近年來,許多在網路上的應用程式以及網路服務蓬勃發展,使得網路人口爆炸性的增加,熱門的網站在每秒可能有很多用戶同時瀏覽,造成後端伺服器無法負荷,並且隨著雲端運算發展迅速,用戶需要的是更多的服務和更好的品質。因此,我們所要思考的是該如何加強雲端運算的整體性能並為用戶提供更滿意,高效的服務。大部份伺服器均利用叢集式的網路架構來解決這個問題,利用很多台伺服器來分享負載跟減少伺服器的反應時間,除此之外,負載均衡是一項重要的技術,可以提高叢集式系統架構的性能。在此論文中,我們提出了基於HTTP redirection架構下的負載平衡演算法,在一般情況以及在突發流量狀況下,使用不同的負載均衡演算法。一般情況下採用weighted round robin演算法,突發流量狀況下採用round robin演算法。實驗的結果顯示我們的方法,能夠使負載平衡以及反應時間得到最佳化。
In recent years, due to the explosive growth of Internet applications and services, the population of Internet increases rapidly. Popular websites may have many users per second while browsing caused the back-end server denial of service. With the rapid development of cloud computing, users need more services and better qualities. Therefore, we have to think for how to enhance the performance of cloud computing and provide users more and better services. Most servers utilize the cluster architecture to solve this problem. Take advantage of many servers to share loading and decrease server response time. In addition to this, load balancing is a very important technology to enhance the performance of cluster architecture. In this thesis, we proposed a load balancing algorithm based on HTTP redirection architecture. Using different algorithms in normal situation and burst situation. We use weighted round robin algorithm in normal situation and use round robin in burst situation. Experiment result shows our proposed scheme can really optimize the back-end server loading and response time.