ICV cloud platform distributed data center-oriented workload optimization method and system

Abstract

The invention provides an ICV cloud platform distributed data center-oriented workload optimization method and system. The method comprises the following steps of: obtaining a total task arrival rate, and related configuration parameters and privacy constraint conditions of a geographic distributed data center; according to the total task arrival rate, related configuration parameters of the geographic distributed data center and privacy constraint conditions, obtaining a workload management strategy based on a privacy perception workload optimization model; wherein the privacy perception workload optimization model is constructed based on an M / M / m queuing model, a privacy model and a power consumption model, mapping from a task to a data center is used as a chromosome coding mode, and an optimal workload strategy is searched under the condition that privacy constraint conditions are met through an evolution process of a genetic algorithm. According to the ICV cloud platform distributed data center-oriented workload optimization method disclosed by the invention, the problems that the existing data security is low, the privacy constraint is not met and the workload management strategy is poor are solved from the source.

Description

Technical Field

[0001] The present invention belongs to the field of workload management, and in particular relates to a workload optimization method and system for an ICV cloud platform distributed data center. Background Art

[0002] ICV, or intelligent connected vehicle, refers to the organic combination of the Internet of Vehicles and smart cars, and will eventually replace human operators in the next generation of vehicles. With the advent of the big data era, the application of geographically distributed data centers has become more and more common due to the complex computing tasks and massive data processing involved.

[0003] Currently, resource utilization, performance optimization, customer satisfaction, market competition and server configuration have been listed as urgent research topics. For ICV cloud platform distributed data centers, how to make full use of available resources, provide users with the best service quality and obtain the best workload management strategy is a huge challenge.

[0004] However, in the geographically distributed environment of ICV cloud platforms, the problem of how to achieve optimal workload management becomes more complicated because it needs to consider multiple factors, including the upper limit of data center processing speed, privacy constraints, and workload. Existing workload optimization methods for distributed data centers of ICV cloud platforms have problems such as low data security, failure to meet privacy constraints, and poor workload management strategies.

[0005] Therefore, how to find the best workload management strategy for distributed data centers on ICV cloud platforms is an urgent problem to be solved in this field. Summary of the invention

[0006] The purpose of the present invention is to provide a workload optimization method and system for distributed data centers of ICV cloud platforms in view of the defects of the prior art. The workload optimization method for distributed data centers of ICV cloud platforms of the present invention obtains a workload management strategy based on a privacy-aware workload optimization model according to the total task arrival rate, relevant configuration parameters of geographically distributed data centers and privacy constraints; wherein the privacy-aware workload optimization model is constructed based on an M / M / m queuing model, a privacy model and a power consumption model, and uses the mapping of tasks to data centers as a chromosome encoding method, and through the evolutionary process of a genetic algorithm, finds the optimal workload strategy while satisfying privacy constraints, thereby fundamentally solving the problems of low data security, non-compliance with privacy constraints and poor workload management strategy in existing workload optimization methods for distributed data centers of ICV cloud platforms.

[0007] In order to achieve the above purpose, the present invention adopts the following technical solutions:

[0008] A workload optimization method for an ICV cloud platform distributed data center, characterized in that it comprises the following steps:

[0009] S1. Obtain the total task arrival rate, relevant configuration parameters of geographically distributed data centers, and privacy constraints;

[0010] S2. According to the total task arrival rate, relevant configuration parameters of the geographically distributed data center and privacy constraints, a workload management strategy is obtained based on a privacy-aware workload optimization model; wherein the privacy-aware workload optimization model is constructed based on an M / M / m queuing model, a privacy model and a power consumption model;

[0011] Step S2 includes:

[0012] S21, input the total task arrival rate, relevant configuration parameters of geographically distributed data centers and privacy constraints;

[0013] S22. Prepare candidate data centers for each type of task; prepare candidate data centers that meet the privacy requirements for each type of task according to the privacy constraints;

[0014] S23, initializing the population and workload distribution plan; specifically including:

[0015] S231. Determine the representation method and fitness function of the problem, and generate an initial population; the population is composed of a group of individuals, and the individuals are composed of chromosomes and fitness values; wherein the chromosomes are encoded as follows: the length of each chromosome is defined as the total task arrival rate, and the tasks are represented by genes; the chromosome encoding method is a mapping of tasks to data centers TasktoDC; in the TasktoDC mapping, each task is assigned to a data center that satisfies its privacy constraints;

[0016] S232, generating an initial workload allocation scheme according to the privacy constraint condition, under which tasks are randomly allocated to a data center that meets its privacy condition and does not exceed the upper limit of the data center processing speed;

[0017] S24, determine whether the preset genetic generation is reached, if so, execute step S210, otherwise, execute step S25;

[0018] S25. Allocate the tasks contained in the chromosome of each individual in the population to a suitable data center; wherein the allocation process satisfies the privacy constraint and does not exceed the upper limit of the processing speed of the data center;

[0019] S26, calculating the workload of each data center and evaluating the individual fitness; wherein the workload of each data center is calculated based on the number of tasks allocated to each data center; and the fitness value of the current individual is calculated based on the workload, and the fitness value is the system performance value under the current allocation scheme;

[0020] S27, select individuals with high fitness values ​​as parents;

[0021] S28, performing mutation operation on the selected individuals;

[0022] S29, generate a new population; the new offspring will be used as a new population, and go to step S24 to continue evolution;

[0023] S210. Select the individual with the highest fitness in the final population as output, thereby obtaining a workload management strategy.

[0024] Furthermore, in step S1, the total task arrival rate represents the total number of tasks arriving at all data centers per unit time in a geographically distributed environment;

[0025] Furthermore, in step S1, the relevant configuration parameters of the geographically distributed data center include the number of servers, task execution speed, utilization, static power consumption, customer satisfaction parameters, and service level agreement.

[0026] Furthermore, in step S1, the privacy constraint condition is to assign the task to a data center with a privacy level not lower than that of the task for execution.

[0027] Furthermore, in step S2, the privacy-aware workload optimization model is constructed based on the M / M / m queuing model, the privacy model, and the power consumption model, including: modeling the data center based on the M / M / m queuing model and simulating the task arrival process.

[0028] Furthermore, the privacy model is a multi-level privacy model, which is established according to the privacy levels of tasks and data centers in various places. The multi-level privacy model consists of two parts: a task privacy level and a data center privacy level.

[0029] Furthermore, the power consumption model: data center power consumption refers to the energy consumed by the data center during operation. In a geographically distributed data center, there are two different processing methods for the power consumption of the data center when it is idle, namely, an idle speed model and a constant speed model.

[0030] The present invention also proposes a workload optimization system for an ICV cloud platform distributed data center, characterized in that the workload optimization system executes the workload optimization method for an ICV cloud platform distributed data center, including: a data acquisition module, a workload management strategy optimization module;

[0031] The data acquisition module acquires the total task arrival rate, relevant configuration parameters of the geographically distributed data center, and privacy constraints:

[0032] The workload management strategy optimization module obtains a workload management strategy based on a privacy-aware workload optimization model according to the total task arrival rate, relevant configuration parameters of the geographically distributed data center and privacy constraints.

[0033] The present invention further provides a computer device, which includes a memory and a processor. The memory stores a computer program, and the processor implements the above method when executing the computer program.

[0034] Compared with the prior art, it has the following beneficial effects:

[0035] The workload optimization method for the distributed data center of the ICV cloud platform of the present invention obtains a workload management strategy based on a privacy-aware workload optimization model according to the total task arrival rate, relevant configuration parameters of the geographically distributed data center and privacy constraints; wherein the privacy-aware workload optimization model is constructed based on the M / M / m queuing model, the privacy model and the power consumption model, and the mapping of tasks to data centers is used as a chromosome encoding method. Through the evolutionary process of the genetic algorithm, the optimal workload strategy for the distributed data center of the ICV cloud platform is found while satisfying the privacy constraints, thereby ensuring data privacy security and obtaining the optimal workload management strategy. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying any creative work.

[0037] Figure 1 A schematic diagram of a workload optimization method for an ICV cloud platform distributed data center provided in an embodiment of the present invention.

[0038] Figure 2 A technical roadmap for a workload optimization method for an ICV cloud platform distributed data center provided in an embodiment of the present invention.

[0039] Figure 3 A flowchart of a privacy-aware workload optimization model provided in an embodiment of the present invention.

[0040] Figure 4 This is a chromosome encoding method provided in an embodiment of the present invention.

[0041] Figure 5 A schematic diagram of a workload optimization system for an ICV cloud platform distributed data center provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0042] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0043] It should be understood that when used in this specification and the appended claims, the terms "include" and "comprises" indicate the presence of described features, integers, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or combinations thereof.

[0044] The present invention will be further described below in conjunction with the accompanying drawings and specific embodiments, but they are not intended to limit the present invention.

[0045] It should also be understood that the terms used in this specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. As used in the specification of the present invention and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" are intended to include plural forms.

[0046] It should be further understood that the term "and / or" used in the present description and the appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0047] Terminology explanation:

[0048] Geographically distributed (geo-distributed) datacenters (DCs): In the field of cloud computing, there are multiple cloud deployment environments, including public cloud, private cloud, hybrid cloud, and geographically distributed data centers. A geographically distributed data center is a distributed architecture consisting of multiple data centers that are located in different regions around the world and interconnected by high-speed networks to form a unified cloud computing service platform. This data center layout allows data, computing, and applications to be deployed and processed in multiple locations to improve system performance and reliability. Geographically distributed data centers are usually established and managed by large cloud computing service providers to meet customer needs for high availability, fault tolerance, and data security.

[0049] Multi-layer privacy and security model: The multi-layer privacy and security model divides data and data centers into different levels. Different data centers may have different privacy requirements and capabilities. In order to ensure data security, tasks involving sensitive data need to be processed in appropriate data centers. Specifically, some data centers may have higher privacy levels, so that the processing of tasks can be more strictly protected to ensure that sensitive data is not leaked. Other data centers may have lower privacy levels and can only process tasks with relatively low privacy requirements. In geographically distributed data centers, a multi-layer privacy model is usually used to meet application security requirements.

[0050] Workload management: Workload management refers to the process of allocating, scheduling, and optimizing workloads. In a geographically distributed environment, there are usually multiple data centers scattered in different geographical locations, which are jointly responsible for processing the workload of the system. When the total task arrival rate at a certain moment is known, our goal is to reasonably allocate these tasks based on the resource status and task requirements of each data center so that each data center can obtain an appropriate task arrival rate, i.e., workload. Workload management in geographically distributed data centers is of great significance to improving the availability, reliability, and flexibility of data centers.

[0051] The present invention proposes a workload optimization method for an ICV cloud platform distributed data center. Figure 1 As shown, the workload optimization method for the ICV cloud platform distributed data center includes the following steps S1 to S2.

[0052] A workload optimization method for an ICV cloud platform distributed data center, characterized in that it comprises the following steps:

[0053] S1. Obtain the total task arrival rate, relevant configuration parameters of geographically distributed data centers, and privacy constraints;

[0054] In step S1, the total task arrival rate represents the total number of tasks arriving at all data centers per unit time in a geographically distributed environment. This concept is usually used to describe and manage issues such as task allocation, load balancing, and resource scheduling in distributed systems. In a geographically distributed environment, data centers located in different geographical locations may have different resources, capabilities, and requirements, so the rate at which tasks arrive may also vary.

[0055] In a specific embodiment, the relevant configuration parameters of the geographically distributed data center include the number of servers, task execution speed, utilization, static power consumption, customer satisfaction parameters, and service level agreement.

[0056] Furthermore, the main symbols and definitions involved in the data center include: the number of data centers m, the i-th data center μ i , μ i The number of servers in h i , μ i The speed at which the server executes tasks i , μ i The task arrival rate v i , total task achievement rate υ, task execution requirement r, μ i The average service rate i , μ i The server utilization θ i , the task is in μ i The execution time x on i , waiting time W i , response time T i .

[0057] μ i The server utilization θ i The calculation formula is: i =υ i / h i ζ i ; where μ i The average service rate i for: Task execution time x i =r / s i . Due to the waiting time W i is a random variable, so μ i The response time is also a random variable, T i =W i +r / s i .

[0058] In addition, the relevant configuration parameters include not only static power consumption but also dynamic power consumption. You can choose one of the two according to the actual situation.

[0059] In a specific embodiment, the privacy constraint condition is to assign the task to a data center with a privacy level not lower than that of the task for execution.

[0060] Specifically, due to the different privacy laws and regulations in different countries and regions, there are also great differences in the strictness of data privacy. According to the location where the input data is generated and the area where the data center is located, the present invention divides the privacy level for it. The privacy constraint condition means that data with a low privacy level can only be transmitted to a data center with a higher privacy level than it. The privacy level of a task is determined by the privacy level of the input data. For example, if a task has two input data, and the privacy requirements are 1 and 2 respectively, the privacy level of the task is 1. That is, the privacy level of the task is consistent with the data with the strictest privacy requirements.

[0061] In a specific embodiment, the privacy level of a task is shown in the following table:

[0062]

[0063] The privacy levels of the data centers are shown in the following table:

[0064]

[0065] Among them, the smaller the number, the higher the privacy level.

[0066] The privacy constraint condition means that the task is assigned to a data center with a privacy level not lower than that of the task for execution. The feasible allocation schemes include:

[0067] (1) Data Center 1: Task 1, Task 2.

[0068] (2) Data Center 1: Task 1; Data Center 2: Task 2.

[0069] S2. According to the total task arrival rate, relevant configuration parameters of the geographically distributed data center and privacy constraints, a workload management strategy is obtained based on a privacy-aware workload optimization model, such as Figure 2 As shown; wherein, the privacy-aware workload optimization model is constructed based on the M / M / m queuing model, the privacy model, and the power consumption model.

[0070] In step S2, the privacy-aware workload optimization model is constructed based on the M / M / m queuing model, the privacy model, and the power consumption model, including: modeling the data center based on the M / M / m queuing model and simulating the task arrival process.

[0071] Specifically, the M / M / m queuing model, also known as the single-queue multi-server parallel service model, is a classic queuing theory model. In this model, "M" represents that both customer arrival and service time obey exponential distribution, and "m" represents the number of service stations. This model is mainly used to describe and analyze a queuing system, in which there is a waiting queue and multiple service stations, the customer arrival and service processes are random, and the service stations are independent of each other. The present invention models the data center based on the M / M / m queuing model and simulates the task arrival process, which conforms to the random characteristics of task arrival and processing in the data center. Using the M / M / m queuing model to model the data center can accurately describe the service process, is easy to understand and analyze, and supports the simulation of data centers of different sizes and configurations.

[0072] The M / M / m queuing model models the data center and is the basis of the workload optimization model. Only by first abstracting the data center into an M / M / m queuing model can we simulate the process of task processing and then optimize it to obtain the workload management strategy.

[0073] In a specific embodiment, the privacy model is a multi-level privacy model, which is established according to the privacy levels of tasks and local data centers. The multi-level privacy model consists of two parts: task privacy level and data center privacy level. The multi-level privacy model can effectively protect data privacy security, aiming to provide a more comprehensive data protection and privacy protection mechanism. Among them, the task privacy level is the basis of the multi-level privacy model, and tasks can be divided into different levels according to the sensitivity and privacy requirements of their input data. Different data centers may have different privacy requirements and capabilities. Some data centers may have a higher privacy level, so that the processing process of the task can be more strictly protected to ensure that sensitive data is not leaked. Other data centers may have a lower privacy level and can only process tasks with relatively low privacy requirements. This difference in privacy levels between different data centers will affect the arrival rate of tasks, because the arrival rate of tasks depends on the processing capacity and available resources of the data center.

[0074] The power consumption model: The power consumption of a data center refers to the energy consumed by a data center during operation. In a geographically distributed data center, there are two different processing methods for the power consumption of a data center when it is idle, namely, an idle speed model and a constant speed model.

[0075] In the idle speed model, when the server is idle, it runs at zero speed, that is, it does not perform any tasks. In this model, the idle server does not consume dynamic power, that is, the power consumption is zero. In the constant speed model, when the server is idle, there is no task to be processed. The server still runs at its fixed speed and generates power consumption, but the server response time can be kept stable and in a high-speed operation state. Power consumption management of the data center is also an important part of the data center operation process.

[0076] In the process of obtaining the workload management strategy based on the privacy-aware workload optimization model, the individual fitness value is used to measure the workload optimization effect. However, the power consumption is comprehensively considered in the fitness value.

[0077] The flowchart of step S2 is as follows Figure 3 As shown, the specific process includes:

[0078] S21. Initialization: Input the total task arrival rate, relevant configuration parameters of the geographically distributed data center, and privacy constraints.

[0079] S22. Prepare candidate data centers for each type of task; prepare candidate data centers that meet the privacy requirements for each type of task according to the privacy constraints.

[0080] S23, initializing the population and workload distribution plan; specifically including:

[0081] S231. Determine the representation method and fitness function of the problem, and generate an initial population; the population consists of a group of individuals, and the individuals consist of chromosomes and fitness values; wherein the chromosomes are encoded as follows: the length of each chromosome is defined as the total task arrival rate, and the tasks are represented by genes; the chromosome encoding method is the mapping of tasks to data centers TasktoDC; in the TasktoDC mapping, each task is assigned to a data center that meets its privacy constraints.

[0082] Chromosome encoding method: In genetic algorithms, chromosome encoding is a key technology that is responsible for converting various possible solutions to the problem to be solved into a form similar to biological chromosomes. This encoding method enables genetic algorithms to perform various operations and optimizations on these solutions just like processing chromosomes in biological evolution. Common chromosome encoding methods include: binary encoding, integer encoding, and real number encoding. Due to the characteristics of geographically distributed data centers, traditional chromosome encoding methods cannot be directly applied to workload management problems. Therefore, by considering all the characteristics of geographically distributed environments, the present invention proposes a new chromosome encoding method. The length of each chromosome is defined as the total task arrival rate, and tasks are represented by genes. Since tasks have different privacy levels, the corresponding privacy constraints need to be met when assigning tasks. Therefore, the chromosome encoding method can be regarded as a mapping of tasks to data centers (TasktoDC).

[0083] In the TasktoDC mapping, each task is assigned to a data center that satisfies its privacy constraints. The gene values ​​in the chromosome represent the serial numbers of the data centers to which the task can be assigned. Figure 4 As shown in the figure, tasks have different privacy levels and are placed in dashed boxes of different colors according to privacy levels of 1, 2, and 3. The different colors of tasks indicate that they are assigned to different data centers. Assume that the privacy levels of four data centers are 3, 2, 1, and 3, respectively. Since the privacy level of task n1 is 1, and there is only DC3 with a privacy level of 1, task n1 is assigned to DC3 with a privacy level of 1. For tasks with a privacy level of 3, all data centers meet their privacy conditions, so they can be assigned to any DC.

[0084] In a genetic algorithm, the fitness value is used to measure the degree of adaptability or the degree of excellence of an individual in the problem-solving space. During its evolution, each individual is assigned a fitness value, and the value is evaluated based on its problem-solving ability. An individual with a high fitness value indicates that it has a better problem-solving ability, while an individual with a low fitness value indicates that it has a poor problem-solving ability. Since the goal of this application is to obtain the best workload management strategy on the basis of satisfying privacy constraints to maximize system performance, the fitness value can be defined as the system performance value. In a genetic algorithm, by selecting individuals with better fitness values, the entire population can be driven to evolve in a better direction. The present invention aims to use a genetic algorithm to solve the workload management problem under constraints. In order to achieve this goal, the above-mentioned chromosome encoding method is adopted to flexibly represent the distribution relationship between tasks and data centers, and to find the best workload management strategy by simulating the population evolution process. Through the iterative evolution process of the genetic algorithm, the gene combination of the chromosome is continuously improved to obtain the optimal workload distribution solution.

[0085] S232. Generate an initial workload distribution plan based on the privacy constraint conditions. Under the initial plan, tasks are randomly distributed to a data center that meets its privacy conditions and does not exceed the upper limit of the data center's processing speed.

[0086] S24, determine whether the preset genetic generation is reached, if so, execute step S210, otherwise, execute step S25.

[0087] S25. Allocate the tasks contained in the chromosome of each individual in the population to a suitable data center; wherein the privacy constraints are satisfied during the allocation process and the upper limit of the processing speed of the data center is not exceeded.

[0088] in, Among them, h k is the number of servers in the kth data center, s k is the speed at which the servers in the kth data center execute tasks, is the average execution requirement for the task.

[0089] S26. Calculate the workload of each data center and evaluate individual fitness; wherein, the workload of each data center is calculated based on the number of tasks allocated to each data center; calculate the fitness value of the current individual based on the workload, and the fitness value is the system performance value under the current allocation scheme.

[0090] Since the chromosomes in the individual record the allocation method of each task, the workload of each data center can be calculated by counting the number of tasks allocated to each data center. The fitness value of the current individual is calculated based on the workload. The fitness value is the system performance value under the current allocation scheme. The system performance value is the value of the best performance of each data center.

[0091] The fitness value is the system performance value under the current allocation scheme. The system performance value is the value at which each data center performs best. The fitness value takes into account the following factors:

[0092] Task processing efficiency: Ensure that tasks can be processed and completed as quickly as possible, reducing waiting time and processing time. The shorter the response time, the higher the efficiency.

[0093] Data center resource utilization: Improve the utilization of data center resources and reduce resource idleness and waste.

[0094] Power efficiency: The lower the power consumption of a data center, the better. So power efficiency can be defined as the inverse of power consumption to meet the maximization goal.

[0095] The individual fitness value is defined as the weighted sum of the above three objective factors, and the specific weights can be adjusted according to actual needs.

[0096] S27. Select individuals with high fitness values ​​as parents.

[0097] In order to avoid individuals with better fitness being over-selected in the early stage of the algorithm, thereby reducing the diversity of the population and causing the algorithm to fall into a local optimal solution, the present invention selects individuals as parents based on the tournament method.

[0098] S28. Perform mutation operation on the selected individuals.

[0099] Specifically, in order to maintain the diversity of the population and explore a wider solution space, it is necessary to select a random value between [0,1) for each individual and determine whether it is less than the mutation probability. If it is less than, a mutation operation is performed, that is, a task in the chromosome is randomly selected and a new data center is selected for it. Otherwise, no mutation operation is performed.

[0100] S29, generate a new population; the new offspring will serve as a new population and go to step S24 to continue evolution.

[0101] S210. Select the individual with the highest fitness in the final population as output, thereby obtaining a workload management strategy.

[0102] Specifically, the workload management strategy is a workload value of each data center.

[0103] Figure 5 This is a workload optimization system for an ICV cloud platform distributed data center provided by an embodiment of the present invention. Figure 5 As shown, the workload optimization system for the ICV cloud platform distributed data center includes a data acquisition module and a workload management strategy optimization module;

[0104] The data acquisition module acquires the total task arrival rate, relevant configuration parameters of the geographically distributed data center, and privacy constraints:

[0105] The workload management strategy optimization module obtains a workload management strategy based on a privacy-aware workload optimization model according to the total task arrival rate, relevant configuration parameters of the geographically distributed data center and privacy constraints.

[0106] The above-mentioned workload optimization system for the ICV cloud platform distributed data center can be implemented in the form of a computer program, which can be run on a computer device.

[0107] The computer device includes a processor, a memory, and a network interface connected via a system bus, wherein the memory may include a non-volatile storage medium and an internal memory.

[0108] The non-volatile storage medium can store an operating system and a computer program. The computer program includes program instructions, and when the program instructions are executed, the processor can execute a workload optimization method for an ICV cloud platform distributed data center.

[0109] The processor is used to provide computing and control capabilities to support the operation of the entire computer equipment.

[0110] The internal memory provides an environment for the operation of the computer program in the non-volatile storage medium. When the computer program is executed by the processor, the processor can execute a workload optimization method for the ICV cloud platform distributed data center.

[0111] The network interface is used to communicate with other devices over the network. Those skilled in the art will appreciate that the above-mentioned computer device structure is only a partial structure related to the present application solution, and does not constitute a limitation on the computer device to which the present application solution is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0112] The processor is used to run a computer program stored in a memory, which implements the workload optimization method for an ICV cloud platform distributed data center as described in Example 1.

[0113] It should be understood that in the embodiments of the present application, the processor may be a central processing unit (CPU), and the processor may also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Among them, the general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0114] It can be understood by those skilled in the art that all or part of the processes in the method for implementing the above embodiment can be completed by instructing the relevant hardware through a computer program. The computer program includes program instructions, and the computer program can be stored in a storage medium, which is a computer-readable storage medium. The program instructions are executed by at least one processor in the computer system to implement the process steps of the embodiment of the above method.

[0115] The present invention also provides a storage medium. The storage medium may be a computer-readable storage medium. The storage medium stores a computer program, wherein when the computer program is executed by a processor, the processor executes a workload optimization method for an ICV cloud platform distributed data center as described in Embodiment 1.

[0116] The storage medium may be a USB flash drive, a mobile hard disk, a read-only memory (ROM), a magnetic disk, or an optical disk, or any other computer-readable storage medium that can store program codes.

[0117] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in terms of function in the above description. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0118] In the several embodiments provided by the present invention, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of each unit is only a logical function division, and there may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed.

[0119] The steps in the method of the embodiment of the present invention can be adjusted in order, combined and deleted according to actual needs. The units in the device of the embodiment of the present invention can be combined, divided and deleted according to actual needs. In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0120] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a storage medium. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, terminal, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention.

[0121] Note that the above are only preferred embodiments of the present invention and the technical principles used. Those skilled in the art will understand that the present invention is not limited to the specific embodiments described herein, and that various obvious changes, readjustments and substitutions can be made by those skilled in the art without departing from the scope of protection of the present invention. Therefore, although the present invention has been described in more detail through the above embodiments, the present invention is not limited to the above embodiments, and may include more other equivalent embodiments without departing from the concept of the present invention, and the scope of the present invention is determined by the scope of the appended claims.

Claims

1. A workload optimization method for an ICV cloud platform distributed data center, characterized in that: Includes steps: S1. Obtain the total task arrival rate, relevant configuration parameters of geographically distributed data centers, and privacy constraints; S2. According to the total task arrival rate, relevant configuration parameters of the geographically distributed data center and privacy constraints, a workload management strategy is obtained based on a privacy-aware workload optimization model; wherein the privacy-aware workload optimization model is constructed based on an M / M / m queuing model, a privacy model and a power consumption model; Step S2 includes: S21, input the total task arrival rate, relevant configuration parameters of geographically distributed data centers and privacy constraints; S22. Prepare candidate data centers for each type of task; prepare candidate data centers that meet the privacy requirements for each type of task according to the privacy constraints; S23, initializing the population and workload distribution plan; specifically including: S231. Determine the representation method and fitness function of the problem, and generate an initial population; the population is composed of a group of individuals, and the individuals are composed of chromosomes and fitness values; wherein the chromosomes are encoded as follows: the length of each chromosome is defined as the total task arrival rate, and the tasks are represented by genes; the chromosome encoding method is a mapping of tasks to data centers TasktoDC; in the TasktoDC mapping, each task is assigned to a data center that satisfies its privacy constraints; S232, generating an initial workload allocation scheme according to the privacy constraint condition, under which tasks are randomly allocated to a data center that meets its privacy condition and does not exceed the upper limit of the data center processing speed; S24, determine whether the preset genetic generation is reached, if so, execute step S210, otherwise, execute step S25; S25. Allocate the tasks contained in the chromosome of each individual in the population to a suitable data center; wherein the allocation process satisfies the privacy constraint and does not exceed the upper limit of the processing speed of the data center; S26, calculating the workload of each data center and evaluating the individual fitness; wherein the workload of each data center is calculated based on the number of tasks allocated to each data center; and the fitness value of the current individual is calculated based on the workload, and the fitness value is the system performance value under the current allocation scheme; S27, select individuals with high fitness values ​​as parents; S28, performing mutation operation on the selected individuals; S29, generate a new population; the new offspring will be used as a new population, and go to step S24 to continue evolution; S210. Select the individual with the highest fitness in the final population as output, thereby obtaining a workload management strategy.

2. The method according to claim 1, characterized in that In step S1, the total task arrival rate represents the total number of tasks arriving at all data centers per unit time in a geographically distributed environment.

3. The method according to claim 1, characterized in that In step S1, the relevant configuration parameters of the geographically distributed data center include the number of servers, task execution speed, utilization, static power consumption, customer satisfaction parameters, and service level agreement.

4. The method according to claim 1, characterized in that: In step S1, the privacy constraint condition is to assign the task to a data center with a privacy level not lower than the task for execution.

5. The method according to claim 1, characterized in that In step S2, the privacy-aware workload optimization model is constructed based on the M / M / m queuing model, the privacy model, and the power consumption model, including: modeling the data center based on the M / M / m queuing model and simulating the task arrival process.

6. The method according to claim 1, characterized in that The privacy model is a multi-level privacy model, which is established according to the privacy levels of tasks and data centers in various places. The multi-level privacy model consists of two parts: task privacy level and data center privacy level.

7. The method according to claim 1, characterized in that The power consumption model: The power consumption of a data center refers to the energy consumed by a data center during operation. In a geographically distributed data center, there are two different processing methods for the power consumption of a data center when it is idle, namely, an idle speed model and a constant speed model.

8. A workload optimization system for distributed data centers on an ICV cloud platform, characterized in that: The workload optimization system executes the workload optimization method for the ICV cloud platform distributed data center as claimed in claim 1, including: a data acquisition module, a workload management strategy optimization module; The data acquisition module acquires the total task arrival rate, relevant configuration parameters of the geographically distributed data center, and privacy constraints; The workload management strategy optimization module obtains a workload management strategy based on a privacy-aware workload optimization model according to the total task arrival rate, relevant configuration parameters of the geographically distributed data center and privacy constraints.

9. A computer device, characterized in that: The device comprises a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the method according to any one of claims 1 to 7 is implemented.

Images