CN102508714A - Green-computer-based virtual machine scheduling method for cloud computing - Google Patents

Green-computer-based virtual machine scheduling method for cloud computing Download PDF

Info

Publication number
CN102508714A
CN102508714A CN2011103425851A CN201110342585A CN102508714A CN 102508714 A CN102508714 A CN 102508714A CN 2011103425851 A CN2011103425851 A CN 2011103425851A CN 201110342585 A CN201110342585 A CN 201110342585A CN 102508714 A CN102508714 A CN 102508714A
Authority
CN
China
Prior art keywords
virtual machine
task
energy consumption
main frame
unit
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN2011103425851A
Other languages
Chinese (zh)
Inventor
程春玲
徐小龙
潘钰
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Nanjing Post and Telecommunication University
Nanjing University of Posts and Telecommunications
Original Assignee
Nanjing Post and Telecommunication University
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 Nanjing Post and Telecommunication University filed Critical Nanjing Post and Telecommunication University
Priority to CN2011103425851A priority Critical patent/CN102508714A/en
Publication of CN102508714A publication Critical patent/CN102508714A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
    • Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management

Landscapes

  • Power Sources (AREA)

Abstract

The invention discloses a green-computing-based virtual machine scheduling method for cloud computing, and belongs to the field of computer networks. The concept of green computing is applied to the cloud computing for virtual machine scheduling. The method comprises the following steps of: sequencing tasks according to priorities, estimating the total energy consumption of each task for each virtual machine, scheduling a virtual machine corresponding to the minimum energy consumption of each task, periodically monitoring a load of a host, and regulating a scheduling plan according to the load. By the method, energy consumption for the virtual machine scheduling is reduced, energy load equalization is taken into account, and the green computing is reflected.

Description

The dispatching method of virtual machine that calculates based on green in a kind of cloud computing
Technical field
The present invention relates to the dispatching method of virtual machine in a kind of cloud computing, and relate to the green thought of calculating, can reduce system energy consumption, belong to Distributed Calculation and applications of computer network field.
Background technology
Cloud computing is one of focus of commercial both at home and abroad at present and scientific research institution's research, and it is distributed in calculation task on the resource pool of great amount of calculation mechanism one-tenth, makes various application systems can obtain computing power, storage space and information service as required.Be accompanied by the growth to computation requirement, the information technoloy equipment in the cloud is more and more, and scale is increasing, and the energy consumption cost problem is also more and more obvious.According to Environmental Protection Agency (US EPA) report, U.S. data center had consumed the electric weight up to 61,000,000,000 kilowatts altogether in 2006.On the other hand, the data from the environmental protection tissue show that the discharge capacity of data center's carbon is also more with the discharge capacity of two countries of Argentina than Holland in the our times scope, accounts for 2% of global CO2 emissions.
Under energy consumption cost and carbon emission amount surprising situation like this, the users of computing machine and server expect a kind of new technology of green energy conservation---green calculating.At present green being calculated does not also have uniform definition, wikipedia to think, green calculating refers to utilize various software/hardware advanced technologies; The operating load of present great amount of calculation machine system is reduced, improve its operation efficiency (like the flop/watt index), reduce computer system quantity; Further reduce system support power supply energy consumption; Simultaneously, improve the design of computer system, improve its resource utilization and the recovery; Reduce carbon dioxide/greenhouse gas emission, thereby reach the purpose of energy-saving and environmental protection and saving.Green calculating is also uncomplicated in fact, and it has mainly embodied theory energy-saving and cost-reducing, the protection environment.
Scheduling virtual machine under the cloud environment is implemented in application layer and virtual resource layer.Cloud computing adopts Intel Virtualization Technology that the mapping resources of main frame is arrived virtual machine layer, carries out user's task.The scheduling virtual machine problem is to be that principle is shone upon with task and resource with certain optimization aim; Cloud computing mechanism has been simplified the coupling of task and resource; Make the embodied of required by task resource with a virtual machine, then the resource searching process is encapsulated as the process of search virtual machine.The scheduling of virtual machine on physical machine has based on priority, based on user expectation with based on the scheduling of load balancing.For example; One piece of Chinese invention patent document (application number is 200910241371.8, and Granted publication number is CN 101706743 A) discloses the dispatching method of virtual machine under a kind of multi-core environment, and this method is when system start-up, cpu resource to be carried out subregion according to the scheduling strategy type; And the cpu load situation of each subregion of monitoring in real time when moving in system; Dynamically adjust the size of subregion cpu resource, and use the virtual machine of identical scheduling strategy in same subregion, to dispatch, improved the efficient of scheduling; Reach the purpose of balancing resource load through the size of dynamic adjustment subregion cpu resource; Realized making full use of of resource, reduced the waste of resource, but the task of not considering is carried out the problem that consumes energy consumption.
The present invention mainly considers the scheduling virtual machine problem the cloud computing from energy-conservation angle, embodies the green theory of calculating.Task scheduling is carried out to virtual machine, needed to be transferred on the computer equipment, so considered the transmission energy consumption among the present invention through network interconnection device; Because computer equipment consumes electric power resource, the present invention has considered the calculating energy consumption.And, carry out scheduling virtual machine calculating energy consumption and the leading indicator transmission energy consumption and the green calculating of conduct.
Summary of the invention
The objective of the invention is the thought that green is calculated is applied to carry out scheduling virtual machine in the cloud computing, proposed the dispatching method of virtual machine that calculates based on green in a kind of cloud computing.
Method of the present invention mainly is at first task according to priority to be sorted; Obtain task list
Figure 2011103425851100002DEST_PATH_IMAGE001
, m representes current general assignment number.The calculating energy consumption of estimation tasks on all available virtual machines and transmission energy consumption sum then; Select the least energy consumption corresponding virtual machine of each task to dispatch; And periodically monitoring host computer load, the adjustment operation plan, the task of treating all executes the back and destroys virtual machine.
The dispatching method of virtual machine detailed design of calculating based on green in a kind of cloud computing:
In cloud computing environment, green calculating mainly embodied through cutting down the consumption of energy.Energy consumption is meant the total amount of system at a period of time internal consumption electric power resource.Energy consumption in the cloud computing is mainly reflected on computer equipment, network interconnection device and the memory device.The energy consumption that produces on the computer equipment is called the calculating energy consumption, and is relevant with power consumption with task size, virtual machine processing speed, calculates with formula (2).The energy consumption that network interconnection device produces is mainly reflected on the position of task and virtual machine, and is relevant with the size of distance and task.The present invention is called the transmission energy consumption to the energy consumption that produces on the network interconnection device, calculates with formula (3).The memory device energy consumption is main relevant with memory device product and storage means; Little with dispatching method of the present invention relation, therefore the energy consumption that on the cloud platform, produces of dispatching method of virtual machine of the present invention has mainly been considered preceding two kinds, but promptly when the state of virtual machine be the time spent; Energy consumption is represented with transmission energy consumption sum with calculating energy consumption; When the state of virtual machine when being unavailable, can consume infinitely great expression, shown in formula (1).
Figure 643398DEST_PATH_IMAGE002
（1）
Figure 2011103425851100002DEST_PATH_IMAGE003
（2）
Figure 609080DEST_PATH_IMAGE004
（3）
At first receive the task that the user submits in the inventive method; Priority according to task; Task to arriving sorts; Obtain the task list
Figure 728345DEST_PATH_IMAGE001
of a need scheduling, m representes current general assignment number.Initialization matrix M CTE according to the number of tasks of submitting to first, creates the virtual machine of equal number, and the parameter of each virtual machine of initialization comprises virtual machine numbering (ID), central processing unit (CPU) quantity, storage allocation size, bandwidth, processing speed and state.Estimation tasks is carried out the energy consumption
Figure 2011103425851100002DEST_PATH_IMAGE005
that consumes on each available virtual machine; To each task in the tabulation; It is distributed to the minimum virtual machine of corresponding energy consumption carries out; From T, delete the task of having distributed; Updating task tabulation, and periodically monitoring host computer load, the adjustment operation plan.Treat that whole tasks are finished, destroy virtual machine.
In sum, it is following to obtain technical scheme of the present invention:
The dispatching method of virtual machine that calculates based on green in a kind of cloud computing may further comprise the steps:
Step 1) receives the task that the user submits to; And according to the priority of task; Task to arriving sorts; Obtain the current task tabulation
Figure 121281DEST_PATH_IMAGE001
of a need scheduling; Wherein m representes current general assignment number, if receive user task first, then carries out next step; Otherwise, turn to step 3);
Step 2) system initialization; Comprise initialization matrix M CTE; According to number of tasks; Create virtual machine, the virtual machine number is identical with the number of tasks that the user submits to, obtains a virtual machine tabulation
Figure 693207DEST_PATH_IMAGE006
; Wherein n representes the virtual machine number; The parameter of each virtual machine of initialization comprises virtual machine numbering (ID), central processing unit (CPU) quantity, storage allocation size, bandwidth, power, processing speed and state then, and the state of each virtual machine is available when initial;
Step 3) judges whether task list is empty, if, then destroy virtual machine, overall process finishes; Otherwise, carry out next step;
Step 4) is according to the least energy consumption strategy, and the minimum virtual machine of the corresponding energy consumption of selection task is dispatched;
Step 5) judges whether the clock period expire, if then carry out next step; Otherwise, turn to step 1);
Step 6) is according to load on host computers, and the adjustment operation plan turns to step 1);
In technical scheme of the present invention, according to the least energy consumption strategy, the minimum virtual machine of the corresponding energy consumption of selection task is dispatched specifically and may further comprise the steps:
Step 401) the matrix M CTE with a m * n representes the energy consumption that m task consumes on n virtual machine; The energy consumption of each row same task of representative on n virtual machine in the matrix; Each row is represented the energy consumption of m task on same virtual machine; Here; I task with
Figure 2011103425851100002DEST_PATH_IMAGE007
among the expression T; J virtual machine among
Figure 217509DEST_PATH_IMAGE008
expression VM,
Figure 456860DEST_PATH_IMAGE005
expression
Figure 83014DEST_PATH_IMAGE007
are gone up at
Figure 142237DEST_PATH_IMAGE008
and are carried out the energy consumption that consumes;
Step 402) reads current task tabulation T; Estimate the energy consumption
Figure 449721DEST_PATH_IMAGE005
that all tasks consume among the T by following formula on each virtual machine; The i that inserts matrix M CTE is capable; The j row; Be MCTE [i, j]=
Figure 605896DEST_PATH_IMAGE005
;
（1）
Figure 949470DEST_PATH_IMAGE003
（2）
Figure 795066DEST_PATH_IMAGE004
（3）
In the formula (1); refers to calculate energy consumption; Be that task
Figure 743430DEST_PATH_IMAGE007
goes up the required energy consumption of operation at certain virtual machine
Figure 649069DEST_PATH_IMAGE008
, unit is joule (J);
Figure 745201DEST_PATH_IMAGE010
refers to transmit energy consumption; Be about to task
Figure 391559DEST_PATH_IMAGE007
and be dispatched to upward required energy consumption of virtual machine
Figure 256747DEST_PATH_IMAGE008
, unit is joule (J); In the formula (2);
Figure 2011103425851100002DEST_PATH_IMAGE011
is the power consumption of virtual machine
Figure 333287DEST_PATH_IMAGE008
, and unit is a watt (W); The execution time of
Figure 854398DEST_PATH_IMAGE012
expression task
Figure 369693DEST_PATH_IMAGE007
on virtual machine
Figure 27071DEST_PATH_IMAGE008
, unit is second (S); The size of
Figure 2011103425851100002DEST_PATH_IMAGE013
expression task
Figure 274512DEST_PATH_IMAGE007
; Weigh with instruction number, unit is 1,000,000 instructions (MI);
Figure 282920DEST_PATH_IMAGE014
is the processing speed of virtual machine
Figure 336326DEST_PATH_IMAGE008
, and unit is million instructions per second (MIPS); In the formula (3);
Figure 2011103425851100002DEST_PATH_IMAGE015
refers to the required energy consumption of unit of transfer's parasang instruction number, and unit is joule every meter (
Figure 113789DEST_PATH_IMAGE016
) of per 1,000,000 instructions;
Figure 2011103425851100002DEST_PATH_IMAGE017
representes task
Figure 532133DEST_PATH_IMAGE007
is transferred to the distance of virtual machine
Figure 24906DEST_PATH_IMAGE008
, and unit is a rice (m);
Step 403) to each task to be allocated
Figure 616425DEST_PATH_IMAGE007
in the tabulation;
Figure 248394DEST_PATH_IMAGE018
; Compare its power consumption values on each available virtual machine; Get least energy consumption
Figure 2011103425851100002DEST_PATH_IMAGE019
; K is hour corresponding virtual machine numbering of task
Figure 103218DEST_PATH_IMAGE007
energy consumption; Note min_energy [i]=
Figure 882955DEST_PATH_IMAGE020
; Wherein min_energy is an one-dimension array that contains m element, representes a least energy consumption corresponding virtual machine set by each task;
Step 404) assigns the task to pairing virtual machine successively by the task list order; Be about to task and distribute to the corresponding virtual machine of value
Figure 30219DEST_PATH_IMAGE020
of min_energy [i]; Deletion
Figure 790365DEST_PATH_IMAGE007
from task list; And upgrade matrix M CTE and task list T, be empty until T.
According to load on host computers, the adjustment operation plan specifically may further comprise the steps:
Step 601) among the present invention; Angle from energy consumption; On the power consumption size of the present main frame of load final body of a certain moment main frame; Here; B main frame of expression with
Figure 2011103425851100002DEST_PATH_IMAGE021
;
Figure 995081DEST_PATH_IMAGE022
; Y is the main frame sum; The power consumption in main frame
Figure 2011103425851100002DEST_PATH_IMAGE023
a certain moment is expression with
Figure 131665DEST_PATH_IMAGE024
; The maximum power dissipation of main frame
Figure 738227DEST_PATH_IMAGE023
is expression with
Figure 2011103425851100002DEST_PATH_IMAGE025
; As the power consumption utilization factor of a certain moment main frame,
Figure 2011103425851100002DEST_PATH_IMAGE027
is the power consumption utilization factor threshold value of main frame with
Figure 666344DEST_PATH_IMAGE026
;
Step 602) to each main frame; if
Figure 358356DEST_PATH_IMAGE028
; is the dormancy factor;
Figure 33051DEST_PATH_IMAGE030
; After treating that then virtual machine on executes task; Destroy virtual machine, and this main frame of dormancy;
Step 603) to each main frame; if ; Then explanation
Figure 861647DEST_PATH_IMAGE023
load is overweight; In this clock period; The Status Flag of all virtual machines that will on this main frame, move is unavailable, i.e. virtual machine allocating task on this main frame again;
Step 604) if host power consumption utilization factor
Figure 316079DEST_PATH_IMAGE031
greater than ; is the overload factor;
Figure 2011103425851100002DEST_PATH_IMAGE033
; Most of main frame then is described, and all load is overweight; Wake the main frame of dormancy up and create virtual machine above that, the number of host that wakes up is no more than the number of host that this clock period transshipped.
Compare prior art, the scheduling virtual machine algorithm that calculates based on green in a kind of cloud computing of the present invention has following advantage:
1) dispatching method of virtual machine is simple among the present invention; With the least energy consumption is regulation goal; The virtual machine that each task is all selected to make it consume the energy consumption minimum matees, and may ultimately reach the total energy consumption minimum, and promptly all tasks are finished needs the calculating energy consumption of consumption and transmission energy consumption sum minimum.This method has reduced the energy consumption of scheduling virtual machine, has embodied the theory of green calculating.
2) operation plan is in time adjusted in periodically monitoring host computer load of the present invention, reaches the balanced effect of energy load.
Description of drawings
Fig. 1 is the cloud computing environment synoptic diagram.
Fig. 2 is the scheduling model in the cloud computing environment.
Fig. 3 is the overall flow synoptic diagram of the inventive method.
Embodiment
Describe for ease, we have following application example at supposition:
Cloud computing environment shown in accompanying drawing 1 is made up of scheduling broker and great amount of calculation machine, memory device and the network interconnection device that is connected them.The user can submit task requests anywhere to; Scheduling broker is the interface between user and the cloud computing environment, is used for realizing the inventive method; On a computing machine, can set up or delete a plurality of virtual machines dynamically and satisfy task requests.
In the scheduling model shown in the accompanying drawing 2, comprise y main frame in the physical host, on y main frame, create and built n virtual machine.X user successively submitted m task to; According to the least energy consumption strategy; That the energy consumption of task
Figure 232400DEST_PATH_IMAGE034
is hour the most corresponding is virtual machine
Figure 2011103425851100002DEST_PATH_IMAGE035
; So being chosen in
Figure 912572DEST_PATH_IMAGE035
,
Figure 899004DEST_PATH_IMAGE034
go up execution; That the energy consumption of task
Figure 410550DEST_PATH_IMAGE036
is hour the most corresponding is virtual machine
Figure 2011103425851100002DEST_PATH_IMAGE037
, goes up execution so
Figure 119880DEST_PATH_IMAGE036
is chosen in
Figure 8201DEST_PATH_IMAGE037
.
In conjunction with accompanying drawing 3, its embodiment is:
Step 1) user
Figure 890707DEST_PATH_IMAGE038
successively submits m task
Figure 2011103425851100002DEST_PATH_IMAGE039
to; According to priority; Task to arriving sorts; Obtain the task list
Figure 118557DEST_PATH_IMAGE001
of a need scheduling; If receive user task first, then carry out next step; Otherwise, turn to step 3);
Step 2) system initialization comprises initialization matrix M CTE, according to number of tasks; Create virtual machine; The virtual machine number is identical with the number of tasks that the user submits to, is n here, then the parameter of each virtual machine of initialization; Comprise virtual machine numbering (ID), central processing unit (CPU) quantity, storage allocation size, bandwidth, power, processing speed and state, the state of each virtual machine is available when initial;
Step 3) judges whether task list is empty, if, then destroy virtual machine, overall process finishes; Otherwise, carry out next step;
Step 4) is according to the least energy consumption strategy, and the minimum virtual machine of the corresponding energy consumption of selection task is dispatched, and specific practice is following:
The first step: the matrix M CTE with a m * n representes the energy consumption that m task consumes on n virtual machine; The energy consumption of each row same task of representative on n virtual machine in the matrix; Each row is represented the energy consumption of m task on same virtual machine; Here; I task with
Figure 61105DEST_PATH_IMAGE007
among the expression T; J virtual machine among
Figure 436723DEST_PATH_IMAGE008
expression VM,
Figure 857340DEST_PATH_IMAGE005
expression
Figure 267592DEST_PATH_IMAGE007
are gone up at
Figure 381042DEST_PATH_IMAGE008
and are carried out the energy consumption that consumes;
Second step: read current task tabulation T; Estimate the energy consumption
Figure 243956DEST_PATH_IMAGE005
that all tasks consume among the T by following formula on each virtual machine; The i that inserts matrix M CTE is capable; The j row; Be MCTE [i, j]=
Figure 468264DEST_PATH_IMAGE005
;
Figure 464514DEST_PATH_IMAGE002
（1）
Figure 686548DEST_PATH_IMAGE003
（2）
Figure 99075DEST_PATH_IMAGE004
（3）
In the formula (1);
Figure 64757DEST_PATH_IMAGE009
refers to calculate energy consumption; Be that task
Figure 246339DEST_PATH_IMAGE007
goes up the required energy consumption of operation at certain virtual machine
Figure 373695DEST_PATH_IMAGE008
, unit is joule (J);
Figure 211201DEST_PATH_IMAGE010
refers to transmit energy consumption; Be about to task
Figure 777312DEST_PATH_IMAGE007
and be dispatched to upward required energy consumption of virtual machine
Figure 16663DEST_PATH_IMAGE008
, unit is joule (J); In the formula (2);
Figure 377237DEST_PATH_IMAGE011
is the power consumption of virtual machine
Figure 702040DEST_PATH_IMAGE008
, and unit is a watt (W); The execution time of
Figure 9524DEST_PATH_IMAGE012
expression task
Figure 900120DEST_PATH_IMAGE007
on virtual machine
Figure 631928DEST_PATH_IMAGE008
, unit is second (S); The size of
Figure 178447DEST_PATH_IMAGE013
expression task
Figure 289622DEST_PATH_IMAGE007
; Weigh with instruction number, unit is 1,000,000 instructions (MI); is the processing speed of virtual machine
Figure 205943DEST_PATH_IMAGE008
, and unit is million instructions per second (MIPS); In the formula (3);
Figure 239758DEST_PATH_IMAGE015
refers to the required energy consumption of unit of transfer's parasang instruction number, and unit is joule every meter (
Figure 889045DEST_PATH_IMAGE016
) of per 1,000,000 instructions;
Figure 754233DEST_PATH_IMAGE017
representes task
Figure 830773DEST_PATH_IMAGE007
is transferred to the distance of virtual machine
Figure 414201DEST_PATH_IMAGE008
, and unit is a rice (m);
The 3rd step: to each task to be allocated
Figure 601600DEST_PATH_IMAGE007
in the tabulation;
Figure 586874DEST_PATH_IMAGE018
; Compare its power consumption values on each available virtual machine; Get least energy consumption
Figure 831386DEST_PATH_IMAGE019
; K is hour corresponding virtual machine numbering of task
Figure 902110DEST_PATH_IMAGE007
energy consumption; Note min_energy [i]= ; Wherein min_energy is an one-dimension array that contains m element, representes a least energy consumption corresponding virtual machine set by each task;
The 4th step: assign the task to pairing virtual machine successively by the task list order; Be about to task and distribute to the corresponding virtual machine of value of min_energy [i]; Deletion
Figure 443764DEST_PATH_IMAGE007
from task list; And upgrade matrix M CTE and task list T, be empty until T;
The 5th step: in this instance; Task
Figure 238544DEST_PATH_IMAGE034
goes up at
Figure 932831DEST_PATH_IMAGE035
carries out the energy consumption minimum that consumes; Task
Figure 522075DEST_PATH_IMAGE036
goes up at
Figure 505075DEST_PATH_IMAGE037
carries out the energy consumption minimum that consumes; Then
Figure 900284DEST_PATH_IMAGE034
distributed to virtual machine
Figure 386760DEST_PATH_IMAGE035
; And deletion
Figure 474802DEST_PATH_IMAGE034
from task list;
Figure 679518DEST_PATH_IMAGE036
distributed to virtual machine
Figure 878418DEST_PATH_IMAGE037
, and deletion from task list;
Step 5) judges whether the clock period expire, if then carry out next step; Otherwise, turn to step 1);
Step 6) is according to load on host computers, and the adjustment operation plan turns to step 1), and specific practice is following:
The first step: among the present invention; Angle from energy consumption; On the power consumption size of the present main frame of load final body of a certain moment main frame; Here; B main frame of expression with ; ; Y is the main frame sum; The power consumption in main frame
Figure 865559DEST_PATH_IMAGE023
a certain moment is expression with
Figure 388945DEST_PATH_IMAGE024
; The maximum power dissipation of main frame
Figure 756472DEST_PATH_IMAGE023
is expression with
Figure 732518DEST_PATH_IMAGE025
; As the power consumption utilization factor of a certain moment main frame,
Figure 792058DEST_PATH_IMAGE027
is the power consumption utilization factor threshold value of main frame with
Figure 210904DEST_PATH_IMAGE026
;
Second step: to each main frame; if
Figure 127225DEST_PATH_IMAGE028
;
Figure 793829DEST_PATH_IMAGE029
is the dormancy factor;
Figure 872644DEST_PATH_IMAGE030
; After treating that then virtual machine on
Figure 308304DEST_PATH_IMAGE023
executes task; Destroy virtual machine, and this main frame of dormancy;
The 3rd step: to each main frame; if
Figure 814372DEST_PATH_IMAGE031
; Then explanation
Figure 965343DEST_PATH_IMAGE023
load is overweight; In this clock period; The Status Flag of all virtual machines that will on this main frame, move is unavailable, i.e. virtual machine allocating task on this main frame again;
The 4th step: if host power consumption utilization factor
Figure 138015DEST_PATH_IMAGE031
greater than
Figure 847848DEST_PATH_IMAGE032
;
Figure 18247DEST_PATH_IMAGE032
is the overload factor; ; Most of main frame then is described, and all load is overweight; Wake the main frame of dormancy up and create virtual machine above that, the number of host that wakes up is no more than the number of host that this clock period transshipped.

Claims (3)

1. the dispatching method of virtual machine that calculates based on green in the cloud computing is characterized in that, may further comprise the steps:
Step 1) receives the task that the user submits to; And according to the priority of task; Task to arriving sorts; Obtain the current task tabulation of a need scheduling; Wherein m representes current general assignment number, if receive user task first, then carries out next step; Otherwise, turn to step 3);
Step 2) system initialization; Comprise initialization matrix M CTE; According to number of tasks; Create virtual machine, the virtual machine number is identical with the number of tasks that the user submits to, obtains a virtual machine tabulation
Figure 152286DEST_PATH_IMAGE002
; Wherein n representes the virtual machine number; The parameter of each virtual machine of initialization comprises virtual machine numbering (ID), central processing unit (CPU) quantity, storage allocation size, bandwidth, power, processing speed and state then, and the state of each virtual machine is available when initial;
Step 3) judges whether task list is empty, if, then destroy virtual machine, overall process finishes; Otherwise, carry out next step;
Step 4) is according to the least energy consumption strategy, and the minimum virtual machine of the corresponding energy consumption of selection task is dispatched;
Step 5) judges whether the clock period expire, if then carry out next step; Otherwise, turn to step 1);
Step 6) is according to load on host computers, and the adjustment operation plan turns to step 1).
2. dispatching method of virtual machine according to claim 1 is characterized in that said step 4) according to the least energy consumption strategy, and the minimum virtual machine of the corresponding energy consumption of selection task is dispatched and may further comprise the steps:
Step 401) the matrix M CTE with a m * n representes the energy consumption that m task consumes on n virtual machine; The energy consumption of each row same task of representative on n virtual machine in the matrix; Each row is represented the energy consumption of m task on same virtual machine; Here; I task with among the expression T; J virtual machine among
Figure 863070DEST_PATH_IMAGE004
expression VM,
Figure 25061DEST_PATH_IMAGE005
expression
Figure 352137DEST_PATH_IMAGE003
are gone up at
Figure 574171DEST_PATH_IMAGE004
and are carried out the energy consumption that consumes;
Step 402) reads current task tabulation T; Estimate the energy consumption
Figure 986697DEST_PATH_IMAGE005
that all tasks consume among the T by following formula on each virtual machine; The i that inserts matrix M CTE is capable; The j row; Be MCTE [i, j]=
Figure 686800DEST_PATH_IMAGE005
;
（1）
Figure 258388DEST_PATH_IMAGE007
（2）
（3）
In the formula (1);
Figure 662005DEST_PATH_IMAGE009
refers to calculate energy consumption; Be that task
Figure 698094DEST_PATH_IMAGE003
goes up the required energy consumption of operation at certain virtual machine
Figure 261931DEST_PATH_IMAGE004
, unit is joule (J);
Figure 649050DEST_PATH_IMAGE010
refers to transmit energy consumption; Be about to task and be dispatched to upward required energy consumption of virtual machine
Figure 847130DEST_PATH_IMAGE004
, unit is joule (J); In the formula (2);
Figure 581868DEST_PATH_IMAGE011
is the power consumption of virtual machine
Figure 128387DEST_PATH_IMAGE004
, and unit is a watt (W); The execution time of expression task on virtual machine
Figure 218199DEST_PATH_IMAGE004
, unit is second (S); The size of
Figure 252014DEST_PATH_IMAGE013
expression task
Figure 698039DEST_PATH_IMAGE003
; Weigh with instruction number, unit is 1,000,000 instructions (MI);
Figure 763560DEST_PATH_IMAGE014
is the processing speed of virtual machine
Figure 902417DEST_PATH_IMAGE004
, and unit is million instructions per second (MIPS); In the formula (3);
Figure 423528DEST_PATH_IMAGE015
refers to the required energy consumption of unit of transfer's parasang instruction number, and unit is joule every meter (
Figure 610927DEST_PATH_IMAGE016
) of per 1,000,000 instructions;
Figure 596201DEST_PATH_IMAGE017
representes task
Figure 843642DEST_PATH_IMAGE003
is transferred to the distance of virtual machine
Figure 914367DEST_PATH_IMAGE004
, and unit is a rice (m);
Step 403) to each task to be allocated
Figure 905456DEST_PATH_IMAGE003
in the tabulation;
Figure 745236DEST_PATH_IMAGE018
; Compare its power consumption values on each available virtual machine; Get least energy consumption
Figure 163579DEST_PATH_IMAGE019
; K is hour corresponding virtual machine numbering of task
Figure 456020DEST_PATH_IMAGE003
energy consumption; Note min_energy [i]=
Figure 250801DEST_PATH_IMAGE020
; Wherein min_energy is an one-dimension array that contains m element, representes a least energy consumption corresponding virtual machine set by each task;
Step 404) assigns the task to pairing virtual machine successively by the task list order; Be about to task
Figure 945088DEST_PATH_IMAGE003
and distribute to the corresponding virtual machine of value
Figure 534332DEST_PATH_IMAGE020
of min_energy [i]; Deletion
Figure 579648DEST_PATH_IMAGE003
from task list; And upgrade matrix M CTE and task list T, be empty until T.
3. dispatching method of virtual machine according to claim 1 is characterized in that said step 6) according to load on host computers, and the adjustment operation plan may further comprise the steps:
Step 601) among the present invention; Angle from energy consumption; On the power consumption size of the present main frame of load final body of a certain moment main frame; Here; B main frame of expression with
Figure 933048DEST_PATH_IMAGE021
; ; Y is the main frame sum; The power consumption in main frame
Figure 507566DEST_PATH_IMAGE023
a certain moment is expression with ; The maximum power dissipation of main frame
Figure 645604DEST_PATH_IMAGE023
is expression with
Figure 252165DEST_PATH_IMAGE025
; As the power consumption utilization factor of a certain moment main frame,
Figure 937542DEST_PATH_IMAGE027
is the power consumption utilization factor threshold value of main frame with
Figure 511109DEST_PATH_IMAGE026
;
Step 602) to each main frame; if ;
Figure 401201DEST_PATH_IMAGE029
is the dormancy factor;
Figure 565466DEST_PATH_IMAGE030
; After treating that then virtual machine on
Figure 744775DEST_PATH_IMAGE023
executes task; Destroy virtual machine, and this main frame of dormancy;
Step 603) to each main frame; if
Figure 285478DEST_PATH_IMAGE031
; Then explanation
Figure 601052DEST_PATH_IMAGE023
load is overweight; In this clock period; The Status Flag of all virtual machines that will on this main frame, move is unavailable, i.e. virtual machine allocating task on this main frame again;
Step 604) if host power consumption utilization factor
Figure 599894DEST_PATH_IMAGE031
greater than
Figure 201798DEST_PATH_IMAGE032
;
Figure 881971DEST_PATH_IMAGE032
is the overload factor;
Figure 379948DEST_PATH_IMAGE033
; Most of main frame then is described, and all load is overweight; Wake the main frame of dormancy up and create virtual machine above that, the number of host that wakes up is no more than the number of host that this clock period transshipped.
CN2011103425851A 2011-11-03 2011-11-03 Green-computer-based virtual machine scheduling method for cloud computing Pending CN102508714A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN2011103425851A CN102508714A (en) 2011-11-03 2011-11-03 Green-computer-based virtual machine scheduling method for cloud computing

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN2011103425851A CN102508714A (en) 2011-11-03 2011-11-03 Green-computer-based virtual machine scheduling method for cloud computing

Publications (1)

Publication Number Publication Date
CN102508714A true CN102508714A (en) 2012-06-20

Family

ID=46220806

Family Applications (1)

Application Number Title Priority Date Filing Date
CN2011103425851A Pending CN102508714A (en) 2011-11-03 2011-11-03 Green-computer-based virtual machine scheduling method for cloud computing

Country Status (1)

Country Link
CN (1) CN102508714A (en)

Cited By (22)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102999385A (en) * 2012-11-06 2013-03-27 苏州懿源宏达知识产权代理有限公司 Multiprocessor co-processing method in computation equipment
CN103118124A (en) * 2013-02-22 2013-05-22 桂林电子科技大学 Cloud computing load balancing method based on layering multiple agents
CN103605567A (en) * 2013-10-29 2014-02-26 河海大学 Cloud computing task scheduling method facing real-time demand change
CN103812949A (en) * 2014-03-06 2014-05-21 中国科学院信息工程研究所 Task scheduling and resource allocation method and system for real-time cloud platform
CN104184806A (en) * 2014-08-14 2014-12-03 合肥工业大学 IAAS virtual machine dynamic migration method balancing energy consumption and service quality
WO2015010640A1 (en) * 2013-07-24 2015-01-29 华为技术有限公司 Method of controlling management device, and related device
CN104820618A (en) * 2015-04-24 2015-08-05 华为技术有限公司 Task scheduling method, task scheduling device and multi-core system
CN105912406A (en) * 2016-05-05 2016-08-31 中国人民解放军国防科学技术大学 Low-energy independent task scheduling and resource configuration method
CN106227600A (en) * 2016-07-25 2016-12-14 湖南大学 A kind of multidimensional virtual resource allocation method based on Energy-aware
CN106775948A (en) * 2016-12-23 2017-05-31 安徽师范大学 A kind of cloud method for scheduling task and device based on priority
CN107145216A (en) * 2017-05-05 2017-09-08 北京景行锐创软件有限公司 A kind of dispatching method
CN107273185A (en) * 2017-06-19 2017-10-20 成都鼎智汇科技有限公司 A kind of control method for equalizing load based on virtual machine
CN107341043A (en) * 2017-06-28 2017-11-10 东北大学 A kind of emulation mode for the consumption of data center for assessing regenerative resource hybrid power supply
CN107479949A (en) * 2017-08-21 2017-12-15 重庆电子工程职业学院 Low energy consumption cloud computing method for scheduling task
CN108279968A (en) * 2017-12-29 2018-07-13 中国联合网络通信集团有限公司 A kind of dispatching method and device of resources of virtual machine
WO2018170683A1 (en) * 2017-03-20 2018-09-27 华平智慧信息技术（深圳）有限公司 Task allocation method and system for cloud service in monitoring system
CN109032756A (en) * 2018-07-04 2018-12-18 中国人民解放军国防科技大学 Scheduling method of virtualized cloud data center
CN110083453A (en) * 2019-04-28 2019-08-02 南京邮电大学 Energy saving resources dispatching method based on Min-Max algorithm under a kind of cloud computing environment
CN112068943A (en) * 2020-09-08 2020-12-11 山东省计算中心（国家超级计算济南中心） Micro-service scheduling method based on complex heterogeneous environment and implementation system thereof
CN112764883A (en) * 2021-01-25 2021-05-07 江苏赞奇科技股份有限公司 Energy management method of cloud desktop system based on software definition
CN113920678A (en) * 2021-10-19 2022-01-11 深圳市盛邦通信有限公司 Remote smoke alarm system on M5311 chip based on middle moving object
CN116405391A (en) * 2023-04-10 2023-07-07 长扬科技(北京)股份有限公司 OpenStack-based virtual machine node screening method, system and storage medium

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1596392A (en) * 2001-01-05 2005-03-16 国际商业机器公司 Computer system having low energy consumption
CN101604264A (en) * 2009-07-08 2009-12-16 深圳先进技术研究院 The method for scheduling task of supercomputer and system
CN101719081A (en) * 2009-12-01 2010-06-02 北京大学 Method for scheduling virtual machines

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1596392A (en) * 2001-01-05 2005-03-16 国际商业机器公司 Computer system having low energy consumption
CN101604264A (en) * 2009-07-08 2009-12-16 深圳先进技术研究院 The method for scheduling task of supercomputer and system
CN101719081A (en) * 2009-12-01 2010-06-02 北京大学 Method for scheduling virtual machines

Cited By (36)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102999385B (en) * 2012-11-06 2016-05-25 国网山东省电力公司枣庄供电公司 Multiprocessor cooperative processing method in computing equipment
CN102999385A (en) * 2012-11-06 2013-03-27 苏州懿源宏达知识产权代理有限公司 Multiprocessor co-processing method in computation equipment
CN103118124A (en) * 2013-02-22 2013-05-22 桂林电子科技大学 Cloud computing load balancing method based on layering multiple agents
CN103118124B (en) * 2013-02-22 2015-08-05 桂林电子科技大学 A kind of cloud computing load balancing method based on the many agencies of layering
US10218578B2 (en) 2013-07-24 2019-02-26 Huawei Technologies Co., Ltd. Method for controlling a management device and related device
WO2015010640A1 (en) * 2013-07-24 2015-01-29 华为技术有限公司 Method of controlling management device, and related device
CN103605567A (en) * 2013-10-29 2014-02-26 河海大学 Cloud computing task scheduling method facing real-time demand change
CN103605567B (en) * 2013-10-29 2017-03-22 河海大学 Cloud computing task scheduling method facing real-time demand change
CN103812949B (en) * 2014-03-06 2016-09-07 中国科学院信息工程研究所 A kind of task scheduling towards real-time cloud platform and resource allocation methods and system
CN103812949A (en) * 2014-03-06 2014-05-21 中国科学院信息工程研究所 Task scheduling and resource allocation method and system for real-time cloud platform
CN104184806B (en) * 2014-08-14 2017-05-17 合肥工业大学 IAAS virtual machine dynamic migration method balancing energy consumption and service quality
CN104184806A (en) * 2014-08-14 2014-12-03 合肥工业大学 IAAS virtual machine dynamic migration method balancing energy consumption and service quality
CN104820618A (en) * 2015-04-24 2015-08-05 华为技术有限公司 Task scheduling method, task scheduling device and multi-core system
CN104820618B (en) * 2015-04-24 2018-09-07 华为技术有限公司 A kind of method for scheduling task, task scheduling apparatus and multiple nucleus system
CN105912406A (en) * 2016-05-05 2016-08-31 中国人民解放军国防科学技术大学 Low-energy independent task scheduling and resource configuration method
CN106227600A (en) * 2016-07-25 2016-12-14 湖南大学 A kind of multidimensional virtual resource allocation method based on Energy-aware
CN106227600B (en) * 2016-07-25 2019-05-17 湖南大学 A kind of multidimensional virtual resource allocation method based on Energy-aware
CN106775948A (en) * 2016-12-23 2017-05-31 安徽师范大学 A kind of cloud method for scheduling task and device based on priority
CN106775948B (en) * 2016-12-23 2020-09-04 安徽师范大学 Cloud task scheduling method and device based on priority
WO2018170683A1 (en) * 2017-03-20 2018-09-27 华平智慧信息技术（深圳）有限公司 Task allocation method and system for cloud service in monitoring system
CN107145216A (en) * 2017-05-05 2017-09-08 北京景行锐创软件有限公司 A kind of dispatching method
CN107273185B (en) * 2017-06-19 2020-09-29 观典防务技术股份有限公司 Load balancing control method based on virtual machine
CN107273185A (en) * 2017-06-19 2017-10-20 成都鼎智汇科技有限公司 A kind of control method for equalizing load based on virtual machine
CN107341043A (en) * 2017-06-28 2017-11-10 东北大学 A kind of emulation mode for the consumption of data center for assessing regenerative resource hybrid power supply
CN107479949A (en) * 2017-08-21 2017-12-15 重庆电子工程职业学院 Low energy consumption cloud computing method for scheduling task
CN108279968B (en) * 2017-12-29 2021-05-11 中国联合网络通信集团有限公司 Virtual machine resource scheduling method and device
CN108279968A (en) * 2017-12-29 2018-07-13 中国联合网络通信集团有限公司 A kind of dispatching method and device of resources of virtual machine
CN109032756A (en) * 2018-07-04 2018-12-18 中国人民解放军国防科技大学 Scheduling method of virtualized cloud data center
CN110083453A (en) * 2019-04-28 2019-08-02 南京邮电大学 Energy saving resources dispatching method based on Min-Max algorithm under a kind of cloud computing environment
CN112068943A (en) * 2020-09-08 2020-12-11 山东省计算中心（国家超级计算济南中心） Micro-service scheduling method based on complex heterogeneous environment and implementation system thereof
CN112068943B (en) * 2020-09-08 2022-11-25 山东省计算中心（国家超级计算济南中心） Micro-service scheduling method based on complex heterogeneous environment and implementation system thereof
CN112764883A (en) * 2021-01-25 2021-05-07 江苏赞奇科技股份有限公司 Energy management method of cloud desktop system based on software definition
CN112764883B (en) * 2021-01-25 2024-01-19 江苏赞奇科技股份有限公司 Energy management method of cloud desktop system based on software definition
CN113920678A (en) * 2021-10-19 2022-01-11 深圳市盛邦通信有限公司 Remote smoke alarm system on M5311 chip based on middle moving object
CN113920678B (en) * 2021-10-19 2023-01-17 深圳市盛邦通信有限公司 Remote smoke alarm system on M5311 chip based on middle moving object
CN116405391A (en) * 2023-04-10 2023-07-07 长扬科技(北京)股份有限公司 OpenStack-based virtual machine node screening method, system and storage medium

Similar Documents

Publication Publication Date Title
CN102508714A (en) Green-computer-based virtual machine scheduling method for cloud computing
Choudhary et al. A critical analysis of energy efficient virtual machine placement techniques and its optimization in a cloud computing environment
CN102780759B (en) Based on the cloud computing resource scheduling method in regulation goal space
Changtian et al. Energy-aware genetic algorithms for task scheduling in cloud computing
CN102662750A (en) Virtual machine resource optimal control method and control system based on elastic virtual machine pool
CN108182115A (en) A kind of virtual machine load-balancing method under cloud environment
CN102158513A (en) Service cluster and energy-saving method and device thereof
CN100576179C (en) A kind of based on energy-optimised gridding scheduling method
CN102521055B (en) Virtual machine resource allocating method and virtual machine resource allocating system
CN104199736A (en) Method for saving energy of data center under cloud environment
CN106951059A (en) Based on DVS and the cloud data center power-economizing method for improving ant group algorithm
Braiki et al. Fuzzy-logic-based multi-objective best-fit-decreasing virtual machine reallocation
US20130167152A1 (en) Multi-core-based computing apparatus having hierarchical scheduler and hierarchical scheduling method
Mao et al. A multi-resource task scheduling algorithm for energy-performance trade-offs in green clouds
CN103365726A (en) Resource management method and system facing GPU (Graphic Processing Unit) cluster
CN102404412A (en) Energy saving method and system for cloud compute data center
CN103297499A (en) Scheduling method and system based on cloud platform
CN106126344A (en) A kind of method for allocating tasks and device
CN102208986A (en) Cluster power consumption distribution and control method
Zhang et al. An energy-aware host resource management framework for two-tier virtualized cloud data centers
CN107070965B (en) Multi-workflow resource supply method under virtualized container resource
CN107479949A (en) Low energy consumption cloud computing method for scheduling task
CN103176850A (en) Electric system network cluster task allocation method based on load balancing
CN103049326A (en) Method and system for managing job program of job management and scheduling system
CN110308991B (en) Data center energy-saving optimization method and system based on random tasks

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
C02 Deemed withdrawal of patent application after publication (patent law 2001)
WD01 Invention patent application deemed withdrawn after publication

Application publication date: 20120620