應用全球伺服器負載平衡技術提升資訊服務—以M公司為例

Abstract

個案公司因為業務量提高導致資訊系統服務負載愈來愈沉重，網路頻寬滿載進而導致客戶與員工時常存取內部資源時發生服務異常，為了改善上述的情況個案公司導入全球伺服器負載平衡解決方案，因應跨國企業世界各地的服務需求，必須保持其各項服務的可用性與反應速度，並且避免單一資料中心任何服務故障發生，而正常服務可自動且快速地引導使用者轉移至鄰近資料中心存取備援資源以達到服務高可用性。

個案公司對於全球伺服器負載平衡解決方案的應用是非常完整的，可成為其他跨國企業的Reference Site，與導入全球伺服器負載平衡解決方案的成功典範公司，且目前導入全球伺服器負載平衡解決方案的台灣企業並不多，故本研究透過深度訪談法與參與觀察法針對個案公司進行個案研究，本個案研究報告說明M公司如何透過導入全球伺服器負載平衡解決方案及調整網路基礎架構、應用程式架構與資料同步方式，得出全球伺服器負載平衡技術提升資訊服務品質效率與效益的Know How，藉以提供欲導入全球伺服器負載平衡技術解決方案的跨國企業做為參考依據或盤點本身已提供服務項目是否已具備導入全球伺服器負載平衡技術之必要條件得以建置高可用性與效能之服務環境。

Due to M company's rising business volume, quality of information access service becomes highly sensitive to bandwidth bottlenecks and congestion. Therefore, scalability, high availability, and accelerated performance of the information system become critical to the success of company's global business. Centralized information system architecture represents big challenges to a global enterprise like M Company which must serve localized content to users in different parts of the world. This kind of architecture often exists a single point failure. If a site loses connectivity to the Internet, it will be inaccessible to users which prompt to significant impact to the business. Global Service Load Balance (GSLB) conquers these problems by intelligently distributing traffic to different geographic locations in the Internet. Users can be automatically directed to the nearest or least loaded servers if network bandwidth bottlenecks have occurred.

By excising the in-depth interview and participant observation, it has come to my conclusion that the practice of deploying the cost effective, high performance load balancing and traffic management solution in M Company has proven to be a successful text-book model in Taiwan. The case study illustrates methodologies and know-how's of network infrastructure architecture layout, application program design, and data synchronization model which support a high level of availability, redundancy, and fault tolerance. The model can serve as a best practice reference to enterprises which are looking for solving their existing data inquiry issues and building their next generation load balancing and traffic management solutions.