CN101593134B - Method and device for allocating CPU resources of virtual machine - Google Patents

Method and device for allocating CPU resources of virtual machine Download PDF

Info

Publication number
CN101593134B
CN101593134B CN2009100880212A CN200910088021A CN101593134B CN 101593134 B CN101593134 B CN 101593134B CN 2009100880212 A CN2009100880212 A CN 2009100880212A CN 200910088021 A CN200910088021 A CN 200910088021A CN 101593134 B CN101593134 B CN 101593134B
Authority
CN
China
Prior art keywords
virtual machine
cpu resource
current period
period
cpu
Prior art date
Application number
CN2009100880212A
Other languages
Chinese (zh)
Other versions
CN101593134A (en
Inventor
胡春明
李博
耿建平
钟亮
怀进鹏
Original Assignee
北京航空航天大学
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 北京航空航天大学 filed Critical 北京航空航天大学
Priority to CN2009100880212A priority Critical patent/CN101593134B/en
Publication of CN101593134A publication Critical patent/CN101593134A/en
Application granted granted Critical
Publication of CN101593134B publication Critical patent/CN101593134B/en

Links

Abstract

The invention discloses a method and a device for allocating CPU resources of a virtual machine. The method comprises the following steps of: monitoring operating conditions of all virtual machines in the current period to obtain the monitoring information which at least includes CPU resource utilization rates of all virtual machines in the current period and the throughputs applied to all virtual machines in the current period; according to the monitoring information, performing feedback adjustment on CPU resource request of all virtual machines in the next period by taking the target CPU resource utilization rate of all virtual machines and the target throughputs applied to all virtual machines as a target; according to the CPU resource request of all virtual machines, deciding the CPU resource allocations of all virtual machines in the next period on a virtual machine management control layer; and according to the CPU resource allocations of all virtual machines, allocating CPU resources for all virtual machines. The method adopts a dynamic mode to allocate CPU resources for all virtual machines, so that all virtual machines on one physical node can be coordinated to share the CPU resources.

Description

Virtual machine cpu resource distribution method and device
Technical field
The present invention relates to the virtual machine technique field, relate in particular to a kind of virtual machine cpu resource distribution method and device.
Background technology
In data center, move needed stabilized power source environment, suitable Temperature and Humidity Control and network bandwidth condition for ensureing server, need carry out centralized and unified management to a large amount of servers.Traditional server application deployment is often taked the mode of private server (Dedicated server).Under this mode, on same physical node, influence each other the assembly that a general physical node is only disposed single application or used to each other for avoiding a plurality of application deployments.But adverse consequencess such as the utilization factor of hardware resource is low excessively, the expense of purchase hardware resource is excessive that this mode will inevitably cause, the while has also increased the cost of management maintenance.
The appearance of virtual machine technique makes that a plurality of operation system examples can parallel running on same physical node, and for the application that is deployed in respectively wherein provides good performance to isolate, has led the trend of Server Consolidation.The application mode of this Server Consolidation has not only improved the resource utilization of server, has reduced the cost of purchasing hardware, and has reduced the loss of the energy and the human cost of management maintenance.
Virtual machine technique is meant on the complete computer hardware platforms of a cover, through the method for software or hardware, one group of hardware invented organize independently hardware more, for the software systems (normally operating system) of a plurality of independent operatings provide independently hardware running environment.Virtual machine technique needs one to be called virtual machine monitor (Virtual Machine Monitor hereinafter to be referred as software systems VMM), manages and dispatches the hardware resource of a plurality of operating systems.It can encapsulate, isolates, monitors and manage a plurality of virtual machine instance, makes on same the physical machine and can not move polytype operating system independently, mutually with conflicting.In addition, virtual machine technique is the distribution of software and disposes the new mode that provides.Software encapsulation in virtual machine image (disk file that has encapsulated the whole operation system is called virtual machine image), is utilized technology such as empty, as to intend machine deployment, migration, can under network environment, realize the distribution and the deployment of software neatly.
In enterprise-level server, the application program of deployment is mainly the final user based on network application service is provided, and this application has the resource requirement characteristic of dynamic change in time.In order to tackle the dynamic change of user's request, server is necessary for to use provides enough resources to ensure its service level.In the server field of having introduced virtual machine technique, how this problem then is converted into carries out resources allocation to a plurality of resource container-virtual machines on the single physical node and is deployed in single application in the virtual machine to the demand of resource with indirect guarantee.
Yet the static resource allocation mode that adopts at present can not satisfy application requirements.Because guarantee the service quality used, must distribute according to a maximum demand of using, a maximum demand of using on the one hand is uncertain, on the other hand, still will face the low excessively problem of resource utilization.
Summary of the invention
The purpose of this invention is to provide a kind of virtual machine cpu resource distribution method and device,, make a plurality of virtual machines on the physical node coordinate to share cpu resource to adopt dynamical fashion for each virtual machine distributes cpu resource.
For realizing above-mentioned purpose, the invention provides a kind of virtual machine cpu resource distribution method, comprising:
The ruuning situation of each virtual machine of monitoring current period obtains monitoring information, and monitoring information comprises at least: the handling capacity of using on the cpu resource utilization factor of each virtual machine of current period and each virtual machine of current period;
According to monitoring information, be aim parameter with the target throughput of using on the target cpu resource utilization factor of each virtual machine and each virtual machine, the cpu resource request amount of feedback adjustment each virtual machine of following one-period;
According to the cpu resource request amount of each virtual machine, in the make a strategic decision cpu resource sendout of each virtual machine of following one-period of Virtual Machine Manager key-course;
Based on the cpu resource sendout of each virtual machine, for each virtual machine distributes cpu resource.
Wherein, said according to said monitoring information, be aim parameter with the target throughput of using on the target cpu resource utilization factor of said each virtual machine and each virtual machine, the cpu resource request amount of feedback adjustment said each virtual machine of following one-period comprises:
The cpu resource utilization factor of current period virtual machine is 1 o'clock, according to formula u k + 1 = u k + λ 1 × v k t Ref × ( t Ref - t k ) The cpu resource request amount of feedback adjustment said each virtual machine of following one-period; Wherein, u kBe the cpu resource sendout of current period virtual machine, u K+1Be the cpu resource request amount of virtual machine of following one-period, v kBe the cpu resource use amount of current period virtual machine, t RefBe the target throughput of using on the current period virtual machine, t kBe the handling capacity of using on the current period virtual machine, λ 1Be that value is the variable between 0 to 2; Said cpu resource utilization factor does
The cpu resource utilization factor of current period virtual machine less than 1 and the current period virtual machine on the handling capacity used during less than said target throughput, according to formula u k + 1 = u k - λ 2 × v k r Ref × ( r Ref - v k u k ) The cpu resource request amount of feedback adjustment said each virtual machine of following one-period; Wherein, u kBe the cpu resource sendout of current period virtual machine, u K+1Be the cpu resource request amount of virtual machine of following one-period, v kBe the cpu resource use amount of current period virtual machine, r RefBe the target cpu resource utilization factor of virtual machine, λ 2Be that value is the variable between 0 to 2; Said cpu resource utilization factor does
The cpu resource utilization factor of current period virtual machine less than 1 and the current period virtual machine on the handling capacity used during more than or equal to said target throughput, the cpu resource request amount of feedback adjustment said virtual machine of following one-period is the cpu resource use amount of the said virtual machine of current period.
For realizing above-mentioned purpose, the invention provides a kind of virtual machine cpu resource distributor, comprising:
Monitoring modular, the ruuning situation that is used to monitor each virtual machine of current period obtains monitoring information, and monitoring information comprises at least: the handling capacity of using on the cpu resource utilization factor of each virtual machine of current period and each virtual machine of current period;
Feedback control module is used for according to monitoring information, is aim parameter with the target throughput of using on the target cpu resource utilization factor of each virtual machine and each virtual machine, the cpu resource request amount of feedback adjustment each virtual machine of following one-period;
Decision-making module is used for the cpu resource request amount according to each virtual machine, in the make a strategic decision cpu resource sendout of each virtual machine of following one-period of Virtual Machine Manager key-course;
Distribution module is used for the cpu resource sendout according to each virtual machine, for each virtual machine distributes cpu resource.
Wherein, said feedback control module comprises:
First feedback control module, the cpu resource utilization factor that is used at the current period virtual machine is 1 o'clock, according to formula u k + 1 = u k + λ 1 × v k t Ref × ( t Ref - t k ) The cpu resource request amount of feedback adjustment said each virtual machine of following one-period; Wherein, u kBe the cpu resource sendout of current period virtual machine, u K+1Be the cpu resource request amount of virtual machine of following one-period, v kBe the cpu resource use amount of current period virtual machine, t RefBe the target throughput of using on the current period virtual machine, t kBe the handling capacity of using on the current period virtual machine, λ 1Be that value is the variable between 0 to 2; Said cpu resource utilization factor does
Second feedback control module, be used for the cpu resource utilization factor of current period virtual machine less than 1 and the current period virtual machine on the handling capacity used during less than said target throughput, according to formula u k + 1 = u k - λ 2 × v k r Ref × ( r Ref - v k u k ) The cpu resource request amount of feedback adjustment said each virtual machine of following one-period; Wherein, u kBe the cpu resource sendout of current period virtual machine, u K+1Be the cpu resource request amount of virtual machine of following one-period, v kBe the cpu resource use amount of current period virtual machine, r RefBe the target cpu resource utilization factor of virtual machine, λ 2Be that value is the variable between 0 to 2; Said cpu resource utilization factor does
The 3rd feedback control module; Be used for the cpu resource utilization factor of current period virtual machine less than 1 and the current period virtual machine on the handling capacity used during more than or equal to said target throughput, the cpu resource request amount of feedback adjustment said virtual machine of following one-period is the cpu resource use amount of the said virtual machine of current period.
Virtual machine cpu resource distribution method of the present invention, device; After the handling capacity of using on the operating position of each virtual machine CPU of monitoring current period and each virtual machine; With the cpu resource utilization factor that monitors and the handling capacity of application is input quantity; With target cpu resource rate and target throughput is aim parameter, the cpu resource request amount of each virtual machine of feedback adjustment.With above-mentioned each cpu resource request amount as CPU sendout according to dynamic decision each virtual machine of following one-period; Thereby; Can make the collaborative cpu resource of sharing of a plurality of virtual machines on the physical node; Have stability and convergence preferably, and can distinguish the priority of different virtual machine, preferentially satisfy the resource request of high priority.
Description of drawings
Fig. 1 is the process flow diagram of deployed environment among the virtual machine cpu resource distribution method embodiment of the present invention;
Fig. 2 is the process flow diagram of virtual machine cpu resource distribution method embodiment of the present invention;
Fig. 3 is the process flow diagram of feedback adjustment among the virtual machine cpu resource distribution method embodiment of the present invention;
Fig. 4 is the structural representation of virtual machine cpu resource distributor embodiment of the present invention;
Fig. 5 is the structural representation of feedback control module among the virtual machine cpu resource distributor embodiment of the present invention;
Fig. 6 is the structural representation of decision-making module among the virtual machine cpu resource distributor embodiment of the present invention.
Embodiment
Through accompanying drawing and embodiment, technical scheme of the present invention is done further detailed description below.
The present invention is an example with the dummy machine system that apache2 (a kind of web server software) is installed on each virtual machine, and the solution of virtual machine cpu resource distribution method, device is provided.
Fig. 1 is the process flow diagram of deployed environment among the virtual machine cpu resource distribution method embodiment of the present invention.Using before this sends out virtual machine cpu resource distribution method preceding, needing to create virtual machine, in preliminary works such as virtual machine deploy application, as shown in Figure 1, the process of deployed environment is following among the virtual machine cpu resource distribution method embodiment of the present invention:
Step 10: monitor of virtual machine is installed, is created a plurality of virtual machines;
Monitor of virtual machine is installed, and is utilized the command interface of monitor of virtual machine to create a plurality of VME operating system mirror image instances, set name (vmname), IP address, memory size, priority (priority) and the weight (weight) of virtual machine.
Step 11: in the Virtual Machine Manager key-course, dispose monitoring modular, feedback control module and decision-making module;
Monitoring modular is carried out initialization, target setting cpu busy percentage and target throughput.Monitoring modular is used to monitor the ruuning situation of each virtual machine.Feedback control module is used for the virtual machine operation information of monitoring module monitors is carried out FEEDBACK CONTROL; Decision-making module is according to the CPU sendout of each virtual machine of the output of feedback control module decision-making.
Step 12: use apache2 in each virtual machine deploy;
Get into each virtual machine image, apache2 is installed, and start its status module and be used to monitor change in throughput.
Step 13: start each virtual machine.
Start each virtual machine, each virtual machine is added in the FEEDBACK CONTROL on the monitor of virtual machine automatically.After starting virtual machine, adopting virtual machine cpu resource distribution method of the present invention is each virtual machine distribution cpu resource in the operational process.
Fig. 2 is the process flow diagram of virtual machine cpu resource distribution method embodiment of the present invention, and as shown in Figure 2, this method comprises:
Step 20: the ruuning situation of each virtual machine of monitoring current period obtains monitoring information, and monitoring information comprises at least: the handling capacity of using on the cpu resource utilization factor of each virtual machine of current period and each virtual machine of current period;
The cpu resource utilization factor is and uses irrelevant common index that can not accurate performance guarantee be provided for application, and handling capacity is application oriented performance index, therefore the handling capacity with cpu resource utilization factor and application combines as the performance index of virtual machine.
Step 21: according to monitoring information, be aim parameter with the target throughput of using on the target cpu resource utilization factor of each virtual machine and each virtual machine, the cpu resource request amount of feedback adjustment each virtual machine of following one-period;
Target throughput to use on the target cpu resource utilization factor of each virtual machine and each virtual machine is an aim parameter, and monitoring information is carried out feedback adjustment, obtains the cpu resource request amount of each virtual machine of following one-period;
Handling capacity to use on the cpu resource utilization factor that monitors each virtual machine of current period and each virtual machine is the input quantity of FEEDBACK CONTROL; Target cpu resource utilization factor and target throughput with prior setting are the aim parameter of FEEDBACK CONTROL, and feedback adjustment is exported the cpu resource request amount of each virtual machine of following one-period.Concrete feedback procedure can be as shown in Figure 3, and Fig. 3 is the process flow diagram of feedback adjustment among the virtual machine cpu resource distribution method embodiment of the present invention.
Step 22: according to the cpu resource request amount of each virtual machine, in the make a strategic decision cpu resource sendout of each virtual machine of following one-period of Virtual Machine Manager key-course;
When the defining virtual machine, the given priority and the weight of a virtual machine, when being used to resource contention occur, mechanism provides decision-making foundation for resource allocation decisions.When the cpu resource request amount summation of each virtual machine during, satisfy the cpu resource request of each virtual machine of following one-period less than actual cpu resource amount; When the cpu resource request amount summation of each virtual machine during more than or equal to actual cpu resource amount; Under the situation of the minimum cpu resource request amount that satisfies each virtual machine; According to the cpu resource sendout of high priority principle of priority decision-making each virtual machine of following one-period, the virtual machine of same priority is according to the weight proportion Resources allocation.
Step 23: according to the cpu resource sendout of each virtual machine, for each virtual machine distributes cpu resource.
As shown in Figure 3, the feedback adjustment strategy is following among the virtual machine cpu resource distribution method embodiment of the present invention:
Wherein, u kBe the cpu resource sendout of current period virtual machine, u K+1Be the cpu resource request amount of virtual machine of following one-period, v kBe the cpu resource use amount of current period virtual machine, t RefBe the target throughput of using on the virtual machine, t kBe the handling capacity of using on the current period virtual machine, r RefBe the target cpu resource utilization factor of virtual machine, λ 2Be that value is the variable between 0 to 2, along with t kVariation λ 2Get different value, λ 1Be that value is the variable between 0 to 2, along with v k/ u kVariation λ 1Get different value.
The cpu resource utilization factor of current period virtual machine is 1 o'clock, explains that virtual machine is in overload,, in such cases, further judge the handling capacity t that uses on the current period virtual machine kWhether less than target throughput t Ref, work as t kLess than t RefThe time, λ 1Get 0.5; Work as t kGreater than t RefThe time, λ 1Get 1.Then, according to formula u k + 1 = u k + λ 1 × v k t Ref × ( t Ref - t k ) The cpu resource request amount of feedback adjustment said each virtual machine of following one-period.
The cpu resource utilization factor of current period virtual machine less than 1 and the current period virtual machine on the handling capacity used less than target throughput t RefThe time, whether the cpu resource utilization factor of further judging the current period virtual machine is less than target cpu resource utilization factor, as (v k/ u k) less than r RefThe time, virtual machine is in light condition, λ 2Get 0.5; As (v k/ u k) greater than r RefThe time, virtual machine is in heavy condition, λ 2Get 0.8.Then, according to formula u k + 1 = u k - λ 2 × v k r Ref × ( r Ref - v k u k ) The cpu resource request amount of feedback adjustment each virtual machine of following one-period.
The cpu resource utilization factor of current period virtual machine less than 1 and the current period virtual machine on the handling capacity used more than or equal to target throughput t RefThe time, the cpu resource request amount of feedback adjustment virtual machine of following one-period is the cpu resource use amount of current period virtual machine.
Virtual machine cpu resource distribution method of the present invention; After the handling capacity of on the operating position of each virtual machine CPU of client layer monitoring current period and each virtual machine, using; With the cpu resource utilization factor that monitors and the handling capacity of application is input quantity; With target cpu resource rate and target throughput is aim parameter, and feedback adjustment goes out the cpu resource request amount of each virtual machine.With above-mentioned each cpu resource request amount as CPU sendout according to dynamic decision each virtual machine of following one-period; Thereby; Can make the collaborative cpu resource of sharing of a plurality of virtual machines on the physical node; Have stability and convergence preferably, and can distinguish the priority of different virtual machine, preferentially satisfy the resource request of high priority.
Fig. 4 is the structural representation of virtual machine cpu resource distributor embodiment of the present invention, and as shown in Figure 4, this device comprises: monitoring modular 410, feedback control module 420, decision-making module 430 and distribution module 440.
The ruuning situation of monitoring modular 410 each virtual machine of monitoring current period obtains monitoring information, and monitoring information comprises at least: the handling capacity of using on the cpu resource utilization factor of each virtual machine of current period and each virtual machine of current period.The monitoring information that feedback control module 420 obtains according to monitoring modular is an aim parameter with the target throughput of using on the target cpu resource utilization factor of each virtual machine and each virtual machine, the cpu resource request amount of feedback adjustment each virtual machine of following one-period.Decision-making module 430 is according to the cpu resource request amount of each virtual machine of feedback control module output, and the priority of each virtual machine of setting, in the make a strategic decision cpu resource sendout of each virtual machine of following one-period of Virtual Machine Manager key-course.
Fig. 5 is the structural representation of feedback control module among the virtual machine cpu resource distributor embodiment of the present invention, and as shown in Figure 5, feedback control module 420 comprises: first feedback control module 421, second feedback control module 422 and the 3rd feedback control module 423.
After the cpu resource utilization factor that first feedback control module is judged the current period virtual machine is 1, explain that virtual machine is in overload, further judge the handling capacity t that uses on the current period virtual machine kWhether less than target throughput t Ref, work as t kLess than t RefThe time, set λ 1Be 0.5; Work as t kGreater than t RefThe time, set λ 1Be 1.Then, first feedback control module is according to formula u k + 1 = u k + λ 1 × v k t Ref × ( t Ref - t k ) The cpu resource request amount of feedback adjustment said each virtual machine of following one-period.Wherein, u kBe the cpu resource sendout of current period virtual machine, u K+1Be the cpu resource request amount of virtual machine of following one-period, v kBe the cpu resource use amount of current period virtual machine, t RefBe the target throughput of using on the virtual machine, t kBe the handling capacity of using on the current period virtual machine, λ 1Be that value is the variable between 0 to 2, along with t kVariation get different value.
The cpu busy percentage of judging the current period virtual machine in second feedback control module less than 1 and the current period virtual machine on the handling capacity used less than said target throughput after; Whether the cpu resource utilization factor of further judging the current period virtual machine is less than target cpu resource utilization factor, as (v k/ u k) less than r RefThe time, virtual machine is in light condition, λ 2Get 0.5; As (v k/ u k) greater than r RefThe time, virtual machine is in heavy condition, λ 2Get 0.8.Then, second feedback control module is according to formula u k + 1 = u k - λ 2 × v k r Ref × ( r Ref - v k u k ) The cpu resource request amount of feedback adjustment each virtual machine of following one-period.Wherein, u kBe the cpu resource sendout of current period virtual machine, u K+1Be the cpu resource request amount of virtual machine of following one-period, v kBe the cpu resource use amount of current period virtual machine, r RefBe the target cpu resource utilization factor of virtual machine, λ 2Be that value is the variable between 0 to 2, along with v k/ u kVariation get different value;
The cpu busy percentage of current period virtual machine less than 1 and the current period virtual machine on the handling capacity used during more than or equal to said target throughput, the 3rd following one-period of control module feedback adjustment the cpu resource request amount of said virtual machine be the cpu resource use amount of the said virtual machine of current period.
The structural representation of decision-making module among Fig. 6 virtual machine cpu resource of the present invention distributor embodiment, as shown in Figure 6, decision-making module 430 comprises: first decision-making module 431 and second decision-making module 432.When the cpu resource request amount summation of each virtual machine during less than actual cpu resource amount, first decision-making module satisfies the cpu resource request of each virtual machine of following one-period.When the cpu resource request amount summation of said each virtual machine during more than or equal to actual cpu resource amount; Under the situation of the minimum cpu resource request amount that satisfies said each virtual machine, second decision-making module is according to the cpu resource sendout of high priority principle of priority decision-making said each virtual machine of following one-period.
Virtual machine cpu resource distributor of the present invention; Through monitoring modular after the handling capacity of using on the operating position of each virtual machine CPU of client layer monitoring current period and each virtual machine; With the cpu resource utilization factor that monitors and the handling capacity of application is the FEEDBACK CONTROL input quantity; With target cpu resource rate and target throughput is the FEEDBACK CONTROL aim parameter, and feedback adjustment goes out the cpu resource request amount of each virtual machine.With above-mentioned each cpu resource request amount as foundation; The CPU sendout of decision module dynamic decision each virtual machine of following one-period; Thereby, can make the collaborative cpu resource of sharing of a plurality of virtual machines on the physical node, have stability and convergence preferably; And can distinguish the priority of different virtual machine, preferentially satisfy the resource request of high priority.
What should explain at last is: above embodiment is only in order to technical scheme of the present invention to be described but not limit it; Although the present invention has been carried out detailed explanation with reference to preferred embodiment; Those of ordinary skill in the art is to be understood that: it still can make amendment or be equal to replacement technical scheme of the present invention, also can not make amended technical scheme break away from the spirit and the scope of technical scheme of the present invention and these are revised or be equal to replacement.

Claims (8)

1. a virtual machine cpu resource distribution method is characterized in that, comprising:
The ruuning situation of each virtual machine of monitoring current period obtains monitoring information, and said monitoring information comprises at least: the handling capacity of using on the cpu resource utilization factor of each virtual machine of current period and each virtual machine of current period;
According to monitoring information, be aim parameter with the target throughput of using on the target cpu resource utilization factor of said each virtual machine and said each virtual machine, the cpu resource request amount of feedback adjustment said each virtual machine of following one-period;
According to the cpu resource request amount of said each virtual machine, in the make a strategic decision cpu resource sendout of said each virtual machine of following one-period of Virtual Machine Manager key-course;
According to the cpu resource sendout of said each virtual machine, for said each virtual machine distributes cpu resource;
Said according to said monitoring information, be aim parameter with the target throughput of using on the target cpu resource utilization factor of said each virtual machine and each virtual machine, the cpu resource request amount of feedback adjustment said each virtual machine of following one-period comprises:
The cpu resource utilization factor of current period virtual machine is 1 o'clock, according to formula The cpu resource request amount of feedback adjustment said each virtual machine of following one-period; Wherein, u kBe the cpu resource sendout of current period virtual machine, u K+1Be the cpu resource request amount of virtual machine of following one-period, v kBe the cpu resource use amount of current period virtual machine, t RefBe the target throughput of using on the current period virtual machine, t kBe the handling capacity of using on the current period virtual machine, λ 1Be that value is the variable between 0 to 2; Said cpu resource utilization factor does
The cpu resource utilization factor of current period virtual machine less than 1 and the current period virtual machine on the handling capacity used during less than said target throughput, according to formula The cpu resource request amount of feedback adjustment said each virtual machine of following one-period; Wherein, u kBe the CPU resources allocation amount of current period virtual machine, u K+1Be the cpu resource request amount of virtual machine of following one-period, v kBe the cpu resource use amount of current period virtual machine, r RefBe the target cpu resource utilization factor of virtual machine, λ 2Be that value is the variable between 0 to 2; Said cpu resource utilization factor does
The cpu resource utilization factor of current period virtual machine less than 1 and the current period virtual machine on the handling capacity used during more than or equal to said target throughput, the cpu resource request amount of feedback adjustment said virtual machine of following one-period is the cpu resource use amount of the said virtual machine of current period.
2. virtual machine cpu resource distribution method according to claim 1; It is characterized in that; The cpu busy percentage of said current period virtual machine is 1 o'clock; Cpu resource request amount according to formula feedback adjustment said each virtual machine of following one-period comprises:
Judge that the handling capacity of using on the current period virtual machine whether less than said target throughput, sets said formula according to judged result Middle λ 1Value.
3. virtual machine cpu resource distribution method according to claim 1; It is characterized in that; The cpu resource utilization factor of said current period virtual machine less than 1 and the current period virtual machine on the handling capacity used during less than said target throughput; Cpu resource request amount according to formula feedback adjustment said each virtual machine of following one-period comprises:
Whether the cpu resource utilization factor of judging the current period virtual machine according to judged result sets said formula less than said target cpu resource utilization factor Middle λ 2Value.
4. virtual machine cpu resource distribution method according to claim 1 is characterized in that, said cpu resource request amount according to said each virtual machine in the make a strategic decision cpu resource sendout of said each virtual machine of following one-period of Virtual Machine Manager key-course, comprising:
When the cpu resource request amount summation of said each virtual machine during, satisfy the cpu resource request amount of said each virtual machine of following one-period less than actual cpu resource amount;
When the cpu resource request amount summation of said each virtual machine during more than or equal to said actual cpu resource amount; Under the situation of the minimum cpu resource request amount that satisfies said each virtual machine, according to the make a strategic decision cpu resource sendout of said each virtual machine of following one-period of high priority virtual machine priority principle.
5. according to the said virtual machine cpu resource of claim 1 distribution method, it is characterized in that, before the ruuning situation of said each virtual machine of monitoring current period obtains monitoring information, also comprise:
Monitor of virtual machine is installed, is created a plurality of virtual machines;
In the Virtual Machine Manager key-course, dispose monitoring modular, feedback control module and decision-making module;
Use in each virtual machine deploy;
Start each virtual machine.
6. a virtual machine cpu resource distributor is characterized in that, comprising:
Monitoring modular, the ruuning situation that is used to monitor each virtual machine of current period obtains monitoring information, and said monitoring information comprises at least: the handling capacity of using on the cpu resource utilization factor of each virtual machine of current period and each virtual machine of current period;
Feedback control module is used for according to monitoring information, is aim parameter with the target throughput of using on the target cpu resource utilization factor of said each virtual machine and said each virtual machine, the cpu resource request amount of feedback adjustment said each virtual machine of following one-period;
Decision-making module is used for the cpu resource request amount according to said each virtual machine, in the make a strategic decision cpu resource sendout of said each virtual machine of following one-period of Virtual Machine Manager key-course;
Distribution module is used for the cpu resource sendout according to said each virtual machine, for said each virtual machine distributes cpu resource;
Said feedback control module comprises:
First feedback control module, the cpu resource utilization factor that is used at the current period virtual machine is 1 o'clock, according to formula The cpu resource request amount of feedback adjustment said each virtual machine of following one-period; Wherein, u kBe the cpu resource sendout of current period virtual machine, u K+1Be the cpu resource request amount of virtual machine of following one-period, v kBe the cpu resource use amount of current period virtual machine, t RefBe the target throughput of using on the current period virtual machine, t kBe the handling capacity of using on the current period virtual machine, λ 1Be that value is the variable between 0 to 2; Said cpu resource utilization factor does
Second feedback control module, be used for the cpu resource utilization factor of current period virtual machine less than 1 and the current period virtual machine on the handling capacity used during less than said target throughput, according to formula The cpu resource request amount of feedback adjustment said each virtual machine of following one-period; Wherein, u kBe the cpu resource sendout of current period virtual machine, u K+1Be the cpu resource request amount of virtual machine of following one-period, v kBe the cpu resource use amount of current period virtual machine, r RefBe the target cpu resource utilization factor of virtual machine, λ 2Be that value is the variable between 0 to 2; Said cpu resource utilization factor does
The 3rd feedback control module; Be used for the cpu resource utilization factor of current period virtual machine less than 1 and the current period virtual machine on the handling capacity used during more than or equal to said target throughput, the cpu resource request amount of feedback adjustment said virtual machine of following one-period is the cpu resource use amount of the said virtual machine of current period.
7. according to the said virtual machine cpu resource of claim 6 distributor; It is characterized in that; Said first feedback control module also is used for: the cpu resource utilization factor of current period virtual machine is 1 o'clock; Judge that the handling capacity of using on the current period virtual machine whether less than said target throughput, sets said formula according to judged result Middle λ 1Value;
Said second feedback control module also is used for: the cpu busy percentage of current period virtual machine less than 1 and the current period virtual machine on the handling capacity used during less than said target throughput; Whether the cpu busy percentage of judging the current period virtual machine according to judged result sets said formula less than said target cpu resource utilization factor Middle λ 2Value.
8. according to the said virtual machine cpu resource of claim 6 distributor, it is characterized in that said decision-making module comprises:
First decision-making module is used for satisfying the cpu resource request of said each virtual machine of following one-period when the cpu resource request amount summation of said each virtual machine during less than actual cpu resource amount;
Second decision-making module; Be used for when the cpu resource request amount summation of said each virtual machine during more than or equal to actual cpu resource amount; Under the situation of the minimum cpu resource request amount that satisfies said each virtual machine, according to the cpu resource sendout of high priority principle of priority decision-making said each virtual machine of following one-period.
CN2009100880212A 2009-06-29 2009-06-29 Method and device for allocating CPU resources of virtual machine CN101593134B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN2009100880212A CN101593134B (en) 2009-06-29 2009-06-29 Method and device for allocating CPU resources of virtual machine

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN2009100880212A CN101593134B (en) 2009-06-29 2009-06-29 Method and device for allocating CPU resources of virtual machine

Publications (2)

Publication Number Publication Date
CN101593134A CN101593134A (en) 2009-12-02
CN101593134B true CN101593134B (en) 2012-05-30

Family

ID=41407803

Family Applications (1)

Application Number Title Priority Date Filing Date
CN2009100880212A CN101593134B (en) 2009-06-29 2009-06-29 Method and device for allocating CPU resources of virtual machine

Country Status (1)

Country Link
CN (1) CN101593134B (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105843678A (en) * 2016-03-17 2016-08-10 国网山东省电力公司信息通信公司 Optimization theory-based theory and system for scheduling and migrating resources of virtual machine

Families Citing this family (24)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101986284B (en) * 2010-11-04 2012-05-23 华中科技大学 Dynamic recovery system for waste space of virtual machine image file
CN102141930A (en) * 2010-12-20 2011-08-03 华为技术有限公司 Method and device for distributing resources to virtual machines
CN102654841A (en) * 2011-03-02 2012-09-05 中国电信股份有限公司 Method and device for allocating computing resource of virtual machine based on fine granularity
CN102681899B (en) * 2011-03-14 2015-06-10 金剑 Virtual computing resource dynamic management system of cloud computing service platform
EP2698711B1 (en) 2011-06-30 2015-08-05 Huawei Technologies Co., Ltd. Method for dispatching central processing unit of hotspot domain virtual machine and virtual machine system
US9176788B2 (en) * 2011-08-16 2015-11-03 Esds Software Solution Pvt. Ltd. Method and system for real time detection of resource requirement and automatic adjustments
CN102521052B (en) * 2011-12-06 2013-12-25 北京航空航天大学 Resource allocation method of virtualized data center and virtual machine monitor
CN103186450A (en) * 2011-12-28 2013-07-03 英业达集团(天津)电子技术有限公司 Method and system for monitoring virtual resource consumption
CN103530185B (en) * 2012-07-02 2018-12-04 南京中兴新软件有限责任公司 Method for optimizing resources and device
CN103235735A (en) * 2012-11-16 2013-08-07 佳都新太科技股份有限公司 Resource control model based on utilization rate in XEN virtual machine system architecture
CN103885812B (en) 2012-12-21 2018-03-27 华为技术有限公司 Virtual machine specification method of adjustment and device
WO2014121485A1 (en) * 2013-02-07 2014-08-14 华为技术有限公司 Method and system for managing virtual machines
CN103246570A (en) * 2013-05-20 2013-08-14 百度在线网络技术(北京)有限公司 Hadoop scheduling method and system and management node
CN103336722B (en) * 2013-07-16 2017-09-26 上海大学 A kind of CPU resources of virtual machine monitoring and dynamic allocation method
CN104331328B (en) * 2013-07-22 2018-06-12 中国电信股份有限公司 Schedule virtual resources method and schedule virtual resources device
CN104518897A (en) * 2013-09-30 2015-04-15 中国电信股份有限公司 Resource management optimization processing method and resource management optimization processing device for virtual firewalls
CN103678151A (en) * 2013-12-20 2014-03-26 汉柏科技有限公司 Configuration method and configuration device of block storage device of virtual machine
CN105320559B (en) * 2014-07-30 2019-02-19 中国移动通信集团广东有限公司 A kind of dispatching method and device of cloud computing system
CN104102531B (en) * 2014-08-06 2017-11-14 汉柏科技有限公司 Multi-internet integration system and method based on virtual machine
CN104281496B (en) * 2014-10-27 2018-02-02 深圳中兴网信科技有限公司 The distribution method and distributor of virtual resource
CN104702680B (en) * 2015-02-13 2019-01-04 深圳易科声光科技股份有限公司 A kind of monitoring resource method and system
CN106201655B (en) * 2016-06-30 2019-07-02 中国联合网络通信集团有限公司 Virtual machine distribution method and virtual machine distribution system
CN109918198B (en) * 2019-02-18 2020-02-11 中国空间技术研究院 Simulation cloud platform load scheduling system and method based on user characteristic prediction
CN110764900A (en) * 2019-09-06 2020-02-07 深圳壹账通智能科技有限公司 Data distribution test method and device under high concurrency, terminal and storage medium

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101183315A (en) * 2007-12-10 2008-05-21 华中科技大学 Paralleling multi-processor virtual machine system
CN101403983A (en) * 2008-11-25 2009-04-08 北京航空航天大学 Resource monitoring method and system for multi-core processor based on virtual machine

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101183315A (en) * 2007-12-10 2008-05-21 华中科技大学 Paralleling multi-processor virtual machine system
CN101403983A (en) * 2008-11-25 2009-04-08 北京航空航天大学 Resource monitoring method and system for multi-core processor based on virtual machine

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
陆晓雯.《虚拟机资源监测调整机制研究》.《华中科技大学硕士学位论文》.2008,全文. *

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105843678A (en) * 2016-03-17 2016-08-10 国网山东省电力公司信息通信公司 Optimization theory-based theory and system for scheduling and migrating resources of virtual machine

Also Published As

Publication number Publication date
CN101593134A (en) 2009-12-02

Similar Documents

Publication Publication Date Title
US10789106B2 (en) Server consolidation using virtual machine resource tradeoffs
US20200334023A1 (en) Self-moving operating system installation in cloud-based network
US20210034432A1 (en) Virtual systems management
US11003502B2 (en) Resource manager for managing the sharing of resources among multiple workloads in a distributed computing environment
US10635496B2 (en) Thread pool management
US9661071B2 (en) Apparatus, systems and methods for deployment and management of distributed computing systems and applications
US9813335B2 (en) System, method, and computer program for augmenting a physical system utilizing a network function virtualization orchestrator (NFV-O)
Qu et al. Delay-aware scheduling and resource optimization with network function virtualization
US9647955B2 (en) Systems, methods, and devices for dynamic resource monitoring and allocation in a cluster system
RU2648610C2 (en) System and method of zoning in software defined networks
CN103810023B (en) The intelligent deployment method of Distributed Application and system in a kind of cloud platform
US9806979B1 (en) System, method, and computer program for optimizing a chain of virtual network functions in a network based on network function virtualization (NFV)
US8978032B2 (en) Host naming application programming interface
US9843531B2 (en) Asynchronous framework for management of IaaS
CN105740074B (en) A kind of virtual machine load-balancing method based on cloud computing
WO2018028581A1 (en) Statement regarding federally sponsored research or development
US10505796B2 (en) Network function virtualization
US9158586B2 (en) Systems and methods for managing cloud computing resources
JP5544967B2 (en) Virtual machine management program and virtual machine management apparatus
US9563479B2 (en) Brokering optimized resource supply costs in host cloud-based network using predictive workloads
US8190740B2 (en) Systems and methods for dynamically provisioning cloud computing resources
US9229711B2 (en) Optimizing on demand allocation of virtual machines using a stateless preallocation pool
US10102018B2 (en) Introspective application reporting to facilitate virtual machine movement between cloud hosts
US8141090B1 (en) Automated model-based provisioning of resources
US8799895B2 (en) Virtualization-based resource management apparatus and method and computing system for virtualization-based resource management

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
C14 Grant of patent or utility model
GR01 Patent grant
CF01 Termination of patent right due to non-payment of annual fee
CF01 Termination of patent right due to non-payment of annual fee

Granted publication date: 20120530

Termination date: 20170629