CN110888713A - Trusted virtual machine migration algorithm for heterogeneous cloud data center - Google Patents
Trusted virtual machine migration algorithm for heterogeneous cloud data center Download PDFInfo
- Publication number
- CN110888713A CN110888713A CN201911117015.5A CN201911117015A CN110888713A CN 110888713 A CN110888713 A CN 110888713A CN 201911117015 A CN201911117015 A CN 201911117015A CN 110888713 A CN110888713 A CN 110888713A
- Authority
- CN
- China
- Prior art keywords
- physical machine
- utilization rate
- machine
- physical
- virtual machine
- Prior art date
- Legal status (The legal status 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 status listed.)
- Pending
Links
- 238000013508 migration Methods 0.000 title claims abstract description 58
- 230000005012 migration Effects 0.000 title claims abstract description 57
- 238000013507 mapping Methods 0.000 claims abstract description 29
- 238000005265 energy consumption Methods 0.000 claims abstract description 17
- 238000005457 optimization Methods 0.000 claims abstract description 14
- 238000000034 method Methods 0.000 claims abstract description 10
- 230000008569 process Effects 0.000 claims abstract description 7
- 239000011159 matrix material Substances 0.000 claims description 12
- 238000012163 sequencing technique Methods 0.000 claims description 10
- 238000012545 processing Methods 0.000 claims description 3
- 230000009467 reduction Effects 0.000 claims description 3
- 230000003044 adaptive effect Effects 0.000 claims description 2
- 239000013618 particulate matter Substances 0.000 description 16
- 238000005516 engineering process Methods 0.000 description 5
- 230000010354 integration Effects 0.000 description 5
- 238000011160 research Methods 0.000 description 4
- 230000000694 effects Effects 0.000 description 3
- 230000002035 prolonged effect Effects 0.000 description 2
- 230000003068 static effect Effects 0.000 description 2
- 230000006978 adaptation Effects 0.000 description 1
- 230000002411 adverse Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 238000010792 warming Methods 0.000 description 1
- 239000002699 waste material Substances 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/44—Arrangements for executing specific programs
- G06F9/455—Emulation; Interpretation; Software simulation, e.g. virtualisation or emulation of application or operating system execution engines
- G06F9/45533—Hypervisors; Virtual machine monitors
- G06F9/45558—Hypervisor-specific management and integration aspects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/46—Multiprogramming arrangements
- G06F9/48—Program initiating; Program switching, e.g. by interrupt
- G06F9/4806—Task transfer initiation or dispatching
- G06F9/4843—Task transfer initiation or dispatching by program, e.g. task dispatcher, supervisor, operating system
- G06F9/485—Task life-cycle, e.g. stopping, restarting, resuming execution
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/46—Multiprogramming arrangements
- G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
- G06F9/5083—Techniques for rebalancing the load in a distributed system
- G06F9/5088—Techniques for rebalancing the load in a distributed system involving task migration
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/44—Arrangements for executing specific programs
- G06F9/455—Emulation; Interpretation; Software simulation, e.g. virtualisation or emulation of application or operating system execution engines
- G06F9/45533—Hypervisors; Virtual machine monitors
- G06F9/45558—Hypervisor-specific management and integration aspects
- G06F2009/4557—Distribution of virtual machine instances; Migration and load balancing
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Landscapes
- Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Stored Programmes (AREA)
Abstract
A trusted virtual machine migration algorithm for a heterogeneous cloud data center comprises the following steps: (1) according to the initial mapping condition of the virtual machine, selecting an overload physical machine (2) with the utilization rate exceeding the upper limit, and determining the dimension of a migration algorithm according to the dimensions of the virtual machine and the physical machine parameters; (3) integrating until the utilization rate of all dimensions of the physical machine is less than the upper limit; (4) judging whether an unoptimized overload physical machine (5) exists or not, and selecting a light-load physical machine with the utilization rate smaller than the lower limit according to the initial mapping condition of the virtual machine; (6) determining the dimension of a migration algorithm; (7) sequentially integrating virtual machines deployed on the light-load physical machine; (8) judging whether an unoptimized light-load physical machine exists, sequentially finishing the optimization of all light-load physical machines, finishing the migration process and generating a final mapping result; the method and the device reduce the energy consumption of the cloud data center and improve the success rate of virtual machine migration.
Description
Technical Field
The invention relates to the technical field of high-performance computing and cloud computing, in particular to a trusted virtual machine migration algorithm for a heterogeneous cloud data center.
Background
In order to meet the continuously improved performance requirements of cloud users, cloud service providers further expand the scale of cloud data centers. However, large-scale cloud data centers inevitably bring high energy consumption, which means high cost for enterprises. With global energy shortages and climate warming, energy consumption of cloud computing has become one of the bottlenecks that restrict the development of cloud computing.
The energy-saving mode of cloud computing can be mainly divided into a static energy-saving mode and a dynamic energy-saving mode. Static energy saving is mainly performed on computer hardware equipment, and energy consumption factors are generally considered at the beginning of design. The dynamic energy saving mainly comprises a dynamic voltage frequency adjustment (DVFS) technology and an optimized scheduling algorithm, and the improvement of the scheduling algorithm is a good research approach and mainly comprises cloud task scheduling, virtual machine scheduling and virtual machine integration. None of the previous migration algorithms consider the following problems. Firstly, most of the algorithms take isomorphic cloud data centers as research objects. However, in the actual situation, the cloud data center mostly adopts heterogeneous clusters composed of cheap machines, each physical machine can show different performances due to different configurations of a CPU, an internal memory and the like, the previous algorithm is used for solving the mapping problem of the homogeneous clusters, and the heterogeneity of the clusters is not considered. Secondly, the attributes of the virtual machine and the physical machine are generally composed of multidimensional factors such as CPU utilization rate, memory size, storage capacity and bandwidth size, the prior algorithm is to normalize the attributes, and in the normalization process, the great deviation of the result is caused by the difference of the obtained weight parameters. Therefore, the multidimensional attributes of the virtual machine and the physical machine should be considered comprehensively in a real environment. Thirdly, since the previous research does not consider the performance loss caused by the migration of the virtual machine, the performance loss caused by a large number of migrations is a significant factor which is not negligible when the migrations are performed. In the past, the aim of improving the utilization rate of a physical machine is mostly to the maximum, but the excessively high utilization rate of the physical machine can aggravate the competition of a virtual machine running on the physical machine for certain shared resources, and can adversely affect the performance of the virtual machine, so that the performance of the virtual machine is reduced, and the execution time of a cloud task is prolonged. Therefore, an upper limit should be set on the utilization rate of the physical machine in order to guarantee the performance of the virtual machine. Meanwhile, because the power consumption of the physical machine in the idle state is still about 70% of the full load power consumption, the energy consumption can be reduced in the true sense only by closing the idle and low-utilization physical machine. Finally, in order to reduce the power consumption of the cloud data center, the load balancing effect is often poor after the virtual machine integration is performed, and a good load balancing effect is a premise for improving the efficiency of the cloud data center, so that the load balancing must be comprehensively considered while the virtual machine is migrated.
Disclosure of Invention
In order to solve the problems in the prior art, the invention aims to provide a trusted virtual machine migration algorithm for a heterogeneous cloud data center, and a cloud computing correlation technology, a virtual machine migration technology and a scheduling optimization technology of the cloud heterogeneous data center are applied to construct a cloud data center energy consumption optimization, performance analysis and automatic migration method, so that automatic identification of overloaded and underloaded physical machines of the cloud data center, automatic migration of virtual machines and automatic integration of physical machines can be realized, the energy consumption of the cloud data center is reduced to the maximum extent, and the success rate of virtual machine migration is improved.
In order to achieve the purpose, the invention adopts the technical scheme that:
a trusted virtual machine migration algorithm for a heterogeneous cloud data center comprises the following steps:
(1) according to the initial mapping condition of the virtual machine, selecting an overload physical machine with the utilization rate exceeding the upper limit, wherein the upper limit of the utilization rate is 90%;
(2) determining the dimensionality of a migration algorithm according to the dimensionality of the parameters of the virtual machine and the physical machine, if the dimensionality of the migration algorithm is the one-dimensional virtual machine and the physical machine, directly sequencing overloaded physical machines according to the parameters from high utilization rate to low utilization rate, and otherwise, sequencing according to the CPU utilization rate;
when a multi-dimensional heterogeneous physical machine is considered, migrating an overloaded physical machine set according to the utilization rate of a CPU (central processing unit); because the migration loss is considered, the CPU requirement of each migrated virtual machine is 1.1 times of the original CPU requirement;
(3) sequentially integrating the virtual machines deployed on the overloaded physical machine, wherein the virtual machines with the minimum capacity on the physical machine start to migrate each time until the utilization rate of all dimensions of the physical machine is less than the upper limit;
sequencing and sequentially migrating a virtual machine set of an overloaded physical machine; selecting a target physical machine according to the CPU utilization rate by utilizing an optimal adaptive BF algorithm; updating the mapping matrix of the virtual machine and judging whether the mapping matrix is smaller than the utilization rate upper limit or not every time migration is completed, if so, finishing the optimization of the next physical machine, and otherwise, continuing migration until the mapping matrix is smaller than the utilization rate upper limit;
(4) judging whether an unoptimized overload physical machine exists or not, if so, repeating the step (3), otherwise, turning to the step (5);
sequentially completing the optimization of all the physical machines to ensure that the utilization rate of the physical machines is less than the upper limit, thereby completing the migration of all overloaded physical machines and ensuring that the virtual machines on the physical machines cannot cause performance reduction due to the contention of certain common resources;
(5) selecting a light-load physical machine with the utilization rate less than the lower limit according to the initial mapping condition of the virtual machine, wherein the lower limit of the utilization rate is 60%;
(6) determining the dimension of a migration algorithm according to the dimensions of the parameters of the virtual machine and the physical machine; if the virtual machine and the physical machine are one-dimensional, sorting the light-load physical machines from small to large according to the utilization rate of the parameters, and otherwise, sorting according to the CPU utilization rate;
when a multi-dimensional heterogeneous physical machine is considered, a light-load physical machine set is migrated according to the CPU utilization rate, and the CPU requirement of each migrated virtual machine is 1.1 times of the original CPU requirement due to the consideration of migration loss;
(7) sequentially integrating the virtual machines deployed on the light-load physical machine, wherein the virtual machine with the minimum capacity on the physical machine starts to be migrated each time until all the virtual machines of the physical machine are migrated;
then sorting and sequentially migrating the virtual machine set of the light-load physical machine, and selecting a target physical machine according to the CPU utilization rate by using a BF algorithm; until all the virtual machines of the physical machine are migrated, finally, closing the idle physical machine, thereby reducing the energy consumption of the cloud data center;
(8) judging whether an unoptimized light-load physical machine exists, if so, repeating the step (7), otherwise, ending the whole scheduling algorithm;
and sequentially finishing the optimization of all the light-load physical machines, finishing the migration process and generating a final mapping result.
Compared with the prior art, the invention has the beneficial effects that:
firstly, a plurality of factors influencing the energy consumption of the physical machine and the overall architecture of the cloud data center are analyzed. Secondly, the influence of the utilization rate of the physical machine on the energy consumption of the cloud data center and the performance loss caused by the migration of the virtual machine are researched, and the upper limit and the lower limit of the utilization rate of the physical machine are determined. Finally, the multidimensional characteristic of the virtual machine and the heterogeneous characteristic of the cloud data center are researched, and the load balance of the cloud data center is considered.
Drawings
Fig. 1 is an example of a cloud data center scheduling system.
Fig. 2 is a cloud data center virtual machine mapping matrix.
Fig. 3 is a cloud datacenter virtual machine migration example.
Detailed Description
In order to make the features, processes and advantages of the present invention more apparent, the present invention will be described in further detail below with reference to the accompanying drawings and examples. It is to be understood that the embodiments described herein are merely illustrative of the basic idea of the invention and are not intended to limit the invention.
The invention relates to a trusted virtual machine migration algorithm for a heterogeneous cloud data center, which is applied to a cloud data center platform. The cloud data center comprises a host computer cluster, wherein PM is { PM ═ PM1,pm2,...,pmnN represents the number of physical machines in the cloud data center. Each physical machine may be described as pmj＝{cj,mj,hj,bjForm (b) of (b), wherein cj,mj,hj,bjRespectively representing the CPU utilization, memory, hard disk, and bandwidth of the physical machine. Each physical machine can accommodate a set of virtual machines, VMsj＝{vm1,vm2,...,vmmWhere m represents the total number of virtual machines contained on the physical machine, each virtual machine also containing CPU utilizationThe parameters of rate, memory, hard disk and bandwidth, etc. can be expressed as vmi＝{cji,mji,hji,bji}. In order to ensure the performance of the physical machines, the sum of the utilization rate of the virtual machine CPU, the memory, the hard disk and the bandwidth contained in each physical machine cannot exceed the capacity of the physical machine.
In order to maximize resource utilization rate, a virtualization technology is utilized in modern data to virtualize hardware resources of a data center into a plurality of virtual machines, services such as computing, storage and the like are provided for cloud users by taking the virtual machines as units, and meanwhile, cloud providers can complete dynamic configuration of the user resources by further adjusting a scheduling scheme of the virtual machines. In order to meet the user requirements or reasonably utilize resources, the scheduling system can also establish, migrate and destroy the virtual machine according to the condition of the cloud task queue. The scheduling system herein is illustrated in fig. 1.
The cloud data center receives cloud tasks from multiple users, and the tasks can be divided into various forms such as calculation intensive type, memory intensive type, storage intensive type and bandwidth intensive type according to different requirements, and are specifically represented by different sizes of parameters such as CPU utilization rate, memory, hard disk and bandwidth. When a user submits a corresponding cloud task to the cloud data center, the scheduling system deploys a corresponding virtual machine allocated to the cloud task according to the characteristics of the cloud task and the attributes of the virtual machine. However, each virtual machine is mapped on a corresponding physical machine, and usually, a corresponding matrix is used to represent the mapping relationship between the virtual machine and the physical machine. As shown in FIG. 2, vijRepresenting virtual machine ViIn physical machine PjIn the above mapping case, the value is 1 if mapping, and 0 if not mapping, so the sum of each row of the matrix must be 1. Many previous studies have focused on the optimization of mapping algorithms, and most of them are aimed at improving the utilization rate of physical machines. The study herein is the virtual machine integration for further optimization after the mapping is completed.
Based on the platform, the trusted virtual machine migration algorithm for the heterogeneous cloud data center provided by the invention comprises the following steps:
(1) according to the initial mapping condition of the virtual machine, selecting an overloaded physical machine with the utilization rate exceeding the upper limit, wherein the upper limit of the utilization rate is 90%.
Ideally, resources can be sufficiently saved by 100% of utilization rate, but previous research shows that when a physical machine is fully loaded, competition of the virtual machine for some shared resources is aggravated, so that performance of the virtual machine is reduced, and therefore execution time of cloud tasks is prolonged or the physical machine is down. In general, the upper limit of the utilization rate is 90% in order to ensure the physical machine performance.
(2) Determining the dimensionality of a migration algorithm according to the dimensionality of the parameters of the virtual machine and the physical machine, if the dimensionality of the migration algorithm is the one-dimensional virtual machine and the physical machine, directly sequencing overloaded physical machines according to the parameters from high utilization rate to low utilization rate, and otherwise, sequencing according to the CPU utilization rate;
when a multi-dimensional heterogeneous physical machine is considered, migrating an overloaded physical machine set according to the utilization rate of a CPU (central processing unit); because the migration loss is considered, the CPU requirement of each migrated virtual machine is 1.1 times of the original CPU requirement. The one-dimensional and isomorphic cloud data center is only a special case of a multi-dimensional heterogeneous physical machine, and only the dimension and the structure of the cloud data center are simplified.
(3) Sequentially integrating the virtual machines deployed on the overloaded physical machine, wherein the virtual machines with the minimum capacity on the physical machine start to migrate each time until the utilization rate of all dimensions of the physical machine is less than the upper limit;
sequencing and sequentially migrating a virtual machine set of the overloaded physical machine, and selecting a target physical machine according to the CPU utilization rate by using an optimal adaptation BF algorithm; and updating the mapping matrix of the virtual machine and judging whether the mapping matrix is smaller than the utilization rate upper limit or not every time the migration is completed, if so, finishing the optimization of the next physical machine, and otherwise, continuing the migration until the mapping matrix is smaller than the utilization rate upper limit.
(4) And (5) judging whether an unoptimized overload physical machine exists, if so, repeating the step (3), and otherwise, turning to the step (5).
And sequentially finishing the optimization of all the physical machines to ensure that the utilization rate of the physical machines is less than the upper limit, thereby finishing the migration of all the overloaded physical machines and ensuring that the virtual machines on the physical machines do not cause performance reduction due to the contention of certain common resources.
(5) And selecting a light-load physical machine with the utilization rate less than the lower limit according to the initial mapping condition of the virtual machine, wherein the lower limit of the utilization rate is 60%.
It is generally considered that the lower the load of the physical machine, the lower the energy consumption should be, but previous studies have found that the physical machine of the data center consumes more than 50% of the full load, and particularly about 50% -70%, even when it is unloaded. Statistics shows that the utilization rate of the global physical machine is only about 10%, which causes great waste on energy consumption, but also provides a basis for further optimizing the energy consumption of the physical machine. In order to improve the utilization rate of the physical machine and reasonably control the energy consumption of the physical machine, in this document, the lower limit of the utilization rate of the physical machine is taken as 60%, and all virtual machines deployed on the physical machine with the utilization rate lower than the lower limit should be integrated, so that the utilization rate of the physical machine is greater than the lower limit.
(6) And determining the dimension of the migration algorithm according to the dimensions of the parameters of the virtual machine and the physical machine. And if the virtual machine and the physical machine are one-dimensional, sorting the light-load physical machines from small to large according to the utilization rate of the parameters, and otherwise, sorting according to the CPU utilization rate.
When a multi-dimensional heterogeneous physical machine is considered, a light-load physical machine set is migrated according to the CPU utilization rate, and the CPU requirement of each migrated virtual machine is 1.1 times of the original CPU requirement due to the consideration of migration loss. The one-dimensional and isomorphic cloud data center is only a special case of a multi-dimensional heterogeneous physical machine, and only the dimension and the structure of the cloud data center are simplified.
(7) Sequentially integrating the virtual machines deployed on the light-load physical machine, wherein the virtual machine with the minimum capacity on the physical machine starts to be migrated each time until all the virtual machines of the physical machine are migrated;
and then sequencing the virtual machine sets of the light-load physical machines, sequentially migrating the virtual machine sets, and selecting a target physical machine according to the CPU utilization rate by using a BF algorithm. And finally, closing the idle physical machine until all the virtual machines of the physical machine are migrated, thereby reducing the energy consumption of the cloud data center.
(8) And (4) judging whether an unoptimized light-load physical machine exists, if so, repeating the step (7), otherwise, ending the whole scheduling algorithm.
The optimization of all light-load physical machines is completed in sequence, the migration process is completed, the final mapping result is generated, the migration success rate of the virtual machines and the efficiency of the physical machines are guaranteed, and the energy consumption of the cloud data center is reduced finally.
As shown in fig. 3, the following illustrates a process of virtual machine migration.
Suppose that the user request contains 4 physical machines PM together1、PM2、PM3And PM4The parameters respectively comprise { CPU utilization rate, memory, storage and bandwidth }, and the units are respectively GHz, GB and 102GB and 101Mbps, with values of { 12121212 }, { 11121212 }, { 1212911 }, and { 11121212 }. PM (particulate matter)1The deployed virtual machine comprises a VM1AAnd VM1BAnd the parameter values are { 9999 } and { 2222 }. PM (particulate matter)2The deployed virtual machine comprises a VM2AAnd VM2BThe parameter values are { 9899 } and { 2321 }. PM (particulate matter)3The deployed virtual machine comprises a VM3AAnd the parameter value is { 8857 }. PM (particulate matter)4The deployed virtual machine comprises a VM4AAnd the parameter value is { 6788 }. As shown in FIG. 3, each cell in the figure sequentially represents 1GHz CPU, 1GB memory, 10 GHz from left to right2GB memory and 101Mbps bandwidth. And the first area is an initial condition, and according to a maximum overload priority rule, firstly, the physical machines with the utilization rate larger than the upper limit are sorted in a descending order according to the CPU utilization rate, and the result is a second area. Second, even PM2Capacity ratio PM of CPU1Small, but its CPU utilization is high, so PM is first treated2The virtual machines are integrated to integrate the VM2BPerforming migration, selecting to migrate to PM according to CPU utilization3However, it has been observed that migration to PM can be detected3Upper PM3Exceeds an upper limit, and thus, selectively migrates to the PM4At this time, PM is shown in the third region2By usingThe rate is lower than the upper limit and the integration is completed. Thirdly, repeating the above steps to PM1The virtual machines are integrated to integrate the VM1BMigration to PM3At this time PM1Is below the upper limit, as shown in the fourth region. And finishing the migration of all the virtual machines so that the utilization rate of each dimension of all the physical machines is lower than the upper limit and higher than the lower limit. The performance of the virtual machine is not reduced or down due to the competition of the virtual machine for the common resources.
According to the method and the device, the deployment result of the cloud task is changed by determining the upper limit and the lower limit of the utilization rate of the physical machine and comparing the loss according to the influence of the virtual machine migration on the performance of the virtual machine and the physical machine, so that the migration success rate of the virtual machine and the utilization rate of the physical machine are improved, the number of used physical machines is reduced, and the load balancing effect is improved. The project group where the applicant is located has completed the implementation of the invention, and compared with other cloud task scheduling algorithms, the performance of the algorithm is improved by about 17% on average.
Claims (1)
1. A trusted virtual machine migration algorithm for a heterogeneous cloud data center is characterized by comprising the following steps:
(1) according to the initial mapping condition of the virtual machine, selecting an overload physical machine with the utilization rate exceeding the upper limit, wherein the upper limit of the utilization rate is 90%;
(2) determining the dimensionality of a migration algorithm according to the dimensionality of the parameters of the virtual machine and the physical machine, if the dimensionality of the migration algorithm is the one-dimensional virtual machine and the physical machine, directly sequencing overloaded physical machines according to the parameters from high utilization rate to low utilization rate, and otherwise, sequencing according to the CPU utilization rate;
when a multi-dimensional heterogeneous physical machine is considered, migrating an overloaded physical machine set according to the utilization rate of a CPU (central processing unit); because the migration loss is considered, the CPU requirement of each migrated virtual machine is 1.1 times of the original CPU requirement;
(3) sequentially integrating the virtual machines deployed on the overloaded physical machine, wherein the virtual machines with the minimum capacity on the physical machine start to migrate each time until the utilization rate of all dimensions of the physical machine is less than the upper limit;
sequencing and sequentially migrating a virtual machine set of an overloaded physical machine; selecting a target physical machine according to the CPU utilization rate by using the optimal adaptive BF; updating the mapping matrix of the virtual machine and judging whether the mapping matrix is smaller than the utilization rate upper limit or not every time migration is completed, if so, finishing the optimization of the next physical machine, and otherwise, continuing migration until the mapping matrix is smaller than the utilization rate upper limit;
(4) judging whether an unoptimized overload physical machine exists or not, if so, repeating the step (3), otherwise, turning to the step (5);
sequentially completing the optimization of all the physical machines to ensure that the utilization rate of the physical machines is less than the upper limit, thereby completing the migration of all overloaded physical machines and ensuring that the virtual machines on the physical machines cannot cause performance reduction due to the contention of certain common resources;
(5) selecting a light-load physical machine with the utilization rate less than the lower limit according to the initial mapping condition of the virtual machine, wherein the lower limit of the utilization rate is 60%;
(6) determining the dimension of a migration algorithm according to the dimensions of the parameters of the virtual machine and the physical machine; if the virtual machine and the physical machine are one-dimensional, sorting the light-load physical machines from small to large according to the utilization rate of the parameters, and otherwise, sorting according to the CPU utilization rate;
when a multi-dimensional heterogeneous physical machine is considered, a light-load physical machine set is migrated according to the CPU utilization rate, and the CPU requirement of each migrated virtual machine is 1.1 times of the original CPU requirement due to the consideration of migration loss;
(7) sequentially integrating the virtual machines deployed on the light-load physical machine, wherein the virtual machine with the minimum capacity on the physical machine starts to be migrated each time until all the virtual machines of the physical machine are migrated;
then sorting and sequentially migrating the virtual machine set of the light-load physical machine, and selecting a target physical machine according to the CPU utilization rate by using a BF algorithm; until all the virtual machines of the physical machine are migrated, finally, closing the idle physical machine, thereby reducing the energy consumption of the cloud data center;
(8) judging whether an unoptimized light-load physical machine exists, if so, repeating the step (7), otherwise, ending the whole scheduling algorithm;
and sequentially finishing the optimization of all the light-load physical machines, finishing the migration process and generating a final mapping result.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911117015.5A CN110888713A (en) | 2019-11-15 | 2019-11-15 | Trusted virtual machine migration algorithm for heterogeneous cloud data center |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911117015.5A CN110888713A (en) | 2019-11-15 | 2019-11-15 | Trusted virtual machine migration algorithm for heterogeneous cloud data center |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110888713A true CN110888713A (en) | 2020-03-17 |
Family
ID=69747578
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911117015.5A Pending CN110888713A (en) | 2019-11-15 | 2019-11-15 | Trusted virtual machine migration algorithm for heterogeneous cloud data center |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110888713A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111580966A (en) * | 2020-04-30 | 2020-08-25 | 西安石油大学 | Cloud task scheduling method based on memory utilization rate |
| CN112256387A (en) * | 2020-10-12 | 2021-01-22 | 麒麟软件有限公司 | Container migration method in container cloud platform |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130073731A1 (en) * | 2011-09-20 | 2013-03-21 | Infosys Limited | System and method for optimizing migration of virtual machines among physical machines |
| CN103294546A (en) * | 2013-04-03 | 2013-09-11 | 华中科技大学 | Multi-dimensional resource performance interference aware on-line virtual machine migration method and system |
| CN105159751A (en) * | 2015-09-17 | 2015-12-16 | 河海大学常州校区 | Energy-efficient virtual machine migration method in cloud data center |
| CN109213595A (en) * | 2017-07-07 | 2019-01-15 | 中兴通讯股份有限公司 | Load equilibration scheduling method, device and computer readable storage medium |
-
2019
- 2019-11-15 CN CN201911117015.5A patent/CN110888713A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130073731A1 (en) * | 2011-09-20 | 2013-03-21 | Infosys Limited | System and method for optimizing migration of virtual machines among physical machines |
| CN103294546A (en) * | 2013-04-03 | 2013-09-11 | 华中科技大学 | Multi-dimensional resource performance interference aware on-line virtual machine migration method and system |
| CN105159751A (en) * | 2015-09-17 | 2015-12-16 | 河海大学常州校区 | Energy-efficient virtual machine migration method in cloud data center |
| CN109213595A (en) * | 2017-07-07 | 2019-01-15 | 中兴通讯股份有限公司 | Load equilibration scheduling method, device and computer readable storage medium |
Non-Patent Citations (4)
| Title |
|---|
| 关启明等: "基于遗传蚁群算法的虚拟机整合", 《黑龙江科技信息》 * |
| 张汉林等: "基于能耗感知的虚拟机迁移管理软件" * |
| 朱亚会等: "云数据中心基于能耗感知的虚拟机调度算法", 《计算机与现代化》 * |
| 龚素文等: "基于迁移技术的云资源动态调度策略研究" * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111580966A (en) * | 2020-04-30 | 2020-08-25 | 西安石油大学 | Cloud task scheduling method based on memory utilization rate |
| CN112256387A (en) * | 2020-10-12 | 2021-01-22 | 麒麟软件有限公司 | Container migration method in container cloud platform |
| CN112256387B (en) * | 2020-10-12 | 2023-06-27 | 麒麟软件有限公司 | Container migration method in container cloud platform |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Jiang et al. | DataABC: A fast ABC based energy-efficient live VM consolidation policy with data-intensive energy evaluation model | |
| US20140082202A1 (en) | Method and Apparatus for Integration of Virtual Cluster and Virtual Cluster System | |
| CN108196935B (en) | Cloud computing-oriented virtual machine energy-saving migration method | |
| WO2015051685A1 (en) | Task scheduling method, device and system | |
| CN107220108B (en) | Method and system for realizing load balance of cloud data center | |
| CN104219318A (en) | Distributed file storage system and method thereof | |
| Li et al. | An energy-aware scheduling algorithm for big data applications in Spark | |
| Guo et al. | A container scheduling strategy based on neighborhood division in micro service | |
| Fan et al. | Improving MapReduce performance by balancing skewed loads | |
| CN112492032B (en) | Workflow cooperative scheduling method under mobile edge environment | |
| CN104182278A (en) | Method and device for judging busy degree of computer hardware resource | |
| CN110888713A (en) | Trusted virtual machine migration algorithm for heterogeneous cloud data center | |
| CN115718644A (en) | Computing task cross-region migration method and system for cloud data center | |
| CN109976879B (en) | Cloud computing virtual machine placement method based on resource usage curve complementation | |
| CN112068959A (en) | Self-adaptive task scheduling method and system and retrieval method comprising method | |
| He et al. | Energy-efficient framework for virtual machine consolidation in cloud data centers | |
| Jiang et al. | An energy-aware virtual machine migration strategy based on three-way decisions | |
| CN110308973A (en) | A kind of container dynamic migration method based on energy optimization | |
| CN108388471B (en) | Management method based on double-threshold constraint virtual machine migration | |
| Huang et al. | Fuzzy clustering with feature weight preferences for load balancing in cloud | |
| WO2024055809A1 (en) | Cloud-computing virtual-resource scheduling method based on clustering ensemble algorithm | |
| CN110308991B (en) | Data center energy-saving optimization method and system based on random tasks | |
| Thiam et al. | An energy-efficient VM migrations optimization in cloud data centers | |
| CN111813512B (en) | High-energy-efficiency Spark task scheduling method based on dynamic partition | |
| CN114138102A (en) | Cloud computing data center energy consumption optimization method based on multi-region division |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20200317 |
|
| RJ01 | Rejection of invention patent application after publication |