Virtual machine scheduling method and device

Abstract

The present application provides a virtual machine scheduling method and device, which includes: obtaining the power consumption type of a target virtual machine of a cloud computing platform, wherein the cloud computing platform includes multiple physical servers; determining a target physical server corresponding to the target virtual machine from the multiple physical servers according to the power consumption type of the target virtual machine and the power consumption type of each of the physical servers, so as to schedule the target virtual machine to the corresponding target physical server, so that the target physical server provides computing resources for the target virtual machine, thereby implementing a virtual machine scheduling method based on power consumption matching, improving the degree of power consumption matching between the virtual machine and the assigned physical server, thereby improving the power consumption utilization rate of the physical server, and reducing the risk of overload of the rack where the server is located.

Description

Technical Field

[0001] The present application relates to the field of cloud computing technology, and in particular to a virtual machine scheduling method and device. Background Art

[0002] Cloud computing has been widely adopted in various fields due to its high flexibility, high scalability, and high cost-effectiveness. Cloud computing platforms process data through distributed physical server clusters, providing efficient data services to businesses and consumers.

[0003] When allocating resources of a cloud computing platform to user instances (virtual machines), it is usually based on static parameters such as memory and number of processor cores of the physical server of the cloud computing platform.

[0004] With the above scheduling method, after the user instance is running, it is easy for the actual load to be inconsistent with the power consumption allocated when the physical server is put on the rack, resulting in constant resource waste or overloading of the rack where the physical server is located, accelerating hardware aging and reducing the service life of the server. Summary of the Invention

[0005] The present application provides a virtual machine scheduling method and device, which implements a virtual machine scheduling strategy based on power consumption type, improves the power consumption matching degree between the virtual machine and the assigned physical server, thereby improving the power consumption utilization rate of the physical server and reducing the risk of overload of the rack where the server is located.

[0006] In a first aspect, the present application provides a virtual machine scheduling method, comprising:

[0007] Obtaining a power consumption type of a target virtual machine of a cloud computing platform, wherein the cloud computing platform includes a plurality of physical servers;

[0008] According to the power consumption type of the target virtual machine and the power consumption type of each of the physical servers, a target physical server corresponding to the target virtual machine is determined from the multiple physical servers, so that the target physical server provides computing resources for the target virtual machine.

[0009] In a second aspect, the present application provides a virtual machine scheduling device, comprising:

[0010] a power consumption type acquisition module, configured to acquire a power consumption type of a target virtual machine of a cloud computing platform, wherein the cloud computing platform includes a plurality of physical servers;

[0011] The server allocation module is used to determine the target physical server corresponding to the target virtual machine from the multiple physical servers according to the power consumption type of the target virtual machine and the power consumption type of each physical server, so that the target physical server provides computing resources for the target virtual machine.

[0012] In a third aspect, the present application provides an electronic device, comprising:

[0013] a processor, and a memory communicatively connected to the processor;

[0014] The memory stores computer-executable instructions;

[0015] The processor executes the computer-executable instructions stored in the memory to implement the virtual machine scheduling method provided in the first aspect of the present application.

[0016] In an eighth aspect, the present application provides a computer-readable storage medium, in which computer execution instructions are stored. When the computer execution instructions are executed by a processor, they are used to implement the virtual machine scheduling method provided in the first aspect of the present application.

[0017] In a ninth aspect, the present application provides a computer program product, comprising a computer program, which, when executed by a processor, implements the virtual machine scheduling method provided in the first aspect of the present application.

[0018] The virtual machine scheduling method and device provided in the present application are aimed at the scenario of user instance processing based on the cloud computing platform. The power consumption type is set in advance for each physical server of the cloud computing platform. Based on the power consumption type of the virtual machine corresponding to the user instance and the power consumption type of each physical server, the target physical server of the virtual machine corresponding to the user instance is determined from multiple physical servers of the cloud computing platform, so that the target physical server calculates the task corresponding to the virtual machine, and implements a server scheduling strategy based on the power consumption type, thereby preferentially allocating physical servers with matching power consumption types to virtual machines, improving the power consumption utilization rate of the physical server, and avoiding slow server operation due to excessive load on the server, accelerating server hardware aging, and reducing server service life. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0020] Figure 1 A schematic diagram of an application scenario of an embodiment of the present application;

[0021] Figure 2 A flowchart of a virtual machine scheduling method provided in an embodiment of the present application;

[0022] Figure 3 A flowchart of a virtual machine scheduling method provided in another embodiment;

[0023] Figure 4 For this application Figure 3 A schematic flow chart of step S305 in the illustrated embodiment;

[0024] Figure 5 For this application Figure 3 A schematic flow chart of step S306 in the illustrated embodiment;

[0025] Figure 6 A schematic diagram of the structure of a virtual machine scheduling system provided in an embodiment of the present application;

[0026] Figure 7 A schematic diagram of the structure of a virtual machine scheduling device provided in an embodiment of the present application;

[0027] Figure 8 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application.

[0028] The above drawings illustrate specific embodiments of the present application, which will be described in more detail below. These drawings and the textual description are not intended to limit the scope of the present application in any way, but rather to illustrate the concepts of the present application to those skilled in the art by reference to specific embodiments. DETAILED DESCRIPTION

[0029] Exemplary embodiments will be described in detail herein, with examples illustrated in the accompanying drawings. In the following description, when referring to the drawings, identical numerals in different figures represent identical or similar elements, unless otherwise indicated. The embodiments described in the following exemplary embodiments are not intended to represent all embodiments consistent with the present application. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the present application, as detailed in the appended claims.

[0030] First, the terms used in this application are explained:

[0031] User instance or instance: A virtual machine created by a user.

[0032] Physical server: A physical server used by cloud computing vendors to host user instances.

[0033] Rack: A rack in a computer room used to house physical servers or host machines, with a rated power consumption limit.

[0034] Power consumption utilization: The ratio of the actual peak power consumption of the server to the power consumption allocated to the server rack.

[0035] Figure 1 This is a schematic diagram of an application scenario of an embodiment of the present application, such as Figure 1As shown, the cloud computing platform is a distributed system composed of multiple physical servers 102, each of which is connected via a network. Figure 1 In this example, a cloud computing platform including 6 physical servers is taken.

[0036] When a user submits a user instance (or virtual machine) to be processed through a user terminal 104, such as a mobile phone, tablet, computer, etc., the scheduling device 106 performs server scheduling based on the needs of the user instance and the memory, CPU (Central Processing Unit) parameters, GPU (Graphics Processing Unit) parameters, etc. of each physical server 102, and allocates a physical server 102 to the user instance, thereby scheduling the user instance to the allocated physical server 102 to meet the computing needs of the user instance.

[0037] In one embodiment, the scheduling device 106 may be a terminal connected to a cloud computing platform, such as a server, a computer, etc.

[0038] In one embodiment, the scheduling device 106 may be a physical server in a cloud computing platform.

[0039] The method of scheduling virtual machines to physical server 102 based on static parameters of the physical server such as memory, CPU parameters, and GPU parameters does not take into account the matching relationship between the load corresponding to the user instance and the power consumption of the physical server 102. This may easily lead to a mismatch between the power consumption of the physical server 102 on which the user instance runs and the load of the user instance. For example, a low-load user instance may be assigned to a high-power physical server 102, resulting in low server utilization, or a high-load user instance may be assigned to a low-power physical server 102, resulting in a risk of overpowering the rack where the physical server 102 is located.

[0040] In order to improve the matching degree between user instances and assigned physical servers, an embodiment of the present application provides a virtual machine scheduling method based on power consumption profiling. The main idea of ​​the method is: based on the power consumption type of the virtual machine and the power consumption type of the physical server, the physical server is scored, and cloud computing resources, i.e., the physical server, are allocated to the virtual machine based on the scoring result, thereby implementing a scheduling strategy that matches the power consumption of the physical server with the load of the processed user instance, improving the power consumption utilization of the physical server, and reducing the risk of server overload.

[0041] The following specific embodiments describe in detail the technical solution of the present application and how the technical solution of the present application solves the above-mentioned technical problems. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments. The embodiments of the present application will be described below in conjunction with the accompanying drawings.

[0042] Figure 2 This is a flow chart of a virtual machine scheduling method provided in an embodiment of the present application. The method provided in this embodiment can be applied to a cloud computing platform, which is a distributed system consisting of multiple physical servers. The virtual machine scheduling method can be executed by any device with data processing capabilities, which can be in the form of a computer, server, etc. Figure 1 The scheduling device 106 in.

[0043] like Figure 2 As shown, the virtual machine scheduling method includes the following steps:

[0044] Step S201: Obtain the power consumption type of a target virtual machine of a cloud computing platform.

[0045] A virtual machine can also be referred to as a virtual machine instance or cloud server, and one virtual machine corresponds to one user instance. The target virtual machine is the virtual machine that needs to be allocated cloud computing platform resources, namely, a physical server. It can be a newly created virtual machine (VM) or a virtual machine that needs to be rescheduled or migrated. The target virtual machine is the virtual machine that needs to be allocated a physical server, or it can be called a virtual machine to be scheduled. There can be one or more target virtual machines.

[0046] Specifically, a user can create a virtual machine instance of the cloud computing platform through a user terminal, that is, create a target virtual machine, and initiate a resource scheduling request to the scheduling device or cloud computing platform based on the created target virtual machine. The scheduling device or cloud computing platform executes the virtual machine scheduling method provided in this application based on the resource scheduling request, thereby allocating a physical server to the target virtual machine.

[0047] For example, a user may create a virtual machine through virtual machine software on a user terminal, such as VMware, Virtual Box, etc.

[0048] Exemplarily, the user may create a target virtual machine by accessing a virtual machine creation page.

[0049] Specifically, the power consumption type of the target virtual machine can be determined based on the target virtual machine's basic configuration or instance specifications. Basic configuration can include the virtual machine's memory, processor configuration (including processor model, main frequency, number of cores, etc.), disk, etc. Instance specifications define the basic configuration of the instance. Cloud computing platforms can provide a variety of instance specifications for users to choose from when creating a virtual machine.

[0050] Specifically, a correspondence between instance specifications and power consumption types may be pre-established, and the power consumption type of the target virtual machine may be determined based on the correspondence and the instance specifications of the target virtual machine.

[0051] Specifically, the power consumption or CPU utilization corresponding to the target virtual machine when it is running can be estimated based on the power consumption or CPU utilization corresponding to virtual machines with the same basic configuration or instance specifications as the target virtual machine that have been running in the cloud computing platform in the past, and the power consumption type of the target virtual machine can be determined based on the estimated power consumption or CPU utilization of the target virtual machine.

[0052] Furthermore, the power consumption type of the target virtual machine may be determined based on a comparison result of the estimated power consumption or CPU utilization of the target virtual machine with one or more preset thresholds.

[0053] For example, if the estimated CPU utilization of the target virtual machine is higher than a preset threshold, the power consumption type of the target virtual machine is determined to be a high power consumption type, which means that the load corresponding to the target virtual machine is heavier; otherwise, the power consumption type of the target virtual machine is determined to be a low power consumption type, which means that the load corresponding to the target virtual machine is lighter.

[0054] Exemplarily, the value range of the preset threshold is between 0 and 1, and the preset threshold may be 60% (or 0.6), 50% (or 0.5), or values ​​thereof.

[0055] Optionally, obtain the power consumption type of the target virtual machine on the cloud computing platform, including:

[0056] The power consumption type of the target virtual machine is determined according to the user and instance type corresponding to the target virtual machine.

[0057] The user corresponding to the target virtual machine can be the user who created the target virtual machine, or the user logged in when the target virtual machine was created. The instance type can be determined based on the instance specifications or basic configuration of the instance, or based on the product used by the instance, where the product is a cloud computing package or instance template provided by the cloud computing platform.

[0058] Specifically, a correspondence between instance types and power consumption types can be established in advance for each user. Based on the user corresponding to the target virtual machine, such as the user identifier, the correspondence applicable to the target virtual machine can be determined from multiple correspondences between instance types and power consumption types. Based on the correspondence and the instance type of the target virtual machine, the power consumption type of the target virtual machine can be determined.

[0059] By customizing the power consumption type determination strategy of virtual machines for different users, the accuracy of determining the power consumption type of virtual machines is improved, providing a basis for subsequent server scheduling.

[0060] Step S202 : determining a target physical server corresponding to the target virtual machine from the multiple physical servers according to the power consumption type of the target virtual machine and the power consumption type of each of the physical servers, so as to schedule the target virtual machine to the corresponding target physical server.

[0061] Furthermore, after the target virtual machine is dispatched to its corresponding target physical server, the target physical server provides computing resources for the target virtual machine, thereby completing the task corresponding to the target virtual machine.

[0062] The target physical server may also be referred to as the host of the target virtual machine. The target physical server may be a physical server in the cloud computing platform whose power consumption type matches that of the target virtual machine. Matching power consumption types may mean identical or similar power consumption types. Similar power consumption types may mean adjacent power consumption types or power consumption types with a small deviation.

[0063] Exemplarily, the target physical server may be a physical server whose power consumption type is the same as or similar to that of the target virtual machine.

[0064] For example, if the power consumption type includes three types, such as low power consumption type, medium power consumption type, and high power consumption type, the medium power consumption type and the high power consumption type can be set to similar power consumption types. That is, if the power consumption type of the target virtual machine is medium power consumption type, the power consumption type of the target physical server can be medium power consumption type or high power consumption type.

[0065] Specifically, the first or any physical server having the same power consumption type as the target virtual machine may be determined as the target physical server.

[0066] If there is no physical server with the same power consumption type as the target virtual machine, the first or any physical server that can be determined to have a power consumption type one level higher than the power consumption type of the target virtual machine will be the target physical server, and so on. The higher the power consumption type level, the heavier the load on the auxiliary or physical server of the corresponding virtual machine at the corresponding rack power consumption.

[0067] Specifically, if there are multiple physical servers whose power consumption types are the same as the target virtual machine's, the first scores of each physical server whose power consumption type is the same as the target virtual machine's can be calculated based on the basic parameters of the physical server, such as processor parameters, memory, etc., and the physical server with the highest first score, or the physical server whose first score matches the user level of the target virtual machine, is determined to be the target physical server.

[0068] The virtual machine scheduling method provided in this embodiment is aimed at the scenario of user instance processing based on the cloud computing platform. The power consumption type is set in advance for each physical server of the cloud computing platform. Based on the power consumption type of the virtual machine corresponding to the user instance and the power consumption type of each physical server, the target physical server of the virtual machine corresponding to the user instance is determined from multiple physical servers of the cloud computing platform, so that the target physical server calculates the task corresponding to the virtual machine, and implements a server scheduling strategy based on the power consumption type, thereby preferentially allocating physical servers with matching power consumption types to virtual machines, improving the power consumption utilization of the physical servers, and avoiding slow server operation due to excessive load on the server, accelerated server hardware aging, and shortened server service life.

[0069] Figure 3 A flowchart of a virtual machine scheduling method provided in another embodiment of the present application. Figure 2 On the basis of the embodiment shown, step S201 and step S202 are further refined and a step of determining the power consumption type of the physical server is added before step S202, such as Figure 3 As shown, the virtual machine scheduling method provided in this embodiment may include the following steps:

[0070] Step S301: Obtain the power consumption type of each physical server in the multiple physical servers.

[0071] Specifically, the power consumption type may be set in advance for each physical server in the cloud computing platform.

[0072] Specifically, the power consumption type of each of the multiple physical servers may be determined based on performance parameters of the physical server, wherein the performance parameters of the physical server may include rack power consumption of the physical server, which is the power consumption allocated to the physical server by the rack.

[0073] Specifically, the power consumption dictionary can be combined with the rack power consumption of the physical server to determine the load of the physical server under the corresponding rack power consumption, and the power consumption type of the physical server can be determined based on the load of the physical server under the corresponding rack power consumption. The power consumption dictionary is the correspondence between power consumption and load, and load and rack, where the load can be the utilization rate of the CPU time slice.

[0074] Exemplarily, if the load corresponding to the physical server under the rack power consumption is greater than a preset threshold, the power consumption type of the physical server is determined to be a high power consumption type; otherwise, the power consumption type of the physical server is determined to be a low power consumption type. Multiple preset thresholds can also be set to divide the power consumption type of the physical server into 3 or more types. Taking 5 power consumption types as an example, the order from low to high in terms of load is low power consumption type, relatively low power consumption type, medium power consumption type, relatively high power consumption type, and high power consumption type. Thus, the power consumption type of the physical server is determined based on the comparison results of the load of the physical server under the corresponding rack power consumption and the multiple preset thresholds.

[0075] In one embodiment, a power consumption profile of the physical server may be determined based on the rack power consumption of the physical server and a power consumption dictionary, and a power consumption type of the physical server may be determined based on the power consumption profile of the physical server.

[0076] Step S302: Obtain the virtual machine usage record of the user corresponding to the target virtual machine.

[0077] The virtual machine usage record is used to record the processor utilization of physical servers allocated to virtual machines of various instance types created for the same user at various time points in the cloud computing platform's history. A user's virtual machine usage record can include the instance type and processor utilization of each virtual machine created by the user over time.

[0078] Specifically, the configuration parameters such as the name, instance type, network configuration parameters, etc. of the virtual machines created by each user can be stored, and after the virtual machine is started, the processor utilization of the host machine when processing the tasks corresponding to the virtual machine can be obtained, so as to generate a virtual machine usage record for each user based on the configuration parameters of each virtual machine created by the user and the processor utilization corresponding to each virtual machine.

[0079] Furthermore, after each virtual machine (including the target virtual machine) is started, the processor utilization of the host machine running the virtual machine at each time point can be collected at a certain period, such as one hour, three hours, or 12 hours. For each virtual machine, based on the configuration parameters of the virtual machine and the processor utilization corresponding to the virtual machine at each time point, a usage record of the virtual machine is generated. The usage records of each virtual machine corresponding to the same user constitute the virtual machine usage record of the user.

[0080] In one embodiment, a power consumption profile of the user's virtual machine corresponding to the target virtual machine can be generated based on the virtual machine usage record of the user, and the power consumption type of the target virtual machine can be determined based on the power consumption profile of the user's virtual machine, or the processor utilization corresponding to the target virtual machine can be predicted based on the power consumption profile of the user's virtual machine, thereby determining the power consumption type of the target virtual machine based on the predicted processor utilization.

[0081] Specifically, the power consumption type of the target virtual machine can be determined by determining a model and a power consumption profile based on pre-trained power consumption types.

[0082] Step S303 : predicting the processor utilization corresponding to the target virtual machine according to the instance type of the target virtual machine and the processor utilization corresponding to each virtual machine recorded in the virtual machine usage record.

[0083] Specifically, the processor utilization corresponding to each virtual machine of the same instance type as the target virtual machine can be obtained from the virtual machine usage record of the user corresponding to the target virtual machine, recorded as the first processor utilization, and the processor utilization corresponding to the target virtual machine can be predicted based on each first processor utilization.

[0084] Exemplarily, an average value of the utilization rates of the first processors may be determined as the processor utilization rate corresponding to the target virtual machine.

[0085] Exemplarily, based on a pre-established prediction model, such as a trained neural network model, the processor utilization of the target virtual machine created by the user at the current time can be predicted according to each first processor utilization and its corresponding time node.

[0086] Specifically, the processor utilization corresponding to the target virtual machine may be predicted based on the processor utilization corresponding to each virtual machine created by the user and having an instance type similar to that of the target virtual machine, which is recorded as the second processor utilization.

[0087] The instance types being similar may refer to the instance types belonging to the same general category, such as general-purpose, computing, and big data types. Each general category may correspond to multiple instance types.

[0088] Specifically, a weight coefficient can be set for each instance type based on the similarity with the instance type of the target virtual machine, and a weighted average of the second processor utilizations is calculated based on the weight coefficients, and the weighted average is determined as the processor utilization corresponding to the target virtual machine.

[0089] Furthermore, for each virtual machine created by the user that is similar to the instance type of the target virtual machine, the weight coefficient of the virtual machine can be determined based on parameters such as the instance type, creation time, and startup time of the virtual machine. Based on the weight coefficient, the weighted average of the utilization rates of each second processor is calculated, and the weighted average is determined to be the processor utilization corresponding to the target virtual machine.

[0090] For example, the higher the similarity between the instance type of the virtual machine and the instance type of the target virtual machine, the more similar the creation time of the virtual machine is to the creation time of the target virtual machine, and the more similar the startup time of the virtual machine is to the startup time of the target virtual machine, the higher the weight coefficient of the virtual machine.

[0091] Step S304: determining the power consumption type of the target virtual machine according to the predicted processor utilization.

[0092] Specifically, the power consumption type of the target virtual machine may be determined based on a comparison result of the predicted processor utilization and one or more preset thresholds.

[0093] Step S305 : Calculate a scheduling score of the power consumption characteristics of each physical server according to the power consumption type of the target virtual machine and the power consumption type of each physical server.

[0094] Specifically, the scheduling score for the power consumption characteristics of each physical server in the cloud computing platform can be calculated based on the power consumption class of the target virtual machine and the power consumption type of the physical server. The higher the similarity between the power consumption type of the physical server and the power consumption type of the target virtual machine, the higher the scheduling score of the power consumption characteristics of the physical server.

[0095] Optionally, calculating the scheduling score of the power consumption characteristics of each physical server according to the power consumption type of the target virtual machine and the power consumption type of each physical server includes:

[0096] For each physical server, determine whether the power consumption type of the physical server matches that of the target virtual machine; if they match, determine the first score as the scheduling score of the power consumption characteristic of the physical server.

[0097] The power consumption type matching may be the same power consumption type.

[0098] For example, the first score may be 0, 100, 1000 or other values. The first score is a higher score.

[0099] Specifically, the scheduling scores of the power consumption characteristics of the physical servers that match the power consumption type of the target virtual machine are higher than the scheduling scores of the power consumption characteristics of the physical servers that do not match the power consumption type of the target virtual machine.

[0100] In one embodiment, the power consumption type can be divided into two types, low power consumption type and high power consumption type, by a preset threshold. The load corresponding to the low power consumption type is less than or equal to the preset threshold, and the load corresponding to the high power consumption type is greater than the preset threshold.

[0101] Optionally, if the power consumption types of the physical server and the target virtual machine do not match, the method further includes:

[0102] If the power consumption type of the target virtual machine is a low power consumption type and the power consumption type of the physical server is a high power consumption type, the fourth score is determined as the scheduling score of the power consumption characteristics of the physical server; if the power consumption type of the target virtual machine is a high power consumption type and the power consumption type of the physical server is a low power consumption type, the fifth score is determined as the scheduling score of the power consumption characteristics of the physical server; wherein, the fifth score is less than the fourth score, and the fourth score is less than the first score.

[0103] Among them, the load of the low-power target virtual machine or the load of the physical server under the corresponding rack power consumption is less than or equal to the preset threshold, and the load of the high-power target virtual machine or the load of the physical server under the corresponding rack power consumption is greater than the preset threshold.

[0104] Specifically, the scheduling score of the power consumption characteristic of the physical server whose power consumption type does not match the power consumption type of the target virtual machine may be a negative score or a penalty score.

[0105] For example, the fourth score may be -10, -20, -1000, -2000, -200000, etc., and the fifth score may be -50, -60, -5000, -6000, -600000, etc.

[0106] By determining the power consumption type matching, the scheduling scores of the power consumption characteristics of physical servers of various power consumption types are calculated, and a higher scheduling score is assigned to the physical servers with matching power consumption types. This allows virtual machines to be assigned physical servers with matching power consumption types first, and then physical servers with higher power consumption types are matched, thereby improving the power consumption utilization of the physical servers and effectively reducing the risk of overload of the racks where the physical servers are located.

[0107] In one embodiment, to further improve the accuracy of server scheduling, the power consumption types can be divided into three or more types based on multiple preset thresholds. When the power consumption types of the physical server and the target virtual machine do not match, the scheduling score of the physical server power consumption characteristics can be determined based on the deviation between the power consumption types of the physical server and the target virtual machine. The greater the deviation, the lower the scheduling score.

[0108] Optionally, if the power consumption types of the physical server and the target virtual machine do not match, the method further includes:

[0109] If the power consumption type of the physical server is higher than the power consumption type of the target virtual machine, the scheduling score of the power consumption characteristic of the physical server is determined based on the first relationship and the deviation between the power consumption types of the physical server and the target virtual machine, wherein the maximum value of the scheduling score obtained by the first relationship is less than the second score, and the second score is less than the first score; if the power consumption type of the physical server is lower than the power consumption type of the target virtual machine, the scheduling score of the power consumption characteristic of the physical server is determined based on the second relationship and the difference between the power consumption types of the physical server and the target virtual machine, wherein the maximum value of the scheduling score output by the second relationship is less than the third score, and the third score is less than the second score.

[0110] The deviation between the power consumption type of the physical server and the target virtual machine can be the difference between the class or level of the power consumption type of the physical server and the class or level of the power consumption type of the target virtual machine. The greater the deviation corresponding to the physical server, the lower the scheduling score of the power consumption characteristics of the physical server. The higher the class or level of the power consumption type, the higher the power consumption type.

[0111] The first and second equations represent different correspondences, and may be linear or nonlinear, which is not limited in this application. The second score may be any value less than the first score, such as a negative score, such as -200,000. The third score may be any value less than the second score, such as -600,000.

[0112] Exemplarily, the first relational expression and the second relational expression may both be linear relations that decay as the deviation increases, that is, the slopes of the first relational expression and the second relational expression are both negative, and the slope of the first relational expression is greater than the slope of the second relational expression.

[0113] When the power consumption types do not match, based on the mismatched types, the corresponding relationship is selected based on the power consumption types of the physical server and the target virtual machine to calculate the scheduling score of the physical server power consumption characteristics, thereby improving the accuracy of the scheduling score calculation and thus improving the accuracy of the virtual machine scheduling.

[0114] Step S306 : determining a target physical server corresponding to the target virtual machine from the multiple physical servers according to the scheduling score of the power consumption characteristic of the physical server, so as to schedule the target virtual machine to the corresponding target physical server.

[0115] Specifically, a physical server with the highest scheduling score among the multiple physical servers may be determined as the target physical server corresponding to the target virtual machine, and the target virtual machine may be scheduled to the target physical server.

[0116] When there are multiple physical servers with the highest scheduling scores, we can further determine the physical server with the best performance or the performance that is most suitable for the target virtual machine from the physical servers with the highest scheduling scores based on the basic parameters of the physical servers, such as memory and number of processor cores, and use it as the target physical server corresponding to the target virtual machine.

[0117] Furthermore, when scheduling the target virtual machine to the target physical server, the user's virtual machine usage record or the power consumption profile of the user's virtual machine can be updated based on the processor utilization of the target physical server at each time node, that is, the incremental data of the target virtual machine.

[0118] In this embodiment, based on the user's virtual machine usage record and the instance type of the target virtual machine currently created by the user, the processor utilization corresponding to the target virtual machine is predicted based on the processor utilization corresponding to each virtual machine with the same instance type in the virtual machine usage record, and then the power consumption type of the target virtual machine is determined based on the predicted processor utilization of the target virtual machine, thereby realizing the prediction of the virtual machine power consumption type with high prediction accuracy; and then based on the predetermined power consumption type of the physical server and the power consumption type of the target virtual machine, the scheduling score of the power consumption characteristics of each physical server is calculated, the greater the similarity of the power consumption type or the smaller the deviation, the higher the scheduling score of the power consumption characteristics of the physical server, and based on the scheduling score, the host machine of the target virtual machine is determined from multiple physical servers, thereby realizing a server scheduling method based on power consumption matching, which has low complexity, is easy to implement, and has high accuracy.

[0119] Optional, Figure 4 For this application Figure 3 The flowchart of step S305 in the embodiment shown is as follows: Figure 4 As shown, the method for determining the scheduling score of the power consumption characteristics of each physical server in step S305 may include the following steps:

[0120] Step S401 , determining whether the power consumption type of the target virtual machine is a high power consumption type; if so, executing step S402 ; if not, jumping to step S405 .

[0121] Step S402, determining whether the power consumption type of the physical server is a high power consumption type; if so, executing step S403; if not, executing step S404;

[0122] Step S403: Determine that the scheduling score of the power consumption characteristic of the physical server is a first score.

[0123] Step S404: Determine that the scheduling score of the power consumption characteristic of the physical server is a fifth score.

[0124] Step S405 , determining whether the power consumption type of the physical server is a low power consumption type; if so, jumping to step S403 ; if not, executing step S406 .

[0125] Step S406: Determine that the scheduling score of the power consumption characteristic of the physical server is a fourth score.

[0126] In one embodiment, the power consumption scheduler can determine and output the power consumption type of the physical server and the power consumption type of the target virtual machine, such as outputting it to the scorer, which then performs Figure 4 The method for determining the scheduling score shown in the figure outputs the scheduling score of the physical server power consumption characteristic to the power consumption scheduler, and the power consumption scheduler allocates the target physical server to the target virtual machine based on the scheduling score.

[0127] Optionally, determining a target physical server corresponding to the target virtual machine from the multiple physical servers according to the scheduling score of the power consumption characteristic of the physical server includes:

[0128] A target physical server corresponding to the target virtual machine is determined from the multiple physical servers according to the scheduling score of the power consumption characteristic of the physical server and basic parameters of the physical server.

[0129] Among them, the basic parameters include the memory of the physical server and the number of cores of the processor, and the processor can include the CPU and GPU.

[0130] Specifically, the service score of the physical server can be determined based on the basic parameters of the physical server; and the physical server with the highest sum of the scheduling score and the server score is determined to be the target physical server of the target virtual machine.

[0131] For example, the more cores a processor has and the larger the memory is, the higher the service score is.

[0132] Specifically, when there are multiple physical servers (referred to as matching physical servers) with a scheduling score of the first score, the matching physical server with the best service performance can be determined from the multiple matching physical servers as the target physical server of the target virtual machine based on the basic parameters of each matching cloud service.

[0133] Physical servers are scheduled using scheduling scores and multi-dimensional parameters of basic parameters, further improving the accuracy of virtual machine scheduling.

[0134] Optional, Figure 5 For this application Figure 3The flow chart of step S306 in the embodiment shown is as follows: Figure 5 Step S306 may include the following steps:

[0135] Step S501 : for each physical server among the multiple physical servers, determine a server score of the physical server according to various basic parameters of the physical server.

[0136] Specifically, various basic parameters of the physical server may be converted into scores of the servers based on a preset conversion relationship.

[0137] Step S502 : Calculate the total scheduling score of the physical server according to the scheduling score of the power consumption characteristic of the physical server and the server score of the physical server.

[0138] The total scheduling score of the physical server may be the sum of the scheduling score of the power consumption characteristic of the physical server and the server score, or a weighted average of the scheduling score and the service score.

[0139] Step S503 : determining a target physical server corresponding to the target virtual machine from the multiple physical servers according to the total scheduling score.

[0140] Specifically, the physical server with the highest total scheduling score may be determined as the target physical server for the target virtual machine.

[0141] Specifically, the first or any physical server whose total scheduling score is higher than a preset score may be determined as the target physical server of the target virtual machine.

[0142] The score of the physical server is determined based on the scheduling score corresponding to power consumption and the server score corresponding to basic parameters. The score is determined with high accuracy, thereby improving the accuracy of physical server determination. Physical servers are allocated to virtual machines using a scoring system, and virtual machines are scheduled to the physical server with the highest score, which improves the efficiency of virtual machine scheduling.

[0143] In one embodiment, the virtual machine scheduling method provided in the above embodiments of the present application can be executed by a virtual machine scheduling system, which includes a power consumption portrait module, a scorer and a scheduler. The power consumption portrait module is used to determine the power consumption portraits of each physical server and each virtual machine respectively. The scorer is used to determine the power consumption type of the physical server and the virtual machine based on the power consumption portrait, or set a power consumption type label for the physical server and the virtual machine, and determine the scheduling score of the power consumption characteristics of each physical server relative to one of the virtual machines to be scheduled based on the power consumption type of the physical server and the virtual machine; the scheduler is used to allocate a target physical server to the virtual machine to be scheduled, i.e., the target virtual machine, based on the scheduling score.

[0144] Figure 6 A schematic diagram of the structure of a virtual machine scheduling system provided in an embodiment of the present application is shown in FIG. Figure 6 As shown, the virtual machine scheduling system includes a power consumption profile generating module 610, a power consumption profile updating module 620, a scheduling module 630 and a scorer (not shown in the figure).

[0145] The power consumption profile generation module 610 is configured to: determine the power consumption profile of the physical server based on the physical server's rack power consumption and the power consumption dictionary; determine the historical power consumption profile of the virtual machine based on the virtual machine's historical load; synchronize the physical server's power consumption profile with the scheduling module 630; send the historical power consumption profile to the power consumption profile update module 620; and receive the power consumption profile of the virtual machine returned by the scheduling module 630. The power consumption profile update module 620 is configured to update the virtual machine's power consumption profile based on incremental data of the virtual machine's power consumption profile. Specifically, it generates the latest power consumption profile of the virtual machine based on the incremental data of the virtual machine's power consumption profile and the historical power consumption profile, and sends the latest power consumption profile of the virtual machine to the scheduling module 630. The scheduling module 630 is configured to return the latest power consumption profile of the virtual machine to the power consumption profile generation module 610; set power consumption labels for the virtual machine and the physical server based on the virtual machine's power consumption profile and the physical server's power consumption profile, respectively, where the power consumption label is used to characterize the power consumption type; send the power consumption labels of the virtual machine and the physical server to the scorer, and assign a target physical server to the virtual machine based on the scheduling score of the physical server's power consumption characteristics output by the scorer. The scorer is used to calculate the scheduling score of the power consumption characteristics of each physical server according to the power consumption types of the virtual machines and physical servers.

[0146] In one embodiment, the scorer may be provided in the scheduling module 630 .

[0147] Figure 7 This is a structural diagram of a virtual machine scheduling device provided in an embodiment of the present application. The virtual machine scheduling device includes: a power consumption type acquisition module 710 and a server allocation module 720.

[0148] Among them, the power consumption type acquisition module 710 is used to obtain the power consumption type of the target virtual machine of the cloud computing platform, and the cloud computing platform includes multiple physical servers; the server allocation module 720 is used to determine the target physical server corresponding to the target virtual machine from the multiple physical servers according to the power consumption type of the target virtual machine and the power consumption type of each of the physical servers, so as to schedule the target virtual machine to the corresponding target physical server, so that the target physical server provides computing resources for the target virtual machine.

[0149] Optionally, the device further includes:

[0150] The second power consumption type obtaining module is configured to obtain the power consumption type of each physical server in the plurality of physical servers.

[0151] Optionally, the second power consumption type obtaining module is specifically configured to:

[0152] The power consumption type of each physical server in the plurality of physical servers is determined according to the performance parameters of the physical server.

[0153] Optionally, the second power consumption type obtaining module is specifically configured to:

[0154] For each physical server, the power consumption profile of the physical server is determined based on the on-shelf power consumption and power consumption dictionary of the physical server. The power consumption dictionary is used to describe the correspondence between the power consumption and load of the physical server; based on the power consumption profile of the physical server, the power consumption type of the physical server is determined.

[0155] Optionally, the power consumption type acquisition module 710 is specifically configured to:

[0156] The power consumption type of the target virtual machine is determined according to the user and instance type corresponding to the target virtual machine.

[0157] Optionally, the power consumption type acquisition module 710 is specifically configured to:

[0158] Determine the power consumption profile of the user's virtual machine based on the user and instance type corresponding to the target virtual machine; and determine the power consumption type of the target virtual machine based on the power consumption profile of the user's virtual machine.

[0159] Optionally, the power consumption type acquisition module 710 is specifically configured to:

[0160] Obtain a virtual machine usage record of the user corresponding to the target virtual machine, wherein the virtual machine usage record is used to record the processor utilization of the physical server allocated to virtual machines of various instance types corresponding to the user at various time nodes of the cloud computing platform within a historical time; predict the processor utilization corresponding to the target virtual machine based on the instance type of the target virtual machine and the processor utilization corresponding to each virtual machine recorded in the virtual machine usage record; determine the power consumption type of the target virtual machine based on the predicted processor utilization.

[0161] Optionally, the server allocation module 720 includes:

[0162] A score calculation unit is used to calculate the scheduling score of the power consumption characteristics of each physical server according to the power consumption type of the target virtual machine and the power consumption type of each physical server; a server allocation unit is used to determine the target physical server corresponding to the target virtual machine from the multiple physical servers according to the scheduling score of the power consumption characteristics of the physical server.

[0163] Optional score calculation unit, specifically used for:

[0164] For each physical server, determine whether the power consumption type of the physical server matches that of the target virtual machine; if they match, determine the first score as the scheduling score of the power consumption characteristic of the physical server.

[0165] Optionally, if the power consumption types of the physical server and the target virtual machine do not match, the score calculation unit is further configured to:

[0166] If the power consumption type of the physical server is higher than the power consumption type of the target virtual machine, the scheduling score of the power consumption characteristic of the physical server is determined based on the first relationship and the deviation between the power consumption types of the physical server and the target virtual machine, wherein the maximum value of the scheduling score obtained by the first relationship is less than the second score, and the second score is less than the first score; if the power consumption type of the physical server is lower than the power consumption type of the target virtual machine, the scheduling score of the power consumption characteristic of the physical server is determined based on the second relationship and the difference between the power consumption types of the physical server and the target virtual machine, wherein the maximum value of the scheduling score output by the second relationship is less than the third score, and the third score is less than the second score.

[0167] Optionally, if the power consumption types of the physical server and the target virtual machine do not match, the score calculation unit is further configured to:

[0168] If the power consumption type of the target virtual machine is a low power consumption type and the power consumption type of the physical server is a high power consumption type, the fourth score is determined as the scheduling score of the power consumption characteristics of the physical server; if the power consumption type of the target virtual machine is a high power consumption type and the power consumption type of the physical server is a low power consumption type, the fifth score is determined as the scheduling score of the power consumption characteristics of the physical server; wherein, the fifth score is less than the fourth score, and the fourth score is less than the first score.

[0169] Optionally, a server allocation unit is used to:

[0170] A target physical server corresponding to the target virtual machine is determined from the multiple physical servers according to the scheduling score of the power consumption characteristics of the physical server and basic parameters of the physical server; wherein the basic parameters include the memory and the number of cores of the processor of the physical server.

[0171] Optionally, a server allocation unit is used to:

[0172] For each physical server among the multiple physical servers, determine the server score of the physical server based on the various basic parameters of the cloud computing service; calculate the total scheduling score of the physical server based on the scheduling score of the physical server's power consumption characteristics and the server score of the physical server; and determine the target physical server corresponding to the target virtual machine from the multiple physical servers based on the server's total scheduling score.

[0173] The virtual machine scheduling device provided in the embodiment of the present application can be used to perform the above Figures 2 to 5 The technical solutions provided in any corresponding embodiment have similar implementation principles and technical effects, which will not be described in detail in this embodiment.

[0174] Figure 8 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application is shown in FIG. Figure 8 As shown, the electronic device provided in this embodiment includes:

[0175] At least one processor 810; and a memory 820 communicatively connected to the at least one processor; wherein the memory 820 stores computer-executable instructions; the at least one processor 810 executes the computer-executable instructions stored in the memory so that the electronic device executes the virtual machine scheduling method provided in any of the aforementioned embodiments.

[0176] Optionally, the memory 820 may be independent or integrated with the processor 810 .

[0177] The implementation principle and technical effects of the electronic device provided in this embodiment can be found in the aforementioned embodiments and will not be described in detail here.

[0178] An embodiment of the present application further provides a computer-readable storage medium, in which computer-executable instructions are stored. When the computer-executable instructions are executed by a processor, the virtual machine scheduling method provided in any of the aforementioned embodiments can be implemented.

[0179] An embodiment of the present application also provides a computer program product, including a computer program, which, when executed by a processor, implements the virtual machine scheduling method provided in any of the aforementioned embodiments.

[0180] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of modules is merely a logical function division. In actual implementation, other division methods may be used. For example, multiple modules may be combined or integrated into another system, or some features may be ignored or not implemented.

[0181] The above-mentioned integrated module implemented in the form of a software functional module can be stored in a computer-readable storage medium. The above-mentioned software functional module is stored in a storage medium and includes a number of instructions for causing a computer device (which can be a personal computer, server, or network device, etc.) or a processor to perform some steps of the method described in each embodiment of the present application.

[0182] It should be understood that the above-mentioned processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), etc. The general-purpose processor may be a microprocessor or any conventional processor. The steps of the method disclosed in the application may be directly embodied as being executed by a hardware processor, or may be executed by a combination of hardware and software modules in the processor. The memory may include high-speed RAM memory, and may also include non-volatile storage NVM, such as at least one disk memory, and may also be a USB flash drive, a mobile hard disk, a read-only memory, a magnetic disk or an optical disk, etc.

[0183] The storage medium may be implemented by any type of volatile or non-volatile memory device, or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. The storage medium may be any available medium that can be accessed by a general-purpose or special-purpose computer.

[0184] An exemplary storage medium is coupled to a processor so that the processor can read information from the storage medium and write information to the storage medium. Of course, the storage medium can also be an integral part of the processor. The processor and the storage medium can be located in an application specific integrated circuit (ASIC). Of course, the processor and the storage medium can also exist as discrete components in an electronic device or a main control device.

[0185] It should be noted that, in this document, the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, article, or apparatus comprising a series of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of other identical elements in the process, method, article, or apparatus comprising the element.

[0186] The serial numbers of the above embodiments of the present application are for description only and do not represent the advantages or disadvantages of the embodiments.

[0187] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus the necessary general hardware platform, and of course can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes a number of instructions for enabling a terminal device (which can be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods provided in each embodiment of the present application.

[0188] Those skilled in the art will readily appreciate other embodiments of the present application after considering the specification and practicing the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of the present application that follow the general principles of the present application and include common knowledge or customary techniques in the art not disclosed herein. The description and examples are to be considered as exemplary only, and the true scope and spirit of the present application are indicated by the following claims.

[0189] It should be understood that the present application is not limited to the exact structure described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present application is limited only by the appended claims.

Claims

1. A virtual machine scheduling method, characterized in that: include: Obtaining a power consumption type of a target virtual machine of a cloud computing platform, wherein the cloud computing platform includes a plurality of physical servers; Determining, from the plurality of physical servers, a target physical server corresponding to the target virtual machine according to the power consumption type of the target virtual machine and the power consumption type of each of the physical servers, so as to schedule the target virtual machine to the corresponding target physical server; Determining a target physical server corresponding to the target virtual machine from the multiple physical servers according to the power consumption type of the target virtual machine and the power consumption type of each of the physical servers includes: For each physical server, determining whether the power consumption type of the physical server matches that of the target virtual machine; If there is a match, determining the first score as the scheduling score of the power consumption characteristic of the physical server; If there is no match, the method further includes: If the power consumption type of the physical server is higher than the power consumption type of the target virtual machine, the scheduling score of the power consumption characteristic of the physical server is determined based on the first relational expression and the deviation between the power consumption types of the physical server and the target virtual machine, wherein the maximum value of the scheduling score obtained by the first relational expression is less than the second score, and the second score is less than the first score; if the power consumption type of the physical server is lower than the power consumption type of the target virtual machine, the scheduling score of the power consumption characteristic of the physical server is determined based on the second relational expression and the difference between the power consumption types of the physical server and the target virtual machine, wherein the maximum value of the scheduling score output by the second relational expression is less than the third score, and the third score is less than the second score; or, If the power consumption type of the target virtual machine is a low power consumption type and the power consumption type of the physical server is a high power consumption type, determining a fourth score as the scheduling score of the power consumption characteristics of the physical server; if the power consumption type of the target virtual machine is a high power consumption type and the power consumption type of the physical server is a low power consumption type, determining a fifth score as the scheduling score of the power consumption characteristics of the physical server; wherein the fifth score is less than the fourth score, and the fourth score is less than the first score; A target physical server corresponding to the target virtual machine is determined from the multiple physical servers according to the scheduling score of the power consumption characteristic of the physical server.

2. The method according to claim 1, characterized in that Obtain the power consumption type of the target virtual machine on the cloud computing platform, including: The power consumption type of the target virtual machine is determined according to the user and instance type corresponding to the target virtual machine.

3. The method according to claim 2, characterized in that Determining the power consumption type of the target virtual machine according to the user and instance type corresponding to the target virtual machine includes: Determining a power consumption profile of the user's virtual machine based on the user and instance type corresponding to the target virtual machine; The power consumption type of the target virtual machine is determined according to the power consumption profile of the user's virtual machine.

4. The method according to claim 2, characterized in that Determining the power consumption type of the target virtual machine according to the user and instance type corresponding to the target virtual machine includes: Obtaining a virtual machine usage record of a user corresponding to the target virtual machine, wherein the virtual machine usage record is used to record processor utilization of physical servers allocated to virtual machines of various instance types corresponding to the user at various time points in the cloud computing platform over a historical period; Predicting a processor utilization rate corresponding to the target virtual machine based on the instance type of the target virtual machine and the processor utilization rates corresponding to the respective virtual machines recorded in the virtual machine usage record; The power consumption type of the target virtual machine is determined according to the predicted processor utilization.

5. The method according to any one of claims 1 to 4, characterized in that The method further comprises: For each physical server, determine a power consumption profile of the physical server based on the rack power consumption of the physical server and a power consumption dictionary, where the power consumption dictionary is used to describe the corresponding relationship between the power consumption and load of the physical server; The power consumption type of the physical server is determined according to the power consumption profile of the physical server.

6. The method according to any one of claims 1 to 4, characterized in that Determining a target physical server corresponding to the target virtual machine from the multiple physical servers according to the scheduling score of the power consumption characteristic of the physical server includes: Determining a target physical server corresponding to the target virtual machine from the multiple physical servers according to the scheduling score of the power consumption characteristics of the physical server and the basic parameters of the physical server; Among them, basic parameters include the memory of the physical server and the number of processor cores.

7. A virtual machine scheduling device, characterized in that: include: a power consumption type acquisition module, configured to acquire a power consumption type of a target virtual machine of a cloud computing platform, wherein the cloud computing platform includes a plurality of physical servers; A server allocation module is configured to determine, from the plurality of physical servers, a target physical server corresponding to the target virtual machine according to the power consumption type of the target virtual machine and the power consumption type of each of the physical servers, so as to schedule the target virtual machine to the corresponding target physical server; Server allocation module, including: Score calculation unit, used to: For each physical server, determining whether the power consumption type of the physical server matches that of the target virtual machine; If there is a match, determining the first score as the scheduling score of the power consumption characteristic of the physical server; If no match is found, also includes: If the power consumption type of the physical server is higher than the power consumption type of the target virtual machine, the scheduling score of the power consumption characteristic of the physical server is determined based on the first relational expression and the deviation between the power consumption types of the physical server and the target virtual machine, wherein the maximum value of the scheduling score obtained by the first relational expression is less than the second score, and the second score is less than the first score; if the power consumption type of the physical server is lower than the power consumption type of the target virtual machine, the scheduling score of the power consumption characteristic of the physical server is determined based on the second relational expression and the difference between the power consumption types of the physical server and the target virtual machine, wherein the maximum value of the scheduling score output by the second relational expression is less than the third score, and the third score is less than the second score; or, If the power consumption type of the target virtual machine is a low power consumption type and the power consumption type of the physical server is a high power consumption type, determining a fourth score as the scheduling score of the power consumption characteristics of the physical server; if the power consumption type of the target virtual machine is a high power consumption type and the power consumption type of the physical server is a low power consumption type, determining a fifth score as the scheduling score of the power consumption characteristics of the physical server; wherein the fifth score is less than the fourth score, and the fourth score is less than the first score; The server allocation unit is configured to determine a target physical server corresponding to the target virtual machine from the multiple physical servers according to the scheduling score of the power consumption characteristic of the physical server.

Images