yùn)行的穩(wěn)定性、準(zhǔn)確性與實(shí)時(shí)性以及如何通過計(jì)量計(jì)費(fèi)提高云服務(wù)的質(zhì)量已經(jīng)成為云計(jì)算領(lǐng)域亟待解決的問題。本文提出了一種動(dòng)態(tài)的計(jì)量計(jì)費(fèi)模型,可依據(jù)集群當(dāng)前資源稀缺度動(dòng)態(tài)調(diào)整資源價(jià)格。資源的稀缺程度在一定程度上反映了當(dāng)前的供求關(guān)系,因此通過動(dòng)態(tài)調(diào)整稀缺資源的價(jià)格可以平衡資源的供需關(guān)系。在動(dòng)態(tài)調(diào)價(jià)的基礎(chǔ)上,本文設(shè)計(jì)并實(shí)現(xiàn)了基于浮動(dòng)利率的Paa S平臺(tái)計(jì)量計(jì)費(fèi)系統(tǒng),并在日任務(wù)量千萬級(jí)的條件下對任務(wù)實(shí)例的資源用量進(jìn)行分析計(jì)算,給出用戶賬單。首先,本文介紹了云計(jì)算計(jì)量計(jì)費(fèi)的相關(guān)理論,包括進(jìn)程級(jí)虛擬化技術(shù)Docker、流式計(jì)算技術(shù)、分布式存儲(chǔ)以及相關(guān)平臺(tái)的計(jì)費(fèi)策略等。其次,本文設(shè)計(jì)并實(shí)現(xiàn)了基于浮動(dòng)利率的Paa S平臺(tái)計(jì)量計(jì)費(fèi)系統(tǒng)。該系統(tǒng)分為計(jì)量監(jiān)測模塊、流式計(jì)費(fèi)模塊、計(jì)費(fèi)策略管理模塊、賬單管理模塊以及展示模塊。其中計(jì)量監(jiān)測模塊包括指標(biāo)形式化描述、指標(biāo)采集器、指標(biāo)準(zhǔn)確性評(píng)估以及流式發(fā)布等功能,計(jì)量是計(jì)費(fèi)的基礎(chǔ),準(zhǔn)確可靠的計(jì)量系統(tǒng)可為計(jì)費(fèi)提供穩(wěn)定的數(shù)據(jù)。流式計(jì)費(fèi)模塊由實(shí)時(shí)計(jì)算和離線計(jì)算兩部分組成,在賬單處理上采用實(shí)時(shí)計(jì)算的方法保證賬單的實(shí)時(shí)性;在浮動(dòng)利率的生成上采用線性回歸方法分析浮動(dòng)利率,并在離線計(jì)算中完成。計(jì)費(fèi)策略管理模塊為整個(gè)計(jì)費(fèi)過程提供價(jià)格數(shù)據(jù),包括對計(jì)費(fèi)項(xiàng)目配置、影響因子調(diào)節(jié)以及操作記錄等功能。賬單管理模塊維護(hù)所有用戶賬單并提供賬單存取接口。展示模塊為用戶和管理員提供可視化的操作界面。最后,本文從指標(biāo)采集效率、計(jì)費(fèi)效率以及資源消耗等方面對基于浮動(dòng)利率的Paa S平臺(tái)計(jì)量計(jì)費(fèi)系統(tǒng)進(jìn)行了測試,測試結(jié)果驗(yàn)證了系統(tǒng)的可用性和高效性,滿足系統(tǒng)設(shè)計(jì)要求。
[Abstract]:With the arrival of cloud computing era, various service operators have launched a lightweight Paa S-based platform based on container technology, which provides efficient and convenient service for business application deployment. In the current cloud computing service mode, metering and billing system is an essential means to realize the economic benefits of cloud resources. In addition, how to ensure the stability of metering and billing system in cloud environment, Accuracy, real-time and how to improve cloud service quality through metering and billing have become the urgent problems in cloud computing field. This paper presents a dynamic metering and charging model, which can dynamically adjust the resource price according to the current resource scarcity of the cluster. The scarcity degree of resources reflects the current relationship between supply and demand to a certain extent, so dynamic adjustment of the price of scarce resources can balance the supply and demand relationship of resources. On the basis of dynamic price adjustment, this paper designs and implements the Paa S platform metering and charging system based on floating interest rate, and analyzes and calculates the resource consumption of the task instance under the condition of thousands of daily tasks, and gives the user bill. Firstly, this paper introduces the related theories of cloud computing metering and billing, including process level virtualization technology Docker, streaming computing technology, distributed storage and accounting policies of related platforms and so on. Secondly, this paper designs and implements a Paa S platform metering and billing system based on floating interest rate. The system is divided into measurement monitoring module, flow billing module, billing strategy management module, bill management module and display module. Measurement and monitoring module includes formal description of indicators, index collector, index accuracy evaluation and flow publishing functions. Measurement is the basis of accounting. Accurate and reliable measurement system can provide stable data for accounting. The flow charging module is composed of real-time calculation and off-line calculation. The real-time calculation method is used to ensure the real time of the bill, and the linear regression method is used to analyze the floating rate in the generation of floating interest rate. And completed in the offline calculation. The charging policy management module provides the pricing data for the whole charging process, including the functions of configuring the charging items, adjusting the influence factors and recording the operation. The billing management module maintains all user bills and provides billing access interfaces. The presentation module provides a visual interface for users and administrators. Finally, this paper tests the Paa S platform metering and charging system based on floating interest rate from the aspects of index collection efficiency, accounting efficiency and resource consumption. The test results verify the availability and efficiency of the system, and meet the system design requirements.
【學(xué)位授予單位】：哈爾濱工業(yè)大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2017
【分類號(hào)】：TP393.09

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 張晉芳;王清心;丁家滿;劉彥君;黃心;;一種云計(jì)算環(huán)境下大數(shù)據(jù)動(dòng)態(tài)遷移策略[J];計(jì)算機(jī)工程;2016年05期

2 單莘;祝智崗;張龍;付長冬;魏書曉;;基于流處理技術(shù)的云計(jì)算平臺(tái)監(jiān)控方案的設(shè)計(jì)與實(shí)現(xiàn)[J];計(jì)算機(jī)應(yīng)用與軟件;2016年04期

3 段春梅;;云計(jì)算分布式緩存技術(shù)在海量數(shù)據(jù)處理平臺(tái)中的應(yīng)用[J];智能計(jì)算機(jī)與應(yīng)用;2016年01期

4 伍陽;;基于Docker的虛擬化技術(shù)研究[J];信息技術(shù);2016年01期

5 蘇宇;沈蘇彬;;一種云計(jì)算資源使用的采集記賬方案[J];計(jì)算機(jī)技術(shù)與發(fā)展;2015年11期

6 馬億e，

本文編號(hào)：1998170