CN110362383B - P-E balanced VM migration method for seasonal non-stationary concurrency - Google Patents
P-E balanced VM migration method for seasonal non-stationary concurrency Download PDFInfo
- Publication number
- CN110362383B CN110362383B CN201910627610.7A CN201910627610A CN110362383B CN 110362383 B CN110362383 B CN 110362383B CN 201910627610 A CN201910627610 A CN 201910627610A CN 110362383 B CN110362383 B CN 110362383B
- Authority
- CN
- China
- Prior art keywords
- physical server
- resource
- plist
- migrated
- migration
- 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.)
- Active
Links
- 230000005012 migration Effects 0.000 title claims abstract description 60
- 238000013508 migration Methods 0.000 title claims abstract description 60
- 238000000034 method Methods 0.000 title claims abstract description 32
- 230000001932 seasonal effect Effects 0.000 title claims abstract description 16
- 238000005265 energy consumption Methods 0.000 claims abstract description 31
- 230000004044 response Effects 0.000 claims abstract description 24
- 239000011159 matrix material Substances 0.000 claims abstract description 23
- 230000003068 static effect Effects 0.000 claims abstract description 20
- 238000013468 resource allocation Methods 0.000 claims description 22
- 230000003247 decreasing effect Effects 0.000 claims description 6
- 230000001174 ascending effect Effects 0.000 claims description 3
- 238000013178 mathematical model Methods 0.000 claims description 3
- 230000006870 function Effects 0.000 claims 1
- 238000004422 calculation algorithm Methods 0.000 abstract description 18
- 230000008569 process Effects 0.000 abstract description 5
- 230000008859 change Effects 0.000 description 5
- 238000010586 diagram Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 230000010354 integration Effects 0.000 description 3
- 238000004088 simulation Methods 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000013507 mapping Methods 0.000 description 2
- 230000004931 aggregating effect Effects 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000001052 transient effect Effects 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
- G06F9/4856—Task life-cycle, e.g. stopping, restarting, resuming execution resumption being on a different machine, e.g. task migration, virtual machine 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
- G06F9/46—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
Landscapes
- Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides a seasonal non-stationary concurrency amount oriented P-E balanced VM migration method, and relates to the technical field of cloud computing. The method comprises two parts of VM static deployment and VM dynamic migration; firstly, according to the memory and CPU resources provided by the physical server and required by each VM, the VM is statically deployed, in the process, the VM is deployed on the physical host under the condition of meeting the customer requirements, and meanwhile, the number of the physical host is reduced as much as possible, so that the purpose of reducing energy consumption is achieved. And after the VM is statically deployed, the VM is migrated by adopting a VM dynamic migration strategy. The first part predicts the average response time of each VM by using an RBF algorithm and selects the VM to be migrated according to a set threshold; the second part is the selection of a target server, and the target server is selected through the updated resource demand matrix to complete the migration of the VM of one stage; and the first part and the second part are circularly carried out, and the whole VM migration process is completed.
Description
Technical Field
The invention relates to the technical field of cloud computing, in particular to a seasonal non-stationary concurrency amount oriented P-E balanced VM migration method.
Background
Virtualization technology can effectively utilize existing software and hardware resources, and Virtual Machines (VMs) can also be created using virtualization software. The VMs are independent entities on the network, facilitating sharing of hardware-related resources together. Under appropriate conditions, the overall performance, energy consumption and flexibility of the cloud service related system can be obviously improved by migrating a certain VM from one physical host to another physical host. The virtualization technology can help cloud service providers to realize ordered on-demand resource deployment, and an effective solution is provided for flexible resource management and reduction of energy consumption. For cloud services with virtualization, one of its main tasks is infrastructure as a service IaaS, such as Amazon EC 2: the tenant pays a fee to rent the VM. Because different resource utilization rates are caused by different mapping relationships between VMs and PMs (Physical machines), for a cloud service provider, a main problem is how to place multiple VMs required by a tenant on a Physical server to achieve the purposes of load balancing, resource utilization optimization, and the like. At this point, a VM migration is required to implement. The migration of VMs can be implemented by two methods, offline migration and live migration, respectively. Offline migration is a pause/resume operation, and therefore, offline migration has considerable downtime. Live migration employs a pre-copy approach as opposed to suspend/resume operations for offline migration. These two traditional VM migration methods have high cost, and the migration cost includes Service interruption, network traffic increase, and possibly Service Level Agreement (SLA) violation. Frequent migration of large numbers of VMs is neither practical nor feasible, and a reasonably efficient VM migration system can greatly reduce migration costs.
Disclosure of Invention
The technical problem to be solved by the present invention is to provide a seasonal non-stationary concurrency amount oriented P-E balanced VM migration method for realizing migration of a virtual machine in a cloud service, in view of the above deficiencies in the prior art.
In order to solve the technical problems, the technical scheme adopted by the invention is as follows: the VM migration method facing P-E balance of the seasonal non-stationary concurrency comprises two parts of VM static deployment and VM dynamic migration;
the VM static deployment part comprises the following steps:
step 1: n VMs are allocated in the cloud service data center, and the set of VMs is V ═ V1,v2,...,vnAnd if m physical servers, namely hosts, exist in the cloud data center, the host set is H ═ H1,h2,...,hm}; and the requirement that the resource applied by each VM is less than the resource provided by a single physical server is met;
step 1.1: setting the ith virtual machine viThe required memory resource and CPU resource are v respectivelyi m,vi c,i=1,…,n;
Step 1.2: setting the jth physical server hjThe owned memory resource and CPU resource are respectively positioned in hj m,hj c,j=1,…,m;
Step 2: setting a resource allocation matrix X, X ═ XijI is more than or equal to 1 and less than or equal to n, j is more than or equal to 1 and less than or equal to m, wherein x ij1 stands for the ith virtual machine viIs distributed to physical server hjIn, x ij0 denotes the ith virtual machine viIs not assigned to a physical server hjThe preparation method comprises the following steps of (1) performing;
and step 3: representing physical server h using identifier yjIs allocated a VM, thereby indicating whether the physical server is to enter a task execution state: let yjDenotes physical server h as 1jOne or more VMs have been deployed, and the physical server enters a task execution state; let y j0 denotes a physical server hjNo VM is deployed, so that a task execution state does not need to be entered;
and 4, step 4: according to the target and the purpose of VM static deployment, abstracting the problem of VM static deployment into a mathematical model, wherein the following formula is shown:
i∈{1,2,...,n},j∈{1,2,...,m},yj∈{0,1},xij∈{0,1}
wherein E isjRepresenting the energy consumption of deploying the VM by the jth physical server;
the aim of the VM static deployment is to deploy m or less VMs to a plurality of physical servers under the condition of meeting the basic requirements of customers; the purpose of VM static deployment is to reduce the starting number of the servers as much as possible under the condition of meeting the requirements of customers, thereby reducing the resource consumption of the physical servers and achieving the purpose of reducing energy consumption;
the VM dynamic migration part comprises the following steps:
step S1: selecting a VM needing to be migrated, namely setting a double threshold according to the average response time of the current VM, and classifying the running VMs according to the threshold, wherein the specific method comprises the following steps:
step S1.1: obtaining an average response time T for each running VMiAnd are sorted in a non-decreasing order;
step S1.2: calculating the maximum average response time T of each VMmaxThe following formula shows:
Tmax=Tmin×(1+α)
wherein, TminRepresenting the minimum average response time, wherein alpha is a number greater than zero and is set by a cloud service provider according to requirements;
step S1.3: average response time TiLess than TminPuts the VM of (1) into the running VM setSorting according to ascending order;
step S1.4: average response time TiGreater than TmaxPuts the VM of (1) into the running VM setAnd sorting in descending order;
step S1.3: using RBF algorithm to assemble VMAndpredicting the next average response time of each VM, and still enabling the prediction result to be smaller than TminAnd is greater than TmaxThe VM executes the steps S1.3-S1.4 again to obtain a VM set to be migratedAnd
step S2: selecting a target server; if all the VMs on a certain physical server are selected to be required to be migrated, the physical server is closed to save energy consumption, and when the placement of the VMs required to be migrated cannot be met in the physical server set in the running state, a new physical server is activated to perform resource expansion;
step S2.1: obtaining the energy consumption value of each physical server and putting the energy consumption value into the running physical server set PListEAnd according to a rule of notSorting in descending order;
step S2.2: sequentially fetching VM sets to be migratedIn each VM, the required resources of each VM are sequentially matched with PListEComparing the residual resources of the physical server, if the resource demand of the VM to be migrated is less than the residual resources of the physical server, migrating, and updating PListEAnd a resource allocation matrix X; if the resource requirement of the VM to be migrated is greater than or equal to the residual resources of the physical server, sequentially performing PListEComparing the residual resources of the physical servers until a migratable physical server is found, and updating the PListEAnd a resource allocation matrix X; if PListEIf all the physical servers in the system can not meet the resource requirement of the VM to be migrated, starting a new physical server, migrating the VM to the new physical server, and updating PListEAnd a resource allocation matrix X;
step S2.3: sequentially fetching VM sets to be migratedIn each VM, the required resources of each VM are sequentially matched with PListEComparing the residual resources of the physical server, if the resource demand of the VM to be migrated is less than the residual resources of the physical server, migrating, and updating PListEAnd a resource allocation matrix X; otherwise, sequentially mixing with PListEComparing the residual resources of the physical servers until a migratable physical server is found, and updating the PListEAnd a resource allocation matrix X; if PListEIf all the physical servers in the system can not meet the resource requirement of the VM to be migrated, starting a new physical server, migrating the VM to the new physical server, and updating PListEAnd a resource allocation matrix X.
Advantageous effects
Adopt the produced beneficial effect of above-mentioned technical scheme to lie in: according to the method for migrating the VMs based on the P-E balance of the seasonal unstable concurrency quantity, which is provided by the invention, the initial configuration of each VM in the server is carried out at high efficiency, and the VM and the target server are selected according to the average response time and the resource allocation matrix in a circulating manner, so that the balance between the performance and the energy consumption in a cloud service system is realized, and the high flexibility of a cloud computing environment is improved. In the process of aiming at the seasonal non-stable concurrency, the performance and the energy consumption are balanced, a reasonable VM migration strategy is generated, the defects of a traditional VM scheduling mode are overcome, the migration cost is reduced, and the cloud service system achieves load balance when providing services.
Drawings
FIG. 1 is a diagram of the CloudSim architecture provided by an embodiment of the present invention;
FIG. 2 is a flowchart of a method for VM migration that is oriented to P-E tradeoff of seasonal non-stationary concurrency according to an embodiment of the present invention;
fig. 3 is a diagram for comparing energy consumption of the VM migration method of the present invention with that of other VM migration methods provided by the embodiment of the present invention;
fig. 4 is a diagram comparing SLA default rates of the VM migration method of the present invention and other migration methods provided in the embodiment of the present invention.
Detailed Description
The following detailed description of embodiments of the present invention is provided in connection with the accompanying drawings and examples. The following examples are intended to illustrate the invention but are not intended to limit the scope of the invention.
In this embodiment, as shown in fig. 1, a CloudSim cloud computing simulation platform deployed in an HPZ820 workstation is taken as an example, and the VM migration method facing P-E tradeoff of seasonal non-stationary concurrency is used to migrate the virtual machine in the cloud computing center.
A P-E weighted VM migration method facing seasonal non-stationary concurrency quantity, as shown in fig. 2, includes two parts, i.e., VM static deployment and VM dynamic migration;
the VM static deployment part comprises the following steps:
step 1: n VMs are allocated in the cloud service data center, and the set of VMs is V ═ V1,v2,...,vnAnd if m physical servers, namely hosts, exist in the cloud data center, the host set is H ═ H1,h2,...,hm}; and the requirement that the resource applied by each VM is less than the resource provided by a single physical server is met;
step 1.1: setting the ith virtual machine viThe required memory resource and CPU resource are v respectivelyi m,vi c,i=1,…,n;
Step 1.2: setting a jth physical server hjThe owned memory resource and CPU resource are respectively positioned in hj m,hj c,j=1,…,m;
Step 2: setting a resource allocation matrix X, X ═ XijI is more than or equal to 1 and less than or equal to n, j is more than or equal to 1 and less than or equal to m, wherein x ij1 stands for the ith virtual machine viIs distributed to physical server hjIn, xij0 denotes the ith virtual machine viIs not assigned to a physical server hjPerforming the following steps;
and step 3: representing physical server h using identifier yjIs allocated a VM, thereby indicating whether the physical server is to enter a task execution state: let yjDenotes physical server h as 1jOne or more VMs have been deployed, and the physical server enters a task execution state; let y j0 denotes a physical server hjNo VM is deployed, so that a task execution state does not need to be entered;
and 4, step 4: the aim of VM static deployment is to deploy m or less VMs to a plurality of physical servers under the condition of meeting the basic requirements of customers, and the aim of VM static deployment is to reduce the starting number of the servers as much as possible under the condition of meeting the requirements of customers, thereby reducing the resource consumption of the physical servers and achieving the aim of reducing the energy consumption; therefore, according to the objective and purpose of VM static deployment, the problem of VM static deployment is abstracted into a mathematical model, as shown in the following formula:
i∈{1,2,...,n},j∈{1,2,...,m},yj∈{0,1},xij∈{0,1}
wherein E isjRepresenting the energy consumption of deploying the VM by the jth physical server;
the VM dynamic migration part comprises the following steps:
step S1: selecting a VM needing to be migrated, namely setting a double threshold according to the average response time of the current VM, and classifying the running VMs according to the threshold, wherein the specific method comprises the following steps:
step S1.1: obtaining an average response time T for each running VMiAnd are sorted in a non-decreasing order;
step S1.2: calculating the maximum average response time T of each VMmaxThe following formula shows:
Tmax=Tmin×(1+α)
wherein, TminRepresenting the minimum average response time, wherein alpha is a number greater than zero and is set by a cloud service provider according to requirements;
step S1.3: average response time TiLess than TminPuts the VM of (1) into the running VM setSorting according to ascending order;
step S1.4: average response time TiGreater than TmaxPuts the VM of (1) into the running VM setAnd sorting in descending order;
step S1.3: aggregating VMs using RBF algorithmAndpredicting the next average response time of each VM, and still enabling the prediction result to be smaller than TminAnd is greater than TmaxThe VM executes the steps S1.3-S1.4 again to obtain a VM set to be migratedAndthis is to avoid some only transient SLA violations by some VMs due to temporary load surges, resulting in unnecessary migration, additional migration costs and negative impact on other hosts due to VM migration.
Step S2: selecting a target server; if all VMs on a certain physical server are selected to need to be migrated, the physical server is shut down to save energy consumption, and the process can be understood as resource integration; resource expansion is opposite to resource integration, and when the placement of a VM needing migration cannot be met in a physical server set in a running state, a new physical server is activated to perform resource expansion;
step S2.1: obtaining the energy consumption value of each physical server and putting the energy consumption value into the running physical server set PListEAnd are sorted in a non-decreasing order; the physical servers are sorted according to the energy consumption value, so that the physical server with the largest energy consumption residual amount is selected, and the resource is enabled to have the largest integration effect.
Step S2.2: sequentially fetching VM sets to be migratedIn each VM, the required resources of each VM are sequentially matched with PListEComparing the residual resources of the physical server, if the resource demand of the VM to be migrated is less than the residual resources of the physical serverIf the residual resources are available, migration is carried out, and PList is updatedEAnd a resource allocation matrix X; if the resource requirement of the VM to be migrated is greater than or equal to the residual resources of the physical server, sequentially performing PListEComparing the residual resources of the physical servers until a migratable physical server is found, and updating the PListEAnd a resource allocation matrix X; if PListEIf all the physical servers can not meet the resource requirement of the VM to be migrated, starting a new physical server, migrating the VM to the new physical server, and updating PListEAnd a resource allocation matrix X;
step S2.3: sequentially fetching VM sets to be migratedIn each VM, the required resources of each VM are sequentially matched with PListEComparing the residual resources of the physical server, if the resource demand of the VM to be migrated is less than the residual resources of the physical server, migrating, and updating PListEAnd a resource allocation matrix X; otherwise, sequentially mixing with PListEComparing the residual resources of the physical servers until a migratable physical server is found, and updating the PListEAnd a resource allocation matrix X; if PListEIf all the physical servers in the system can not meet the resource requirement of the VM to be migrated, starting a new physical server, migrating the VM to the new physical server, and updating PListEAnd a resource allocation matrix X.
The CloudSim cloud computing simulation platform adopted in the embodiment not only supports modeling and simulation of a large cloud computing infrastructure, but also provides mapping from a physical host to a Virtual Machine (VM), can monitor resources, supports energy consumption modeling, and has the capability of simulating load change along with time change. The data center simulated in the embodiment has 60 physical servers, and table 1 shows the settings of relevant parameters of 6 types of servers. In addition, the number of VM requests received by the data center is 120.
Table 1 six types of physical server parameter settings
In this embodiment, the arrival of the user request concurrency amount is simulated on the CloudSim platform, and the VM static deployment strategy and the dynamic VM migration strategy are added on the CloudSim platform according to the method of the present invention. With the continuous change of the concurrency, the number of VMs executing the task is increased and decreased. In order to evaluate the performance of the VM migration strategy provided by the invention, the feasibility of the method provided by the invention is verified by taking the average energy consumption and the SLA default rate as evaluation indexes.
And loading the simulated seasonal non-stationary concurrency quantity to a VM executing a task in a physical server, and respectively executing deployment and migration of the VM by adopting the VM migration strategy, the IP algorithm and the NSGA-II migration strategy provided by the invention. Fig. 3 shows the VM migration strategy and the average energy consumption change condition of the IP and NSGA-II algorithms proposed by the method of the present invention as the VM number changes continuously.
As can be seen from fig. 3, when the number of VMs is less than 58, the IP algorithm with the lowest average power consumption is the lowest, because the IP algorithm can give a more accurate solution when the number of VMs is small. At this time, the average energy consumption obtained by the VM migration policy provided by the present invention is between NSGA-II and the IP algorithm, which also indicates that the solution accuracy of the VM migration policy provided by the present invention is between NSGA-II and the IP algorithm when the number of VMs is less than 58. However, as the number of VMs increases, the energy consumption generated by VM migration performed by the IP algorithm increases faster than that of the other two algorithms, and at this time, the energy consumption generated by executing the VM migration policy proposed by the present invention is significantly lower than that generated by executing the NSGA-II and IP algorithms, and the energy consumption increases more slowly. The calculation shows that the energy consumption generated by the execution of the strategy provided by the invention is averagely saved by 3.16 percent compared with that of NSGA-II; the average savings compared to the IP algorithm is 4.15%.
In the process of VM migration, the computing resources on the physical server also need to be migrated and integrated, and an SLA breach may occur when a VM executing a task in the physical machine cannot obtain sufficient memory resources and CPU resources. The VM load is a Web application request, the load type is a seasonal non-stationary concurrent request, and in order to ensure the user experience effect, the response time of the SLA request is set to be less than 0.25s in this embodiment. Fig. 4 shows the change of SLA default rate with the load per unit time increased by using the VM migration policy and IP algorithm, NSGA-II algorithm proposed by the present invention.
As shown in fig. 4, when the VM deployment and migration policy proposed by the present invention is used to execute tasks, the SLA default rates are lower than those of the IP and NSGA-II algorithms. The calculation shows that the SLA default rate generated by adopting the strategy provided by the invention is reduced by 10.09% on average compared with the IP algorithm and is reduced by 4.09% on average compared with the NSGA-II algorithm.
Finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it should be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some or all of the technical features may be equivalently replaced; such modifications and substitutions do not depart from the spirit of the corresponding technical solutions and scope of the present invention as defined in the appended claims.
Claims (3)
1. A method for migrating a VM facing P-E trade-off of seasonal non-stationary concurrency, characterized by: the method comprises two parts of VM static deployment and VM dynamic migration;
the VM static deployment part comprises the following steps:
step 1: n VMs are allocated in the cloud service data center, and the set of VMs is V ═ V1,v2,...,vnMeanwhile, m physical servers, namely hosts, exist in the cloud service data center, and the host set is H ═ H1,h2,...,hm}; and the resource applied by each VM is less than the resource provided by a single physical server;
step 1.1: setting the ith virtual machine viThe required memory resource and CPU resource are v respectivelyi m,vi c,i=1,…,n;
Step 1.2: setting a jth physical server hjOwned byMemory resource and CPU resource are respectively positioned in hj m,hj c,j=1,…,m;
Step 2: setting a resource allocation matrix X, X ═ XijI is more than or equal to 1 and less than or equal to n, j is more than or equal to 1 and less than or equal to m, wherein xij1 stands for the ith virtual machine viIs distributed to physical server hjIn, xij0 denotes the ith virtual machine viIs not assigned to a physical server hjPerforming the following steps;
and 3, step 3: representing physical server h using identifier yjIs allocated a VM, thereby indicating whether the physical server is to enter a task execution state: let yj1 denotes physical server hjOne or more VMs have been deployed, and the physical server enters a task execution state; let yj0 denotes physical server hjNo VM is deployed, so that a task execution state does not need to be entered;
and 4, step 4: according to the target and the purpose of VM static deployment, abstracting the problem of VM static deployment into a mathematical model, wherein the following formula is shown:
i∈{1,2,...,n},j∈{1,2,...,m},yj∈{0,1},xij∈{0,1}
wherein E isjRepresenting the energy consumption of deploying the VM by the jth physical server;
the VM dynamic migration part comprises the following steps:
step S1: selecting a VM that requires migration, i.e., based on the average of the current VMsSetting double thresholds according to the response time, and classifying the running VMs according to the thresholds to obtain a VM set to be migratedAnd
step S2: selecting a target server; if all the VMs on a certain physical server are selected to be required to be migrated, the physical server is closed to save energy consumption, and when the placement of the VMs required to be migrated cannot be met in the physical server set in the running state, a new physical server is activated to perform resource expansion;
step S2.1: obtaining the energy consumption value of each physical server and putting the energy consumption value into the running physical server set PListEAnd are sorted in a non-decreasing order;
step S2.2: sequentially fetching VM sets to be migratedIn each VM, the required resources of each VM are sequentially matched with PListEComparing the residual resources of the physical server, if the resource demand of the VM to be migrated is less than the residual resources of the physical server, migrating, and updating PListEAnd a resource allocation matrix X; if the resource requirement of the VM to be migrated is greater than or equal to the residual resources of the physical server, sequentially performing PListEComparing the residual resources of the physical servers until a migratable physical server is found, and updating the PListEAnd a resource allocation matrix X; if PListEIf all the physical servers in the system can not meet the resource requirement of the VM to be migrated, starting a new physical server, migrating the VM to the new physical server, and updating PListEAnd a resource allocation matrix X;
step S2.3: sequentially fetching VM sets to be migratedIn each VM, the required resources of each VM are sequentially matched with PListEComparing the residual resources of the physical server, if the resource demand of the VM to be migrated is less than the residual resources of the physical server, migrating, and updating PListEAnd a resource allocation matrix X; otherwise, sequentially mixing with PListEComparing the residual resources of the physical servers until a migratable physical server is found, and updating the PListEAnd a resource allocation matrix X; if PListEIf all the physical servers in the system can not meet the resource requirement of the VM to be migrated, starting a new physical server, migrating the VM to the new physical server, and updating PListEAnd a resource allocation matrix X.
2. The seasonal non-stationary concurrency oriented P-E weighted VM migration method of claim 1, wherein: the aim of the VM static deployment is to deploy m or less VMs to a plurality of physical servers under the condition of meeting the basic requirements of customers; the purpose of VM static deployment is to reduce the starting number of the servers as much as possible under the condition of meeting the requirements of customers, thereby reducing the resource consumption of the physical servers and achieving the purpose of reducing energy consumption.
3. The method of seasonal non-stationary concurrency oriented P-E weighted VM migration according to claim 1, wherein: the specific method of step S1 is as follows:
step S1.1: obtaining an average response time T for each running VMiAnd are sorted in a non-decreasing order;
step S1.2: calculating the maximum average response time T of each VMmaxThe following formula shows:
Tmax=Tmin×(1+α)
wherein, TminRepresenting the minimum average response time, wherein alpha is a number greater than zero and is set by a cloud service provider according to requirements;
step S1.3: average response time TiLess than TminPuts the running VM set into the VM ofCombination of Chinese herbsSorting according to ascending order;
step S1.4: average response time TiGreater than TmaxPut into a running VM setSorting according to descending order;
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910627610.7A CN110362383B (en) | 2019-07-12 | 2019-07-12 | P-E balanced VM migration method for seasonal non-stationary concurrency |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910627610.7A CN110362383B (en) | 2019-07-12 | 2019-07-12 | P-E balanced VM migration method for seasonal non-stationary concurrency |
Publications (2)
Publication Number | Publication Date |
---|---|
CN110362383A CN110362383A (en) | 2019-10-22 |
CN110362383B true CN110362383B (en) | 2022-06-24 |
Family
ID=68218967
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910627610.7A Active CN110362383B (en) | 2019-07-12 | 2019-07-12 | P-E balanced VM migration method for seasonal non-stationary concurrency |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110362383B (en) |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115150402B (en) * | 2022-08-30 | 2022-11-18 | 京华信息科技股份有限公司 | Cloud resource allocation method and system |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102236582A (en) * | 2011-07-15 | 2011-11-09 | 浙江大学 | Method for balanced distribution of virtualization cluster load in a plurality of physical machines |
CN106775987A (en) * | 2016-12-30 | 2017-05-31 | 南京理工大学 | A kind of dispatching method of virtual machine for improving resource efficiency safely in IaaS cloud |
Family Cites Families (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US8468230B2 (en) * | 2007-10-18 | 2013-06-18 | Fujitsu Limited | Method, apparatus and recording medium for migrating a virtual machine |
US8924560B2 (en) * | 2010-11-29 | 2014-12-30 | At&T Intellectual Property I, L.P. | Optimized game server relocation environment |
JP6819131B2 (en) * | 2016-08-18 | 2021-01-27 | 富士通株式会社 | Information processing equipment, information processing methods, information processing programs and information processing systems |
-
2019
- 2019-07-12 CN CN201910627610.7A patent/CN110362383B/en active Active
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102236582A (en) * | 2011-07-15 | 2011-11-09 | 浙江大学 | Method for balanced distribution of virtualization cluster load in a plurality of physical machines |
CN106775987A (en) * | 2016-12-30 | 2017-05-31 | 南京理工大学 | A kind of dispatching method of virtual machine for improving resource efficiency safely in IaaS cloud |
Non-Patent Citations (3)
Title |
---|
IaaS环境中一种保证SLA的资源调度策略;吴海双等;《计算机工程》;20130715(第07期);第51-54页 * |
一种迁移成本可感知的虚拟机整合算法;窦晖 等;《中国科技论文》;20170731;第12卷(第14期);第1640-1645、1669页 * |
云环境下基于冷点虚拟机迁移的热点消除方法;郭军 等;《清华大学学报(自然科学版)》;20161231;第56卷(第11期);第1232-1236页 * |
Also Published As
Publication number | Publication date |
---|---|
CN110362383A (en) | 2019-10-22 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Ranjbari et al. | A learning automata-based algorithm for energy and SLA efficient consolidation of virtual machines in cloud data centers | |
Ding et al. | Fault-tolerant elastic scheduling algorithm for workflow in cloud systems | |
Esfandiarpoor et al. | Structure-aware online virtual machine consolidation for datacenter energy improvement in cloud computing | |
Mann et al. | Remedy: Network-aware steady state VM management for data centers | |
CN107273185B (en) | Load balancing control method based on virtual machine | |
US9116181B2 (en) | Method, apparatus, and system for virtual cluster integration | |
Cao et al. | Dynamic VM consolidation for energy-aware and SLA violation reduction in cloud computing | |
CN105302630B (en) | A kind of dynamic adjusting method and its system of virtual machine | |
Sayadnavard et al. | A reliable energy-aware approach for dynamic virtual machine consolidation in cloud data centers | |
Xu et al. | Migration cost and energy-aware virtual machine consolidation under cloud environments considering remaining runtime | |
Zakarya et al. | An energy aware cost recovery approach for virtual machine migration | |
Al-Dulaimy et al. | Type-aware virtual machine management for energy efficient cloud data centers | |
Maurya et al. | Energy conscious dynamic provisioning of virtual machines using adaptive migration thresholds in cloud data center | |
Guzek et al. | HEROS: Energy-efficient load balancing for heterogeneous data centers | |
Sekhar et al. | A survey on energy efficient server consolidation through vm live migration | |
US20240036937A1 (en) | Workload placement for virtual gpu enabled systems | |
Razali et al. | Virtual machine migration implementation in load balancing for cloud computing | |
Mosa et al. | Dynamic virtual machine placement considering CPU and memory resource requirements | |
Liang et al. | A low-power task scheduling algorithm for heterogeneous cloud computing | |
Alnowiser et al. | Enhanced weighted round robin (EWRR) with DVFS technology in cloud energy-aware | |
Zakarya | An extended energy-aware cost recovery approach for virtual machine migration | |
Zhou et al. | Strategy optimization of resource scheduling based on cluster rendering | |
Lin et al. | Novel resource allocation model and algorithms for cloud computing | |
Xu et al. | Energy-Aware Resource Management in Cloud Computing Considering Load Balance. | |
CN110362383B (en) | P-E balanced VM migration method for seasonal non-stationary concurrency |
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 | ||
GR01 | Patent grant | ||
GR01 | Patent grant | ||
EE01 | Entry into force of recordation of patent licensing contract | ||
EE01 | Entry into force of recordation of patent licensing contract |
Application publication date: 20191022 Assignee: Shenyang Zhizhi Technology Co.,Ltd. Assignor: Northeastern University Contract record no.: X2023210000209 Denomination of invention: A VM Migration Method for P-E Balance of Seasonal Non stationary Concurrency Granted publication date: 20220624 License type: Common License Record date: 20231127 |