WO2020134364A1

WO2020134364A1 - Virtual machine migration method, cloud computing management platform, and storage medium

Info

Publication number: WO2020134364A1
Application number: PCT/CN2019/110999
Authority: WO
Inventors: 童遥; 孔鹏; 李华; 申光
Original assignee: 中兴通讯股份有限公司
Priority date: 2018-12-28
Filing date: 2019-10-14
Publication date: 2020-07-02
Also published as: CN111381928B; CN111381928A

Abstract

A virtual machine migration method, comprising: if a first load of a first physical machine is greater than a first threshold at the current first moment, predicting a second load of the first physical machine at a second moment, wherein the second moment is after the first moment; if the second load is greater than a second threshold, selecting, from the virtual machines on the first physical machine, a virtual machine to be migrated, wherein the second threshold is greater than the first threshold; determining a second physical machine, migrating the virtual machine to be migrated to the second physical machine, wherein the second physical machine is different from the first physical machine. A cloud computing management platform and a storage medium are further comprised.

Description

Virtual machine migration method, cloud computing management platform and storage medium

Technical field

The embodiments of the present invention relate to the field of virtual machine resource scheduling in a cloud computing data center, and in particular, to a virtual machine migration method, a cloud computing management platform, and a storage medium.

Background technique

The cloud computing management platform is a complex large-scale system, and the virtual machine is an important part of the cloud computing management platform. It has the characteristics of cross-system, resource isolation, and migration. Efficient virtual machine cluster deployment and dynamic migration operations can decentralize virtual It is an important part of the cloud computing management platform that the machine is integrated and meets the needs of large-scale applications. It can improve the operation, maintenance and service quality of the cloud computing management platform.

However, because the cloud computing management platform has the characteristics of huge resource scale, complex structure, and wide geographical distribution, and its application has the characteristics of flexible and variable requests, different resource types, and dynamic load changes, this is a virtual machine cluster deployment and dynamic The migration strategy presents a huge challenge. Dynamic migration occurs because of the need for load balancing. In the cloud computing environment, some computers may be overloaded and have been overloaded, while other calculations are often idle, which greatly reduces The overall resource utilization of the system. Load balancing is to adjust tasks among multiple physical resources to obtain the best overall resource utilization.

The migration triggering strategy in the related art is based on a fixed threshold. When a certain type of performance load of a node is greater than the fixed threshold, the system will trigger a migration; however, due to various types of sudden Therefore, once the sudden load at any instant reaches a fixed threshold, the virtual machine migration will be triggered, which leads to frequent virtual machine migration, unnecessary migration overhead, and system reliability cannot be guaranteed.

Summary of the invention

In view of this, the embodiments of the present invention are expected to provide a virtual machine migration method, cloud computing management platform, and storage medium, which solves that the sudden burst load in the related art will trigger the virtual machine migration when it reaches a fixed threshold, which leads to frequent Virtual machine migration causes unnecessary migration overhead, avoids the interference of the sudden load on the system, and improves the reliability of the system.

To achieve the above objective, according to an aspect of an embodiment of the present invention, a virtual machine migration method is provided, which is applied to a cloud computing management platform. The method includes:

If the first load of the first physical machine is greater than the first threshold at the current first moment, the second load of the first physical machine at the second moment is predicted; wherein, the second moment is after the first moment;

If the second load is greater than the second threshold, select the virtual machine to be migrated from the virtual machines on the first physical machine; wherein, the second threshold is greater than the first threshold;

Determine a second physical machine, and migrate the virtual machine to be migrated to the second physical machine; wherein the second physical machine is different from the first physical machine.

Optionally, if the first load of the first physical machine at the current first moment is greater than the first threshold, predicting the second load of the first physical machine at the second moment includes:

If the first load is greater than the first threshold, obtain a third load at least two moments within a preset time period; wherein the preset time period includes the first time and the time before the first time At least one third moment, the third load at each moment in the at least two moments includes at least two loads with different attributes; the loads with different attributes represent different types of resources occupied in the first physical machine;

Calculating the fourth load at each moment according to the third load at each moment and the weight value of the third load at each moment;

The second load is calculated based on the fourth load at each moment.

Optionally, the calculating the second load according to the fourth load at each moment includes:

The average value of the fourth load at multiple times is calculated to obtain the second load.

Optionally, if the first load is greater than the first threshold, before acquiring a plurality of different types of load data at each time among the multiple times before the first time, the method further includes :

Receiving a user request, and obtaining the request type of the user request; wherein, the request type includes a static request or a dynamic request;

Correspondingly, before calculating the fourth load at each moment according to the third load at each moment and the weight value of the third load at each moment, the method further includes:

Acquiring the attribute of each load in the third load at each moment;

According to the request type and the attribute of each load in the third load at each moment, the weight value of each load in the third load at each moment is determined.

Optionally, the third load at each moment further includes a distance, wherein the distance represents the distance between the first physical machine and the data center that receives the user request, and the cloud computing management platform further includes In the data center, the data center is used to manage a third physical machine, and the third physical machine includes the first physical machine.

Optionally, before calculating the fourth load at each moment according to the third load at each moment and the weight value of the third load at each moment, the method further includes:

Acquiring the attribute of each load in the third load at each moment;

According to the attribute of each load in the third load at each moment, the weight value of each load in the third load at each moment is determined.

Optionally, the number of virtual machines on the first physical machine is multiple, and if the second load is greater than a second threshold, selecting the virtual machine to be migrated from the virtual machines on the first physical machine includes: :

If the second load is greater than the second threshold, obtain the resource utilization rate of the first physical machine;

The virtual machine to be migrated is selected from multiple virtual machines on the first physical machine according to the resource utilization rate.

Optionally, the selecting a virtual machine to be migrated from multiple virtual machines on the first physical machine according to the resource utilization rate includes:

If the resource utilization rate is greater than or equal to a third threshold, predict a fifth load of each virtual machine on the first physical machine at the third moment; wherein the third moment is after the first moment;

Select the virtual machine corresponding to the maximum load among the fifth loads from the virtual machines on the first physical machine, and determine the virtual machine to be migrated.

If the resource utilization rate is less than or equal to a fourth threshold, it is determined that each virtual machine on the first physical machine is the virtual machine to be migrated; wherein, the fourth threshold is less than the third threshold.

Optionally, the determining the second physical machine includes:

Acquiring the sixth load of each physical machine in the cloud computing management platform at the first moment;

The physical machine with the sixth load less than the fifth threshold is selected from the physical machines in the cloud computing management platform, and determined as the third physical machine; wherein the third physical machine is different from the first physical machine, so The number of the third physical machine is multiple;

Predicting the seventh load of the virtual machine on each third physical machine at the fourth moment; wherein, the fourth moment is after the first moment;

Calculating the sum of the seventh load of each third physical machine and the load of the virtual machine to be migrated to obtain multiple target loads;

Determining a target load less than a sixth threshold among the plurality of target loads;

Selecting the third physical machine corresponding to the target load less than the sixth threshold from the third physical machine and determining it as the fourth physical machine; wherein the number of the fourth physical machine is at least one;

The second physical machine is determined from at least one fourth physical machine.

Optionally, the determining the second physical machine from at least one fourth physical machine includes:

Calculating the value of the seventh load of each fourth physical machine minus the value of the sixth load of each fourth physical machine to obtain the resource demand expansion and contraction amount of each fourth physical machine;

Calculating the value of the total resources of each fourth physical machine minus the value of the sixth load of each fourth physical machine to obtain the amount of unallocated resources of each fourth physical machine;

Calculating the value of the unallocated resources of each fourth physical machine minus the amount of resource demand expansion and contraction of each fourth physical machine to obtain the predicted remaining resource amount of each fourth physical machine;

The second physical machine is determined from the fourth physical machine according to the predicted remaining resource amount of each of the fourth physical machine.

Optionally, the determining the second physical machine from the fourth physical machine according to the predicted remaining resource amount of the fourth physical machine includes:

It is determined that the fourth physical machine corresponding to the maximum value in the predicted remaining resource amount of the fourth physical machine is the second physical machine.

Optionally, a virtual machine cluster is established on the physical machine of the cloud computing management platform, the number of the virtual machines to be migrated is at least two, the number of the virtual machines to be migrated and the number of the second physical machines The same; before migrating the virtual machine to be migrated to the second physical machine, the method further includes:

Obtaining the load of each of the virtual machines to be migrated;

Determining a target physical machine corresponding to each virtual machine to be migrated from a plurality of second physical machines; wherein the plurality of second physical machines include the target physical machine;

Acquiring loads of multiple of the target physical machines;

According to the load of each of the virtual machines to be migrated and the load of each of the target physical machines, sort the multiple target physical machines to obtain a priority queue.

Optionally, according to the load of each virtual machine to be migrated and the load of each target physical machine, sorting the plurality of second physical machines to obtain a priority queue includes:

Calculating the sum of the load of each virtual machine to be migrated and the load of the target physical machine corresponding to each virtual machine to be migrated to obtain a transmission load;

Sort multiple target physical machines according to the transmission load to obtain a priority queue.

Optionally, the migrating the virtual machine to be migrated to the second physical machine includes:

Determine the priority of each of the target physical machines according to the priority queue;

Mirroring the virtual machines to be migrated corresponding to each of the target physical machines in order of high priority to low priority; wherein the transmission load corresponding to the target physical machine with high priority is less than that with low priority The transmission load corresponding to the target physical machine.

According to another aspect of the embodiments of the present invention, a cloud computing management platform is also provided. The cloud computing management platform includes:

Memory, used to store executable instructions;

The processor is configured to execute the executable instructions stored in the memory to implement the steps in the above virtual machine migration method.

According to another aspect of the embodiments of the present invention, a storage medium is also provided, where the storage medium stores one or more programs, and the one or more programs may be executed by one or more processors to implement, for example, The steps of the above virtual machine migration method.

In the virtual machine migration method, cloud computing management platform, and storage medium provided in the embodiments of the present invention, if the first load of the first physical machine at the first moment is greater than the first threshold, the second load of the first physical machine at the second moment is predicted ; Where the second moment is after the first moment; if the second load is greater than the second threshold, select the virtual machine to be migrated from the virtual machines on the first physical machine; wherein, the second threshold is greater than the first threshold; determine the second physical Machine, and migrate the virtual machine to be migrated to the second physical machine; where the second physical machine is different from the first physical machine; that is to say, in the embodiment of the present invention, the burst of the first physical machine is determined at the first moment When the sexual load is greater than the first threshold, the load of the first physical machine at the future moment is predicted, and when it is determined that the load at the future moment is higher than the second threshold greater than the first threshold, a migration is triggered, which is effective This avoids the interference of the sudden load on the system, and solves the problem that the instantaneous sudden load in the related technology will trigger the migration of the virtual machine when it reaches a fixed threshold. This leads to the problem of frequent virtual machine migration and unnecessary migration overhead. The interference of the sudden load on the system is avoided, and the reliability of the system is improved.

BRIEF DESCRIPTION

1 is a schematic diagram of a functional architecture of a cloud computing platform management system provided by an embodiment of the present invention;

2 is a schematic diagram of an implementation architecture of a cloud computing platform management system provided by an embodiment of the present invention;

3 is a schematic diagram of an agent implementation architecture provided by an embodiment of the present invention;

4 is a schematic diagram of a virtual machine migration model provided by an embodiment of the present invention;

5 is a schematic flowchart of a virtual machine migration method according to an embodiment of the present invention;

6 is a schematic flowchart of another virtual machine migration method according to an embodiment of the present invention;

7 is a schematic flowchart of another virtual machine migration method according to an embodiment of the present invention;

8 is a schematic diagram of a virtual machine cluster deployment architecture based on resource classification provided by an embodiment of the present invention;

9 is a schematic flowchart of a virtual machine migration method according to another embodiment of the present invention;

10 is a schematic flowchart of another virtual machine migration method according to another embodiment of the present invention;

11 is a schematic structural diagram of a cloud computing platform management system according to an embodiment of the present invention.

detailed description

The technical solutions in the embodiments of the present invention will be described clearly and completely in conjunction with the drawings in the embodiments of the present invention.

It should be understood that the “embodiments of the present invention” or “previous embodiments” mentioned throughout the specification mean that specific features, structures, or characteristics related to the embodiments are included in at least one embodiment of the present invention. Therefore, “in the embodiments of the present invention” or “in the foregoing embodiments” appearing throughout the specification does not necessarily refer to the same embodiments. In addition, these specific features, structures, or characteristics may be combined in one or more embodiments in any suitable manner. In various embodiments of the present invention, the size of the sequence numbers of the above processes does not mean that the execution order is sequential, and the execution order of each process should be determined by its function and inherent logic, and should not correspond to the implementation process of the embodiments of the present invention Constitute any limitation. The sequence numbers of the above embodiments of the present invention are for description only, and do not represent the advantages and disadvantages of the embodiments.

Before describing the embodiments of the present invention in further detail, the virtual machine migration technology in the related art will be described.

However, because the cloud computing management platform has the characteristics of huge resources, complex structure, and wide geographical distribution, and its application has the characteristics of flexible requests, different resource types, and dynamic changes in load, etc., it deploys and dynamically migrates virtual machine clusters. The strategy presents a huge challenge. Dynamic migration occurs because of the need for load balancing. In the cloud computing environment, some computers may be overloaded and have been overloaded, while other calculations are often idle, which greatly reduces The overall resource utilization rate of the system. Load balancing is to adjust tasks among multiple physical resources to obtain the best overall resource utilization.

According to the characteristics of cloud computing management platform resources, through analysis, we can know that the main difficulties of dynamic migration are the following three aspects: the response to sudden loads, the huge number of system virtual machines, and the guarantee of system reliability.

In the embodiments of the present invention, the above three aspects are briefly introduced:

First, the response to sudden loads

With the widespread use of cloud computing management platforms, the size of users may become very large at a certain moment, which is one of the sudden loads. Taking November 11, 2017 as an example, the number of visits to a website can be reached within a short period of time, which is much higher than the average. This sudden load behavior poses a major challenge to system performance analysis and dynamic migration of virtual machines.

Second, the number of system virtual machines is huge

The cloud computing management platform has thousands of servers, and each physical machine may have dozens of virtual machines. The resource types of these virtual machines vary, and performance data is being generated all the time. Faced with such large-scale servers and data, it is difficult to directly analyze the performance of the cloud computing management platform using traditional methods.

The third aspect, the guarantee of system reliability

The reliability of the system requires that on the one hand, dynamic migration cannot make the system overhead too large to interfere with the normal operation of the system; on the other hand, dynamic migration cannot interfere with applications deployed in virtual machines. This requires dynamic migration to select virtual machines that really need to be migrated, and physical nodes that will not increase the load excessively for a period of time in the future, so as to ensure that virtual machine migration will not interfere with existing virtual machines and applications.

It should be noted that, in addition to the above three aspects, we should also pay attention to the prevention of shock after the virtual machine migration. Because multiple nodes may select the same physical node machine as the target node based on the current load information, the load of the selected target node increases sharply. After exceeding the threshold set by the administrator, it will cause two or even three migrations. Migration shocks.

In related technologies, some systems use operating system virtualization technology, such as non-downtime migration technology, which migrates the virtual machine to another lightly loaded server without the user's awareness. It provides regular management and maintenance of the system. And error recovery provides a brand new experience. Although the system consumes less performance when migrating, the system cannot effectively apply sudden loads.

In the related art, some systems do not need to modify the operating system and kernel system during migration, so that the service can run without interruption; but during the migration process, the migration entity is the operating system image and application, the migration efficiency is low, and the migration shock is not resolved The problem.

In related technologies, there is a seamless migration technology between current servers based on Intel processors and future technologies. This technology has good backward compatibility, but the migration system does not fully consider the performance relationship before and after migration, and migration may be shocked. Increase system overhead.

It can be known from the virtual machine migration technology adopted in the above-mentioned related technologies that the current dynamic migration of virtual machines still has defects, such as:

(1) The migration strategy of most load balancing methods is only based on a fixed threshold, that is, the load of the node machine exceeds the upper limit of the load threshold, then the node machine triggers a migration, so that the peak burst load will be instantaneous Triggering the migration of virtual machines can easily cause unnecessary migration overhead.

(2) The current load prediction method either predicts the load of the virtual machine or the load of the application, and some methods of load prediction of the physical host. The load index range is narrow, and the singular value algorithm is used in the prediction method. Incremental weighting, etc., but the purpose of the prediction is to select physical machines that meet the conditions and migrate all virtual machines on it. During the migration process, the same physical node machine may be selected as the target node, resulting in a sharp increase in the load of the selected target node, resulting in Migration shocks.

(3) The current method of virtual machine migration involves cluster deployment. The virtual machine and the physical machine are sorted according to the central processing unit (CPU) occupancy rate or memory occupancy rate, and then one by one is matched, or the virtual machine is used. The ratio between the CPU usage and the memory usage is sorted. The sorting result matches the sorting result of the physical machine. The reference index is narrow and the improvement of the migration efficiency of the virtual machine is not significant.

(4) The current image transfer method involved in virtual machine migration is based on the sequential copying of the cache of the virtual machine image to the target physical host. When the size of the virtual machine cluster to be migrated is relatively large, there is a defect of low efficiency.

(5) For the consideration of distance factors in the current virtual machine migration, only the single data center scenario is considered, and the closest physical host is considered, and the multiple data center scenario is not considered, that is, there is no means to prevent the migration of virtual machines across data centers. .

According to the foregoing embodiments, the embodiments of the present invention provide a virtual machine migration method, which is applied to a cloud computing management platform. The cloud computing management platform may also be referred to as a cloud computing management platform system, which may also be referred to as a system hereinafter; the system uses a multi-level distribution The architecture has multiple data centers, and each data center has multiple physical machines and a data center server. The data center is connected to the system server through the network.

The client accesses the system server. A system server manages all the data centers. Each data center is equipped with a data center server. The data center server manages all physical machines in the data center and is installed on each physical machine. The agent is responsible for processing the commands sent by the data center server and responding to the operation results. This distributed multi-layer management structure can easily realize the expansion of the system, and is beneficial to reduce the pressure on the server.

The following describes an exemplary functional architecture of a cloud computing management platform system that implements an embodiment of the present invention. Referring to FIG. 1, the cloud computing management platform system 10 includes an image file management module 110, a cluster management module 120, a performance monitoring module 130, and a resource management module 140, described separately below.

The image file management module (Image Manager) 110 can upload and download images, classify image resource types, and copy management. The image resource type is initially specified by the user, and can be revised later by the system.

The virtual machine cluster management module (Cluster Manager) 120 can implement cluster editing, cluster deployment and cluster deletion.

Performance monitoring module (PerformanceMonitor) 130 can realize physical machine performance monitoring, virtual machine performance monitoring and system load performance prediction.

Resource management module (Resource Manager) 140 can realize physical resource addition, physical resource deletion, physical resource editing and image library management.

The implementation architecture of the cloud computing management platform system is shown in FIG. 2 and includes a presentation layer 210, a logic layer 220, and a data persistence layer 230, which are described below.

The presentation layer 210 is a display interface for user interaction with the system, and displays system performance, operation results, and cluster operation status through tables, graphics, and text.

The logic layer 220 is the core layer of the cluster management strategy. It is responsible for responding to the operation commands sent by the presentation layer, and provides access interfaces to display the system operation status. It provides five sub-modules, which are the Resource Manager module and performance management. (Performance Manager) module, virtualization engine (Virtualization Engine) module, agent management (Agent Manager) module, front-end interactive interface API Interface module.

The Agent Manager module is mainly deployed on each physical machine and virtual machine of the system. The agent on the physical machine is mainly used to respond to the operation commands sent by the data center server, such as virtual machine creation, start, stop, destruction, migration Wait, it has to collect performance data and submit to the system. The Agent in the virtual machine is mainly used to send the current performance status of the virtual machine to the system. The collected performance data mainly includes CPU utilization, network bandwidth utilization, memory utilization, and file system status over a period of time. , Cluster deployment files, cluster startup and other operations, need to install Agent on each physical machine to participate, Agent is also responsible for sending machine performance status information to the server, the implementation architecture of Agent is shown in Figure 3, including image files The processing module 310, the virtualization event processing module 320, and the performance monitoring module 330 will be described separately below.

The image file processing module 310 can realize the reception of the image file image and the transmission of the image file.

The virtualization event processing module 320 can implement virtual machine startup, virtual machine creation, virtual machine stop, and virtual machine deployment.

The performance monitoring module 330 can implement performance monitoring on database programs, file transfer programs, and Extensible Markup Language XML programs.

The Performance Manager module is responsible for analyzing the data collected by the Agent Manager to determine the current operating status of the entire and each physical. The Performance Manager module can predict the performance of the system that may occur in the next time. When the load is too heavy, it can issue an overload warning to the system to remind managers to increase physical resources or adjust the load through load balancing. This module provides decision support for the deployment and migration of virtual machines and users viewing the current system operating status.

The Virtualization Engine module is the core module of the system. It can accept user requests to implement a series of virtualization operations, including the creation of virtual machines, the start of virtual machines, the stop of virtual machines, the destruction of virtual machines and other events. After the cluster deployment request is received, after XML parsing of the cluster deployment request, if it is determined that the cluster can be deployed, the cluster deployment process is performed, such as passing the image, configuring the cluster deployment file, starting the cluster, etc. It is sent to the Agent Manager module in the form of commands for specific operations. It also accepts the information from the Performance Manager module. When a machine is overloaded, it migrates some virtual machines to a lighter-loaded machine.

The Resource Manager module is mainly responsible for managing physical resources and monitoring the current physical resource usage. These physical resources mainly include machine CPU, memory, file system, and network. This module also manages the resources of each mirror library.

The data persistence layer 230 provides various methods for data persistence, the database program (Database) Provider provides support for database access, and the file service program (File-Server Provider) persists large data in the form of files, such as mirror files , XML program (XML Provider) is mainly used to support the reading and writing of XML files.

In the embodiment of the present invention, the cloud computing management platform system 10 further includes a virtual machine dynamic migration controller (not shown in FIG. 1).

The virtual machine dynamic migrator is composed of two large modules, a workload prediction module and a migration control module, as shown in FIG. 4, and the migration control module may include a migration trigger module, a virtual machine selection module to be migrated, and a target physical machine selection module. Part; Among them, the migration trigger module is set to determine the timing of migration, that is, determine when a virtual machine should be migrated, to ensure that the performance of the entire physical machine always maintains a good state; the virtual machine selection module to be migrated is set to determine which virtual The machine should be migrated; the target physical machine selection module is set to determine to which target physical machine the virtual machine to be migrated should be migrated.

Based on the foregoing embodiments, embodiments of the present invention provide a virtual machine migration method, which is applied to a cloud computing management platform; as shown in FIG. 5, the method includes the following steps:

Step 101: If the first load of the first physical machine at the first moment is greater than the first threshold, predict the second load of the first physical machine at the second moment.

Among them, the second moment is after the first moment.

In the embodiment of the present invention, the cloud computing management platform determines that the first load of the first physical machine at the first moment is greater than the first threshold, and does not directly trigger the migration of the virtual machine on the first physical machine, but predicts the future at the second moment The second load of the first physical machine at the moment, so that the second load is used to further determine whether to trigger the migration of the virtual machine.

In the embodiments of the present invention, due to the diversification of applications in virtual machines, there are also huge differences in the balance of virtual machine events, resource usage, and load, and different migration strategies will have huge differences in migration costs. In order to support the decision-making system to select the correct virtual machine and migration destination to be migrated, it is necessary to predict the key indicator of workload.

In the embodiment of the present invention, the system can collect the load data of the physical machine in real time or periodically collect the load data of the physical machine, which lays a foundation for predicting the load data of the physical machine. Exemplarily, in practical applications, a total of 288 sample points can be collected within 24 hours with a sampling period of 5 minutes.

In the embodiment of the present invention, after the load data of the physical machine is collected in real time, the load data is saved, and further, the CPU utilization data of various services on the physical machine can be used as the dynamic requirements of the CPU parameters of the cloud task executed by the virtual machine of the physical machine the amount.

In the embodiment of the present invention, the load at the second time T2 may be calculated based on the load data at the first time T1 and the load data at multiple times within a specified time period before T1.

In another embodiment of the present invention, when it is determined that a sudden load occurs on the first physical machine at time T1, the load data at multiple times within a specified time period before T1 may be used to calculate the load at the second time T2.

As shown in FIG. 6, the migration controller in the system completes the migration under the guidance of the control module in the migration controller through the cooperation between the migration trigger module, the virtual machine selection module to be migrated, and the target physical machine selection module. Decision, selection of virtual machines to be migrated, migration of virtual machines to be migrated to target physical machines. In the embodiment of the present invention, the virtual machine dynamic migration controller adopts a performance prediction algorithm to determine the migration timing and select the migration target physical machine.

Step 102: If the second load is greater than the second threshold, select the virtual machine to be migrated from the virtual machines on the first physical machine.

Among them, the second threshold is greater than the first threshold.

In the embodiment of the present invention, the cloud computing management platform determines that the second load is greater than the second threshold, that is, when it is determined that the load in the future is higher than the second threshold greater than the first threshold, a migration is triggered; and, once the virtual machine is determined When the migration is triggered, select some or all virtual machines from the first physical machine for migration.

As can be seen from the above, in the cloud computing management platform of the embodiment of the present invention, through the migration trigger module, the system can determine the performance load status of a physical machine node and determine the time to trigger the migration. The migration triggering strategy in the related art is based on a fixed threshold, and when certain types of performance load of a physical machine node exceeds this specified threshold, the system will trigger a migration. However, in the embodiment of the present invention, it is considered that due to the presence of various types of sudden loads, this method of specifying the threshold will inevitably lead to frequent virtual machine migration; therefore, in the embodiment of the present invention, the first physical machine is determined at the first moment When the first load is greater than the first threshold, the virtual machine migration is not directly carried out, but the load of the past period and the current time is comprehensively considered, and then the future load is predicted, only when the future load meets the preset conditions. Trigger the migration of virtual machines. In this way, you can effectively prevent the interference of the sudden load on the system and reduce the performance loss of the system due to unnecessary migration.

In practical applications, combining the prediction results and current load information, the system will determine whether to trigger a virtual machine migration. In practical applications, first, it is analyzed whether the current load of a physical machine exceeds the threshold k1, and if it is exceeded, the next step is performed; second, in order to verify whether the current load information is a sudden load, it can be based on the physical machine's The load data is used to predict the load at the future time; then, after obtaining the load at the future time, if the load at the future time is greater than the threshold k1, the system will trigger a migration. On the contrary, the system will not trigger the migration, so it can effectively avoid the interference of the sudden load on the system.

It should be noted that, according to the method for determining the timing of virtual machine migration in the embodiment of the present invention, as long as a virtual machine migration command is triggered, certain virtual machines on this physical machine must be selected for migration. These virtual machines that need to be migrated The machine is called a candidate virtual machine, that is, a virtual machine to be migrated. The selection of the virtual machine to be migrated must meet a series of requirements. For example, the migration is short-lived and does not cause system migration shocks.

In the embodiment of the present invention, when the system selects the virtual machine to be migrated through the virtual machine to be migrated selection module, the system can be implemented through the following two steps:

Step B1: According to the predicted result, the resource type with the highest utilization rate is analyzed.

Step B2: Select a virtual machine with the highest utilization rate for the resource type analyzed in step B1. This virtual machine is the virtual machine to be migrated.

In another embodiment of the present invention, when the physical machine is overloaded, that is, the CPU utilization of the physical machine exceeds the first parameter, and when the physical machine is overloaded, that is, the CPU utilization of the physical machine is lower than the second parameter, It is necessary to migrate some or all of the virtual machines on the host to other hosts according to the current state; where the first parameter is greater than the second parameter, for example, the first parameter is 80% and the second parameter is 20%.

In the embodiment of the present invention, when in an overload state, the criterion for selecting the virtual machine to be migrated may be that the virtual machine with the largest predicted load on the physical machine is first migrated out of the current working host, that is, the current physical machine, and the current physical machine can satisfy the remaining Future resource requirements for virtual machines. When in a low-load state, all virtual machines are migrated out of the current physical machine, and then the virtual machine is deleted to save computing resources.

Step 103: Determine the second physical machine, and migrate the virtual machine to be migrated to the second physical machine.

Among them, the second physical machine is different from the first physical machine, and the second physical machine belongs to the cloud computing management platform.

In the embodiment of the present invention, when the virtual machine to be migrated has been determined, the next step is to select the migration target host, that is, the destination physical machine. The selection of the migration destination host needs to consider that it can provide sufficient resources to the virtual machine to be migrated, and ensure that after the migration is completed, the load of the destination host will not exceed the preset threshold and cause two or more migrations.

In the embodiment of the present invention, when the system selects the migration target host through the target physical machine selection module, the following three steps can be implemented:

Step C1: Select the destination host with less load information.

Step C2: When the number of destination hosts is multiple, as long as the destination host with the smallest load information is selected, the workload prediction method is used to predict the load of the destination host in the future. If the load of the destination host in the future is still the smallest compared to other destination hosts, the next step is executed, otherwise it returns to step C1 to reselect the destination host.

Step C3: If the load value of the virtual machine to be migrated plus the predicted value of step C2 will not exceed the preset threshold, the destination host will be selected as the migration target of the virtual machine, otherwise return to step C1.

In the virtual machine migration method provided by the embodiment of the present invention, if the first load of the first physical machine at the first moment is greater than the first threshold, the second load of the first physical machine at the second moment is predicted; After a moment; if the second load is greater than the second threshold, select the virtual machine to be migrated from the virtual machines on the first physical machine; wherein, the second threshold is greater than the first threshold; determine the second physical machine, and transfer the virtual machine to be migrated Migrate to the second physical machine; where the second physical machine is different from the first physical machine; that is, in the embodiment of the present invention, when it is determined that the burst load of the first physical machine at the first moment is greater than the first threshold, Predict the load of the first physical machine at the future moment, and when it is determined that the load at the future moment is higher than the second threshold greater than the first threshold, trigger a migration, thus effectively avoiding sudden load pairs The interference of the system solves the problem that the sudden burst load in the related technology will trigger the migration of the virtual machine when it reaches a fixed threshold, which leads to frequent virtual machine migration, causing unnecessary migration overhead, and avoiding the sudden load on the system. Interference improves the reliability of the system.

Based on the foregoing embodiments, embodiments of the present invention provide a virtual machine migration method, which is applied to a cloud computing management platform; a virtual machine cluster is established on a physical machine of the cloud computing management platform, and the number of virtual machines to be migrated is at least two. In the embodiments of the present invention, the virtual machine migration process can be optimized by using cluster deployment technology.

In the embodiment of the present invention, the main function to be completed by the virtual machine cluster deployment method based on resource classification is to be able to select N physical nodes, and the resources of these nodes are enough to deploy a cluster of n virtual machines, so that the entire physical The resources of the computer system are relatively balanced and comprehensively utilized. As shown in FIGS. 7 and 8, the method includes:

Step 201: If the first load of the first physical machine at the first moment is greater than the first threshold, predict the second load of the first physical machine at the second moment.

Among them, the second moment is after the first moment.

In the embodiment of the present invention, in step 201, if the first load of the first physical machine at the first moment is greater than the first threshold, before predicting the second load of the first physical machine at the second moment, the following steps are further included:

Step S1: The cloud computing management platform analyzes the resource type of the virtual machine cluster and distinguishes the resource type. There are many ways to analyze the resource type of the cluster: it can be specified by the user; it can be analyzed according to the historical data of the image used by the cluster.

Step S2: The cloud computing management platform filters the cluster deployment request. If it is analyzed that the deployment system is unable to complete the cluster deployment due to CPU, memory, file capacity, network bandwidth and other limitations, the deployment request should be filtered and submitted to customer feedback.

Step 202: If the second load is greater than the second threshold, select the virtual machine to be migrated from the virtual machines on the first physical machine.

Among them, the second threshold is greater than the first threshold.

In the embodiment of the present invention, in the process of selecting virtual machines to be migrated, it can be assumed that the number of virtual machines is N, and the number of virtual machines to be deployed for initial operation of each physical machine node is 0, and a capacity of N is created. Priority queue, the priority queue represents the virtual machine resource load priority queue.

Step 203: Determine the second physical machine.

In the embodiment of the present invention, in the process of determining the second physical machine, all physical machine nodes can be traversed in cycles, and it can be calculated whether the physical machine node can also deploy a new virtual machine according to the following formula. When the load information value L<L1, It indicates that the current load of the physical node is relatively low and can be used to deploy virtual machines; when the load information value L1<L<L2, it indicates that the current load of the physical machine node is in the best state, and the virtual machine can also be deployed, but the new virtual machine cannot The physical machine node is overloaded; when the load information value L>L2, it indicates that the physical machine node is currently overloaded, and the physical machine node should be reselected.

In the embodiment of the present invention, the load information is determined according to the resource type of the cluster, and the cluster is divided into a computing-intensive cluster, a storage-intensive cluster, and a traffic-intensive cluster. Furthermore, determining the virtual machine to be migrated based on the cluster can improve the positioning efficiency.

In the embodiment of the present invention, after calculating the load value according to the weight of different resources according to the resource type of the cluster, the virtual machine resource load priority queue is inserted until the physical nodes are selected for all virtual machines.

Step 204: Obtain the load of each virtual machine to be migrated.

Step 205: Determine a target physical machine corresponding to each virtual machine to be migrated from multiple second physical machines.

Wherein, the plurality of second physical machines include target physical machines.

Step 206: Acquire the load of multiple target physical machines.

Step 207: Sort multiple target physical machines according to the load of each virtual machine to be migrated and the load of each target physical machine to obtain a priority queue.

In the embodiment of the present invention, step 207 sorts multiple target physical machines according to the load of each virtual machine to be migrated and the load of each target physical machine to obtain a priority queue, which can be achieved by the following steps:

Step 207a: Calculate the sum of the load of each virtual machine to be migrated and the load of the target physical machine corresponding to each virtual machine to be migrated to obtain the transmission load.

Step 207b: Sort the multiple target physical machines according to the transmission load to obtain a priority queue.

Step 208: Determine the priority of each target physical machine according to the priority queue.

Step 209: Perform image transmission on the virtual machine to be migrated corresponding to each target physical machine according to the order of priority from high to low.

The transmission load corresponding to the target physical machine with a high priority is less than the transmission load corresponding to the target physical machine with a low priority.

In the embodiment of the present invention, migrating the virtual machine to be migrated from the first physical machine to the target physical machine in the second physical machine means that the virtual machine needs to be restored. This process consumes the resources of the physical machine. If the load is high, the data will be transmitted again; that is, if the target physical machine has a low load, the image will be transmitted first, and then the virtual machine will be restored based on the image. Finally, start all the virtual machines in the cluster to complete the cluster deployment.

Based on the foregoing embodiment, an embodiment of the present invention provides a virtual machine migration method, which is applied to a cloud computing management platform. It should be noted that the purpose of this embodiment is to use a virtual machine cluster image to express expressly during the process of virtual machine migration. Technology to help improve the efficiency of virtual machine dynamic migration.

Because the virtual machine image file contains cluster applications, operating systems, etc., its file size is often in the tens of millions of bytes, and can even reach hundreds of millions of bytes. Therefore, it may take a long time to transfer an image file from the image library to the physical machine, and a virtual machine cluster may be composed of dozens of virtual machines. Deploying and migrating such a cluster requires a greater time cost. The fast image transfer method uses the image and copy of the image library as the root node. After the first copy is completed, the image file that has just been transferred in the entire cloud platform system has twice the number of images and copies, and then the image library The mirror and copy of, and the mirror of the first transfer serve as the parent node of the second mirror transfer. After the second mirror transfer, the mirror file in the system will have four times the number of mirrors and copies. This cycle is passed until the number of virtual machines required by the cluster can be met. Exemplarily, before the first copy in FIG. 8, the system only includes the image files such as image1 and image3, image2 and image6, and image4 and image5 in the solid line box; after the first copy is completed, the system includes the dotted box The image files such as image2 and image4, image1 and image5, and image3 and image6, that is, after the first copy is completed, the image files just transferred in the entire cloud platform system have twice the number of images and copies.

In the embodiment of the present invention, adopting the virtual machine cluster image fast express technology may include the following steps:

First: Assuming that there are M backups for an image in the image library, the image transfer strategy will select M physical machines from the selected physical machine, and then transfer the image file to the selected physical machine, which is basically a mirror backup pair A physical machine.

Secondly: After this transfer is complete, the M backups in the image library are called the root node, and the M physical machines that have received the image are called the first-generation child nodes.

Again: If the deployment request of the virtual machine node is still not completed, then M backup nodes are the root node, plus M first-generation child nodes, and send mirror images to 2M second-generation child nodes, respectively.

Finally: Loop the above steps until the deployment request is completed.

It can be seen from the above that the rapid express technology of virtual machine cluster mirroring can provide strategic support for the multi-channel parallel transmission of mirror images, which greatly improves the transmission speed of the virtual machine cluster mirror file.

It should be noted that the description of the same steps and the same content in this embodiment and other embodiments can refer to the descriptions in other embodiments, and are not repeated here.

Based on the foregoing embodiments, embodiments of the present invention provide a virtual machine migration method, which is applied to a cloud computing management platform. The method includes the following steps:

Step 301: Receive a user request and obtain the request type of the user request.

Among them, the request type includes a static request or a dynamic request.

In practical applications, various types of server cluster systems have their own load characteristics, and the load indicators that should be considered are also different. For example: in a Web server cluster system, for a request for static content, the server only needs to read the corresponding from the disk The content can be sent to the browser; when requesting dynamic content, it often needs to be read, compiled and processed to generate the requested file before it can be sent to the client.

As shown in FIG. 9, in the embodiment of the present invention, user requests can be divided into two types. Among them, requests for static content. The main load that such requests bring to the server is the network bandwidth load. In web applications, such requests are mainly requests for static web pages. For requests for dynamic content, the load that such requests bring to the server is mainly the occupation of CPU and memory resources. In web applications, such requests are mainly requests for dynamic web pages and database access.

It should be noted that when determining the request type requested by the user, it can be determined according to the file type requested by the user.

Step 302: If the first load is greater than the first threshold, obtain third loads at least two moments within a preset time period.

The preset time period includes the first time and at least one third time before the first time, and the third load at each time in the at least two times includes at least two loads with different attributes; the loads with different attributes represent the first The types of occupied resources in physical machines are different.

In the embodiment of the present invention, at least two loads with different attributes include: the utilization rate of the central processor of the first physical machine, the utilization rate of the memory of the first physical machine, and the utilization rate of the input interface/output interface of the first physical machine And at least two of the utilization rates of the network bandwidth of the first physical machine.

Step 303: Acquire the attribute of each load in the third load at each moment.

In the embodiment of the present invention, the loads with different attributes represent different types of resources occupied in the first physical machine.

In the embodiment of the present invention, the third load at each moment may include the utilization rate of the central processor of the first physical machine, the utilization rate of the memory of the first physical machine, and the utilization rate of the input interface/output interface of the first physical machine And at least two of the utilization rates of the network bandwidth of the first physical machine.

Step 304: Determine the weight value of each load in the third load at each moment according to the request type and the attribute of each load in the third load at each moment.

In the embodiment of the present invention, the workload forecast value at the next moment is calculated by weighted average according to the current and previous load data. When determining the weight of the load data, it is necessary to first determine the type of user request, if it is a request for static content , The weight of the network bandwidth item is increased by 20%, and the weight of CPU and memory are both reduced by 10%; if it is a request for dynamic content, the weight of CPU and memory are increased by 10%, and the weight of the network bandwidth item is increased. The weight is reduced by 20%.

Exemplarily, when the user request is a static request, the weight value of each load in the third load is determined according to the request type and the attribute of each load in the third load at a certain time, for example, the CPU utilization weight value is 15% , The memory utilization weight value is 15%, the input interface/output interface utilization weight value is 25%, and the network utilization bandwidth weight value is 45%.

When the user request is a dynamic request, the weight value of each load in the third load is determined according to the request type and the attribute of each load in the third load at a certain time, such as the CPU utilization weight value is 35%, the memory utilization rate The weight value is 35%, the input interface/output interface utilization weight value is 25%, and the network utilization bandwidth weight value is 5%.

Step 305: Calculate the fourth load at each moment according to the third load at each moment and the weight value of the third load at each moment.

In the embodiment of the present invention, the fourth load at each moment can be calculated using Formula 1.

L＝a1×ucpu+a2×umem+a3×uio+a4×unet (Formula 1)

Among them, L represents the fourth load at each moment; ucpu represents CPU utilization, umem represents memory utilization, uio represents I/O utilization, and unet represents network bandwidth utilization; a1 represents CPU utilization weight value and a2 represents memory Utilization weight value, a3 indicates the I/O utilization weight value, a4 indicates the network bandwidth utilization weight value, a1+a2+a3+a4=1.

Step 306: Calculate the second load according to the fourth load at each moment.

In the embodiment of the present invention, the average value of the fourth load at multiple times is calculated to obtain the second load.

Step 307: If the second load is greater than the second threshold, select the virtual machine to be migrated from the virtual machines on the first physical machine.

Among them, the second threshold is greater than the first threshold.

In the embodiment of the present invention, if the second load is greater than the second threshold in step 307, selecting the virtual machine to be migrated from the virtual machines on the first physical machine may include the following steps:

Step 307a1, if the second load is greater than the second threshold, obtain the resource utilization rate of the first physical machine.

Step 307b1: Select a virtual machine to be migrated from multiple virtual machines on the first physical machine according to the resource utilization rate.

In the embodiment of the present invention, step 307b1 selects a virtual machine to be migrated from multiple virtual machines on the first physical machine according to the resource utilization rate, and may further include the following steps:

In the first step, if the resource utilization rate is greater than or equal to the third threshold, the fifth load of each virtual machine on the first physical machine at the third moment is predicted.

Among them, the third moment is after the first moment.

In the embodiment of the present invention, if the resource utilization rate is greater than or equal to the third threshold, it is determined that the physical machine is overloaded, and then the fifth load of each virtual machine on the first physical machine at the third moment is predicted.

In the second step, the virtual machine corresponding to the maximum load among the fifth loads is selected from the virtual machines on the first physical machine, and determined as the virtual machine to be migrated.

In the embodiment of the present invention, if the second load is greater than the second threshold in step 307, selecting the virtual machine to be migrated from the virtual machines on the first physical machine may further include the following steps:

Step 307a2: If the resource utilization rate is less than or equal to the fourth threshold, determine that each virtual machine on the first physical machine is a virtual machine to be migrated.

Among them, the fourth threshold is smaller than the third threshold.

In the embodiment of the present invention, if the resource utilization rate is less than or equal to the fourth threshold, it is determined that the physical machine has a low load condition, and then all virtual machines on the physical machine are determined to be virtual machines to be migrated.

In practical applications, when the physical machine is overloaded, such as CPU utilization exceeding 80% and low load, such as CPU utilization being less than 20%, some or all of the virtual machines on the physical machine need to be migrated to other according to the current state Run on the physical machine. When in an overload state, the virtual machines are sorted in descending order according to the amount of future workload, and the virtual machines with the highest workload in the future are selected for migration. The selection criterion is to migrate the virtual machine with the largest predicted load out of the current physical machine first, and enable the host to meet the future resource requirements of the remaining virtual machines. When in a low-load state, all virtual machines are migrated out of the current physical machine.

Step 308: Determine the second physical machine, and migrate the virtual machine to be migrated to the second physical machine.

Among them, the second physical machine is different from the first physical machine.

In the embodiment of the present invention, determining the second physical machine in step 308 may be implemented by the following steps:

Step 308a: Acquire the sixth load of each physical machine in the cloud computing management platform at the first moment.

Step 308b: Select the physical machine with the sixth load less than the fifth threshold from the physical machines in the cloud computing management platform, and determine it as the third physical machine.

The third physical machine is different from the first physical machine, and the number of the third physical machine is multiple.

In the embodiment of the present invention, when determining the third physical machine, the system may also select the physical machine with the smallest load from the physical machines in the cloud computing management platform and determine the first and third physical machines. Further, based on the first and third physical machines, step 308c is executed. If the predicted load corresponding to the first and third physical machines, that is, the seventh load, is not the smallest, then reselect the physical machine in the cloud computing management platform except the first The smallest physical machine other than the third physical machine is the second and third physical machines, and the subsequent steps are performed.

Step 308c: Predict the seventh load of the virtual machine on each third physical machine at the fourth moment.

Among them, the fourth moment is after the first moment.

Step 308d: Calculate the sum of the seventh load of each third physical machine and the load of the virtual machine to be migrated to obtain multiple target loads.

Step 308e: Determine a target load among the multiple target loads that is less than the sixth threshold.

Step 308f: Select the third physical machine corresponding to the target load less than the sixth threshold from the third physical machine, and determine it as the fourth physical machine.

Among them, the number of the fourth physical machine is at least one.

Step 308g: Determine a second physical machine from at least one fourth physical machine.

In the embodiment of the present invention, the step 308g determining the second physical machine from the at least one fourth physical machine may be implemented by the following steps:

First, calculate the value of the seventh load of each fourth physical machine minus the sixth load of each fourth physical machine to obtain the amount of resource demand expansion and contraction of each fourth physical machine.

In the embodiment of the present invention, the amount of resource demand expansion and contraction is a negative value, indicating that the future load is decreasing. The amount of resource demand expansion and contraction is positive, indicating that the future load is increasing.

Secondly, the value of each fourth physical machine minus the sixth load of each fourth physical machine is calculated to obtain the unallocated resource amount of each fourth physical machine.

Thirdly, the value of the unallocated resource amount of each fourth physical machine minus the resource demand expansion and contraction amount of each fourth physical machine is calculated to obtain the predicted remaining resource amount of each fourth physical machine.

Finally, according to the predicted remaining resource amount of each fourth physical machine, the second physical machine is determined from the fourth physical machine.

In the embodiment of the present invention, determining the second physical machine from the fourth physical machine according to the predicted remaining resource amount of each fourth physical machine may be achieved by the following steps: determining the maximum of the predicted remaining resource amount of the fourth physical machine The fourth physical machine corresponding to the value is the second physical machine.

Step 401: If the first load is greater than the first threshold, obtain third loads at least two moments within a preset time period.

The preset time period includes the first time and at least one third time before the first time, and the third load at each time in the at least two times includes at least two loads with different attributes.

As shown in FIG. 10, in the embodiment of the present invention, the third load at each moment includes a distance, where the distance represents the distance between the first physical machine and the data center that receives the user request, and the cloud computing management platform also includes data The data center is used to manage the third physical machine. The third physical machine includes the first physical machine. At the same time, the third load at each moment also includes the utilization rate of the central processor of the first physical machine and the At least two of the utilization rate of the memory, the utilization rate of the input interface/output interface of the first physical machine, and the utilization rate of the network bandwidth of the first physical machine.

It should be noted that a large cloud computing platform usually includes multiple data centers, commonly known as data centers in the same city. Multiple data centers are generally not located in a building, but are distributed in data center computer rooms in different locations. The distance between them is 5-10 kilometers. If the virtual machines are migrated across data centers, the service level agreement of the business will be affected due to the distance. Therefore, the migration of virtual machines across data centers is restricted.

In the embodiment of the present invention, according to the load data of the current first time and a plurality of second times before the first time, the weighted average is used to calculate the predicted workload value at the next time. When calculating the weighted average, except for CPU and memory , I/O and network bandwidth are four indicators, and the distance indicator is added, and the unit can be kilometers. The effect achieved in this way is that the predicted load of the local data center is low and the load of the remote data center is high, which effectively prevents the virtual machine from migrating to the remote data center.

Step 402: Acquire the attribute of each load in the third load at each moment.

In the embodiment of the present invention, loads with different attributes represent different types of resources occupied in the first physical machine.

Step 403: Determine the weight value of each load in the third load at each moment according to the attribute of each load in the third load at each moment.

As shown in FIG. 10, for example, the CPU utilization weight value is 20%, the memory utilization weight value is 20%, the input interface/output interface utilization weight value is 20%, and the network utilization bandwidth weight value is 20. %.

Step 404: Calculate the fourth load at each moment according to the third load at each moment and the weight value of the third load at each moment.

Step 405: Calculate the second load according to the fourth load at each moment.

Step 406: If the second load is greater than the second threshold, select the virtual machine to be migrated from the virtual machines on the first physical machine.

Among them, the second threshold is greater than the first threshold. In practical applications, when the physical machine is overloaded, such as CPU utilization exceeding 80% and low load, such as CPU utilization being less than 20%, some or all of the virtual machines on the physical machine need to be migrated to other according to the current state Run on the physical machine. When in an overload state, the virtual machines are sorted in descending order according to the amount of future workload, and the virtual machines with the highest workload in the future are selected for migration. The selection criterion is to migrate the virtual machine with the largest predicted load out of the current physical machine first, and enable the host to meet the future resource requirements of the remaining virtual machines. When in a low-load state, all virtual machines are migrated out of the current physical machine.

Step 407: Determine the second physical machine, and migrate the virtual machine to be migrated to the second physical machine.

In summary, the embodiments of the present invention have the following beneficial effects:

(1) A workload prediction method is proposed. The load comprehensively considers multiple factors such as CPU usage, memory usage, disk IO, network bandwidth, and distance. By predicting the future load of tasks performed by the virtual machine, and by The classification of business requests adjusts the weight of the load index, so as to accurately select the virtual machines that need to be migrated and install the hosts, reduce the number of virtual machine migrations, and avoid migration shocks.

(2) A virtual machine cluster deployment method based on resource classification is proposed. Different virtual machine clusters are divided according to resource types, which can quickly select the virtual machines to be migrated and the target physical host, and help improve the efficiency of virtual machine dynamic migration.

(3) A fast image transfer method based on image copy split is proposed. This method uses the image and copy of the image library as the root node, and through multiple cyclic copies, the virtual machine image can be M, 2M, 4M... Fast copy speed, help improve the efficiency of virtual machine dynamic migration.

(4) A method to prevent the migration of virtual machines across data centers is proposed. By introducing the data center distance index into the calculation of load prediction, the migration of virtual machines in multiple data center scenarios is effectively prevented, thereby ensuring the service level agreement of the business.

Based on the foregoing embodiments, embodiments of the present invention provide a cloud computing management platform that can be applied to a virtual machine migration method provided in the embodiments corresponding to FIGS. 5 and 7. Referring to FIG. 11, The cloud computing management platform 11 includes: a processor 1101, a memory 1102, and a communication bus 1103, where:

The communication bus 1103 is used to implement a communication connection between the processor 1101 and the memory 1102.

The processor 1101 is used to execute a virtual machine migration program stored in the memory 1102 to implement the following steps:

If the first load of the first physical machine at the current first moment is greater than the first threshold, the second load of the first physical machine at the second moment is predicted; where the second moment is after the first moment;

Determine the second physical machine, and migrate the virtual machine to be migrated to the second physical machine; where the second physical machine is different from the first physical machine.

In other embodiments of the present invention, the processor 1101 is used to execute in the memory 1102 if the first load of the first physical machine at the first moment is greater than the first threshold, predict the second load of the first physical machine at the second moment , Can be achieved by the following steps:

If the first load is greater than the first threshold, obtain a third load at least two moments within a preset time period; wherein the preset time period includes the first moment and at least one third moment before the first moment, at least two moments The third load at each moment in includes at least two loads with different attributes; the loads with different attributes represent different types of resources occupied in the first physical machine;

Calculate the fourth load at each moment according to the third load at each moment and the weight value of the third load at each moment;

The second load is calculated based on the fourth load at each moment.

In other embodiments of the present invention, when the processor 1101 is used to execute the calculation of the second load according to the fourth load at each moment in the memory 1102, the following steps may be implemented:

In other embodiments of the present invention, the processor 1101 is used to execute a virtual machine migration program in the memory 1102 to implement the following steps:

Receive user requests and obtain the request types requested by users; where the request types include static requests or dynamic requests;

Obtain the attributes of each load in the third load at each moment;

In other embodiments of the present invention, the third load at each moment also includes a distance, where the distance represents the distance between the first physical machine and the data center that receives the user request, and the cloud computing management platform also includes a data center, The data center is used to manage the third physical machine, and the third physical machine includes the first physical machine.

Obtain the attributes of each load in the third load at each moment;

In other embodiments of the present invention, the processor 1101 is used to execute a plurality of virtual machines on the first physical machine in the memory 1102. If the second load is greater than the second threshold, select from the virtual machines on the first physical machine When you want to migrate a virtual machine, you can use the following steps:

In other embodiments of the present invention, the processor 1101 is used to execute the selection of a virtual machine to be migrated from multiple virtual machines on the first physical machine in the memory 1102 according to the resource utilization rate by the following steps:

If the resource utilization rate is greater than or equal to the third threshold, predict the fifth load of each virtual machine on the first physical machine at the third moment; where the third moment is after the first moment;

The virtual machine corresponding to the maximum load in the fifth load is selected from the virtual machines on the first physical machine, and determined as the virtual machine to be migrated.

If the resource utilization rate is less than or equal to the fourth threshold, it is determined that each virtual machine on the first physical machine is a virtual machine to be migrated; wherein, the fourth threshold is less than the third threshold.

In other embodiments of the present invention, when the processor 1101 is used to execute the determination of the second physical machine in the memory 1102, the following steps may be implemented:

Obtain the sixth load of each physical machine in the cloud computing management platform at the first moment;

Select the physical machine with the sixth load less than the fifth threshold from the physical machines in the cloud computing management platform and determine it as the third physical machine; where the third physical machine is different from the first physical machine, the number of third physical machines is Multiple

Predict the seventh load of the virtual machine on each third physical machine at the fourth moment; where the fourth moment is after the first moment;

Calculate the sum of the seventh load of each third physical machine and the load of the virtual machine to be migrated to obtain multiple target loads;

Determine a target load that is less than the sixth threshold among multiple target loads;

The third physical machine corresponding to the target load less than the sixth threshold is selected from the third physical machines, and determined as the fourth physical machine; wherein the number of the fourth physical machine is at least one;

In other embodiments of the present invention, when the processor 1101 is used to execute determining the second physical machine from the at least one fourth physical machine in the memory 1102, the following steps may be implemented:

Calculate the value of the seventh load of each fourth physical machine minus the sixth load of each fourth physical machine to obtain the resource demand expansion and contraction amount of each fourth physical machine;

Calculate the total resource of each fourth physical machine minus the value of the sixth load of each fourth physical machine to obtain the unallocated resource amount of each fourth physical machine;

Calculate the value of the unallocated resources of each fourth physical machine minus the resource demand expansion and contraction of each fourth physical machine to obtain the predicted remaining resource amount of each fourth physical machine;

The second physical machine is determined from the fourth physical machine according to the predicted remaining resource amount of each fourth physical machine.

In other embodiments of the present invention, the processor 1101 is used to execute the determination of the second physical machine from the fourth physical machine according to the predicted remaining resource amount of the fourth physical machine in the memory 1102 by the following steps:

In other embodiments of the present invention, a virtual machine cluster is established on the physical machine of the cloud computing management platform, the number of virtual machines to be migrated is at least two, and the number of virtual machines to be migrated is the same as the number of second physical machines. The device 1101 is used to execute a virtual machine migration program in the storage 1102 to implement the following steps:

Obtain the load of each virtual machine to be migrated;

Determining the target physical machine corresponding to each virtual machine to be migrated from the plurality of second physical machines; wherein the plurality of second physical machines include the target physical machine;

Obtain the load of multiple target physical machines;

According to the load of each virtual machine to be migrated and the load of each target physical machine, the multiple target physical machines are sorted to obtain a priority queue.

In other embodiments of the present invention, the processor 1101 is used to execute, in the memory 1102, sort multiple second physical machines according to the load of each virtual machine to be migrated and the load of each target physical machine to obtain a priority queue Can be achieved by the following steps:

Calculate the sum of the load of each virtual machine to be migrated and the load of the target physical machine corresponding to each virtual machine to be migrated to obtain the transmission load;

In other embodiments of the present invention, when the processor 1101 is used to execute the root migration of the virtual machine to be migrated to the second physical machine in the memory 1102, the following steps may be implemented:

Determine the priority of each target physical machine according to the priority queue;

Mirror the virtual machines corresponding to each target physical machine to be migrated in order of high priority to low priority; where the transmission load corresponding to the target physical machine with high priority is less than that corresponding to the target physical machine with low priority load.

It should be noted that, for the specific implementation process of the steps performed by the processor in this embodiment, reference may be made to the implementation process in the virtual machine migration method provided in the embodiments corresponding to FIGS. 5 and 7, and details are not described here.

Based on the foregoing embodiments, the embodiments of the present invention provide a storage medium that stores one or more programs, and the one or more programs may be executed by one or more processors to implement Figures 5 and 7 Steps in the method for migrating a virtual machine provided by the corresponding embodiment.

It should be noted that the above-mentioned computer storage media may be read-only memory (Read Only Memory, ROM), programmable read-only memory (Programmable Read-Only Memory, PROM), erasable programmable read-only memory (Erasable Programmable Read-memory) Only Memory (EPROM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Magnetic Random Access Memory (Ferromagnetic Random Access Memory, FRAM), Flash Memory (Flash) Memory, Magnetic Surface memory, compact disc, or read-only compact disc (Compact Disc Read-Only Memory, CD-ROM) and other memories; can also be one of the above memories or any combination of various electronic devices, such as mobile phones, computers, tablet devices, Personal digital assistant, etc.

It should be noted that in this article, the terms "include", "include" or any other variant thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device that includes a series of elements includes not only those elements, It also includes other elements that are not explicitly listed, or include elements inherent to this process, method, article, or device. In the absence of more restrictions, the element defined by the sentence "include one..." does not exclude that there are other identical elements in the process, method, article or device that includes the element.

The sequence numbers of the above embodiments of the present invention are for description only, and do not represent the advantages and disadvantages of the embodiments.

Through the description of the above embodiments, those skilled in the art can clearly understand that the methods in the above embodiments can be implemented by means of software plus a necessary general hardware platform, and of course, can also be implemented by hardware, but in many cases the former is better Implementation. Based on this understanding, the technical solution of the present invention can be embodied in the form of a software product in essence or part that contributes to the existing technology, and the computer software product is stored in a storage medium (such as ROM/RAM, magnetic disk, The CD-ROM includes several instructions to enable a terminal device (which may be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in the embodiments of the present invention.

The present invention is described with reference to flowcharts and/or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each flow and/or block in the flowchart and/or block diagram and a combination of the flow and/or block in the flowchart and/or block diagram may be implemented from computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, special-purpose computer, embedded processing machine, or other programmable data processing device to produce a machine that enables the generation of instructions executed by the processor of the computer or other programmable data processing device A device for realizing the functions specified in one block or multiple blocks of one flow or multiple flows of a flowchart and/or one block or multiple blocks of a block diagram.

These computer program instructions may also be stored in a computer readable memory that can guide a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture including an instruction device, the instructions The device implements the functions specified in one block or multiple blocks of the flowchart one flow or multiple flows and/or block diagrams.

These computer program instructions can also be loaded onto a computer or other programmable data processing device, so that a series of operating steps are performed on the computer or other programmable device to produce computer-implemented processing, which is executed on the computer or other programmable device The instructions provide steps for implementing the functions specified in one block or multiple blocks of the flowchart one flow or multiple flows and/or block diagrams.

The above are only the preferred embodiments of the present invention and do not limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made by the description and drawings of the present invention, or directly or indirectly used in other related technical fields The same reason is included in the patent protection scope of the present invention.

Industrial applicability

The embodiment of the present invention solves the problem that the sudden burst load in the related art will trigger the migration of the virtual machine when it reaches a fixed threshold, thus causing frequent virtual machine migration, causing unnecessary migration overhead, and avoiding the sudden load The interference of the system improves the reliability of the system.

Claims

A virtual machine migration method is applied to a cloud computing management platform. The method includes:

If the first load of the first physical machine is greater than the first threshold at the current first moment, the second load of the first physical machine at the second moment is predicted; wherein, the second moment is after the first moment;

If the second load is greater than the second threshold, select the virtual machine to be migrated from the virtual machines on the first physical machine; wherein, the second threshold is greater than the first threshold;

Determine a second physical machine, and migrate the virtual machine to be migrated to the second physical machine; wherein the second physical machine is different from the first physical machine.
The method according to claim 1, wherein, if the first load of the first physical machine at the current first moment is greater than a first threshold, predicting the second load of the first physical machine at the second moment includes:

If the first load is greater than the first threshold, obtain a third load at least two moments within a preset time period; wherein the preset time period includes the first time and the time before the first time At least one third moment, the third load at each moment in the at least two moments includes at least two loads with different attributes; the loads with different attributes represent different types of resources occupied in the first physical machine;

Calculating the fourth load at each moment according to the third load at each moment and the weight value of the third load at each moment;

The second load is calculated based on the fourth load at each moment.
The method according to claim 2, wherein the calculating the second load according to the fourth load at each moment includes:

The average value of the fourth load at multiple times is calculated to obtain the second load.
The method according to claim 2, wherein, if the first load is greater than the first threshold, acquiring a plurality of different kinds of load data at each time in a plurality of times before the first time Previously, the method also included:

Receiving a user request, and obtaining the request type of the user request; wherein, the request type includes a static request or a dynamic request;

Correspondingly, before calculating the fourth load at each moment according to the third load at each moment and the weight value of the third load at each moment, the method further includes:

Acquiring the attribute of each load in the third load at each moment;

According to the request type and the attribute of each load in the third load at each moment, the weight value of each load in the third load at each moment is determined.
The method according to claim 2, wherein the third load at each moment further comprises a distance, wherein the distance characterizes the distance between the first physical machine and the data center receiving the user request, so The cloud computing management platform further includes the data center, and the data center is used to manage a third physical machine, and the third physical machine includes the first physical machine.
The method according to claim 5, wherein before calculating the fourth load at each moment according to the weight value of the third load at each moment and the third load at each moment, the The method also includes:

Acquiring the attribute of each load in the third load at each moment;

According to the attribute of each load in the third load at each moment, the weight value of each load in the third load at each moment is determined.
The method according to claim 1, wherein the number of virtual machines on the first physical machine is multiple, and if the second load is greater than a second threshold, from the virtual machines on the first physical machine Select the virtual machine to be migrated, including:

If the second load is greater than the second threshold, obtain the resource utilization rate of the first physical machine;

The virtual machine to be migrated is selected from multiple virtual machines on the first physical machine according to the resource utilization rate.
The method according to claim 7, wherein the selecting a virtual machine to be migrated from a plurality of virtual machines on the first physical machine according to the resource utilization rate includes:

If the resource utilization rate is greater than or equal to a third threshold, predict a fifth load of each virtual machine on the first physical machine at the third moment; wherein the third moment is after the first moment;

Select the virtual machine corresponding to the maximum load among the fifth loads from the virtual machines on the first physical machine, and determine the virtual machine to be migrated.
The method according to claim 7, wherein the selecting a virtual machine to be migrated from a plurality of virtual machines on the first physical machine according to the resource utilization rate includes:

If the resource utilization rate is less than or equal to a fourth threshold, it is determined that each virtual machine on the first physical machine is the virtual machine to be migrated; wherein, the fourth threshold is less than the third threshold.
The method of claim 1, wherein the determining the second physical machine comprises:

Acquiring the sixth load of each physical machine in the cloud computing management platform at the first moment;

The physical machine with the sixth load less than the fifth threshold is selected from the physical machines in the cloud computing management platform, and determined as the third physical machine; wherein the third physical machine is different from the first physical machine, so The number of the third physical machine is multiple;

Predicting the seventh load of the virtual machine on each third physical machine at the fourth moment; wherein, the fourth moment is after the first moment;

Calculating the sum of the seventh load of each third physical machine and the load of the virtual machine to be migrated to obtain multiple target loads;

Determining a target load less than a sixth threshold among the plurality of target loads;

Selecting the third physical machine corresponding to the target load less than the sixth threshold from the third physical machine and determining it as the fourth physical machine; wherein the number of the fourth physical machine is at least one;

The second physical machine is determined from at least one fourth physical machine.
The method of claim 10, wherein the determining the second physical machine from at least one of the fourth physical machines comprises:

Calculating the value of the seventh load of each fourth physical machine minus the value of the sixth load of each fourth physical machine to obtain the resource demand expansion and contraction amount of each fourth physical machine;

Calculating the value of the total resources of each fourth physical machine minus the value of the sixth load of each fourth physical machine to obtain the amount of unallocated resources of each fourth physical machine;

Calculating the value of the unallocated resources of each fourth physical machine minus the amount of resource demand expansion and contraction of each fourth physical machine to obtain the predicted remaining resource amount of each fourth physical machine;

The second physical machine is determined from the fourth physical machine according to the predicted remaining resource amount of each of the fourth physical machine.
The method according to claim 11, wherein the determining the second physical machine from the fourth physical machine according to the predicted remaining resource amount of the fourth physical machine comprises:

It is determined that the fourth physical machine corresponding to the maximum value in the predicted remaining resource amount of the fourth physical machine is the second physical machine.
The method according to claim 1, wherein a virtual machine cluster is established on a physical machine of the cloud computing management platform, the number of the virtual machines to be migrated is at least two, and the number and the number of the virtual machines to be migrated are The number of the second physical machines is the same; before migrating the virtual machine to be migrated to the second physical machine, the method further includes:

Obtaining the load of each of the virtual machines to be migrated;

Determining a target physical machine corresponding to each virtual machine to be migrated from a plurality of second physical machines; wherein the plurality of second physical machines include the target physical machine;

Acquiring loads of multiple of the target physical machines;

According to the load of each of the virtual machines to be migrated and the load of each of the target physical machines, sort the multiple target physical machines to obtain a priority queue.
The method according to claim 13, wherein, according to the load of each of the virtual machines to be migrated and the load of each of the target physical machines, sorting the plurality of second physical machines to obtain a priority queue, include:

Calculating the sum of the load of each virtual machine to be migrated and the load of the target physical machine corresponding to each virtual machine to be migrated to obtain a transmission load;

Sort multiple target physical machines according to the transmission load to obtain a priority queue.
The method according to claim 14, wherein the migrating the virtual machine to be migrated to the second physical machine comprises:

Determine the priority of each of the target physical machines according to the priority queue;

Mirroring the virtual machines to be migrated corresponding to each of the target physical machines in order of high priority to low priority; wherein the transmission load corresponding to the target physical machine with high priority is less than that with low priority The transmission load corresponding to the target physical machine.
A cloud computing management platform. The cloud computing management platform includes:

Memory, used to store executable instructions;

A processor, configured to execute executable instructions stored in the memory, to implement the virtual machine migration method according to any one of claims 1 to 15.
A storage medium storing one or more programs, the one or more programs may be executed by one or more processors, so as to realize the virtual according to any one of claims 1 to 15. Steps of the machine migration method.