CN107423114A - A kind of dynamic migration of virtual machine method based on classification of service - Google Patents
A kind of dynamic migration of virtual machine method based on classification of service Download PDFInfo
- Publication number
- CN107423114A CN107423114A CN201710582510.8A CN201710582510A CN107423114A CN 107423114 A CN107423114 A CN 107423114A CN 201710582510 A CN201710582510 A CN 201710582510A CN 107423114 A CN107423114 A CN 107423114A
- Authority
- CN
- China
- Prior art keywords
- virtual machine
- service
- internal memory
- utilization rate
- migrated
- 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
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/44—Arrangements for executing specific programs
- G06F9/455—Emulation; Interpretation; Software simulation, e.g. virtualisation or emulation of application or operating system execution engines
- G06F9/45533—Hypervisors; Virtual machine monitors
- G06F9/45558—Hypervisor-specific management and integration aspects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/44—Arrangements for executing specific programs
- G06F9/455—Emulation; Interpretation; Software simulation, e.g. virtualisation or emulation of application or operating system execution engines
- G06F9/45533—Hypervisors; Virtual machine monitors
- G06F9/45558—Hypervisor-specific management and integration aspects
- G06F2009/4557—Distribution of virtual machine instances; Migration and load balancing
Landscapes
- Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Debugging And Monitoring (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The present invention provides a kind of dynamic migration of virtual machine method based on classification of service, and its step includes:1) the synthesis utilization rate of virtual machine is obtained;2) when the synthesis utilization rate of above-mentioned virtual machine is more than threshold value, the synthesis utilization rate that virtual machine respectively services is assessed, each service is classified according to division demand, and according to classification results selection target virtual machine;Wherein described threshold value and division demand are defined by keeper according to different scenes;3) related resource of the above-mentioned partial service by classification is moved into target virtual machine.The present invention considers to analyze virtual machine service condition, only migrates partial service and user.Therefore, the inventive method can not only improve overall migration performance, and can improve the service efficiency of resource.
Description
Technical field
The present invention relates to virtual machine migration method, more particularly to a kind of dynamic migration of virtual machine side based on classification of service
Method.
Background technology
In recent years, a large amount of popularization and applications of cloud computing, with the increase of usage time, the consumption of resource and scarcity are also increasingly bright
It is aobvious.At present, most of method for using virtual machine (vm) migration, entirely virtual machine (VM) will be migrated.But if cloud platform
Middle resources of virtual machine is limited, and the resource that possibly can not provide abundance carries out bulk migration.And virtual machine is as in cloud computing environment
Important component, using and running for user's related resource is carry, also as the core of transition process.Virtual machine (vm) migration mistake
Cold migration and thermophoresis are broadly divided into journey, thermophoresis is also referred to as dynamic migration.Cold migration is integrally moved after virtual machine is shut down
Move, but influence the use of user;Dynamic migration is the migration that virtual machine is completed in the case where user does not discover, and it ensures
In transition process, still smoothness run, user will not perceive any difference to virtual machine, therefore as the weight of research and application
Point.
Existing dynamic migration mode mainly includes premigrates, rear migration etc..Wherein premigrates is most commonly used mode.
Premigrates is primarily referred to as in transition process, and by constantly circulating, the internal storage data of source virtual machine is sent to target virtual machine.
Pre-copy mechanism greatly reduces the internal storage data amount that the copy stage needs of shutdown transmit, so as to which downtime be contracted significantly
It is small.However, for the very fast memory part of renewal speed, each cyclic process all can be dirty, it is necessary to repeat to copy, while
Cause cycle-index very more, the time of migration is elongated.After migrate and mainly first stop and by CPU state in source virtual machine and energy
The minimum working condition for making virtual machine resume operation is sent to target virtual machine, starts target virtual machine afterwards and brings into operation, and
The page of missing is progressively migrated.But this method due in target virtual machine page faults have a strong impact on maneuverability
Can, cause service performance to decline, therefore use few.
During actual migration, same virtual machine provides different services for multiple users, but when user's visit capacity increases
Greatly, it is necessary to be migrated to resource during inadequate resource.Traditional method is that virtual machine bulk migration is bigger to a resource
In virtual machine, but satisfactory virtual machine on the one hand may be not present, on the other hand migration causes the time longer completely.
The content of the invention
In view of the above-mentioned problems, the present invention considers to analyze virtual machine service condition, only partial service and user are moved
Move.Therefore, the present invention provides a kind of dynamic migration of virtual machine method based on classification of service, and this method can not only improve overall
Migration performance, and the service efficiency of resource can be improved.
Fig. 1 is refer to, the present invention mainly on the basis of premigrates, is effectively divided for service in transition process, will
The relatively independent resource in part is migrated in service, and can effectively be combined with existing method, improves dynamic migration
During efficiency.Wherein, the VMM in the figure represents monitor of virtual machine.
For the above-mentioned purpose, the technical solution adopted in the present invention is:
A kind of dynamic migration of virtual machine method based on classification of service, its step include:
1) the synthesis utilization rate of virtual machine is obtained;
2) when the synthesis utilization rate of above-mentioned virtual machine is more than threshold value, the synthesis utilization rate that virtual machine respectively services is assessed, according to
Each service is classified according to division demand, and according to classification results selection target virtual machine;Wherein described threshold value and division need
Ask and defined by keeper according to different scenes;
3) related resource of the above-mentioned partial service by classification is moved into target virtual machine.
Further, also include before step 1):Each virtual machine in the scheduled time is monitored by monitor of virtual machine
Related resource service condition;Wherein described virtual machine related resource service condition includes CPU, internal memory, the size of hard disk and predetermined
The average service rate of CPU, internal memory, hard disk in time.
Further, comprehensive utilization rate Φ=α Cr+ β Mr+ γ Hr described in step 1), wherein α, β, γ is shadow
Weight is rung, and is defined by keeper according to different scenes;Cr, Mr, Hr refer respectively to CPU in the scheduled time, internal memory, hard disk it is flat
Equal utilization rate.
Further, step 2) also includes:If without suitable target virtual machine, each service is divided again
Class.
Further, the CPU of target virtual machine described in step 2), internal memory, the free space size of hard disk are both greater than treated
σ times of migration service set utilization rate;Wherein σ >=2, the set of service to be migrated refer to wait to move according to the division of division demand
The set of service of shifting.
Further, in step 2) when the synthesis utilization rate of above-mentioned virtual machine is more than threshold value, following steps are performed:
2-1) calculate the synthesis utilization rate that respectively services in virtual machine, and it is descending be ranked up, according to ranking results and
Division demand is classified each service;
2-2) according to above-mentioned classification results selection target virtual machine.
Further, migrated in step 3) according to the order of the synthesis utilization rate of each service after sequence, and root
According to the related resource situation of change serviced in the scheduled time of monitoring, it is determined that the sequencing of the related resource migration of service.
Further, when the related resource of service does not change, then directly migrated;When the correlation money of service
When source changes, then the service to change is monitored, and be ranked up according to service change speed, according to service
The principle of related resource change from less to more is migrated;Wherein described related resource includes host name (id), CPU total amounts
(Cs), CPU usage amounts (Cu), memory amount (Ms), internal memory usage amount (Mu), hard disk total amount (Hs), hard disk usage amount (Hu).
Further, step 3) also includes:Hard disk is arranged to shared storage.
Further, step 3) also includes:Fast transferring is concentrated using pre-copy method to internal memory;It is wherein described pre- to copy
Shellfish method concentrates fast transferring to refer to:Internal memory changing unit is marked using shadow EMS memory, and internal memory changing unit is copied
Shellfish to free memory region is constantly migrated, until the content of internal memory changing unit can be migrated disposably in the τ times;Wherein
τ is defined by keeper.
The beneficial effects of the present invention are:The present invention provides a kind of dynamic migration of virtual machine method based on classification of service,
This method was after virtual machine related resource service condition was collected, it is determined whether needs to migrate and need the target migrated.If
Need to migrate, then different services are classified, the related resource migration for effectively selecting suitable virtual machine to be serviced, significantly
Improve the efficiency of migration.And hard disk related data is arranged to shared storage, facilitates data sharing by this method in transition process
With storage;In terms of internal memory, internal memory is integrally migrated, then by shadow EMS memory, internal memory changing unit be marked,
And internal memory changing unit is constantly copied to free memory region, monoblock type migration is carried out afterwards, until last synchronous.
Brief description of the drawings
Fig. 1 is that the related resource of service provided by the invention migrates schematic diagram.
Fig. 2 is a kind of dynamic migration of virtual machine method flow diagram based on classification of service provided by the invention.
Fig. 3 maps to free core for the internal memory changing unit and internal memory changing unit marked in one embodiment of the invention
Domain schematic diagram.
Embodiment
To enable the features described above of the present invention and advantage to become apparent, special embodiment below, and coordinate institute's accompanying drawing to make
Describe in detail as follows.
The present invention provides a kind of dynamic migration of virtual machine method based on classification of service, refer to Fig. 2, this method step bag
Include:
1st, the synthesis utilization rate of virtual machine is obtained.
The service condition in each virtual machine scheduled time is carried out by the monitor of virtual machine of virtual machine monitoring platform
Monitoring, to obtain virtual machine related resource service condition, and obtains the comprehensive of virtual machine according to virtual machine related resource service condition
Close utilization rate.The wherein described virtual machine related resource service condition of obtaining mainly is included to CPU in virtual machine, internal memory, hard disk
The size and feature such as average service rate of CPU, internal memory, hard disk is extracted in the scheduled time, build the characteristic set of correlation, side
Just made a decision for following resource service condition.
2nd, when the synthesis utilization rate of above-mentioned virtual machine is more than threshold value, the synthesis utilization rate that virtual machine respectively services is assessed, according to
Each service is classified according to division demand, and according to classification results selection target virtual machine (i.e. suitable free virtual machine).
The synthesis utilization rate of virtual machine to be obtained according to virtual machine related resource service condition sets corresponding threshold value, when
, it is necessary to assess the synthesis utilization rate that virtual machine respectively services during more than threshold value, each service is classified according to division demand;Wherein
The threshold value and division demand are defined by keeper according to different scenes.Meanwhile free virtual machine resource is assessed, to phase
Resource is closed to be migrated.
3rd, transition process.
According to the related resource situation of change serviced in the scheduled time of monitoring, determine that the priority of related resource migration is suitable
Sequence.In transition process, when the related resource of service does not change, then directly migrated;When the related resource of service
, it is necessary to be monitored to the service to change when changing, and it is ranked up according to service change speed, according to service
The principle of related resource change from less to more is migrated successively, until terminating.In specific transition process, in terms of hard disk,
Hard disk related data is arranged to shared storage, facilitates data sharing and storage;In terms of internal memory, internal memory is integrally moved
Move, then by shadow EMS memory, internal memory changing unit is marked, and internal memory changing unit is constantly copied to free memory
Region, monoblock type migration is carried out afterwards, until last synchronous.
Name a specific embodiment and further explain the inventive method.The step of embodiment, includes:
1st, virtual machine related resource service condition is obtained, and then obtains the synthesis utilization rate of virtual machine.Due to virtual frame
In structure, a monitor of virtual machine is usually constructed with to realize centralized management, the raising service to physics and virtual IT foundation structures
Device utilization rate and across multiple virtual platforms optimize dynamic resource.Therefore, the present invention can be read by monitor of virtual machine
Each virtual machine related resource service condition Ε={ ξ } in platform.Libvirt provides a kind of API, and master is run on for safety management
Client operating system on machine.It is a kind of DLL for realizing virtual machine chemical industry tool, supports various virtual machine monitors,
Such as Xen, KVM, QEMU.Meanwhile libvirt can receive user instruction and call or monitoring virtual machine in corresponding instruction and
Service.The present invention, with reference to C and Python, writes correlative code and checks virtual machine phase mainly by calling libvirt interfaces
Close resource service condition.
First, present invention is primarily concerned with virtual machine related resource include host name (id), CPU total amounts (Cs), CPU use
Measure (Cu), memory amount (Ms), internal memory usage amount (Mu), hard disk total amount (Hs), hard disk usage amount (Hu) etc..Due to only extracting
The service condition at CPU, internal memory, hard disk a certain moment can not represent the situation in the one section time, such as may be sometime
Peak value, but remaining time do not change.Therefore, the present invention choose moment t1 and t2 obtain respectively Cu1, Cu2, Mu1, Mu2,
Hu1, Hu2, and calculate its average service rate and represent the virtual machine related resource service condition Cr, Mr, Hr in a period of time.Typically
T1 to t2 time interval is fixed (t1 to t2 time interval to be preset, when being collectively referred to as default below
Between), in this embodiment, its time interval default setting is 30 seconds.If specific demand, keeper can be right according to demand
Time interval is adjusted.Then calculate t1 the and t2 times in average service rate Cr=f (Cu1, Cu2)/Cs, Mr=f (Mu1,
Mu2)/Ms, Hr=f (Hu1, Hu2)/Hs, wherein f (x, y)=| x+y |/2.Because hard disk total amount and memory amount are transported in program
It will not be changed between the departure date, so not taking the mean.By features described above construction feature vector ξt=<id,Cs,Cr,Ms,Mr,
Hs,Hr>.In subsequent descriptions, Cr, Mr, Hr are the average service rate in the scheduled time, no longer carry out repeat specification.
2nd, when the embodiment mainly considers no satisfactory target virtual machine Direct Transfer, considering will be to be migrated virtual
Related service in machine is divided, and is carried out part migration, original virtual machine (virtual machine i.e. to be migrated) is met
The service requirement of part of service, and another part service can also normally transport in other satisfactory target virtual machines
OK, migration is realized while not influenceing to service normal operation.Service is divided into two parts (m=2) by the embodiment, will wherein
A part moves to a virtual machine instance and illustrated, and can also migrate if needed to multiple virtual machines, here without
It is described in detail.
The synthesis utilization rate Φ of virtual machine is calculated first.Consider the service condition of CPU, internal memory and hard disk, synthesis makes
With rate Φ=α Cr+ β Mr+ γ Hr, wherein α, β, γ is weighing factor, i.e. pass of the keeper to CPU, internal memory and hard disk
Note degree, it can be defined according to different scenes by keeper;Generally α, β, γ ∈ [0,1], and alpha+beta+γ=1.By
Typically easily extended in hard disk, so hard disk can set relatively low weight;And internal memory and CPU be using more, so internal memory and CPU
It is the core for influenceing dynamic migration of virtual machine whole efficiency, therefore higher weights can be set.In this embodiment, α, β, γ are set
The default value put is respectively α=0.4, β=0.4, γ=0.2.Meanwhile keeper defines threshold value Ψ, generally Ψ ∈
[0.5,1], i.e., when a virtual machine related resource service condition exceedes half, the virtual machine is in larger load condition.Work as Φ>
During Ψ, then need to start migration flow.WhereinAndIt is respectively to be serviced in virtual machine
Synthesis utilization rate, be calculated according to Φ computational methods.
Before migration starts, it is necessary to be carried out to the synthesis utilization rate of each service in virtual machine to be migrated in partition process
Statistics, is calculated according to Φ computational methodsThen it is descending to be ranked up.In assorting process
In, the present invention follows the principle for making migration service number less as far as possible, therefore adds up since being taken the maximum virtual machine of resource,
It is more than half using total number resource and when being number minimum when taking resource, then is divided into two classes, i.e.,With The present invention selects less service to be migrated, that is, works as i>During n/2, then Re=T2 is migrated, work as i
During≤n/2, then Re=T1 is migrated.Wherein Re is ready-portioned set of service to be migrated.If the target virtual machine of migration is more than 2
When individual, then the class number of service is divided into corresponding virtual machine number, I will not elaborate.
Determine whether suitable free virtual machine as target virtual machine, the characteristic vector ξ of the target virtual machinet'=<
Id ', Cs ', Cr ', Ms ', Mr ', Hs ', Hr '>, it is necessary to meet { Cs ' (1-Cr ')>σ·Cs-Re·Cr-Re∧ { Ms ' (1-
Mr ')>σ·Ms-Re·Mr-Re∧ { Hs ' (1-Hr ')>σ·Hs-Re·Hr-Re, i.e. target virtual machine free time CPU, free time in
Deposit, the size of free hard disk need to be more than virtual machine to be migrated in set of service (set of service i.e. to be migrated) utilization rate σ (σ >=
2) it is just eligible above again.Wherein Cs-Re、Cr-Re、Ms-Re、Mr-Re、Hs-Re、Hr-ReSet of service Re to be migrated is represented respectively
Middle CPU, internal memory and hard disk total amount and utilization rate.If suitable virtual machine is then migrated, if without suitable void
Plan machine migrates again, it is necessary to (service is divided into m+1 classes) after classifying again to service.
3rd, it is necessary to treat CPU that migration service set Re={ s1, s2 ..., sn } is related to, interior in specific transition process
Deposit, the related resource such as hard disk carries out comprehensively monitoring, and the order of migration is determined according to situation of change.
Judge firstly the need of the related resource situation of change to being serviced in the scheduled time, service si (i=1,2 ... n)
Related resource situation of change Θ si=<Cr-si,Mr-si,Hr-si>, wherein Cr-si、Mr-si、Hr-siRepresent service si correlations
CPU, internal memory and hard disk average service rate, its computational methods are identical with step 1.If Θ si=Ф, CPU, internal memory and hard
Disk does not change, then carries out Direct Transfer.If the related resource of service changes, comprehensive according to related resource becomes
Rate is ranked up from small to large, and the service Re ' after sequence are migrated successively.Certainly, the present invention assigns to parameters
Different weights, such as when the influence for thinking internal memory rate of change is bigger than hard disk, then higher weights are assigned to internal memory, in terms of this
Corresponding value is calculated to be compared.The present invention will use the weighing factor for the synthesis utilization rate Φ that virtual machine is calculated in step 2, also just
In comprehensive change rate fc (Θ si)=α Cc+ β Mc+ γ Hc of the front and rear related resource for being uniformly compared, that is, servicing.
In transition process, hard disk is first arranged to shared storage, facilitate reduced in data migration process interaction times and
Time;Because CPU state only accounts for the very small part of migration, therefore the main transition process that internal memory is discussed.In terms of internal memory
Using pre-copy method, but in combination with shadow EMS memory, monoblock type fast transferring is carried out to internal memory to be migrated, improves migration effect
Rate.Shadow EMS memory is carried out corresponding mainly by one section of space of distribution with internal memory appropriate section in the present invention, to internal memory change section
Divide and be marked.Shadow EMS memory is mainly relevant with internal memory change monitoring granularity with actual memory corresponding relation, if in migration
Depositing is monitored and migrates for unit according to byte, then 1 byte also corresponds to the 1bit in shadow EMS memory, ratio 8:1.When
So, in order to improve efficiency, the present invention is monitored according to the size of page, then is the corresponding shadow of 1 page (such as 4KB memory headrooms)
1bit in sub- internal memory.When the internal memory unit of monitoring changes, corresponding shadow EMS memory corresponding part is labeled as 1, otherwise
For 0.In subsequent migration the part for being will be marked to carry out copy migration in shadow EMS memory.It is implemented as follows:
First, overall be copied in target virtual machine is carried out to internal memory to be migrated.Then, (wait to move in original virtual machine
Move virtual machine) in the corresponding shadow EMS memory of division the internal memory changing unit since migration is marked.Drawn in memory part
A part of free memory is separated, the memory part content to change is uniformly pointed into free memory, then migrated once
Afterwards, the mark change memory part of division is disposably migrated.Meanwhile then is pointed to again to internal memory changing unit the free time
Internal memory, circulate always, until internal memory to be migrated is sufficiently small, can have been migrated within the τ times, then suspend virtual machine, disposably move
After shifting, whole transition process is completed.Wherein, τ can be defined by keeper, usually minimum time span, such as 0.01s.
Fig. 3 is refer to, after internal memory starts migration, the internal memory to change is marked by shadow EMS memory, such as
0x23F910, Ox23F914, Ox23FF2C ... ....After bulk migration is complete, the internal memory that these change is mapped to interior
The continuum that middle clear area 0xFFFF00 starts is deposited, is then again migrated the memory part to change.With this
Meanwhile also the region to be changed in former internal memory is marked and is mapped to clear area again and be migrated.This mistake
Cheng Xunhuan is carried out, until the memory part to change is sufficiently small, can disposably have been migrated within the τ times.
In summary, present invention mainly solves when resources of virtual machine deficiency, and the idle sufficient virtual machine of neither one supplies
During migration, a kind of dynamic migration of virtual machine method based on classification of service is proposed.This method makes in acquisition virtual machine related resource
On the basis of situation, the synthesis utilization rate of virtual machine is monitored;When the synthesis utilization rate of virtual machine is more than given threshold value
When, each service in virtual machine is classified, and suitable virtual machine is selected, the related resource of partial service is entered into Mobile state and moved
Move, meet user's use demand.In addition, when there is sufficiently large free virtual machine, can also be directly using this method by virtually
Machine dynamic migration.
Implement to be merely illustrative of the technical solution of the present invention rather than be limited above, the ordinary skill people of this area
Technical scheme can be modified by member or equivalent substitution, without departing from the spirit and scope of the present invention, this hair
Bright protection domain should be to be defined described in claims.
Claims (10)
1. a kind of dynamic migration of virtual machine method based on classification of service, its step include:
1) the synthesis utilization rate of virtual machine is obtained;
2) when the synthesis utilization rate of above-mentioned virtual machine is more than threshold value, the synthesis utilization rate that virtual machine respectively services is assessed, foundation is drawn
Point demand is classified to each service, and according to classification results selection target virtual machine;Wherein described threshold value and division demand by
Keeper defines according to different scenes;
3) related resource of the above-mentioned partial service by classification is moved into target virtual machine.
2. the method as described in claim 1, it is characterised in that also include before step 1):Supervised by monitor of virtual machine
Control each virtual machine related resource service condition in the scheduled time;Wherein described virtual machine related resource service condition include CPU,
Internal memory, the size of hard disk and in the scheduled time CPU, internal memory, hard disk average service rate.
3. the method as described in claim 1, it is characterised in that comprehensive utilization rate Φ=α Cr+ β Mr+ described in step 1)
γ Hr, wherein α, β, γ are weighing factor, and are defined by keeper according to different scenes;Cr, Mr, Hr refer respectively to pre- timing
Interior CPU, internal memory, the average service rate of hard disk.
4. the method as described in claim 1, it is characterised in that step 2) also includes:If without suitable target virtual machine,
Then each service is classified again.
5. the method as described in claim 1, it is characterised in that CPU, internal memory, the hard disk of target virtual machine described in step 2)
Free space size be both greater than σ times of set of service utilization rate to be migrated;Wherein σ >=2, the set of service to be migrated refer to
Set of service to be migrated according to division demand division.
6. the method as described in claim 1, it is characterised in that the synthesis utilization rate in step 2) when above-mentioned virtual machine is more than threshold
During value, following steps are performed:
2-1) calculate the synthesis utilization rate that respectively services in virtual machine, and it is descending be ranked up, according to ranking results and division
Demand is classified each service;
2-2) according to above-mentioned classification results selection target virtual machine.
7. method as claimed in claim 6, it is characterised in that according to the synthesis utilization rate of each service after sequence in step 3)
Order migrated, and the related resource situation of change serviced in the scheduled time according to monitoring, it is determined that the related money of service
The sequencing of source migration.
8. method as claimed in claim 7, it is characterised in that when the related resource of service does not change, then directly enter
Row migration;When the related resource of service changes, then the service to change is monitored, and it is fast according to service change
Slowly it is ranked up, is migrated according to the principle of the related resource of service change from less to more;Wherein described related resource includes
Host name, CPU total amounts, CPU usage amounts, memory amount, internal memory usage amount, hard disk total amount, hard disk usage amount.
9. method as claimed in claim 3, it is characterised in that step 3) also includes:Hard disk is arranged to shared storage.
10. method as claimed in claim 3, it is characterised in that step 3) also includes:Internal memory is concentrated using pre-copy method
Fast transferring;Wherein described pre-copy method concentrates fast transferring to refer to:Rower is entered to internal memory changing unit using shadow EMS memory
Note, and internal memory changing unit is copied to free memory region and constantly migrated, when the content of internal memory changing unit is in τ
Between can disposably migrate;Wherein τ is defined by keeper.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710582510.8A CN107423114B (en) | 2017-07-17 | 2017-07-17 | Virtual machine dynamic migration method based on service classification |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710582510.8A CN107423114B (en) | 2017-07-17 | 2017-07-17 | Virtual machine dynamic migration method based on service classification |
Publications (2)
Publication Number | Publication Date |
---|---|
CN107423114A true CN107423114A (en) | 2017-12-01 |
CN107423114B CN107423114B (en) | 2020-12-22 |
Family
ID=60430105
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710582510.8A Active CN107423114B (en) | 2017-07-17 | 2017-07-17 | Virtual machine dynamic migration method based on service classification |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107423114B (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108549572A (en) * | 2018-04-08 | 2018-09-18 | 郑州云海信息技术有限公司 | A kind of heterogeneous server virtual management system and method for data center's total management system |
CN108874535A (en) * | 2018-05-14 | 2018-11-23 | 中国平安人寿保险股份有限公司 | A kind of task adjusting method, computer readable storage medium and terminal device |
CN112948067A (en) * | 2019-12-11 | 2021-06-11 | 北京金山云网络技术有限公司 | Service scheduling method and device, electronic equipment and storage medium |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20110239216A1 (en) * | 2010-03-23 | 2011-09-29 | Nec Corporation | Service providing system, a virtual machine server, a service providing method, and a program thereof |
CN105335234A (en) * | 2015-10-29 | 2016-02-17 | 贵州电网有限责任公司电力调度控制中心 | Method for immediately migrating virtual machine |
CN105607947A (en) * | 2015-12-11 | 2016-05-25 | 西北工业大学 | Novel cloud environment virtual machine scheduling method |
CN106250210A (en) * | 2016-08-03 | 2016-12-21 | 上海交通大学 | Dispatching method of virtual machine under cloud environment |
CN106406976A (en) * | 2016-07-21 | 2017-02-15 | 柏科数据技术(深圳)股份有限公司 | Method and apparatus for identifying IO intensive application in cloud computing environment |
-
2017
- 2017-07-17 CN CN201710582510.8A patent/CN107423114B/en active Active
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20110239216A1 (en) * | 2010-03-23 | 2011-09-29 | Nec Corporation | Service providing system, a virtual machine server, a service providing method, and a program thereof |
CN105335234A (en) * | 2015-10-29 | 2016-02-17 | 贵州电网有限责任公司电力调度控制中心 | Method for immediately migrating virtual machine |
CN105607947A (en) * | 2015-12-11 | 2016-05-25 | 西北工业大学 | Novel cloud environment virtual machine scheduling method |
CN106406976A (en) * | 2016-07-21 | 2017-02-15 | 柏科数据技术(深圳)股份有限公司 | Method and apparatus for identifying IO intensive application in cloud computing environment |
CN106250210A (en) * | 2016-08-03 | 2016-12-21 | 上海交通大学 | Dispatching method of virtual machine under cloud environment |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108549572A (en) * | 2018-04-08 | 2018-09-18 | 郑州云海信息技术有限公司 | A kind of heterogeneous server virtual management system and method for data center's total management system |
CN108874535A (en) * | 2018-05-14 | 2018-11-23 | 中国平安人寿保险股份有限公司 | A kind of task adjusting method, computer readable storage medium and terminal device |
CN108874535B (en) * | 2018-05-14 | 2022-06-10 | 中国平安人寿保险股份有限公司 | Task adjusting method, computer readable storage medium and terminal device |
CN112948067A (en) * | 2019-12-11 | 2021-06-11 | 北京金山云网络技术有限公司 | Service scheduling method and device, electronic equipment and storage medium |
Also Published As
Publication number | Publication date |
---|---|
CN107423114B (en) | 2020-12-22 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Priya et al. | Resource scheduling algorithm with load balancing for cloud service provisioning | |
CN106502792B (en) | A kind of multi-tenant priority scheduling of resource method towards different type load | |
CN103810016B (en) | Realize method, device and the group system of virtual machine (vm) migration | |
CN109324875B (en) | Data center server power consumption management and optimization method based on reinforcement learning | |
US8402140B2 (en) | Methods and apparatus for coordinated energy management in virtualized data centers | |
CN102236582B (en) | Method for balanced distribution of virtualized cluster load in a plurality of physical machines | |
CN101488098B (en) | Multi-core computing resource management system based on virtual computing technology | |
CN104102543B (en) | The method and apparatus of adjustment of load in a kind of cloud computing environment | |
CN107168782A (en) | A kind of concurrent computational system based on Spark and GPU | |
CN106598733A (en) | Three-dimensional virtual resource scheduling method of cloud computing energy consumption key | |
CN104636187B (en) | Dispatching method of virtual machine in NUMA architecture based on load estimation | |
WO2014029219A1 (en) | Method and device for integrating virtualized cluster, and virtualized cluster system | |
CN104503838A (en) | Method for scheduling virtual CPU (Central Processing Unit) | |
CN103365700A (en) | Cloud computing virtualization environment-oriented resource monitoring and adjustment system | |
CN103823714A (en) | Virtualization-based method and device for adjusting QoS (quality of service) of node memory of NUMA (non uniform memory access architecture) | |
CN102270159A (en) | Access controlling and load balancing method for virtualized environment | |
CN107423114A (en) | A kind of dynamic migration of virtual machine method based on classification of service | |
CN106775949A (en) | A kind of Application of composite feature that perceives migrates optimization method online with the virtual machine of the network bandwidth | |
CN109558216B (en) | Single root I/O virtualization optimization method and system based on online migration | |
CN106201700A (en) | The dispatching method that a kind of virtual machine migrates online | |
CN105740059B (en) | A kind of population dispatching method towards Divisible task | |
CN103106112A (en) | Method and device based on maximum load and used for load balancing scheduling | |
Emara et al. | Genetic-Based Multi-objective Task Scheduling Algorithm in Cloud Computing Environment. | |
CN104008001B (en) | Virtual machine dynamic migrating method applied to mass data support | |
CN105930202B (en) | A kind of virtual machine migration method of three threshold values |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |