DC 欄位 |
值 |
語言 |
DC.contributor | 資訊工程學系在職專班 | zh_TW |
DC.creator | 李壹維 | zh_TW |
DC.creator | Yi-wei Li | en_US |
dc.date.accessioned | 2016-3-15T07:39:07Z | |
dc.date.available | 2016-3-15T07:39:07Z | |
dc.date.issued | 2016 | |
dc.identifier.uri | http://ir.lib.ncu.edu.tw:88/thesis/view_etd.asp?URN=101552031 | |
dc.contributor.department | 資訊工程學系在職專班 | zh_TW |
DC.description | 國立中央大學 | zh_TW |
DC.description | National Central University | en_US |
dc.description.abstract | 雲端運算在近年來已經越來越普及,帶動了網路流量與儲存資料的爆炸性成長。使用者不需要了解雲端中基礎設施的專業知識,也不需要自己管理、控制,使用者只需要專注於所需的資源與服務。而雲端運算的應用方式通常是以虛擬化的形式,把資訊技術,包括運算、儲存及網路頻寬,以服務的方式透過網路提供給使用者。如何能夠有效率的管理分配虛擬資源來滿足使用者需求則成為了雲端計算一項重要的課題。
CloudReports是根據雲端運算模式的一個模擬分散式運算環境的圖形化工具。本研究中我們實施了一些常用的演算法在CloudReports上,並仿照Amazon EC2執行環境進行了有關完成時間 (makespan) 和決策時間的幾個模擬,查看在同質和在異質環境中的性能表現。我們的結果顯示,在同質環境中Max-min及Round-robin會是比較好的選擇,但其他啟發式演算法表現上就沒有特別出色,且決策時間花費也比較長。異質環境方面,在平均表現上Max-min會比較好,但如果在任務數量及長度最多的時候,螞蟻演算法卻擁有較好的性能表現,當然也花費了相當長的決策時間去尋找最佳解。另外異質環境中在有限的VM時,基因演算法能夠在任務數量及長度最多時表現最佳。 | zh_TW |
dc.description.abstract | Cloud computing has become increasingly popular in recent years. One of the important issues in cloud computing is resource allocation for different kinds of tasks. Choosing a good scheduling algorithm for different kinds of computing jobs is the key to utilize resources efficiently. To this end, this study aims to investigate how different scheduling algorithms perform on different kinds of virtual environment, which may consist of heterogeneous virtual machines or homogeneous virtual machines. We have implemented several scheduling algorithms on CloudReports, which is a graphical tool for simulation of distributed computing environments based on the cloud computing model. The algorithms to be evaluated include random scheduling algorithms, heuristic scheduling algorithms, and meta-heuristics-based algorithms. We have conducted several simulations to evaluate the performance of various scheduling algorithms in terms of makespan of tasks and decision time of scheduling, given different kinds of system and task configurations. Our results show that, Max-Min and Round-Robin would be better choices in a homogeneous environment. Heuristic algorithms are not favored in a homogeneous environment since they need long decision time and may not achieve a good makespan. Generally, Max-Min would be a better choice in a heterogeneous environment. However, as the number of tasks and the length of each task become large, meta-heuristic algorithms tend to outperform Max-Min in makespan. | en_US |
DC.subject | 雲端運算 | zh_TW |
DC.subject | 雲端模擬 | zh_TW |
DC.subject | 排程演算法 | zh_TW |
DC.subject | 啟發式演算法 | zh_TW |
DC.subject | Cloud Computing | en_US |
DC.subject | Cloud Simulation | en_US |
DC.subject | Task Scheduling | en_US |
DC.subject | Heuristic Algorithms | en_US |
DC.subject | Meta-Heuristic Algorithms | en_US |
DC.title | 基於不同分配策略針對雲端環境中的任務排程及比較 | zh_TW |
dc.language.iso | zh-TW | zh-TW |
DC.title | Task scheduling for the comparison of different allocation strategies in the cloud environment | en_US |
DC.type | 博碩士論文 | zh_TW |
DC.type | thesis | en_US |
DC.publisher | National Central University | en_US |