High Availability Real-Time Migration Technology for Hyper-V and VMware (1)

Microsoft's Hyper-V virtualization Management Program has aroused widespread attention in the industry since its launch. This document compares the high availability and fast migration capabilities of Hyper-V with the real-time migration capabilities of VMware's VMotion. Before conducting a detailed analysis, we also need to understand why high availability plays an essential role in virtualization applications.

Virtualization is an awesome technology. It has great advantages in reducing the total cost of ownership of enterprises. The most significant advantage is that it saves energy consumption. Taking a data center with 10,000 servers as an example, if we use the virtualization technology at a ratio of 2 to 1 to reduce the number of servers by half, the energy consumption and cost saved by the retired 5,000 servers are incredible.

The monthly energy bill reflects the problem most. In fact, it is very simple to achieve a integration ratio. Our internal IT department has been using virtual machines for many years. Currently, the number of virtual machines used exceeds 2,500, making IT easy to achieve proportional integration. With Hyper-V, the proportion of server consolidation will also rise rapidly ).

However, virtualization is not perfect. The main problem caused by virtualization is spof. In the past, users could run 20 workloads on each of their physical servers. When one of the servers goes down, although it is terrible, it will not affect the entire server network.

In a virtualized environment, if 20 workloads are running at the top of a server, what will happen if the physical server goes down? That means all the workloads will be paralyzed. This is not just a disaster. In fact, a large number of virtualization users reflect this single point of white space:

Virtualization is an amazing technology, but it cannot be deployed without a high availability solution. If the virtual machine goes down and there is no supporting high availability solution, the data center manager will be unemployed.

The above is a summary of the important role of high availability in virtualization deployment.

With the concept of high availability, we know that we need to provide solutions for planned and unexpected downtime. Planned downtime is a relatively easy solution among the two types of downtime, because it is also very common in the plan. Generally, hardware devices add memory, storage devices, or BIOS upgrades) or software patches are included in the planned downtime column. Most users reserve time for downtime early in the morning or on weekends ).

Unexpected downtime is relatively difficult to solve. When a server suffers a sudden power failure, the user wants the virtual machine running on the server to automatically restart other servers without user intervention.

Hyper-V hypervisor is combined with Failover tertering in the Windows Server 2008 operating system. Even if the user unplugs the power, all virtual machines can be automatically restarted on other nodes without user operations. In addition, the Windows Server 2008 Operating System Enterprise Edition and data center edition both provide this function without additional charges ).