具負載平衡機制與資源調適之雲端服務中介平台設計與實作

A Cloud Service Middleware with Load Balance Mechanism and Resource Adaptation: Design and Implementation

許育倫(Yu-Lun Hsu)

指導教授 ： 龔旭陽

國立屏東科技大學/管理學院/資訊管理系所/碩士(2013年)

https://doi.org/10.6346/NPUST.2013.00057

摘要

資訊科技快速演進，從傳統分散式運算(Distributed Computing)、網格運算(Grid Computing)至現今雲端運算(Cloud Computing)，其中虛擬化(Virtualization)技術使硬體資源更具彈性(Flexible)，隨著需求增加或減少，可輕易達到快速部署與動態資源調度，更是雲端運算能夠快速發展之關鍵因素，雲端運算服務提供使用者經由網際網路即能遠端存取高效能(High Performance)運算、分散式儲存(Distributed Storage)與通訊等服務，在近幾年雲端運算已逐漸成為各家資訊科技業者發展之重點目標。
有鑑於此，本論文提出一在資源可變下的雲端服務最佳化模型(Optimization Model)，透過使用者願支付成本、不同等級虛擬機器效能以及租賃價格，在使用者需求量的變動下進行資源的調整。系統實作部份，結合具彈性與跨異質(Crossing Heterogeneous)平台之網路服務(Web Service)，提供服務開發者能夠註冊與部署之雲端中介平台，並設計一最小能源負載平衡(Energy-Minimization Load Balance, EM-LB)機制，考量實體機器溫度與功率消耗資訊，以維持實體機器於安全溫度範圍內之低功率使用，減少熱點產生與能源消耗，並且依據虛擬機器可用之運算資源計算服務所需耗費時間，挑選符合在服務水平協議(Service Level Agreement, SLA)下的虛擬機器進行分派，達成確保服務品質之目標。
最後，本論文利用雲端模擬工具CloudSim建構資料中心、大量的實體機器與虛擬機器等雲端環境，藉由模擬及分析EM-LB機制在雲端環境中所造成實體機器的溫度變化、可用性變化與能源消耗結果來驗證此機制之可行性，同時採用虛擬化軟體VirtualBox實現雲端服務中介平台之設計與實作，針對使用者不同需求量變化，透過LINGO軟體來建立雲端服務最佳化模型並求取最佳解。

關鍵字

雲端運算 ； 負載平衡 ； 資源最佳化 ； 服務水平協議 ； 網路服務

並列摘要

Cloud services can provide high performance computing, distributed storage and scalable services to users over a network. Virtualization technology is the key issue of the cloud computing services. The fundamental function of virtualization is to dynamically increase or decrease the computing resource allocation based on the cloud service demands. In this study, we presented a Cloud Service Optimization Model for resource adaptation control, which considers the user's cost, different levels virtual machine of efficacy and rental price. We proposed a Cloud Service Middleware (CSM), which supports flexibility of web services via heterogeneous platforms. Developers can conveniently register and deploy the web services to the middleware. The middleware provides an Energy-Minimization Load Balance (EM-LB) mechanism, which considers the physical machine's temperature, power consumption, and the computing resources of virtual machines. EM-LB mechanism maintains a safe temperature range and low power usage to reduce hotspot and avoid the high consumption of energy. Furthermore, EM-LB mechanism fairly schedules the operations of virtual machines to achieve service quality. The simulation results shows the good efficiency of the proposed Cloud Service Optimization Model and EM-LB mechanism. Finally, we implement the proposed control mechanism to show the feasibility of this study.

並列關鍵字

Cloud Computing ； Load Balance ； Resource Optimization ； Service Level Agreement ； Web Service

參考文獻

[4] 鐘嘉德、高天助、楊嘉栩，雲端運算與產業發展，研考雙月刊，第34卷，第4期，第20-31頁，2010。

[16] L. Ramos and R. Bianchini, "C-Oracle: Predictive Thermal Management for Data Centers", 14th International Symposium on High-Performance Computer Architecture, 2008.

[5] A. Beloglazov, J. Abawajy and R. Buyya, "Energy-aware resource allocation heuristics for efficient management of data centers for Cloud computing", Future Generation Computer Systems, Vol. 28, Issue 5, pp. 755–768, 2012.

[6] A. Beloglazov and R. Buyya, "Optimal Online Deterministic Algorithms and Adaptive Heuristics for Energy and Performance Efficient Dynamic Consolidation of Virtual Machines in Cloud Data Centers", Concurrency and Computation: Practice and Experience, Vol. 24, Issue 13, pp. 1397-1420, 2012.

[9] D. Al-Shammary, I. Khalil, and L. E. George, "Clustering SOAP Web Services on Internet Computing Using Fast Fractals", IEEE International Symposium on Network Computing and Applications, 2011.