WO2023019735A1 - 虚拟机的部署方法、装置、设备以及计算机可读存储介质 - Google Patents

虚拟机的部署方法、装置、设备以及计算机可读存储介质 Download PDF

Info

Publication number
WO2023019735A1
WO2023019735A1 PCT/CN2021/127500 CN2021127500W WO2023019735A1 WO 2023019735 A1 WO2023019735 A1 WO 2023019735A1 CN 2021127500 W CN2021127500 W CN 2021127500W WO 2023019735 A1 WO2023019735 A1 WO 2023019735A1
Authority
WO
WIPO (PCT)
Prior art keywords
antibody
affinity
virtual machine
mapping relationship
antigen
Prior art date
Application number
PCT/CN2021/127500
Other languages
English (en)
French (fr)
Inventor
郭城阳
闫冬冬
Original Assignee
苏州浪潮智能科技有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 苏州浪潮智能科技有限公司 filed Critical 苏州浪潮智能科技有限公司
Publication of WO2023019735A1 publication Critical patent/WO2023019735A1/zh

Links

Images

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00Arrangements for program control, e.g. control units
    • G06F9/06Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/44Arrangements for executing specific programs
    • G06F9/455Emulation; Interpretation; Software simulation, e.g. virtualisation or emulation of application or operating system execution engines
    • G06F9/45533Hypervisors; Virtual machine monitors
    • G06F9/45558Hypervisor-specific management and integration aspects
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06NCOMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00Computing arrangements based on biological models
    • G06N3/002Biomolecular computers, i.e. using biomolecules, proteins, cells
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00Arrangements for program control, e.g. control units
    • G06F9/06Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/44Arrangements for executing specific programs
    • G06F9/455Emulation; Interpretation; Software simulation, e.g. virtualisation or emulation of application or operating system execution engines
    • G06F9/45533Hypervisors; Virtual machine monitors
    • G06F9/45558Hypervisor-specific management and integration aspects
    • G06F2009/4557Distribution of virtual machine instances; Migration and load balancing

Definitions

  • the present application relates to the technical field of cloud computing platforms, and in particular to a method for deploying a virtual machine; it also relates to a device, equipment, and computer-readable storage medium for deploying a virtual machine.
  • Cloud data centers widely use virtualization technology to allocate the resources of a single physical host to multiple virtual machines, thereby improving resource utilization and reducing operating and maintenance costs. How to deploy virtual machines in a reasonable, balanced, and batch manner has become an important task in the operation and maintenance of cloud computing platforms.
  • the methods for determining the best allocation strategy for deploying virtual machines in batches on physical hosts are mainly judgment methods based on resource thresholds and some heuristic intelligent algorithms. Although these methods can determine one or more allocation strategies for deploying virtual machines in batches to physical hosts, they cannot guarantee the reasonable utilization and load balancing of resources in the cloud computing platform.
  • the purpose of this application is to provide a virtual machine deployment method, which can realize rapid batch deployment of virtual machines and achieve optimal load balancing among physical hosts.
  • Another object of the present application is to provide a virtual machine deployment device, device, and computer-readable storage medium, all of which have the above-mentioned technical effects.
  • the present application provides a virtual machine deployment method, including:
  • Evolving the antibodies includes:
  • the antibody is evolved according to the adjusted crossover rate and the mutation rate.
  • setting the mapping relationship between the virtual machine and the physical host includes:
  • mapping relationship between the virtual machine and the physical host is set, the number of the mapping relationships is a preset number, and the mapping relationship satisfies the fact that the physical host under the mapping relationship does not have resource over-allocation.
  • the calculating the affinity of the antibody to the antigen comprises:
  • the affinity of the antibody to the antigen is obtained by taking the reciprocal of the Euclidean distance.
  • the target resource includes: CPU, memory, and hard disk.
  • the constraint condition is that the usage rate of the target resource of the physical host is within a preset range.
  • immunizing the antibody includes:
  • the antibody is replaced with the new antibody according to the probability.
  • the obtaining the probability of replacing the antibody with the new antibody according to the affinity difference and the annealing factor includes:
  • Tg represents the annealing factor
  • P indicates the probability that the new antibody replaces the antibody.
  • the present application also provides a virtual machine deployment device, including:
  • the first setting module is used to set the affinity function and the constraint condition, and use the affinity function and the constraint condition as the antigen;
  • the second setting module is used to set the mapping relationship between the virtual machine and the physical host, and use the mapping relationship as an antibody;
  • a calculation module used to calculate the affinity between the antibody and the antigen
  • a selection module used to select the antibody with the highest affinity from the antibodies
  • the evolution and immunization module is used to evolve the antibody with the highest affinity and immunize the antibody with the highest affinity until the affinity between the evolved antibody and the antigen converges;
  • a deployment module configured to deploy the virtual machine according to the mapping relationship corresponding to the antibody when the affinity converges
  • Evolving the antibodies includes:
  • the antibody is evolved according to the adjusted crossover rate and the mutation rate.
  • the present application also provides a virtual machine deployment device, including:
  • a processor configured to implement the steps of the method for deploying a virtual machine as described in any one of the preceding items when executing the computer program.
  • the present application also provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the virtual The steps of the deployment method of the machine.
  • the virtual machine deployment method provided in this application includes: setting an affinity function and a constraint condition, and using the affinity function and the constraint condition as an antigen; setting a mapping relationship between a virtual machine and a physical host, and using The mapping relationship is used as an antibody; calculate the affinity between the antibody and the antigen; select the antibody with the highest affinity from the antibodies; evolve the antibody with the highest affinity and compare the affinity The highest antibody is vaccinated until the affinity between the evolved antibody and the antigen converges; the virtual machine is deployed according to the mapping relationship corresponding to the antibody when the affinity converges;
  • the antibody includes: adjusting the crossover rate and the mutation rate according to the preset fuzzy subset; and evolving the antibody according to the adjusted crossover rate and the mutation rate.
  • the virtual machine deployment method provided by this application is based on the artificial immune algorithm, abstracts the problem of deploying virtual machines in batches as an optimization problem of determining the mapping relationship between virtual machines and physical hosts, and takes the physical hosts to achieve the most load balance as the optimization problem.
  • the goal is to obtain the optimal solution of the mapping relationship by setting and changing the mapping relationship between the virtual machine and the physical host, that is, to obtain the best deployment method of the virtual machine.
  • Applying this deployment method can achieve the best load balance between physical hosts while realizing rapid batch deployment of virtual machines, and ensure that after the deployment is completed, the cloud computing platform runs with better load balance and better resource utilization.
  • this deployment method introduces fuzzy reasoning theory into the artificial immune process, and realizes the intelligent and adaptive determination of the crossover rate and mutation rate of the immune process.
  • the virtual machine deployment device, device, and computer-readable storage medium provided in the present application all have the above-mentioned technical effects.
  • FIG. 1 is a schematic flowchart of a method for deploying a virtual machine provided in an embodiment of the present application
  • FIG. 2 is a schematic diagram of a membership function provided by the embodiment of the present application.
  • FIG. 3 is a schematic diagram of another membership function provided by the embodiment of the present application.
  • FIG. 4 is a schematic diagram of a virtual machine deployment device provided by an embodiment of the present application.
  • FIG. 5 is a schematic diagram of a virtual machine deployment device provided by an embodiment of the present application.
  • the core of the present application is to provide a virtual machine deployment method, which can realize rapid batch deployment of virtual machines and achieve optimal load balance among physical hosts.
  • Another core of the present application is to provide a virtual machine deployment device, device, and computer-readable storage medium, all of which have the above-mentioned technical effects.
  • FIG. 1 is a schematic flowchart of a virtual machine deployment method provided in the embodiment of the present application. Referring to FIG. 1, the method mainly includes:
  • the present application abstracts the problem of deploying virtual machines in batches into the problem of determining the mapping relationship between virtual machines and physical hosts, so antigens and antibodies are set first.
  • Antigens include affinity functions and constraints.
  • the antibody is the mapping relationship between the virtual machine and the physical host.
  • the optimal mapping relationship is determined by performing subsequent steps.
  • the so-called optimal mapping relationship means that when the physical host deploys the virtual machine according to the mapping relationship, the best load balancing among the physical hosts can be achieved.
  • an antibody represents a virtual machine deployment scheme, that is, a mapping relationship between a virtual machine and a physical host. Assuming that a cloud computing platform initially has M physical hosts, and N virtual machines need to be deployed on these physical hosts, then the antibody obtained by using the above coding method is:
  • v N indicates that the Nth virtual machine is mapped on the vNth physical host.
  • the way of setting the mapping relationship between the virtual machine and the physical host is: setting the mapping relationship between the virtual machine and the physical host, the number of the mapping relationship is a preset number, and the mapping The relationship satisfies that the physical host under the mapping relationship does not have resource over-allocation.
  • the Random function in the Java language can be used to sequentially select integers in [1, M] for each position of the antibody to obtain the initial antibody group, that is, to obtain some initial mapping relationships. Then determine whether the physical hosts under these mapping relationships have resource over-allocation, and delete the mapping relationship with resource over-allocation, that is, eliminate invalid antibodies from the antibody group.
  • mapping relationship between the virtual machine and the physical host finally set in this embodiment is A more appropriate quantity is the preset quantity.
  • the preset number may be 30.
  • the present application abstracts the deployment problem of a virtual machine as an optimal solution problem with constraints. Determine the affinity function and constraint conditions in advance, and on the basis of setting the mapping relationship between the virtual machine and the physical host, under the constraint conditions, calculate the affinity of each antibody according to the affinity function, that is, get the The load balance value of the physical host.
  • the load balance value reflects the load balance between physical hosts under a certain mapping relationship.
  • the calculating the affinity of the antibody to the antigen comprises:
  • the affinity of the antibody to the antigen is obtained by taking the reciprocal of the Euclidean distance.
  • the target resource includes: CPU, memory and hard disk.
  • this embodiment focuses on the CPU, memory, and hard disk of the physical hosts.
  • the load is defined as follows:
  • s ⁇ 1, 2, 3 are used to represent the CPU, memory, and hard disk of the physical host M, respectively.
  • Us represents the usage rate of the s resource. For example, when s takes a value of 1, U 1 represents the usage rate of the CPU of the physical host M.
  • the utilization rate of each resource of each physical host under the mapping relationship can be obtained correspondingly, and then the load of each physical host under the mapping relationship can be obtained according to the above formula .
  • L(M) represents the load of the physical host M.
  • the load balancing between physical hosts is measured using the Euclidean distance, as shown in the following formula:
  • the affinity function is:
  • the load balance value of the physical host under the mapping relationship can be obtained, that is, the antibody Affinity.
  • the constraint condition may be that the usage rate of the target resource of the physical host is within a preset range.
  • the preset interval is [0,0.9]
  • the constraint condition is Us ⁇ [0,0.9].
  • S104 selecting the antibody with the highest affinity from the antibodies
  • the strategy of combining elite selection and roulette selection can be used to select the antibody with the highest affinity and store it in the memory cell library.
  • S106 Deploy a virtual machine according to the mapping relationship corresponding to the antibody when the affinity converges.
  • the evolution of antibodies refers to the transformation of the mapping relationship, and after each evolution and immunization, the affinity of the antibody is calculated until the affinity of the antibody converges, so far the corresponding mapping relationship of the antibody when the affinity converges
  • the final mapping relationship is encapsulated as parameters in Json format, and then the script for automatically deploying virtual machines uses the parameters in Json format to distribute tasks to each physical host, so that each physical host can create and deploy the virtual machine in turn. The allocated virtual machines, and finally quickly complete the batch deployment of virtual machines.
  • the way of evolving antibodies is:
  • the antibody is evolved based on the adjusted crossover rate and the change rate.
  • this embodiment uses an adaptive method to determine the crossover rate and mutation rate in the immune process.
  • this embodiment introduces the fuzzy reasoning theory into the artificial immune process to realize intelligent and adaptive determination of the crossover rate and mutation rate of the immune process.
  • L avg (g) is used to represent the average affinity between the g-generation antibody and the antigen
  • the difference in the average affinity between two consecutive generations of antibodies can represent the evolutionary effect of the antibody, namely:
  • the antibody evolves towards a stronger immune effect from generation to generation. At this time, it is appropriate to increase the crossover rate and mutation rate appropriately. If It means that the antibody degenerates from generation to generation, which may be caused by excessive crossover rate and mutation rate. At this time, consider appropriately reducing the crossover rate and mutation rate.
  • This embodiment uses NB (negative big), NS (negative small), ZE (zero), PS (positive small), PB (positive big), can be in Five fuzzy subsets are defined on the domain of discourse, and their membership functions are shown in Figure 2.
  • the way of inoculating and immunizing the antibody is:
  • the antibody is replaced with the new antibody according to the probability.
  • this example utilizes simulated annealing to optimize the vaccination process.
  • the best antibody is selected from the library, and the antibody is immunized.
  • the so-called immunization is to make a small-scale position exchange on the chain of antibody A, then a new antibody N is generated, and the affinity (ie, load balance value) of the two antibodies is calculated as L(A), respectively.
  • L(N) the difference between the two is Then, according to the Metropolis criterion, the probability P of antibody N replacing antibody A is determined, as shown in the following formula:
  • Tg represents the annealing factor, which is a preset value.
  • Introducing fuzzy logic into the immune process can realize dynamic adjustment of crossover rate and mutation rate.
  • the introduction of simulated annealing and dynamic adjustment of the crossover rate and mutation rate can not only make the entire immune process converge quickly, but also ensure that the optimal antibody is finally obtained, that is, the optimal deployment plan can be obtained, which can effectively improve efficiency and performance.
  • the affinity of the final antibody converges, that is, the final load balance value of the physical host converges, indicating that even if the mapping relationship is changed, the load balance value of the physical host will not have a big difference.
  • the mapping The relationship is optimal.
  • the antibody in the memory unit library at the end of immunization is the optimal antibody, and the final mapping relationship can be obtained by backcoding this antibody.
  • the virtual machine deployment method provided in this application is based on the artificial immune algorithm, abstracting the problem of deploying virtual machines in batches into an optimization problem of determining the mapping relationship between virtual machines and physical hosts, and using physical hosts to achieve the most balanced load
  • the optimal solution of the mapping relationship is obtained, that is, the best deployment mode of the virtual machine is obtained.
  • Applying this deployment method can achieve the best load balance between physical hosts while realizing rapid batch deployment of virtual machines, and ensure that after the deployment is completed, the cloud computing platform runs with better load balance and better resource utilization.
  • FIG. 4 is a schematic diagram of a virtual machine deployment device provided in the embodiment of the present application.
  • the device includes:
  • the first setting module 10 is used to set an affinity function and a constraint condition, and use the affinity function and the constraint condition as an antigen;
  • the second setting module 20 is configured to set the mapping relationship between the virtual machine and the physical host, and use the mapping relationship as an antibody;
  • Calculation module 30 used to calculate the affinity between the antibody and the antigen
  • a selection module 40 configured to select the antibody with the highest affinity from the antibodies
  • the evolution and immunization module 50 is used to evolve the antibody with the highest affinity and immunize the antibody with the highest affinity until the affinity between the evolved antibody and the antigen converges;
  • the deployment module 60 is configured to deploy the virtual machine according to the mapping relationship corresponding to the antibody when the affinity converges.
  • the second setting module 20 is specifically used for:
  • mapping relationship between the virtual machine and the physical host is set, the number of the mapping relationships is a preset number, and the mapping relationship satisfies the fact that the physical host under the mapping relationship does not have resource over-allocation.
  • calculation module 30 is specifically used for:
  • the affinity of the antibody to the antigen is obtained by taking the reciprocal of the Euclidean distance.
  • the target resource includes: CPU, memory, and hard disk.
  • the constraint condition is that the usage rate of the target resource of the physical host is within a preset range.
  • the evolution and immunity module 50 is specifically used for:
  • the antibody is evolved based on the adjusted crossover rate and the change rate.
  • the evolution and immunity module 50 is specifically used for:
  • the antibody is replaced with the new antibody according to the probability.
  • the virtual machine deployment device provided in this application is based on the artificial immune algorithm, abstracts the problem of deploying virtual machines in batches into an optimization problem of determining the mapping relationship between virtual machines and physical hosts, and takes the physical hosts to achieve the most balanced load as the optimization goal. By setting and changing the mapping relationship between the virtual machine and the physical host, the optimal solution of the mapping relationship is obtained, that is, the best deployment mode of the virtual machine is obtained. Applying this deployment method can not only achieve rapid batch deployment of virtual machines, but also achieve the best load balance among physical hosts, and ensure that after the deployment is completed, the cloud computing platform will run with better load balance and better resource utilization.
  • the present application also provides a virtual machine deployment device, as shown in FIG. 5 , the device includes a memory 1 and a processor 2 .
  • Memory 1 used to store computer programs
  • Processor 2 configured to execute a computer program to implement the following steps:
  • the present application also provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the following steps can be realized:
  • the computer-readable storage medium may include: U disk, mobile hard disk, read-only memory (Read-Only Memory, ROM), random access memory (Random Access Memory, RAM), magnetic disk or optical disk, etc., which can store program codes. medium.
  • each embodiment in the description is described in a progressive manner, each embodiment focuses on the difference from other embodiments, and the same and similar parts of each embodiment can be referred to each other.
  • the description is relatively simple, and for related details, please refer to the description of the method part.
  • RAM random access memory
  • ROM read-only memory
  • EEPROM electrically programmable ROM
  • EEPROM electrically erasable programmable ROM
  • registers hard disk, removable disk, CD-ROM, or any other Any other known storage medium.

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Software Systems (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Biophysics (AREA)
  • General Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • Computational Linguistics (AREA)
  • Computing Systems (AREA)
  • Artificial Intelligence (AREA)
  • Data Mining & Analysis (AREA)
  • Evolutionary Computation (AREA)
  • General Health & Medical Sciences (AREA)
  • Molecular Biology (AREA)
  • Biomedical Technology (AREA)
  • Health & Medical Sciences (AREA)
  • Spectroscopy & Molecular Physics (AREA)
  • Mathematical Physics (AREA)
  • Organic Chemistry (AREA)
  • Chemical & Material Sciences (AREA)
  • Peptides Or Proteins (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

一种虚拟机的部署方法、装置、设备以及计算机可读存储介质,其中方法包括:设置亲和度函数与约束条件,并以所述亲和度函数与所述约束条件作为抗原(S101);设置虚拟机与物理主机的映射关系,并以所述映射关系作为抗体(S102);计算所述抗体与所述抗原的亲和度(S103);从所述抗体中挑选出亲和度最高的所述抗体(S104);进化亲和度最高的所述抗体并对亲和度最高的所述抗体进行接种免疫,直到进化后的所述抗体与所述抗原的亲和度收敛(S105);根据亲和度收敛时的所述抗体对应的所述映射关系部署所述虚拟机(S106)。应用该部署方法能够实现快速批量部署虚拟机,并使物理主机间达到最佳负载均衡。

Description

虚拟机的部署方法、装置、设备以及计算机可读存储介质
本申请要求在2021年08月20日提交中国专利局、申请号为202110957630.8、发明名称为“虚拟机的部署方法、装置、设备以及计算机可读存储介质”的中国专利申请的优先权,其全部内容通过引用结合在本申请中。
技术领域
本申请涉及云计算平台技术领域,特别涉及一种虚拟机的部署方法;还涉及一种虚拟机的部署装置、设备以及计算机可读存储介质。
背景技术
云数据中心广泛使用虚拟化技术实现将单一物理主机的资源分配给多个虚拟机,从而提高资源利用率,同时减少运营和维护成本。如何合理、均衡、批量部署虚拟机已成为云计算平台运营与维护过程中的重要任务。当前,确定在物理主机上批量部署虚拟机的最佳分配策略的方法主要是基于资源阈值的判断法和部分启发式智能算法。虽然,这些方法能够确定出一种或多种批量部署虚拟机到物理主机的分配策略,但是却无法保证云计算平台中资源的合理利用与负载均衡。
因此,提供一种能够使物理主机间达到最佳负载均衡的虚拟机批量部署方案已成为本领域技术人员亟待解决的技术问题。
发明内容
本申请的目的是提供一种虚拟机的部署方法,能够实现快速批量部署虚拟机,并使物理主机间达到最佳负载均衡。本申请的另一个目的是提供一种虚拟机的部署装置、设备以及计算机可读存储介质,均具有上述技术效果。
为解决上述技术问题,本申请提供了一种虚拟机的部署方法,包括:
设置亲和度函数与约束条件,并以所述亲和度函数与所述约束条件作为抗原;
设置虚拟机与物理主机的映射关系,并以所述映射关系作为抗体;
计算所述抗体与所述抗原的亲和度;
从所述抗体中挑选出亲和度最高的所述抗体;
进化亲和度最高的所述抗体并对亲和度最高的所述抗体进行接种免疫,直到进化后的所述抗体与所述抗原的亲和度收敛;
根据亲和度收敛时的所述抗体对应的所述映射关系部署所述虚拟机;
进化所述抗体包括:
根据预设模糊子集调整交叉率与变异率;
根据调整后的所述交叉率与所述变异率,进化所述抗体。
可选的,设置虚拟机与物理主机的映射关系包括:
设置虚拟机与所述物理主机的映射关系,所述映射关系的个数为预设数量,且所述映射关系满足在所述映射关系下的所述物理主机不存在资源超配的情况。
可选的,所述计算所述抗体与所述抗原的亲和度包括:
计算所述抗体对应的所述映射关系下的所述物理主机的目标资源的使用率;
根据所述目标资源的使用率得到所述物理主机的负载;
计算各所述物理主机的负载的均值,并计算所述物理主机的负载与所述均值的欧式距离;
将所述欧式距离取倒数得到所述抗体与所述抗原的亲和度。
可选的,所述目标资源包括:CPU、内存以及硬盘。
可选的,所述约束条件为所述物理主机的所述目标资源的使用率位于预设区间。
可选的,对所述抗体进行接种免疫包括:
变异所述抗体得到新的抗体,并计算所述新的抗体与所述抗体的亲和度的差值;
根据所述亲和度差值以及退火因子,得到所述新的抗体替换所述抗体的概率;
根据所述概率将所述抗体替换为所述新的抗体。
可选的,所述根据所述亲和度差值以及退火因子,得到所述新的抗体替换所述抗体的概率包括:
依据公式
Figure PCTCN2021127500-appb-000001
得到所述新的抗体替换所述抗体的概率;
其中,Tg表示退火因子,
Figure PCTCN2021127500-appb-000002
表示亲和度差值,P表示所述新的抗体替换所述抗体的概率。
为解决上述技术问题,本申请还提供了一种虚拟机的部署装置,包括:
第一设置模块,用于设置亲和度函数与约束条件,并以所述亲和度函数与所述约束条件作为抗原;
第二设置模块,用于设置虚拟机与物理主机的映射关系,并以所述映射关系作为抗体;
计算模块,用于计算所述抗体与所述抗原的亲和度;
挑选模块,用于从所述抗体中挑选出亲和度最高的所述抗体;
进化与免疫模块,用于进化亲和度最高的所述抗体并对亲和度最高的所述抗体进行接种免疫,直到进化后的所述抗体与所述抗原的亲和度收敛;
部署模块,用于根据亲和度收敛时的所述抗体对应的所述映射关系部署所述虚拟机;
进化所述抗体包括:
根据预设模糊子集调整交叉率与变异率;
根据调整后的所述交叉率与所述变异率,进化所述抗体。
为解决上述技术问题,本申请还提供了一种虚拟机的部署设备,包括:
存储器,用于存储计算机程序;
处理器,用于执行所述计算机程序时实现如上任一项所述的虚拟机的部署方法的步骤。
为解决上述技术问题,本申请还提供了一种计算机可读存储介质,所述 计算机可读存储介质上存储有计算机程序,所述计算机程序被处理器执行时实现如上任一项所述的虚拟机的部署方法的步骤。
本申请所提供的虚拟机的部署方法,包括:设置亲和度函数与约束条件,并以所述亲和度函数与所述约束条件作为抗原;设置虚拟机与物理主机的映射关系,并以所述映射关系作为抗体;计算所述抗体与所述抗原的亲和度;从所述抗体中挑选出亲和度最高的所述抗体;进化亲和度最高的所述抗体并对亲和度最高的所述抗体进行接种免疫,直到进化后的所述抗体与所述抗原的亲和度收敛;根据亲和度收敛时的所述抗体对应的所述映射关系部署所述虚拟机;进化所述抗体包括:根据预设模糊子集调整交叉率与变异率;根据调整后的所述交叉率与所述变异率,进化所述抗体。
可见,本申请所提供的虚拟机的部署方法,基于人工免疫算法,将批量部署虚拟机问题抽象为确定虚拟机与物理主机的映射关系的寻优问题,以物理主机达到负载最均衡作为寻优目标,通过设置并变换虚拟机与物理主机的映射关系,得到映射关系的最优解,即得到虚拟机的最佳部署方式。应用该部署方法在实现快速批量部署虚拟机的同时,能够使物理主机间达到最佳负载均衡,确保部署完成后,云计算平台在负载均衡较优且资源利用率较好的状态下运行。同时,该部署方法将模糊推理理论引入到人工免疫过程中,实现了智能化自适应地确定免疫过程的交叉率和变异率。
本申请所提供的虚拟机的部署装置、设备以及计算机可读存储介质均具有上述技术效果。
附图说明
为了更清楚地说明本申请实施例中的技术方案,下面将对现有技术和实施例中所需要使用的附图作简单地介绍,显而易见地,下面描述中的附图仅仅是本申请的一些实施例,对于本领域普通技术人员来讲,在不付出创造性劳动的前提下,还可以根据这些附图获得其他的附图。
图1为本申请实施例所提供的一种虚拟机的部署方法的流程示意图;
图2为本申请实施例所提供的一种隶属函数的示意图;
图3为本申请实施例所提供的另一种隶属函数的示意图;
图4为本申请实施例所提供的一种虚拟机的部署装置的示意图;
图5为本申请实施例所提供的一种虚拟机的部署设备的示意图。
具体实施方式
本申请的核心是提供一种虚拟机的部署方法,能够实现快速批量部署虚拟机,并使物理主机间达到最佳负载均衡。本申请的另一个核心是提供一种虚拟机的部署装置、设备以及计算机可读存储介质,均具有上述技术效果。
为使本申请实施例的目的、技术方案和优点更加清楚,下面将结合本申请实施例中的附图,对本申请实施例中的技术方案进行清楚、完整地描述,显然,所描述的实施例是本申请一部分实施例,而不是全部的实施例。基于本申请中的实施例,本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例,都属于本申请保护的范围。
请参考图1,图1为本申请实施例所提供的一种虚拟机的部署方法的流程示意图,参考图1所示,该方法主要包括:
S101:设置亲和度函数与约束条件,并以所述亲和度函数与所述约束条件作为抗原;
S102:设置虚拟机与物理主机的映射关系,并以所述映射关系作为抗体;
具体的,本申请基于人工免疫算法,将批量部署虚拟机问题抽象为确定虚拟机与物理主机的映射关系问题,故首先设置抗原与抗体。抗原包括亲和度函数与约束条件。抗体即虚拟机与物理主机的映射关系。设置多个虚拟机与多个物理主机之间的初始的映射关系,即设置初始抗体群。进而通过执行后续步骤确定最优映射关系。所谓最优映射关系是指物理主机依据该映射关系部署虚拟机时,各物理主机之间可以达到最佳的负载均衡。
设置虚拟机与物理主机的映射关系可抽象为编码抗体。抗体的结构是由多种氨基酸的排列组成的链条。本申请使用氨基酸的位置编号代表所需要部署的虚拟机的编号,而各个位置上的氨基酸的种类代表不同的物理主机。因此,一种抗体就代表着一种虚拟机部署方案,即一种虚拟机与物理主机的映射关系。假设,某云计算平台初始有M个物理主机,需要在这些物理主机上部署N个虚拟机,那么采用上述编码方式的得到抗体为:
A i(v 1,...,v N),v 1,v 2,...,v N∈[1,M]。v N表示第N个虚拟机映 射在第v N个物理主机上。
在一种具体的实施方式中,设置虚拟机与物理主机的映射关系的方式为:设置虚拟机与所述物理主机的映射关系,所述映射关系的个数为预设数量,且所述映射关系满足在所述映射关系下的所述物理主机不存在资源超配的情况。
具体而言,在不考虑资源限制的情况下,N个虚拟机部署到M台物理主机上共有M N种部署方式,即有M N种映射关系,也即有M N种抗体。但是物理主机的资源是有限的,且具有一个最佳的使用率。因此,首先可利用Java语言中的Random随机函数为抗体的每个位置依次选取[1,M]中的整数,得到初始抗体群,即得到初始的一些映射关系。然后判断在这些映射关系下的物理主机是否存在资源超配的情况,并剔除掉存在资源超配的映射关系,即从抗体群中剔除掉无效的抗体。同时,若映射关系的数量过多,则会导致收敛速度过慢,映射关系的数量过少,则会导致无法得到最优解,因此本实施例最终设置的虚拟机与物理主机的映射关系为一个较为合适的数量即预设数量。例如,预设数量可以为30。
S103:计算所述抗体与所述抗原的亲和度;
具体而言,本申请将虚拟机的部署问题抽象为确定具有约束条件的最优求解问题。预先确定好亲和度函数与约束条件,在设置虚拟机与物理主机的映射关系的基础上,在约束条件下,根据亲和度函数计算各抗体的亲和度,即得到各映射关系下的所述物理主机的负载均衡值。负载均衡值反映的是在某种映射关系下的,各物理主机之间的负载均衡情况。
在一种具体的实施方式中,所述计算所述抗体与所述抗原的亲和度包括:
计算所述抗体对应的所述映射关系下的所述物理主机的目标资源的使用率;
根据所述目标资源的使用率得到所述物理主机的负载;
计算各所述物理主机的负载的均值,并计算所述物理主机的负载与所述均值的欧式距离;
将所述欧式距离取倒数得到所述抗体与所述抗原的亲和度。
其中,所述目标资源包括:CPU、内存以及硬盘。
具体而言,由于批量部署虚拟机最理想的方式是在部署完成后,各物理 主机之间负载达到最佳均衡的状况,故本实施例重点关注物理主机的CPU、内存以及硬盘,物理主机的负载定义如下:
Figure PCTCN2021127500-appb-000003
其中,d取3,则s∈1,2,3,用于分别表示物理主机M的CPU、内存与硬盘。Us表示s资源的使用率。例如,s取值1时,U 1表示物理主机M的CPU的使用率。
对于每一种虚拟机与物理主机的映射关系,都可以相应的得到该映射关系下的,各物理主机的各资源的使用率,进而依据上式得到该映射关系下的,各物理主机的负载。L(M)表示物理主机M的负载。
物理主机之间的负载均衡使用欧氏距离进行度量,如下式所示:
Figure PCTCN2021127500-appb-000004
其中,
Figure PCTCN2021127500-appb-000005
为各物理主机的负载的平均值。
亲和度函数为:
Figure PCTCN2021127500-appb-000006
对于每种映射关系,在基于式(1)得到各物理主机的负载之后,进一步基于式(2)与式(3)可以得到该映射关系下的,物理主机的负载均衡值,即得到抗体的亲和度。
进一步,约束条件可以为所述物理主机的所述目标资源的使用率位于预设区间。例如预设区间为[0,0.9],则约束条件为Us∈[0,0.9]。
S104:从所述抗体中挑选出亲和度最高的所述抗体;
具体的,计算得到各映射关系下的物理主机的负载均衡值后,可采用精英选择与轮盘赌选择相结合的策略,从中挑选出亲和度最高的所述抗体,并将其存储到记忆单元库。
S105:进化亲和度最高的所述抗体并对亲和度最高的所述抗体进行接种免疫,直到进化后的所述抗体与所述抗原的亲和度收敛;
S106:根据亲和度收敛时的所述抗体对应的所述映射关系部署虚拟机。
具体的,进化抗体即是指变换映射关系,并在每次进化与免疫后,计算 得到抗体的亲和度,直到抗体的亲和度收敛,至此以亲和度收敛时的抗体对应的映射关系作为最终的映射关系,并将该最终的映射关系封装为Json格式的参数,然后自动部署虚拟机的脚本使用该Json格式的参数,将任务分发给各个物理主机,使各个物理主机依次对创建部署所分配的虚拟机,最终快速完成虚拟机的批量部署。
其中,在一种具体的实施方式中,进化抗体的方式为:
根据预设模糊子集调整交叉率与变异率;
根据调整后的所述交叉率与所述变化率,进化所述抗体。
具体而言,通常情况下,在人工免疫算法中,免疫过程中的交叉率p c和变异率p m是固定不变的。这样虽然人工免疫算法的模型相对简单,但是这两个值取值的不同会严重影响整个免疫过程,例如过大的变异率和交叉率具有较大的跳跃性,可能会影响抗体进化的稳定。较小的变异率和交叉率又不能很好地模拟免疫过程中的多样性。因此,本实施例采用自适应的方法确定免疫过程中的交叉率和变异率。同时考虑到免疫过程的动态性、不确定性,本实施例将模糊推理理论引入到人工免疫过程中,实现智能化自适应地确定免疫过程的交叉率和变异率。
具体而言,使用L avg(g)代表第g代抗体与抗原的平均亲和度,则连续两代抗体平均亲和度的差值就可以表示抗体的进化效果,即:
Figure PCTCN2021127500-appb-000007
Figure PCTCN2021127500-appb-000008
则代表抗体逐代向免疫效果更强的方向进化,此时适合适当地增加交叉率和变异率,若
Figure PCTCN2021127500-appb-000009
则代表抗体在逐代的退化,可能是交叉率和变异率过大造成的,此时考虑适当地减小交叉率和变异率。
本实施例使用NB(negative big),NS(negative small),ZE(zero),PS(positive small),PB(positive big),可以在
Figure PCTCN2021127500-appb-000010
论域上定义五个模糊子集,其隶属函数如图2所示。
同时,定义p c和p m的变化值为
Figure PCTCN2021127500-appb-000011
Figure PCTCN2021127500-appb-000012
使用NH(negative huge),NB(negative big),NS(negative small),ZE(zero),PS(positive small),PB(positive big)和PH(positive huge)表示“负具大”、“负大”、“负小”、“不变”、“正小”、“正大”和“正具大”,也可以将
Figure PCTCN2021127500-appb-000013
论域定义七个模糊子集,其隶属函数如图3所示,
Figure PCTCN2021127500-appb-000014
因此,在本实施例 中的模糊推理规则可如表1所示:
表1
Figure PCTCN2021127500-appb-000015
可以明白的是,对于NB、NS等的具体数值,可以进行差异性设置,本申请在此不做限定。
另外,在一种具体的实施方式中,对所述抗体进行接种免疫的方式为:
变异所述抗体得到新的抗体,并计算所述新的抗体与所述抗体的亲和度的差值;
根据所述亲和度差值以及退火因子,得到所述新的抗体替换所述抗体的概率;
根据所述概率将所述抗体替换为所述新的抗体。
具体而言,本实施例利用模拟退火优化疫苗接种过程。抗体每进化一代,从本待中挑选出最优的抗体,并对该抗体进行接种免疫。所谓接种免疫是使得抗体A中链条上发生一个小范围的位置交换,则产生一个新的抗体N,分别计算出这两个抗体的亲和度(即负载均衡值)分别为L(A),L(N),二者的差值为
Figure PCTCN2021127500-appb-000016
进而根据Metropolis准测确定抗体N替代抗体A的概率P,如下式所示:
Figure PCTCN2021127500-appb-000017
其中,Tg表示退火因子,为预先设定的定值。如果
Figure PCTCN2021127500-appb-000018
则使用抗体N代替抗体A更新至记忆单元;若
Figure PCTCN2021127500-appb-000019
则以P的概率使用抗体N代替抗体A更新至记忆单元,例如,P为1/3,则抗体N代替抗体A的概率为1/3。
将模糊逻辑引入免疫过程中,可以实现动态的调整交叉率与变异率。模拟退火与动态调整交叉率与变异率的引入,既可以使整个免疫过程快速收敛,又能够保证最终得到最优的抗体,即得到最佳的部署方案,可以有效提高效率与性能。
经过多次接种免疫后,最终抗体的亲和度收敛,即最终物理主机的负载均衡值收敛,表明即使再变化映射关系,物理主机的负载均衡值也不会有大的差异,此时的映射关系最优。免疫终止时记忆单元库中的抗体即为最优的抗体,通过反编码此抗体即可得到最终的映射关系。
综上所述,本申请所提供的虚拟机的部署方法,基于人工免疫算法,将批量部署虚拟机问题抽象为确定虚拟机与物理主机的映射关系的寻优问题,以物理主机达到负载最均衡作为寻优目标,通过设置并变换虚拟机与物理主机的映射关系,得到映射关系的最优解,即得到虚拟机的最佳部署方式。应用该部署方法在实现快速批量部署虚拟机的同时,能够使物理主机间达到最佳负载均衡,确保部署完成后,云计算平台在负载均衡较优且资源利用率较好的状态下运行。
本申请还提供了一种虚拟机的部署装置,下文描述的该装置可以与上文描述的方法相互对应参照。请参考图4,图4为本申请实施例所提供的一种虚拟机的部署装置的示意图,结合图4所示,该装置包括:
第一设置模块10,用于设置亲和度函数与约束条件,并以所述亲和度函数与所述约束条件作为抗原;
第二设置模块20,用于设置虚拟机与物理主机的映射关系,并以所述映射关系作为抗体;
计算模块30,用于计算所述抗体与所述抗原的亲和度;
挑选模块40,用于从所述抗体中挑选出亲和度最高的所述抗体;
进化与免疫模块50,用于进化亲和度最高的所述抗体并对亲和度最高的所述抗体进行接种免疫,直到进化后的所述抗体与所述抗原的亲和度收敛;
部署模块60,用于根据亲和度收敛时的所述抗体对应的所述映射关系部署所述虚拟机。
在上述实施例的基础上,可选的,所述第二设置模块20具体用于:
设置虚拟机与所述物理主机的映射关系,所述映射关系的个数为预设数量,且所述映射关系满足在所述映射关系下的所述物理主机不存在资源超配的情况。
在上述实施例的基础上,可选的,所述计算模块30具体用于:
计算所述抗体对应的所述映射关系下的所述物理主机的目标资源的使用率;
根据所述目标资源的使用率得到所述物理主机的负载;
计算各所述物理主机的负载的均值,并计算所述物理主机的负载与所述均值的欧式距离;
将所述欧式距离取倒数得到所述抗体与所述抗原的亲和度。
在上述实施例的基础上,可选的,所述目标资源包括:CPU、内存以及硬盘。
在上述实施例的基础上,可选的,所述约束条件为所述物理主机的所述目标资源的使用率位于预设区间。
在上述实施例的基础上,可选的,所述进化与免疫模块50具体用于:
根据预设模糊子集调整交叉率与变异率;
根据调整后的所述交叉率与所述变化率,进化所述抗体。
在上述实施例的基础上,可选的,所述进化与免疫模块50具体用于:
变异所述抗体得到新的抗体,并计算所述新的抗体与所述抗体的亲和度的差值;
根据所述亲和度差值以及退火因子,得到所述新的抗体替换所述抗体的概率;
根据所述概率将所述抗体替换为所述新的抗体。
本申请所提供的虚拟机的部署装置,基于人工免疫算法,将批量部署虚拟机问题抽象为确定虚拟机与物理主机的映射关系的寻优问题,以物理主机达到负载最均衡作为寻优目标,通过设置并变换虚拟机与物理主机的映射关系,得到映射关系的最优解,即得到虚拟机的最佳部署方式。应用该部署方 法在实现快速批量部署虚拟机的同时,能够使物理主机间达到最佳负载均衡,确保部署完成后,云计算平台在负载均衡较优且资源利用率较好的状态下运行。
本申请还提供了一种虚拟机的部署设备,参考图5所示,该设备包括存储器1和处理器2。
存储器1,用于存储计算机程序;
处理器2,用于执行计算机程序实现如下的步骤:
设置亲和度函数与约束条件,并以所述亲和度函数与所述约束条件作为抗原;设置虚拟机与物理主机的映射关系,并以所述映射关系作为抗体;计算所述抗体与所述抗原的亲和度;从所述抗体中挑选出亲和度最高的所述抗体;进化亲和度最高的所述抗体并对亲和度最高的所述抗体进行接种免疫,直到进化后的所述抗体与所述抗原的亲和度收敛;根据亲和度收敛时的所述抗体对应的所述映射关系部署所述虚拟机。
对于本申请所提供的设备的介绍请参照上述方法实施例,本申请在此不做赘述。
本申请还提供了一种计算机可读存储介质,该计算机可读存储介质上存储有计算机程序,计算机程序被处理器执行时可实现如下的步骤:
设置亲和度函数与约束条件,并以所述亲和度函数与所述约束条件作为抗原;设置虚拟机与物理主机的映射关系,并以所述映射关系作为抗体;计算所述抗体与所述抗原的亲和度;从所述抗体中挑选出亲和度最高的所述抗体;进化亲和度最高的所述抗体并对亲和度最高的所述抗体进行接种免疫,直到进化后的所述抗体与所述抗原的亲和度收敛;根据亲和度收敛时的所述抗体对应的所述映射关系部署所述虚拟机。
该计算机可读存储介质可以包括:U盘、移动硬盘、只读存储器(Read-Only Memory,ROM)、随机存取存储器(Random Access Memory,RAM)、磁碟或者光盘等各种可以存储程序代码的介质。
对于本申请所提供的计算机可读存储介质的介绍请参照上述方法实施例,本申请在此不做赘述。
说明书中各个实施例采用递进的方式描述,每个实施例重点说明的都是与其他实施例的不同之处,各个实施例之间相同相似部分互相参见即可。对于实施例公开的装置、设备以及计算机可读存储介质而言,由于其与实施例公开的方法相对应,所以描述的比较简单,相关之处参见方法部分说明即可。
专业人员还可以进一步意识到,结合本文中所公开的实施例描述的各示例的单元及算法步骤,能够以电子硬件、计算机软件或者二者的结合来实现,为了清楚地说明硬件和软件的可互换性,在上述说明中已经按照功能一般性地描述了各示例的组成及步骤。这些功能究竟以硬件还是软件方式来执行,取决于技术方案的特定应用和设计约束条件。专业技术人员可以对每个特定的应用来使用不同方法来实现所描述的功能,但是这种实现不应认为超出本申请的范围。
结合本文中所公开的实施例描述的方法或算法的步骤可以直接用硬件、处理器执行的软件模块,或者二者的结合来实施。软件模块可以置于随机存储器(RAM)、内存、只读存储器(ROM)、电可编程ROM、电可擦除可编程ROM、寄存器、硬盘、可移动磁盘、CD-ROM、或技术领域内所公知的任意其它形式的存储介质中。
以上对本申请所提供的虚拟机的部署方法、装置、设备以及计算机可读存储介质进行了详细介绍。本文中应用了具体个例对本申请的原理及实施方式进行了阐述,以上实施例的说明只是用于帮助理解本申请的方法及其核心思想。应当指出,对于本技术领域的普通技术人员来说,在不脱离本申请原理的前提下,还可以对本申请进行若干改进和修饰,这些改进和修饰也落入本申请权利要求的保护范围。

Claims (10)

  1. 一种虚拟机的部署方法,其特征在于,包括:
    设置亲和度函数与约束条件,并以所述亲和度函数与所述约束条件作为抗原;
    设置虚拟机与物理主机的映射关系,并以所述映射关系作为抗体;
    计算所述抗体与所述抗原的亲和度;
    从所述抗体中挑选出亲和度最高的所述抗体;
    进化亲和度最高的所述抗体并对亲和度最高的所述抗体进行接种免疫,直到进化后的所述抗体与所述抗原的亲和度收敛;
    根据亲和度收敛时的所述抗体对应的所述映射关系部署所述虚拟机;
    进化所述抗体包括:
    根据预设模糊子集调整交叉率与变异率;
    根据调整后的所述交叉率与所述变异率,进化所述抗体。
  2. 根据权利要求1所述的部署方法,其特征在于,设置虚拟机与物理主机的映射关系包括:
    设置虚拟机与所述物理主机的映射关系,所述映射关系的个数为预设数量,且所述映射关系满足在所述映射关系下的所述物理主机不存在资源超配的情况。
  3. 根据权利要求1所述的部署方法,其特征在于,所述计算所述抗体与所述抗原的亲和度包括:
    计算所述抗体对应的所述映射关系下的所述物理主机的目标资源的使用率;
    根据所述目标资源的使用率得到所述物理主机的负载;
    计算各所述物理主机的负载的均值,并计算所述物理主机的负载与所述均值的欧式距离;
    将所述欧式距离取倒数得到所述抗体与所述抗原的亲和度。
  4. 根据权利要求3所述的部署方法,其特征在于,所述目标资源包括:CPU、内存以及硬盘。
  5. 根据权利要求3所述的部署方法,其特征在于,所述约束条件为所述物理主机的所述目标资源的使用率位于预设区间。
  6. 根据权利要求1所述的部署方法,其特征在于,对所述抗体进行接种免疫包括:
    变异所述抗体得到新的抗体,并计算所述新的抗体与所述抗体的亲和度的差值;
    根据所述亲和度差值以及退火因子,得到所述新的抗体替换所述抗体的概率;
    根据所述概率将所述抗体替换为所述新的抗体。
  7. 根据权利要求6所述的部署方法,其特征在于,所述根据所述亲和度差值以及退火因子,得到所述新的抗体替换所述抗体的概率包括:
    依据公式
    Figure PCTCN2021127500-appb-100001
    得到所述新的抗体替换所述抗体的概率;
    其中,Tg表示退火因子,
    Figure PCTCN2021127500-appb-100002
    表示亲和度差值,P表示所述新的抗体替换所述抗体的概率。
  8. 一种虚拟机的部署装置,其特征在于,包括:
    第一设置模块,用于设置亲和度函数与约束条件,并以所述亲和度函数与所述约束条件作为抗原;
    第二设置模块,用于设置虚拟机与物理主机的映射关系,并以所述映射关系作为抗体;
    计算模块,用于计算所述抗体与所述抗原的亲和度;
    挑选模块,用于从所述抗体中挑选出亲和度最高的所述抗体;
    进化与免疫模块,用于进化亲和度最高的所述抗体并对亲和度最高的所述抗体进行接种免疫,直到进化后的所述抗体与所述抗原的亲和度收敛;
    部署模块,用于根据亲和度收敛时的所述抗体对应的所述映射关系部署所述虚拟机;
    进化所述抗体包括:
    根据预设模糊子集调整交叉率与变异率;
    根据调整后的所述交叉率与所述变异率,进化所述抗体。
  9. 一种虚拟机的部署设备,其特征在于,包括:
    存储器,用于存储计算机程序;
    处理器,用于执行所述计算机程序时实现如权利要求1至7任一项所述的虚拟机的部署方法的步骤。
  10. 一种计算机可读存储介质,其特征在于,所述计算机可读存储介质上存储有计算机程序,所述计算机程序被处理器执行时实现如权利要求1至7任一项所述的虚拟机的部署方法的步骤。
PCT/CN2021/127500 2021-08-20 2021-10-29 虚拟机的部署方法、装置、设备以及计算机可读存储介质 WO2023019735A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202110957630.8A CN113407311A (zh) 2021-08-20 2021-08-20 虚拟机的部署方法、装置、设备以及计算机可读存储介质
CN202110957630.8 2021-08-20

Publications (1)

Publication Number Publication Date
WO2023019735A1 true WO2023019735A1 (zh) 2023-02-23

Family

ID=77689032

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2021/127500 WO2023019735A1 (zh) 2021-08-20 2021-10-29 虚拟机的部署方法、装置、设备以及计算机可读存储介质

Country Status (2)

Country Link
CN (1) CN113407311A (zh)
WO (1) WO2023019735A1 (zh)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113407311A (zh) * 2021-08-20 2021-09-17 苏州浪潮智能科技有限公司 虚拟机的部署方法、装置、设备以及计算机可读存储介质

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20120198447A1 (en) * 2011-01-31 2012-08-02 International Business Machines Corporation Determining an allocation configuration for allocating virtual machines to physical machines
CN108009003A (zh) * 2017-12-01 2018-05-08 湖南女子学院 一种多目标虚拟机自适应位置选择方法、分布式云系统
CN109542623A (zh) * 2018-11-22 2019-03-29 中国人民解放军陆军工程大学 虚拟系统在服务器中的优化部署方法
CN112463291A (zh) * 2020-11-12 2021-03-09 苏州浪潮智能科技有限公司 一种虚拟机部署方法、装置、设备及可读存储介质
CN113407311A (zh) * 2021-08-20 2021-09-17 苏州浪潮智能科技有限公司 虚拟机的部署方法、装置、设备以及计算机可读存储介质

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US10694402B2 (en) * 2010-11-05 2020-06-23 Mark Cummings Security orchestration and network immune system deployment framework
CN109816262B (zh) * 2019-01-31 2023-04-28 贵州大学 采用改良免疫遗传算法的柔性作业车间调度方法

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20120198447A1 (en) * 2011-01-31 2012-08-02 International Business Machines Corporation Determining an allocation configuration for allocating virtual machines to physical machines
CN108009003A (zh) * 2017-12-01 2018-05-08 湖南女子学院 一种多目标虚拟机自适应位置选择方法、分布式云系统
CN109542623A (zh) * 2018-11-22 2019-03-29 中国人民解放军陆军工程大学 虚拟系统在服务器中的优化部署方法
CN112463291A (zh) * 2020-11-12 2021-03-09 苏州浪潮智能科技有限公司 一种虚拟机部署方法、装置、设备及可读存储介质
CN113407311A (zh) * 2021-08-20 2021-09-17 苏州浪潮智能科技有限公司 虚拟机的部署方法、装置、设备以及计算机可读存储介质

Also Published As

Publication number Publication date
CN113407311A (zh) 2021-09-17

Similar Documents

Publication Publication Date Title
US20210133534A1 (en) Cloud task scheduling method based on phagocytosis-based hybrid particle swarm optimization and genetic algorithm
WO2023019735A1 (zh) 虚拟机的部署方法、装置、设备以及计算机可读存储介质
CN112712171B (zh) 深度卷积神经网络的分布式训练方法、设备和存储介质
CN113485826B (zh) 一种边缘服务器负载均衡方法、系统
CN110362380A (zh) 一种面向网络靶场的多目标优化虚拟机部署方法
CN107193640B (zh) 一种基于Rendezvous哈希算法的虚拟机初始放置策略方法
CN104461739B (zh) 一种基于cloudsim平台的虚拟机批量部署方法
CN110362379A (zh) 基于改进蚁群算法的虚拟机调度方法
CN113037546A (zh) 基于深度强化学习的安全性虚拟网络映射方法
WO2022100306A1 (zh) 一种虚拟机部署方法、装置、设备及可读存储介质
CN108108228A (zh) 一种基于差分进化算法的资源分配方法
CN112130927B (zh) 一种可靠性增强的移动边缘计算任务卸载方法
CN109787836A (zh) 基于遗传算法的vnf调度方案的确定方法及装置
CN109347900B (zh) 基于改进狼群算法的云服务系统自适应演化方法
CN113296893B (zh) 一种基于混合正弦余弦粒子群优化算法的云平台低资源损耗虚拟机放置方法
CN111612124A (zh) 一种面向任务智能调度的网络结构自适应优化方法
CN116431281A (zh) 一种基于鲸鱼优化算法的虚拟机迁移方法
CN114020418A (zh) 一种包含轮盘赌算法的粒子群优化的虚拟机放置方法
CN113326135A (zh) 一种多目标下的云资源调度方法
CN111818119B (zh) 基于改进的蜘蛛集群的数据服务资源优化调度方法
CN112379966A (zh) 一种云数据中心虚拟机实时整合方法及系统
CN113301583A (zh) 网络切片的重映射方法、装置以及存储介质
CN111400020B (zh) 一种虚拟机调度方法、装置和计算机可读存储介质
CN116503385B (zh) 基于虚拟全局代理的糖网眼底图像分级方法和设备
CN111580966B (zh) 一种基于内存利用率的云任务调度方法

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 21953971

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 21953971

Country of ref document: EP

Kind code of ref document: A1