Cloud platform task resource allocation method and system capable of maximizing resource utilization rate

By constructing a multi-user resource optimization model and adopting an optimization method, the problem of low resource utilization rate in the cloud computing platform when allocating multi-user resource, the maximum resource utilization rate and efficient configuration of system resources are achieved.

CN118964018APending Publication Date: 2024-11-15JIANGSU CUDATEC TECH CO LTD +1
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411029229.8
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-30
Publication Date
2024-11-15

AI Technical Summary

Technical Problem

The existing cloud computing platform has problems such as low resource utilization, large system resource consumption, and insufficient configuration accuracy and timeliness when allocating multi-user resources.

Method used

By constructing a multi-user resource optimization model, using resource value vectors and user task number vectors, single-resource saturation optimization, multi-resource saturation optimization and multi-resource undersaturation optimization methods, maximizing resource utilization and optimal resource task number, and optimizing Resource allocation.

Benefits of technology

It improves the resource allocation efficiency of the cloud platform, reduces system resource losses, and maximizes resource utilization.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118964018A_ABST
    Figure CN118964018A_ABST
Patent Text Reader

Abstract

The invention provides a cloud platform task resource allocation method and system capable of maximizing the resource utilization rate, and belongs to the field of cloud computing. The method comprises the following steps: acquiring user resource requirements, and constructing a multi-user resource optimization model; single-resource saturation optimization is adopted, and the resource value vector is utilized to obtain an optimal theoretical expression when the utilization rate of a single type of resources is 1; by adopting multi-resource saturation optimization and utilizing a complementary relaxation condition, obtaining an optimal theoretical expression when the normalized utilization rate of all types of resources is 1; multi-resource undersaturation optimization is adopted, and the resource demand vector is utilized to obtain an optimization theory expression under the condition that the resource utilization rates of all types are smaller than 1 and are equal; and comparing each optimization theory expression to obtain a maximum resource utilization rate and an optimal resource task number, and releasing resources according to the optimal resource task number. According to the invention, the resource configuration efficiency of the cloud platform can be effectively improved, and the resource loss of the system is reduced.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of cloud computing, and in particular to a cloud platform task resource allocation method and system for maximizing resource utilization. Background Art

[0002] With the continuous maturity and development of the era of big data and artificial intelligence, cloud computing has been widely used in various industries such as the Internet, industry, agriculture, ocean and education. One of the key core technologies for realizing the development of cloud computing is the optimal management of cloud computing resources.

[0003] Cloud computing uses a resource sharing computing model that is flexible and available anytime, anywhere and on demand. It allocates and optimizes resources through virtualization technology and provides storage and computing resources to support upper-layer applications. However, there are still some problems in implementing resource allocation on resource-constrained cloud computing platforms. In terms of resource control methods, the number of existing cloud platform users has increased dramatically, and resource demand has changed dynamically. The multi-user resource allocation scheme uses a fixed quota to limit the number of resources for each user, resulting in a large number of users occupying a high level of system resources for a long time, and their resource utilization is low. Obviously, this resource allocation scheme brings a lot of resource overhead and consumes a lot of storage resources. In terms of resource allocation methods, with the rapid growth of cloud platform applications, resources are becoming more diversified and complex. Existing methods basically use multi-objective optimization methods, using resource scheduling constraints such as time, cost, energy saving, revenue, service quality, and resource utilization to perform iterative optimization solutions, which also leads to the consumption of system resources. The accuracy and timeliness of configuration need to be optimized urgently.

[0004] Therefore, for cloud computing platforms under resource constraints, it is urgent to establish a multi-user multi-resource allocation method that maximizes resource utilization, so as to achieve coordinated scheduling of multi-dimensional resources in the cloud platform. Summary of the invention

[0005] The purpose of the present invention is to overcome the defects of the prior art and provide a cloud platform task resource allocation method and system for maximizing resource utilization.

[0006] In order to achieve the above objectives and other objectives, the present invention is implemented by including the following technical solutions: As a first aspect, the present invention proposes a cloud platform task resource allocation method for maximizing resource utilization, comprising the steps of:

[0007] S1. Obtain the total number of resource types M, the total number of users N and the normalized resource requirements of each user in the cloud platform, and construct a multi-user resource optimization model;

[0008] S2, determining whether different users have the same normalized resource demand for a certain resource, if so, proceeding to step S3; if not, proceeding to step S4;

[0009] S3, determine whether the resource utilization rate c is 1, if so, proceed to step S5, if not, proceed to step S6;

[0010] S4, adopting single resource saturation optimization and using resource value vector, we can obtain the optimal theoretical expression when the utilization rate of a single type of resource is 1;

[0011] S5, adopting multi-resource saturation optimization and using complementary relaxation conditions, we can obtain the optimal theoretical expression when the normalized utilization of all types of resources is 1;

[0012] S6, adopting multi-resource undersaturation optimization and using resource demand vectors, obtains the optimal theoretical expression when the utilization of all types of resources is less than 1 and equal;

[0013] S7. Compare the optimal theoretical expressions of steps S4, S5 and S6 to obtain the maximum resource utilization and the optimal number of resource tasks, and release resources according to the optimal number of resource tasks.

[0014] In one embodiment, the multi-user resource optimization model is

[0015]

[0016] x≥0,

[0017]

[0018] Among them, g j =[g j,1 ,g j,2 ,...,g j,N ] T ,1≤i≤N,1≤j≤M is the normalized resource requirement of all users’ single tasks for the jth resource; x=[x 1 ,x 2 ,…,x N ] T is the user task number vector, v = [v 1 ,v 2 ,…,v M ] T is the value vector of the resource; is the resource utilization constraint; is the normalized total resource constraint; x≥0 is a non-negative constraint; is the resource balance constraint between different users, where a i =[g 1,i ,g 2,i ,…,g N,i ] T and a l =[g1,l ,g 2,l ,…,g N,l ] T are the resource requirements of the ith and lth users respectively.

[0019] In one embodiment, in step S4, according to and x = γx 1 , and obtain the resource value vector And calculate resource utilization For j traversing from 1 to M, the maximum resource utilization is according to Obtain the user task number vector x=[x 1 ,x 2 ,…,x N ] T .

[0020] In one embodiment, if c j If there is an element greater than 1, then there is

[0021] In one embodiment, in step S5, according to and x = γx 1 , the solution is If and only if g j =g k ,1≤j,k≤M,j≠k, there is an extreme value, and the maximum resource utilization is

[0022] In one embodiment, in step S6, according to If and only if g j =g k ,1≤j,k≤M,j≠k, there is an extreme value, and the maximum resource utilization is

[0023] In one embodiment, in step S7, the maximum resource utilization is

[0024] As a second aspect, the present invention proposes a cloud platform task resource allocation system for maximizing resource utilization, the system is applied to the method described in any one of the first aspects, the system includes

[0025] User resource demand module: used to obtain the total number of resource types and users in the cloud platform, and the normalized resource demand of each user;

[0026] Single resource saturation optimization module: used to use the value vector to obtain the optimal theoretical expression when the utilization rate of only a single type of resource is 1;

[0027] Multi-resource saturation optimization module: used to use complementary relaxation conditions to obtain the optimal theoretical expression when the utilization of all types of resources is normalized to 1;

[0028] Multi-resource undersaturation optimization module: used to use the resource demand vector to obtain the optimal theoretical expression when the utilization of all types of resources is less than 1 and equal;

[0029] Resource task number allocation module: It is used to maximize resource utilization and ultimately obtain the optimal number of resource tasks when the cloud platform releases resources, so as to optimize the resource allocation performance of the cloud platform.

[0030] As a third aspect, the present invention proposes an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program implements any method described in the first aspect when loaded into the processor.

[0031] As a fourth aspect, the present invention proposes a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements any method described in the first aspect.

[0032] One or more technical solutions provided in the embodiments of the present invention have at least the following technical effects or advantages:

[0033] Under the constraint of total resources, the present invention optimizes the multi-dimensional resources of cloud computing by using the resource value vector and the user task number vector, obtains the optimal number of tasks for each user, maximizes the utilization of system resources, thereby effectively improving the resource allocation efficiency of the cloud platform and reducing the resource loss of the system. The present invention can be applied to many fields such as the Internet, big data applications, and manufacturing, and has broad application prospects. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 Shown is a flow chart of a cloud platform task resource allocation method for maximizing resource utilization according to the present invention.

[0035] Figure 2 Shown is a schematic diagram of multi-user cloud computing resource distribution in a specific embodiment of the present invention. DETAILED DESCRIPTION

[0036] The embodiments of the present invention provide a cloud platform task resource allocation method and system that maximizes resource utilization, thereby solving the problems of low resource utilization, large system resource consumption, low configuration accuracy and lack of timeliness when the cloud platform performs multi-user task and multi-resource allocation in the prior art.

[0037] Below, the technical solutions in the embodiments of the present invention will be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments of the present invention. It should be understood that the present invention is not limited to the example embodiments described herein. 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. It should also be noted that, for the convenience of description, only the parts related to the present invention are shown in the accompanying drawings, rather than all of them.

[0038] Embodiment 1

[0039] Please see attached Figure 1 The embodiment of the present invention provides a cloud platform task resource allocation method for maximizing resource utilization, the method comprising the steps of:

[0040] S1. Obtain the total number of resource types, the total number of users and the normalized resource requirements of each user in the cloud platform, and construct a multi-user resource optimization model;

[0041] Specifically, the steps of constructing the multi-user resource optimization model include:

[0042] (1) Assume that the total number of resource types in the cloud platform is M, the total number of users is N, and for the jth resource, the normalized resource requirement of all users’ single tasks is g j =[g j,1 ,g j,2 ,...,g j,N ] T ,1≤i≤N,1≤j≤M;

[0043] (2) Under the resource constraint condition, assuming that the resource utilization rate is c, the multi-user cloud platform resource allocation optimization model is established, which is expressed as:

[0044]

[0045] x≥0,

[0046]

[0047] Where x = [x 1 ,x 2 ,...,x N ] T The vector of user tasks assigned to the cloud platform, v = [v 1 ,v 2 ,…,v M ] T is the value vector of M resources; is the resource utilization constraint; is the normalized total resource constraint; x≥0 is a non-negative constraint; is the resource balance constraint between different users, where a i =[g 1,i ,g 2,i ,…,g N,i ] T and a l =[g 1,l ,g 2,l ,…,g N,l ] T are the resource requirements of the ith and lth users respectively.

[0048] S2, determining whether different users have the same normalized resource demand for a certain resource, if so, proceeding to step S3; if not, proceeding to step S4;

[0049] S3, determine whether the resource utilization rate c is 1, if so, proceed to step S5, if not, proceed to step S6;

[0050] S4, adopting single resource saturation optimization and using resource value vector, we can obtain the optimal theoretical expression when the utilization rate of a single type of resource is 1;

[0051] Specifically, the implementation process of step S4 is as follows:

[0052] Only the utilization rate of the jth resource is 1, and the utilization rates of the remaining M-1 resources are c<1, which can be expressed as

[0053] According to resource balancing constraints between different users x 1 As the benchmark, the user task number vector x can be expressed as

[0054] Among them, γ is the resource value vector, expressed as

[0055] use Then the task number vector x can be expressed as

[0056] The resource utilization can be converted into

[0057] Among them, if c j If there is an element greater than 1, then there is

[0058] So it can be expressed as

[0059] For j, traversing each case from 1 to M, the maximum resource utilization is:

[0060] S5, adopting multi-resource saturation optimization and using complementary relaxation conditions, we can obtain the optimal theoretical expression when the normalized utilization of all types of resources is 1;

[0061] Specifically, the implementation process of step S5 is as follows:

[0062] Using the relaxation condition, when the utilization of all types of resources is 1, it is expressed as

[0063] Using x = γx 1 , so the number of resource tasks for the jth resource can be expressed as

[0064] It can be solved

[0065] So, if and only if g j =g k ,1≤j,k≤M,j≠k, there is an extreme value, and the resource utilization is expressed as

[0066] S6, adopting multi-resource undersaturation optimization and using resource demand vectors, obtains the optimal theoretical expression when the utilization of all types of resources is less than 1 and equal;

[0067] Specifically, the implementation process of step S6 is as follows:

[0068] The utilization of all resources is c, that is:

[0069] If and only if g j =g k ,1≤j,k≤M,j≠k, there is an extreme value, and the resource utilization is expressed as

[0070] S7. Compare the optimal theoretical expressions of steps S4, S5 and S6 to obtain the maximum resource utilization and the optimal number of resource tasks, and release resources according to the optimal number of resource tasks.

[0071] Specifically, the comparison to maximize resource utilization can be expressed as: The cloud platform releases resources and assigns the optimal resource task number x * Allocate to each user to maximize resource utilization.

[0072] Please combine Figure 2 , below, taking the total number of resource types in the cloud platform as 3 and the total number of users as 3 as an example, the implementation process of the multi-user cloud platform resource allocation optimization model is described in detail.

[0073] Assume that the normalized resource requirement of each user's single task is g 1 =[1,0.5,0.5] T , g 2 =[1,0.5,1] T , g 3 =[0.5,1,0.3] T , the normalized value vector of the three resources is v = [0.5, 1, 0.3] T / 1.8; due to g j ≠g k , so the step S4 is executed, but the steps S5 and S6 are not executed.

[0074] According to the resource balancing constraints of the multi-user resource optimization model User task number vector but

[0075] For j, go through each case from 1 to 3:

[0076] When j=1, but

[0077] When j = 2, At this time, there is an element greater than 1, then

[0078] When j=3, At this time, there is an element greater than 1, then

[0079] At this point, we can see that the maximum resource utilization is

[0080] use It can be solved The task number vector x can be expressed as

[0081] Furthermore, the maximum resource utilization can be expressed as c * =0.70;

[0082] In summary, the optimal number of resource tasks can be expressed as x * =[0.43, 0.38, 0.37] T ;

[0083] Then, the cloud platform releases resources according to the optimal number of resource tasks and allocates tasks to each user.

[0084] Embodiment 2

[0085] Based on the same inventive concept as the cloud platform task resource allocation method for maximizing resource utilization in the aforementioned embodiment, the present invention also provides a cloud platform task resource allocation system for maximizing resource utilization, including:

[0086] User resource demand module: used to obtain the total number of resource types and users in the cloud platform, and the normalized resource demand of each user;

[0087] Single resource saturation optimization module: used to use the value vector to obtain the optimal theoretical expression when the utilization rate of only a single type of resource is 1;

[0088] Multi-resource saturation optimization module: used to use complementary relaxation conditions to obtain the optimal theoretical expression when the utilization of all types of resources is normalized to 1;

[0089] Multi-resource undersaturation optimization module: used to use the resource demand vector to obtain the optimal theoretical expression when the utilization of all types of resources is less than 1 and equal;

[0090] Resource task number allocation module: It is used to maximize resource utilization and ultimately obtain the optimal number of resource tasks when the cloud platform releases resources, so as to optimize the resource allocation performance of the cloud platform.

[0091] The foregoing Figure 1 The various variations and specific examples of a cloud platform task resource allocation method for maximizing resource utilization in Example 1 are also applicable to a cloud platform task resource allocation system for maximizing resource utilization in this embodiment. Through the above detailed description of a cloud platform task resource allocation method for maximizing resource utilization, technical personnel in this field can clearly know the implementation method of a cloud platform task resource allocation system for maximizing resource utilization in this embodiment, so for the sake of brevity of the specification, it will not be described in detail here.

[0092] Embodiment 3

[0093] This embodiment provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program implements the method of the first embodiment when loaded into the processor.

[0094] Embodiment 4

[0095] This embodiment provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the method of the first embodiment is implemented.

[0096] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention belong to the scope of the same technology of the present invention, the present invention is also intended to include these modifications and variations.

Claims

1. A cloud platform task resource allocation method to maximize resource utilization, It is characterized in that Includes steps: S1. Obtain the total number of resource types M, the total number of users N and the normalized resource requirements of each user in the cloud platform, and construct a multi-user resource optimization model; S2, determining whether different users have the same normalized resource demand for a certain resource, if so, proceeding to step S3; if not, proceeding to step S4; S3, determine whether the resource utilization rate c is 1, if so, proceed to step S5, if not, proceed to step S6; S4, adopting single resource saturation optimization and using resource value vector, we can obtain the optimal theoretical expression when the utilization rate of a single type of resource is 1; S5, adopting multi-resource saturation optimization and using complementary relaxation conditions, we can obtain the optimal theoretical expression when the normalized utilization of all types of resources is 1; S6, adopting multi-resource undersaturation optimization and using resource demand vectors, obtains the optimal theoretical expression when the utilization of all types of resources is less than 1 and equal; S7. Compare the optimal theoretical expressions of steps S4, S5 and S6 to obtain the maximum resource utilization and the optimal number of resource tasks, and release resources according to the optimal number of resource tasks.

2. The method according to claim 1, It is characterized in that The multi-user resource optimization model is: Among them, g j =[g j,1 ,g j,2 ,…,g j,N ] T ,1≤i≤N,1≤j≤M is the normalized resource requirement of all users’ single tasks for the jth resource; x=[x 1 ,x 2 ,…,x N ] T is the user task number vector, v = [v 1 ,v 2 ,…,v M ] T is the value vector of the resource; is the resource utilization constraint; is the normalized total resource constraint; x≥0 is a non-negative constraint; is the resource balance constraint between different users, where a i =[g 1,i ,g 2,i ,…,g N,i ] T and a l =[g 1,l ,g 2,l ,…,g N,l ] T are the resource requirements of the ith and lth users respectively.

3. The method according to claim 2, It is characterized in that In step S4, according to and x = γx 1 , and obtain the resource value vector And calculate resource utilization 1≤k≤M,k≠j; for j traversing from 1 to M, the maximum resource utilization is according to Obtain the user task number vector x=[x 1 ,x 2 ,...,x N ] T .

4. The method according to claim 3, It is characterized in that If c j If there is an element greater than 1, then there is 5. The method according to claim 3, It is characterized in that In step S5, according to and x = γx 1 , the solution is 1≤j≤M; if and only if g j =g k ,1≤j,k≤M,j≠k, there is an extreme value, and the maximum resource utilization is 6. The method according to claim 2, It is characterized in that In step S6, according to If and only if g j =g k ,1≤j,k≤M,j≠k, there is an extreme value, and the maximum resource utilization is 7. The method according to claim 6, It is characterized in that In step S7, the maximum resource utilization is 8. A cloud platform task resource allocation system that maximizes resource utilization, It is characterized in that The system is applied to the method according to any one of claims 1 to 7, and the system comprises User resource demand module: used to obtain the total number of resource types and users in the cloud platform, and the normalized resource demand of each user; Single resource saturation optimization module: used to use the value vector to obtain the optimal theoretical expression when the utilization rate of only a single type of resource is 1; Multi-resource saturation optimization module: used to use complementary relaxation conditions to obtain the optimal theoretical expression when the utilization of all types of resources is normalized to 1; Multi-resource undersaturation optimization module: used to use the resource demand vector to obtain the optimal theoretical expression when the utilization of all types of resources is less than 1 and equal; Resource task number allocation module: It is used to maximize resource utilization and ultimately obtain the optimal number of resource tasks when the cloud platform releases resources, so as to optimize the resource allocation performance of the cloud platform.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, It is characterized in that When the computer program is loaded into the processor, the method according to any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium having a computer program stored thereon, It is characterized in that When the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.