CN106125888A - The efficient power-economizing method of the utilization of resources based on virtual machine (vm) migration in cloud data center - Google Patents

The efficient power-economizing method of the utilization of resources based on virtual machine (vm) migration in cloud data center Download PDF

Info

Publication number
CN106125888A
CN106125888A CN201610460901.8A CN201610460901A CN106125888A CN 106125888 A CN106125888 A CN 106125888A CN 201610460901 A CN201610460901 A CN 201610460901A CN 106125888 A CN106125888 A CN 106125888A
Authority
CN
China
Prior art keywords
physical host
virtual machine
resource
host
data center
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.)
Granted
Application number
CN201610460901.8A
Other languages
Chinese (zh)
Other versions
CN106125888B (en
Inventor
韩光洁
阙文辉
江金芳
杨璇
王皓
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.)
Changzhou Campus of Hohai University
Original Assignee
Changzhou Campus of Hohai 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 Changzhou Campus of Hohai University filed Critical Changzhou Campus of Hohai University
Priority to CN201610460901.8A priority Critical patent/CN106125888B/en
Publication of CN106125888A publication Critical patent/CN106125888A/en
Application granted granted Critical
Publication of CN106125888B publication Critical patent/CN106125888B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F1/00—Details not covered by groups G06F3/00 - G06F13/00 and G06F21/00
    • G06F1/26—Power supply means, e.g. regulation thereof
    • G06F1/32—Means for saving power
    • G06F1/3203—Power management, i.e. event-based initiation of a power-saving mode
    • G06F1/3234—Power saving characterised by the action undertaken
    • G06F1/329—Power saving characterised by the action undertaken by task scheduling
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/44—Arrangements for executing specific programs
    • G06F9/455—Emulation; Interpretation; Software simulation, e.g. virtualisation or emulation of application or operating system execution engines
    • G06F9/45533—Hypervisors; Virtual machine monitors
    • G06F9/45558—Hypervisor-specific management and integration aspects
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/46—Multiprogramming arrangements
    • G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
    • G06F9/5083—Techniques for rebalancing the load in a distributed system
    • G06F9/5088—Techniques for rebalancing the load in a distributed system involving task migration
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
    • Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Software Systems (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Computer And Data Communications (AREA)
  • Power Sources (AREA)

Abstract

The invention discloses the efficient power-economizing method of the utilization of resources based on virtual machine (vm) migration in cloud data center, comprise overload physical host and process, the adjustment of resource blockage physical host, underloading physical host selects and virtual machine places four modules.The resource using status of first each to the cloud data center physical host of the method is monitored, and for the main frame of overload, selects suitable virtual machine (vm) migration to go out to reduce physical host load；For the physical host of resource blockage, the mode of virtual machine (vm) migration is utilized to carry out adjustment of load；Then, in the physical host that resource utilization is relatively low, select suitable host according to resource requirement situation, the virtual machine that it runs all migrates out and closes this main frame to reduce the quantity of mobile host computers and to reduce energy consumption；Virtual machine placement module is then for needing the virtual machine migrated to select suitable destination host.This method can be effectively improved the level of resources utilization of cloud data center activity physical host, reduces the energy consumption of cloud data center.

Description

The efficient power-economizing method of the utilization of resources based on virtual machine (vm) migration in cloud data center
Technical field
The invention belongs to Internet technical field, particularly relate to a kind of resource based on virtual machine (vm) migration in cloud data center Utilize efficient power-economizing method.
Background technology
Cloud computing is that a kind of high-throughput using the distribution according to need of model to calculate resource based on paying calculates normal form, and it leads to Excessive data center or server cluster provide cloud service to cloud user, i.e. IaaS (infrastructure i.e. services), PaaS (platform I.e. service) and SaaS (software i.e. services).Cloud computing is developer and IT company provides a great convenience, by them from setting The low-level tasks such as server hardware and management system software frees so that they can be absorbed in opening of application program Send out.Meanwhile, cloud user has only to on-demand pay cloud provider operating cost, without buying mounting hardware equipment, greatly drops The low calculating cost of user and enterprise, improves development efficiency.
Virtualize the basic technology as cloud computing, be total to by the way of multiplexing virtual machine on same main frame Enjoy the hardware resource of costliness.The task that multiple cloud users submit to can on the different virtual machine on same physical host simultaneously Run, and the work between each virtual machine is separate, share physical machine hardware resource.This not only makes to extend each neatly The resource requirement of virtual machine is possibly realized, and also the manager for data center provides conveniently.Service based on virtual machine (vm) migration Device aggregation operator has become a Intel Virtualization Technology main application in data center.
Along with the fast development of cloud computing, the quantity of data center is just with unprecedented speed increment, and data center Energy consumption problem the most increasingly get more and more people's extensive concerning.According to estimates, to the year two thousand twenty, the power consumption of cloud data center will account for The 2% of whole world power consumption.The energy consumption that cloud data center is the hugest, not only can increase greenhouse gas emission and then cause environment Problem, also can improve the use cost of cloud user simultaneously.Owing to the resource utilization of current cloud data center only arrives 20% Between 50%, real-time migration of virtual machine technology is utilized to reduce number of mobile host computers to realize server aggregates, reduce data The method of power consumption has become research boom.Real-time migration of virtual machine technology passes through migration process device state, memory content, net The information such as network state, it is possible to the client operating system state on source host is copied to destination host, it is achieved virtual machine is not With transfer real-time on main frame.But, virtual machine (vm) migration can cause increasing the resource consumption of physical machine, taking the network bandwidth and shadow Ring the problems such as service quality；It addition, the virtual machine migrated out is when selecting host, easily cause the resource of host Use unbalanced so that the situation of the wasting of resources occurs because of resource blockage in host.Therefore, how cloud data are being improved Reduce the quantity of virtual machine (vm) migration while center physical machine resource utilization, minimizing number of mobile host computers, meet user and cloud The SLA (SLA) signed between ISP becomes server aggregates operation urgent problem.
Through finding existing literature search, pertinent literature is as follows:
It is entitled that Xin Li delivered on " Mathematical and Computer Modelling " periodical in 2013 “Energy efficient virtual machine placement algorithm with balanced and Improved resource utilization in a data center " in article, for data center physical machine by The unbalanced problem of resource waste caused is used, it is proposed that a hyperspace partitioning model and based on this in multi dimensional resource Model proposes virtual machine Placement (EAGLE).The resource using status of physical machine is divided into by this hyperspace partitioning model Acceptance region (AD), security domain (SD) and static territory (FD) three regions, EAGLE algorithm is accepting virtual machine according to each main frame After the region at resource using status place, select suitable main frame according to the priority of regional.In literary composition, experimental result shows Showing, the method saves the energy consumption of 15% than classical FF (First Fit) algorithm.But, literary composition is only the placement from virtual machine Angle has carried out the checking of virtual machine and host number, does not put from the virtual machine of the data center already at duty Put and effectively analyze in problem.
K.Sunil Rao delivered on " Future Generation Computer Systems " periodical in 2015 Entitled " Heuristics based server consolidation with residual resource Defragmentation in cloud data centers " broken in order to reduce the surplus resources after server aggregates in a literary composition Sheet, it is proposed that RFAware Server Consolidation algorithm.This algorithm includes four-stage, i.e. pretreatment stage, Physical machine reduces the stage, surplus resources goes the fragmentation stage and migrates the scheme optimization stage.Pretreatment stage is opened at aggregation operator Carrying out resource distribution according to the resource requirement that virtual machine is up-to-date before beginning, the mode that physical machine reduces stage use greedy reduces work The quantity of dynamic main frame, surplus resources goes the fragmentation stage to use unbalanced according to the scheduling scheme generated previous stage to resource Main frame carries out adjustment of load, and the migration optimizing phase then determines that the priority of migration and reduces unnecessary virtual machine (vm) migration. Test result indicate that, the virtual machine (vm) migration quantity needed for this algorithm is more, it is possible to efficiently reduce the fragmentation of surplus resources.
In sum, although the resource of data center uses unbalanced and surplus resources fragmentation problem to change Kind, but yet suffer from some problem values and must further investigate:
1. adjust the resource utilization of cloud data center each physical host, not only to consider the equilibrium of various resource utilize with Reduce the wasting of resources, also should improve the resource utilization of each physical host, make full use of the excellent of each host resource consumption equilibrium Gesture, reduces the quantity of mobile host computers, reduces the energy consumption of cloud data center.
The most above-mentioned document is all not involved with the process of physical machine for having there is resource blockage in data center, As it is shown on figure 3, the resource of these physical machine uses unbalanced that be likely due to be originally present within or resources of virtual machine to need Dynamically change is asked to cause.RFAware Server Consolidation algorithm is right after the physical machine gathering stage The resource using status of each physical machine is adjusted, and this adjustment mode by exchange virtual machine is for already at obstruction The effect of the physical machine of state is the most little.If just using unbalanced physical machine to carry out these resources before gathering starts Processing so that they are able to receive that the virtual machine from other physical host, improve their resource utilization, this is for reducing The physical machine quantity of data center's active state has positive meaning.
3., during physical machine is assembled, above-mentioned document all have ignored what resource requirement in cloud data center dynamically changed Feature, the selection for underloading main frame considers only its resource utilization.In the physical machine that current time resource utilization is low Resource requirement at subsequent time may increase.If only physical machine few for resource consumption is moved as underloading main frame Move, then be easy for causing host to transship when the resource requirement of these virtual machines increases, cause frequently moving of virtual machine Move, affect service quality.Therefore, should be in conjunction with to the predictive value of its resource requirement and currently running when selecting underloading main frame The quantity of virtual machine account for reducing the quantity of virtual machine (vm) migration, increase the stability of host.
Summary of the invention
The problems referred to above existed for prior art, the invention provides in a kind of cloud data center based on virtual machine (vm) migration The efficient power-economizing method of the utilization of resources.Described cloud data center includes substantial amounts of movable physical host and runs on main frame Virtual machine, cloud user submit to task perform in different virtual machines, said method comprising the steps of:
Step a: the process of cloud data center overload main frame: make according to the resource of record in the local manager of physical host Use rate daily record, the resource requirement on this physical host is predicted；If this resource requirement has exceeded the capacity of main frame, then select The virtual machine (vm) migration run on it is gone out, until this overload main frame returns to normal operating conditions；Without finding that overload is main Machine is then directly entered step b；
Step b: the adjustment of the physical host that cloud data center resource blocks: according to each main frame in destination host set Resource service condition selects physical host to be adjusted；Then resource consumption is selected to use with its resource from these physical hosts The virtual machine (vm) migration that type is identical is gone out, and improves the harmony that this physical host resource uses；
Step c: the selection of cloud data center underloading main frame: for the physics master that cloud data center resource utilization rate is relatively low Machine, according to the quantity of the virtual machine run in its resource utilization state and its, selects suitable physical host and by thereon All virtual machine (vm) migrations, on the of a relatively high physical host of resource utilization, are then shut off this physical host with minimizing activity master The quantity of machine；If the virtual machine on underloading main frame existing one or more virtual machine can not find available host, then Virtual machine on this main frame does not migrates, and continues to run with；
Step d: the placement of virtual machine to be migrated in cloud data center: the virtual of migration will be needed in above-mentioned steps a, b, c Machine is placed on suitable physical host in order, improves the resource utilization of each movable physical host in cloud data center.
Preferably, the concrete adjustment process of the resource blockage physical host in described step b is as follows:
A, according to the summation of the resource requirement in cloud data center and the capacity of physical host, calculate and meet in theory Quantity N of the physical host required for current resource demand；
B, resource service condition according to physical host, for one score value of each Framework computing in data center, calculate Formula is as follows:
s c o r e = | ( 1 - 2 · d ) · x + ( 1 + 2 · d ) · y 2 |
X in above formula represents the cpu utilization rate of physical host, and y then represents the memory usage of this physical host, and d represents Now the resource using status of physical host deviates the distance of its utilization of resources equilibrium state；Select the top n that score value is bigger Physical host, as destination host during virtual machine (vm) migration, constitutes destination host set；
C, for this N number of physical host, select the resource using status not physical machine in acceptable equalization region ETA, Then the meansigma methods of these physical host resource utilizations deviation equilibrium state is calculated, select resource utilization and deviate it Distance d of equilibrium state is more thanPhysical host as main frame to be adjusted；
D, for these main frames to be adjusted, according to d value order from big to small, it is processed；If this physics master Resource service condition in machine is Pcpu＞ Pmem, wherein PcpuAnd PmemRepresent cpu and the memory usage of physical host respectively, then The resources of virtual machine conditions of demand migrated are selected to should be Vcpu＞ Vmem, wherein VcpuAnd VmemRepresent the cpu of virtual machine and interior respectively Deposit utilization rate, and the resource service condition of this physical host meet following formula:
| newUtilization P - oldUtilization P | < oldUtilization P m
OldUtilization in above formulaPAnd newUtilizationPRepresent this physics before and after this virtual machine (vm) migration respectively The average resource of main frame, m then represents the quantity of the virtual machine transported on this main frame；When the resource of physical host uses feelings Condition is Pcpu＞ PmemTime, disposition is similar to.
Preferably, the detailed process of the underloading Selection of chiller in described step c is as follows:
The size of the score value according to each physical host, for being not at the physical host in destination host set, meter Calculate its SF value:
S F = predictedUtilization P currentUtilization P &times; m
PredictedUtilization in above formulaPAnd currentUtilizationPRepresent respectively on this physical host The meansigma methods of resource utilization prediction and the meansigma methods of currently practical service condition, m then represents the void run on this physical host The quantity of plan machine；So, just physical host less for SF value is judged to underloading main frame.
Preferably, the detailed process of the virtual machine Placement Strategy in described step d is as follows:
A, for resource blockage physical host needs the virtual machine migrated out, preferentially they are placed into treating of other On the physical host adjusted so that the physical host migrating virtual machine and the resource using status of the physical host accepting virtual machine Fall in ETA or make d value reduce；If there is not such physical host, then this virtual machine is placed into not at destination host On physical host in set so that their resource utilization status falls in ETA or makes d value reduce, and by this physics master Machine joins in destination host set；If still can not find the physical host of placement, then abandon the adjustment to this physical host；
B, for the virtual machine migrated out on poor efficiency main frame, from destination host set select main frame, if This host resource can be made to use state to fall in ETA, then this main frame just accepts this virtual machine；Otherwise, just look for so that accepting The physical host that after this virtual machine, d value is minimum is as destination host；When the physical host in destination host set can not receive again After virtual machine, then look for the maximum physical host of SF value as host to receive other physics master in remaining physical host Virtual machine on machine, until not having physical host to be closed.
Beneficial effects of the present invention:
The present invention take into account the physical host in cloud data center, and to there is resource utilization unbalanced, and carries therefrom The problem of the wasting of resources come, before carrying out aggregation operator, just physical host to resource blockage is adjusted improving it and provides Source utilization rate, the quantity of reduction data center mobile host computers.It addition, place the stage at virtual machine, it is contemplated that each physical host Resource using status, to reduce for the purpose of resource uses unbalanced situation to improve the resource utilization of each main frame simultaneously, This algorithm can be effectively improved the resource utilization of each mobile host computers of cloud data center；Simultaneously for host resource equilibrium state Define the placement process that can speed up virtual machine.The characteristic dynamically changed in view of cloud data center resource demand, in underloading The quantity of the virtual machine of host resource conditions of demand and operation is fully taken into account, it is possible to reduce due to load during Selection of chiller The impact that the virtual machine that change brings migrates repeatedly, improves the service quality of cloud data center.
Accompanying drawing explanation
Fig. 1 is the flow chart of the present invention；
Fig. 2 is the division figure of the resource service condition in physical host；
Fig. 3 is the schematic diagram that in physical host, cpu resource excessively uses；
Fig. 4 is the schematic diagram that in physical host, memory source excessively uses.
Detailed description of the invention
As it is shown in figure 1, the efficient power-economizing method of the utilization of resources based on virtual machine (vm) migration in a kind of cloud data center, described Cloud data center comprise substantial amounts of movable physical host and on main frame running virtual machine, the task that user submits to is in difference Virtual machine in perform, as shown in Figure 3 and Figure 4, physical host there will be because resource uses unbalanced and cause the wasting of resources Situation；In order to improve physical host resource utilization, the energy consumption of reduction cloud data center, based on virtual in a kind of cloud data center It is as follows that the efficient power-economizing method of the utilization of resources that machine migrates is embodied as step:
Step a: the process of cloud data center overload main frame: before server aggregates operation starts, according to physical host Local manager in record resource utilization daily record, the resource requirement on this physical host is predicted；If this resource Demand has exceeded the capacity of main frame, then select the virtual machine (vm) migration run on it to go out, until this overload main frame returns to normally Duty；Without finding that overload main frame is then directly entered step b；
Step b: the adjustment of the physical host that cloud data center resource blocks: according to each main frame in destination host set Resource service condition selects physical host to be adjusted；Then resource consumption is selected to use with its resource from these physical hosts The virtual machine (vm) migration that type is identical is gone out, and improves the harmony that this physical host resource uses；
Step c: the selection of cloud data center underloading main frame: for the physics master that cloud data center resource utilization rate is relatively low Machine, according to the quantity of the virtual machine run in its resource utilization state and its, selects suitable physical host and by thereon All virtual machine (vm) migrations, on the of a relatively high physical host of resource utilization, are then shut off this physical host with minimizing activity master The quantity of machine；If the virtual machine on underloading main frame existing one or more virtual machine can not find available host, then Virtual machine on this main frame does not migrates, and continues to run with；
Step d: the placement of virtual machine to be migrated in cloud data center: the virtual of migration will be needed in above-mentioned steps a, b, c Machine is placed on suitable physical host in order, improves the resource utilization of each movable physical host in cloud data center.
Preferably, the concrete adjustment process of the resource blockage physical host in described step b is as follows:
Summation according to the resource requirement in cloud data center and the capacity of physical host, calculate to meet in theory and work as Quantity N of the physical host required for front resource requirement；Then according to the resource service condition of physical host, in data center One score value of each Framework computing, calculating formula is as follows:
s c o r e = | ( 1 - 2 &CenterDot; d ) &CenterDot; x + ( 1 + 2 &CenterDot; d ) &CenterDot; y 2 |
X in above formula represents the cpu utilization rate of physical host, and y then represents the memory usage of this physical host, and d represents Now the resource using status of physical host deviates the distance of its utilization of resources equilibrium state；Select the top n that score value is bigger Physical host, as destination host during virtual machine (vm) migration, constitutes destination host set；
As in figure 2 it is shown, for this N number of physical host, select resource using status not in acceptable equalization region ETA Physical machine, then calculates the meansigma methods of these physical host resource utilizations deviation equilibrium state, select the utilization of resources Rate deviates distance d of its equilibrium state and is more thanPhysical host as main frame to be adjusted；
For the main frame that these are to be adjusted, according to d value order from big to small, it is processed；If this physical host In resource service condition be Pcpu＞ Pmem, wherein PcpuAnd PmemRepresent cpu and the memory usage of physical host respectively, then select The resources of virtual machine conditions of demand selecting migration should be Vcpu＞ Vmem, wherein VcpuAnd VmemRepresent cpu and the internal memory of virtual machine respectively Utilization rate, and the resource service condition of this physical host meets following formula:
| newUtilization P - oldUtilization P | < oldUtilization P m
OldUtilization in above formulaPAnd newUtilizationPRepresent this physics before and after this virtual machine (vm) migration respectively The average resource of main frame, m then represents the quantity of the virtual machine transported on this main frame；When the resource of physical host uses feelings Condition is Pcpu＜ PmemTime, disposition is similar to.
Preferably, the detailed process of the underloading Selection of chiller in described step c is as follows:
The size of the score value according to each physical host, for being not at the physical host in destination host set, meter Calculate its SF value:
S F = predictedUtilization P currentUtilization P &times; m
PredictedUtilization in above formulaPAnd currentUtilizationPRepresent respectively on this physical host The meansigma methods of resource utilization prediction and the meansigma methods of currently practical service condition, m then represents the void run on this physical host The quantity of plan machine；So, just physical host less for SF value is judged to underloading main frame.
Preferably, the detailed process of the virtual machine Placement Strategy in described step d is as follows:
For resource blockage physical host needs the virtual machine migrated out, preferentially they are placed into other wait adjust On whole physical host so that the resource using status of the physical host migrating virtual machine and the physical host accepting virtual machine falls In ETA or make d value reduce；If there is not such physical host, then this virtual machine is placed into not at destination host collection On physical host in conjunction so that their resource utilization status falls in ETA or makes d value reduce, and by this physical host Join in destination host set；If still can not find the physical host of placement, then abandon the adjustment to this physical host；
For the virtual machine migrated out on poor efficiency main frame, from destination host set, select main frame, if energy This host resource is enough made to use state to fall in ETA, then this main frame just accepts this virtual machine；Otherwise, just look for so that accepting to be somebody's turn to do The physical host that after virtual machine, d value is minimum is as destination host；When the physical host in destination host set can not receive void again After plan machine, then look for the maximum physical host of SF value as host to receive other physical host in remaining physical host On virtual machine, until do not have physical host to be closed.
The above is the preferred embodiment of the present invention, it should be pointed out that: for those skilled in the art For, under the premise without departing from the principles of the invention, it is also possible to make some improvements and modifications, these improvements and modifications should also be as It is considered as protection scope of the present invention.

Claims (4)

1. the efficient power-economizing method of the utilization of resources based on virtual machine (vm) migration in a Zhong Yun data center, it is characterised in that described Cloud data center include substantial amounts of movable physical host and on main frame running virtual machine, the task that cloud user submits to is not Same virtual machine performs, including step in detail below:
Step a: the process of cloud data center overload main frame: according to the resource utilization of record in the local manager of physical host Daily record, is predicted the resource requirement on this physical host；If this resource requirement has exceeded the capacity of main frame, then select on it The virtual machine (vm) migration run is gone out, until this overload main frame returns to normal operating conditions；Without finding overload main frame then It is directly entered step b；
Step b: the adjustment of physical host that cloud data center resource blocks: resource according to each main frame in destination host set Service condition selects physical host to be adjusted；Then resource consumption is selected to use type with its resource from these physical hosts Identical virtual machine (vm) migration is gone out, and improves the harmony that this physical host resource uses；
Step c: the selection of cloud data center underloading main frame: for the physical host that cloud data center resource utilization rate is relatively low, root According to the quantity of the virtual machine run in its resource utilization state and its, select suitable physical host and by all void thereon Plan machine moves on the physical host that resource utilization is of a relatively high, is then shut off this physical host to reduce the number of mobile host computers Amount；If the virtual machine on underloading main frame existing one or more virtual machine can not find available host, then this main frame On virtual machine do not migrate, continue to run with；
Step d: the placement of virtual machine to be migrated in cloud data center: by above-mentioned steps a, b, c need migrate virtual machine by Order is placed on suitable physical host, improves the resource utilization of each movable physical host in cloud data center.
2. a kind of efficient energy-saving square of the utilization of resources based on virtual machine (vm) migration in cloud data center as described in claim 1 Method, it is characterised in that the concrete adjustment process of the resource blockage physical host in described step b is as follows:
(2a), according to the summation of the resource requirement in cloud data center and the capacity of physical host, calculate and meet in theory Quantity N of the physical host required for current resource demand；
(2b), according to the resource service condition of physical host, for one score value of each Framework computing in data center, calculate Formula is as follows:
s c o r e = 1 ( 1 - 2 &CenterDot; d ) &CenterDot; x + ( 1 + 2 &CenterDot; d ) &CenterDot; y 2 |
X in above formula represents the cpu utilization rate of physical host, and y then represents the memory usage of this physical host, and d represents now The resource using status of physical host deviates the distance of its utilization of resources equilibrium state；Select the top n physics that score value is bigger Main frame, as destination host during virtual machine (vm) migration, constitutes destination host set；
(2c), for this N number of physical host, the resource using status not physical machine in acceptable equalization region ETA is selected, so After calculate these physical host resource utilizations deviation equilibrium state meansigma methodsSelecting resource utilization, to deviate it equal Distance d of weighing apparatus state is more thanPhysical host as main frame to be adjusted；
(2d), for these main frames to be adjusted, according to d value order from big to small, it is processed；If this physics master Resource service condition in machine is Pcpu＞ Pmem, wherein PcpuAnd PmemRepresent cpu and the memory usage of physical host respectively, then The resources of virtual machine conditions of demand migrated are selected to should be Vcpu＞ Vmem, wherein VcpuAnd VmemRepresent the cpu of virtual machine and interior respectively Deposit utilization rate, and the resource service condition of this physical host meet following formula:
| newUtilization P - oldUtilization P | < oldUtilization P m
OldUtilization in above formulaPAnd newUtilizationPRepresent this physical host before and after this virtual machine (vm) migration respectively Average resource, m then represents the quantity of the virtual machine transported on this main frame；When the resource service condition of physical host is Pcpu＜ PmemTime, disposition is similar to.
3. a kind of efficient energy-saving square of the utilization of resources based on virtual machine (vm) migration in cloud data center as described in claim 1 Method, it is characterised in that the detailed process of the underloading Selection of chiller in described step c is as follows:
The size of the score value according to each physical host, for being not at the physical host in destination host set, calculates it SF value:
S F = predictedUtilization P currentUtilization P &times; m
PredictedUtilization in above formulaPAnd currentUtilizationPRepresent resource on this physical host respectively The meansigma methods of usage forecast and the meansigma methods of currently practical service condition, m then represents the virtual machine run on this physical host Quantity；So, just physical host less for SF value is judged to underloading main frame.
4. a kind of efficient energy-saving square of the utilization of resources based on virtual machine (vm) migration in cloud data center as described in claim 1 Method, it is characterised in that the detailed process of the virtual machine Placement Strategy in described step d is as follows:
(4a), for resource blockage physical host needs the virtual machine migrated out, preferentially they are placed into treating of other On the physical host adjusted so that the physical host migrating virtual machine and the resource using status of the physical host accepting virtual machine Fall in ETA or make d value reduce；If there is not such physical host, then this virtual machine is placed into not at destination host On physical host in set so that their resource utilization status falls in ETA or makes d value reduce, and by this physics master Machine joins in destination host set；If still can not find the physical host of placement, then abandon the adjustment to this physical host；
(4b), for the virtual machine migrated out on poor efficiency main frame, from destination host set, main frame is selected, if energy This host resource is enough made to use state to fall in ETA, then this main frame just accepts this virtual machine；Otherwise, just look for so that accepting to be somebody's turn to do The physical host that after virtual machine, d value is minimum is as destination host；When the physical host in destination host set can not receive void again After plan machine, then look for the maximum physical host of SF value as host to receive other physical host in remaining physical host On virtual machine, until do not have physical host to be closed.
CN201610460901.8A 2016-06-23 2016-06-23 The efficient power-economizing method of the utilization of resources in cloud data center based on virtual machine (vm) migration Active CN106125888B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201610460901.8A CN106125888B (en) 2016-06-23 2016-06-23 The efficient power-economizing method of the utilization of resources in cloud data center based on virtual machine (vm) migration

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201610460901.8A CN106125888B (en) 2016-06-23 2016-06-23 The efficient power-economizing method of the utilization of resources in cloud data center based on virtual machine (vm) migration

Publications (2)

Publication Number Publication Date
CN106125888A true CN106125888A (en) 2016-11-16
CN106125888B CN106125888B (en) 2018-12-25

Family

ID=57268863

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201610460901.8A Active CN106125888B (en) 2016-06-23 2016-06-23 The efficient power-economizing method of the utilization of resources in cloud data center based on virtual machine (vm) migration

Country Status (1)

Country Link
CN (1) CN106125888B (en)

Cited By (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107608754A (en) * 2017-07-31 2018-01-19 郑州云海信息技术有限公司 A kind of dispatching method of virtual machine and device
CN108572868A (en) * 2017-09-25 2018-09-25 北京金山云网络技术有限公司 A kind of physical host selection method, device, equipment and readable storage medium storing program for executing
WO2019007420A1 (en) * 2017-07-07 2019-01-10 中兴通讯股份有限公司 Load balance scheduling method and device, and computer readable storage medium
CN109976879A (en) * 2019-03-29 2019-07-05 南京邮电大学 A kind of cloud computing virtual machine placement method using curve complementation based on resource
CN111443791A (en) * 2020-03-24 2020-07-24 深信服科技股份有限公司 UPS monitoring method, equipment, storage medium and device based on cloud platform
CN112540896A (en) * 2020-12-08 2021-03-23 南京天朗防务科技有限公司 Automatic VxWorks program distinguishing and running method
CN113687779A (en) * 2021-07-29 2021-11-23 济南浪潮数据技术有限公司 Data migration method and device, electronic equipment and readable storage medium
CN116743589A (en) * 2023-08-15 2023-09-12 中移(苏州)软件技术有限公司 Cloud host migration method and device and electronic equipment

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102117226A (en) * 2011-03-18 2011-07-06 浪潮(北京)电子信息产业有限公司 Resource dispatching system and resource dispatching method
CN102571746A (en) * 2011-11-23 2012-07-11 西安交通大学 Virtual machine deployment method oriented to side channel attack defense of cloud computation environment
CN103677958A (en) * 2013-12-13 2014-03-26 华为技术有限公司 Virtualization cluster resource scheduling method and device
US20140164364A1 (en) * 2012-12-06 2014-06-12 Ca, Inc. System and method for event-driven prioritization
CN105159751A (en) * 2015-09-17 2015-12-16 河海大学常州校区 Energy-efficient virtual machine migration method in cloud data center

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102117226A (en) * 2011-03-18 2011-07-06 浪潮(北京)电子信息产业有限公司 Resource dispatching system and resource dispatching method
CN102571746A (en) * 2011-11-23 2012-07-11 西安交通大学 Virtual machine deployment method oriented to side channel attack defense of cloud computation environment
US20140164364A1 (en) * 2012-12-06 2014-06-12 Ca, Inc. System and method for event-driven prioritization
CN103677958A (en) * 2013-12-13 2014-03-26 华为技术有限公司 Virtualization cluster resource scheduling method and device
CN105159751A (en) * 2015-09-17 2015-12-16 河海大学常州校区 Energy-efficient virtual machine migration method in cloud data center

Cited By (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2019007420A1 (en) * 2017-07-07 2019-01-10 中兴通讯股份有限公司 Load balance scheduling method and device, and computer readable storage medium
CN107608754A (en) * 2017-07-31 2018-01-19 郑州云海信息技术有限公司 A kind of dispatching method of virtual machine and device
CN108572868A (en) * 2017-09-25 2018-09-25 北京金山云网络技术有限公司 A kind of physical host selection method, device, equipment and readable storage medium storing program for executing
CN109976879A (en) * 2019-03-29 2019-07-05 南京邮电大学 A kind of cloud computing virtual machine placement method using curve complementation based on resource
CN109976879B (en) * 2019-03-29 2022-09-16 南京邮电大学 Cloud computing virtual machine placement method based on resource usage curve complementation
CN111443791A (en) * 2020-03-24 2020-07-24 深信服科技股份有限公司 UPS monitoring method, equipment, storage medium and device based on cloud platform
CN111443791B (en) * 2020-03-24 2022-04-29 深信服科技股份有限公司 UPS monitoring method, equipment, storage medium and device based on cloud platform
CN112540896B (en) * 2020-12-08 2024-04-19 南京天朗防务科技有限公司 Automatic judging and running method for VxWorks program
CN112540896A (en) * 2020-12-08 2021-03-23 南京天朗防务科技有限公司 Automatic VxWorks program distinguishing and running method
CN113687779A (en) * 2021-07-29 2021-11-23 济南浪潮数据技术有限公司 Data migration method and device, electronic equipment and readable storage medium
CN113687779B (en) * 2021-07-29 2024-02-23 济南浪潮数据技术有限公司 Data migration method, device, electronic equipment and readable storage medium
CN116743589B (en) * 2023-08-15 2023-11-21 中移(苏州)软件技术有限公司 Cloud host migration method and device and electronic equipment
CN116743589A (en) * 2023-08-15 2023-09-12 中移(苏州)软件技术有限公司 Cloud host migration method and device and electronic equipment

Also Published As

Publication number Publication date
CN106125888B (en) 2018-12-25

Similar Documents

Publication Publication Date Title
CN106125888A (en) The efficient power-economizing method of the utilization of resources based on virtual machine (vm) migration in cloud data center
Sun et al. A new technique for efficient live migration of multiple virtual machines
Yadav et al. Managing overloaded hosts for energy-efficiency in cloud data centers
CN103365700B (en) A kind of facing cloud calculates monitoring resource and the adjustment System of virtualized environment
Lu et al. An effective task scheduling algorithm based on dynamic energy management and efficient resource utilization in green cloud computing environment
CN103823718B (en) Resource allocation method oriented to green cloud computing
CN104008018B (en) The online moving method of virtual machine under cloud computing environment
CN104407912A (en) Virtual machine configuration method and device
Chen et al. ERECT: Energy-efficient reactive scheduling for real-time tasks in heterogeneous virtualized clouds
Liang et al. Memory-aware resource management algorithm for low-energy cloud data centers
WO2015032201A1 (en) Virtual machine placing method and device
CN106656555A (en) Dynamic adjustment method of service resources of cloud computing system
WO2015188504A1 (en) Data centre solution method for decomposition-coordination calculation of block bordered diagonal frame model
He et al. Energy-efficient framework for virtual machine consolidation in cloud data centers
Song et al. Server consolidation energy-saving algorithm based on resource reservation and resource allocation strategy
Zhang et al. An energy-aware host resource management framework for two-tier virtualized cloud data centers
Liang et al. A high-applicability heterogeneous cloud data centers resource management algorithm based on trusted virtual machine migration
CN105306547A (en) Data placing and node scheduling method for increasing energy efficiency of cloud computing system
Zhou et al. Stability property of clouds and cooperative scheduling policies on multiple types of resources in cloud computing
Xu et al. Ufalloc: Towards utility max-min fairness of bandwidth allocation for applications in datacenter networks
Lin et al. Energy-efficient vm placement algorithms for cloud data center
Song et al. Hypergraph-partitioning-based online joint scheduling of tasks and data
Li et al. Research on Dynamic Virtual Machine Scheduling Strategy Based on Improved Genetic Algorithm
Yezdani et al. Power and Performance Issues and Management Approaches in Cloud Computing
Sun et al. An algorithm towards energy-efficient scheduling for real-time tasks under cloud computing environment

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant