Openstack運算架構及虛擬機器 高可靠度保護機制;High Availability protection for Openstack Compute Architecture and User VM

NCUIR > College of Electrical Engineering & Computer Science > Software Engineer > Electronic Thesis & Dissertation > Item 987654321/77375

Please use this identifier to cite or link to this item: http://ir.lib.ncu.edu.tw/handle/987654321/77375

Title:	Openstack運算架構及虛擬機器高可靠度保護機制;High Availability protection for Openstack Compute Architecture and User VM
Authors:	李祥麒;Lee, Shiang-Chi
Contributors:	軟體工程研究所
Keywords:	雲端運算;Openstack;高可靠性;軟體定義叢集;虛擬機器保護;Controller保護
Date:	2018-07-25
Issue Date:	2018-08-31 14:35:51 (UTC+8)
Publisher:	國立中央大學
Abstract:	隨著雲端運算技術的普及，越來越多企業或個人用戶都選擇將其網路服務建置在雲端環境運行，得益於雲端環境的彈性和方便性，使用者只需在網路上註冊虛擬機即可立即部屬自己的網路服務，不僅節省了維運機房的成本，也可以更有效的利用硬體資源。Openstack即是建置相關雲端環境的IaaS系統。當企業或個人使用戶註冊了虛擬機，若是虛擬機因為機器故障導致服務中斷，對用戶來說每分鐘的故障意味著每分鐘都在損失金錢，所以在雲端環境提供系統高可靠性(High Availability, HA)即成為很重要的功能。 Openstack雲端環境建置軟體從2010發表至今，並沒有針對虛擬機器HA提供一個有效的方法。因此中央大學資訊工程研究所的平行與分散式運算實驗室提出了一個針對在Openstack下提供軟體定義運算叢集(Software-Defined High Availability Cluster, SDHAC)，透過邏輯性的切割運算資源成不同的SDHAC提供自動偵測運算資源故障同時復原虛擬機器的機制，並搭配IPMI(Intelligent Platform Management Interface)提供硬體感測器的監控。本研究承襲此種架構，提出一種根據錯誤彼此相依性架構的階層式偵測流程，並提供針對Controller HA的保護，確保使用者不會因為Controller的故障而失去Openstack管理的功能，同時對外提供REST API (Representational State Transfer Application Programming Interface, REST API)的支援，方便外部系統透過統一的介面存取HA系統服務。 ;More and more enterprises and users tend to deploy their web services on cloud platform because of the increasing popularity in cloud computing technology. Benefited from the elastic and easy-to-use features on the cloud platform, users only need to register virtual machine by their hardware requirements then they can start to build their own web service. By this way, users can not only save the money from use the cloud platform rather than maintain their own physical machines but also make good use of the physical machines. Openstack is the kind of software that can build a cloud platform. When the physical machine encounters some errors in terms of the hardware and software then the virtual machine will also be inaccessible subsequently which will cause the money lose to the industry owners and users. So, high availability becomes a serious function in building a cloud platform. Openstack do not provide a formal solution for virtual machine HA since its released in 2012. NCU-PDC lab proposed a research about SDHAC(Software-Defined High Availability Cluster) based on Openstack cloud platform which will divide the compute resources to different SDHAC and combined with the IPMI hardware and software services level detection mechanism to decide the compute resource is in failure state or not. Once there are some errors in the compute resource they will start to recover it. This research propose and implement a hierarchically failure detection method based on the SDHAC and IPMI detection and provide the protection for the controller services which make sure that when some errors happens in the controller, the user can keep accessing to the controller services. Besides, this research supports REST API (Representational State Transfer Application Programming Interface) which allows external requests use the REST standard to perform HA functions.
Appears in Collections:	[Software Engineer] Electronic Thesis & Dissertation

Files in This Item:

File	Description	Size	Format
index.html		0Kb	HTML	268	View/Open

社群 sharing

Loading...