KVM虛擬機器容錯系統實作

The Fault Tolerant KVM implementation

陳信宏(Shin-hong Chen)

指導教授 ： 梁德容

國立中央大學/資訊電機學院/軟體工程研究所/碩士(2014年)

摘要

隨著虛擬化技術廣泛性的使用，一台實體機器不再只用於架設一個伺服器，而是將實體機器虛擬成數台虛擬機器來架設多個伺服器，於是製造伺服器的廠商會想將虛擬化技術直接安裝於自家生產的伺服器上來作為統包方案，來增加產品的附加價值，但是虛擬化會衍生出單點故障的問題，即是實體機器故障影響所有執行於其上的虛擬機器，使得可用性下降，只要製造伺服器的廠商可以使用高可用性虛擬機器當成他們的統包方案，就可以提升更多產品的附加價值，於是可以看出高可用性虛擬機器的重要，但是目前商品化的高可用性虛擬機器僅有VMware一家，即是說廠商要繳昂貴的權利金給VMware來換取授權，於是我們實作高可用性虛擬機器於KVM開源虛擬機器專案，做到效能比的上VMware，使這些尋找高可用性虛擬機器統包方案的廠商有一個低成本方案的選擇。

關鍵字

KVM ； 虛擬機器 ； 高可用性 ； 容錯

並列摘要

With the extensive use of virtualization technology, a single physical machine is no longer just set up as a single server. Instead, single physical machine is virtualized into several virtual machines to host multiple servers. So the manufacturers of server want to install virtualization solution on their products as a turnkey solution to increase the added value of products. However, virtualization will derive a single point of failure problem that is a physical machine failure affects all virtual machines running on it. This decreases the availability of virtualization. It can improve lots of value for the server product as long as the manufacturers of server use high-availability virtual machine as their turnkey solutions. So we can see the importance of high-availability virtual machine. Currently there is only VMware which has commercial high-availability virtual machine product. The manufacturers of server need to pay expensive royalties to VMware to get authorization. Here we implement a highly available virtual machines in KVM virtual machine open source project that achieves comparable performance of VMware, so the manufacturers of server looking for these high-availability virtual machine turnkey solution has a choice of a low-cost highly available virtual machine solution.

並列關鍵字

KVM ； virtual machine ； high-availability

參考文獻

[1] R. P. Goldberg, "Survey of virtual machine research," Computer, vol. 7, pp. 34-45, 1974.

[2] G. J. Popek and R. P. Goldberg, "Formal requirements for virtualizable third generation architectures," Communications of the ACM, vol. 17, pp. 412-421, 1974.

[4] W.-C. Feng, "Making a case for efficient supercomputing," Queue, vol. 1, p. 54, 2003.

[9] D. Petrovic and A. Schiper, "Implementing virtual machine replication: A case study using xen and kvm," in Advanced Information Networking and Applications (AINA), 2012 IEEE 26th International Conference on, 2012, pp. 73-80.

[13] QEMU. Available: http://wiki.qemu.org/Main_Page