Management method and platform of virtualization system

Abstract

The invention provides a management method and platform of a virtualization system. The method is applied to a cloud management platform of the virtualization system, and comprises the following steps: firstly monitoring at least two host nodes corresponding to the virtualization system through a monitor system in the cloud management platform to acquire the current state of each host node in at least two host nodes, and then determining the second host node is to be in a normal state from at least two host nodes when the monitor system predictively determines a fact that the first host node with the current state as the normal state in at least two host nodes is to be in abnormal state; and finally migrating at least one logic partition on the first host node to the second host node through a partition management system in the cloud management platform. Through the management method provided by the invention, the ahead prediction and migration of the possibly abnormal node can be realized, the service partition service interruption time is reduced, or the effect of not interrupting the service is achieved.

Description

technical field [0001] The present invention relates to the field of cloud computing, in particular to a management method and platform of a virtualization system. Background technique [0002] The Power minicomputer of International Business Machines Corporation IBM (International Business Machines Corporation) is used by many enterprises to run important and critical business servers. In the financial industry, minicomputers account for more than 80% of the data center. With the development of cloud computing , The private cloud platform built by the PowerVM virtualization technology provided by IBM has gradually become the mainstream architecture of financial enterprise data centers, and how to use PowerVM to build a stable and highly available computing resource cloud platform has become the main problem faced by enterprise data centers. [0003] For the enterprise virtualization environment, PowerVM provides a dual virtual I / O server (VIOS) redundant architecture soluti...

AI Technical Summary

Problems solved by technology

[0003] For the enterprise virtualization environment, PowerVM provides a dual virtual I / O server (VIOS) redundant architecture solution for high availability, which can make up for the lack of customers being unable to distinguish whether they cannot work normally when a single VIOS is down. However, if both VIOS are down , or the physical host where the VIOS is located is due to a hardware failure, and the business partitions running on the physical host will also be interrupted. In some key applications, the computer system is often required to work continuously 24 hours a day, and unexpected failures Or downtime will cause immeasurable losses

[0004] At the same time, the current high availability on the market is not aimed at PowerVM virtualization (IBM minicomputers), but mostly for virtualization such as Vmware, Xen, and Ken (for X86 servers). At the same time, the monitoring systems on the market are nothing more than detecting physical machines. In the event of a network failure, ping is performed to determine whether the network is reachable. When it is unreachable, the virtual machine is migrated using its own algorithm. This kind of monitoring is relatively one-sided due to the monitoring parameters, which often leads to misjudgment, resulting in unnecessary migration and wasting additional resources. The network load affects the quality of the entire system. At the same time, it is impossible to predict whether there is a problem with the physical hardware in advance. Often, the virtual machine is migrated after the hardware fails.

Images