Method and system for resource preprocessing in dispatching and deployment performing process of virtual machine

Abstract

The embodiment of the invention discloses a method and a system for resource preprocessing in the dispatching and deployment performing process of a virtual machine. The method comprises a virtual machine dispatching process and a virtual machine deployment performing process, wherein the virtual machine dispatching process comprises the steps of receiving a current virtual machine deployment request submitted by a user, transmitting the current virtual machine deployment request to a target physical server according to an initial resource logic value of each candidate physical server, adding the current virtual machine deployment request to a virtual machine deployment queue of the target physical server, and updating the initial resource logic value of the target physical server according to memory and hard disk resources consumed by the current virtual machine. Through the adoption of the embodiment of the invention, accurate data support can be provided for virtual machine dispatching decision through tracking a resource actual value and recording available resources of the physical server according to the resource logic value.

Description

technical field [0001] The present application relates to the field of cloud computing, and in particular to a resource preprocessing method and system in the process of virtual machine scheduling and execution deployment in a cloud computing system. Background technique [0002] Cloud computing is the growth, usage, and delivery model of Internet-based related services, usually involving the provision of dynamically scalable and often virtualized resources over the Internet. In the cloud computing system, when the cloud computing system performs virtual machine scheduling, it distributes the virtual machine deployment operation submitted by the user to a specific physical server for execution, and the newly generated virtual machine will consume resources such as memory and hard disk on the physical server. The goal is to make the physical server environment where the virtual machine runs can at least meet its needs, and to make the resources in the physical server cluster ...

AI Technical Summary

Problems solved by technology

[0005] However, the inventor found in the research process that, using the prior art scheme of using one field, when the tasks submitted by different users operate the field at the same time during the execution process, the access to the field is carried out simultaneously, that is to say, the system must Provide a complete mechanism to ensure the consistency of the field value among different threads

There are two problems here: The first aspect is the accuracy of resource data: if the virtual machine deployment queue in the current server is not empty, after any virtual machine deployment operation or virtual machine deletion operation is completed, only By referring to the resource usage parameters in the virtual machine configuration file, the value of this part is logically subtracted or recovered. There is a large or small deviation from the actual resource value. When these deviations accumulate to a certain stage, if the virtual The machine deployment queue is always in a non-empty state, and there will be a big discrepancy with the actual situation of resources, which will affect the accuracy of scheduling decisions

The second aspect is the concurrency and maintenance cost of resource data: the model overhead of maintaining the consistency of a field value is relatively large, and the modification of the field value at the same time appears to be concurrent, and the field is actually at any time There is only one thread for modifying the value, and the probability of resource competition will be greater

Images