一個以LDAP為基礎的雲端儲存系統

Abstract

近年來，雲端科技產業正蓬勃發展，Google公司是廣為人知雲端企業的代表，該公司提供許多的雲端應用服務。使用者能夠透過網路來存取上面的各種應用服務以及關於個人的資訊。雲端除了須提供運算服務之外，還需要有強大儲存的功能，因此雲端儲存之相關技術發展做為未來雲端技術來說相當重要。關於雲端儲存技術，就不能不提到分散式檔案系統，分散式檔案系統在大型應用系統平台及雲端運算環境中，扮演重要角色，它能讓系統管理員簡化使用者對分散在網路中之檔案的存取和管理。

常見的DFS，像Gluster File System雖然具有良好的效能與空間擴充性， 但必須一開始就決定整個叢集式系統架構，並且無法線上擴增空間。 Hadoop也有諸如 Namenode有單一節點故障的問題，使得Namenode如果故障，無法自動回復。

參考上述不同DFS的不同優缺點，本計畫預計建立一個以LDAP為基礎的雲端儲存系統，能夠讓使用者進行線上儲存與備份。本系統能依照使用者所需的虛擬磁碟空間，以及效能要求，來服務使用者。

而本系統的特點是:加上了個人的使用單位以及可以平行存取。透過個人為使用單位來區隔出使用者的區塊，並且可以根據個人需求來設定所需要的服務。而平行存取的部分，則是希望能透過新的NFS4.1的protocol來提升檔案access的速度，透過將檔案拆放在不同的storage上，達到同時平行讀取檔案提升速度。並且透過備份的功能讓系統提升更好的failover效果。

並且在同質性的方面，與個人雲的系統像是owncloud做比較，期望能達到做出不同個人雲的效果。

In recent years, the cloud technology industry is booming, Google is well-known representative of the cloud enterprise, the company provides many cloud application services.Users can access over the network at the various application services as well as information about individuals. Addition to providing computing services, cloud also need a powerful storage , therefore cloud storage technology development as future cloud technology is very important. When it comes to cloud storage technology, can not fail to mention the Distributed File System, Distributed File System plays an important role in the large-scale application platform and cloud computing environments. It allows system administrators to simplify user access and manage files distributed across a network of. Common distributed file system, like Gluster File System has a good performance and spatial scalability, but it must be the beginning it was decided the entire cluster system architecture, and can not be amplified space online. Hadoop like Namenode single node failure problems, makes failure Namenode, can not recover automatically. In response to the different distributed file system have different advantages and disadvantages, this paper is expected to create a personal cloud storage system that allows users to online storage and backup. The system can, in accordance with the amount of disk space required by the user, as well as performance or backup requirements to service users. The system with other systems significant difference is this: with the personal use of units and parallel access. Through personalized to separate the blocks of the user, and can be set according to the personal needs the services they need. The parallel access part, is hoping to improve the speed of file access through the new NFS4.1 the protocol. Through separate files in different storage to read the file in parallel to enhance the speed. And through the backup function allows the system to enhance better failover.