CN102136993B - A kind of methods, devices and systems of Data Migration - Google Patents
A kind of methods, devices and systems of Data Migration Download PDFInfo
- Publication number
- CN102136993B CN102136993B CN201010243657.2A CN201010243657A CN102136993B CN 102136993 B CN102136993 B CN 102136993B CN 201010243657 A CN201010243657 A CN 201010243657A CN 102136993 B CN102136993 B CN 102136993B
- Authority
- CN
- China
- Prior art keywords
- internal memory
- band
- data
- dirty
- dirty data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 230000005012 migration Effects 0.000 title claims abstract description 65
- 238000013508 migration Methods 0.000 title claims abstract description 65
- 238000000034 method Methods 0.000 title claims abstract description 37
- 230000005540 biological transmission Effects 0.000 claims abstract description 22
- 230000011218 segmentation Effects 0.000 claims abstract description 13
- 230000006870 function Effects 0.000 claims 1
- 230000007423 decrease Effects 0.000 abstract description 12
- 230000002123 temporal effect Effects 0.000 abstract description 6
- 238000010586 diagram Methods 0.000 description 9
- 230000003362 replicative effect Effects 0.000 description 3
- 230000001174 ascending effect Effects 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 230000001154 acute effect Effects 0.000 description 1
- 230000033228 biological regulation Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000004806 packaging method and process Methods 0.000 description 1
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The embodiment of the invention discloses the methods, devices and systems of a kind of Data Migration, wherein the realization of method includes: calculate the rewriting rate of each storage band, and described storage band obtains by needing the memory space segmentation carrying out Data Migration；The dirty data in the storage band that rewriting rate is low is preferentially sent to destination host；Send after state reaches preset requirement at dirty data and terminate Data Migration.The memory space segmentation carrying out Data Migration will be needed to obtain storing band by using；Preferentially send the dirty data technological means in the storage band that rewriting rate is low to destination host, thus utilize spatial locality principle and temporal locality principle, postpone to send by the dirty data of memory space higher for rewriting rate, utilization is write merging and is absorbed more internal memory write operation, decrease the transmission of dirty data, thus decrease volume of transmitted data.
Description
Technical field
The present invention relates to technical field of memory, particularly to the methods, devices and systems of a kind of Data Migration.
Background technology
Popular business model has been had become as, during the offer of service, it is ensured that service by network offering services
Seriality most important.In order to ensure the seriality of service, can avoid providing service by the two-node cluster hot backup of application layer
Equipment break down the service disruption caused.But two-node cluster hot backup faces a lot of problem, such as: hot standby machine (produces with the machine of production
Machine refer to provide service main frame, hot standby machine is the backup host of production machine) between Internet protocol (Internet
Protocol, IP) address cannot keep consistent.And Intel Virtualization Technology is to solve this problem to provide new thinking.
Intel Virtualization Technology simulates multiple fictitious host computer (commonly referred to as virtual machine) in a physical host, each virtual
Machine runs operating system and the application program of oneself.Multiple virtual machines share the resource of physical host, the resource bag of physical host
Include processor, internal memory, network etc.；It is mutually isolated between virtual machine, each self-operating program.Server energy based on Intel Virtualization Technology
Enough it is greatly improved the service efficiency of physical resource, improves operation and the maintenance of service.
Packaging effects based on Intel Virtualization Technology, the real-time migration of virtual machine needs do not interrupting the situation that virtual machine runs
Under, by the virtual machine (vm) migration on a physical host to another physical host.Before and after real-time migration of virtual machine, in virtual machine
The every program run and network connect and all keep not interrupting continuously.The real-time migration of virtual machine has regulation physics load on host computers
Balance, improves resource utilization, reduces the advantages such as energy resource consumption, is one of most important technology in system virtualization.
Due to virtual machine during real-time migration the most out of service, therefore, by the state of virtual machine from a physics master
The difficulty that machine copies to another physical host maximum is how to replicate the virutal machine memory being continually changing.Real-time at virtual machine
In migration, mainly have three problems to need to solve: (1) reduce be migrated virtual machine in the downtime of all programs, the most virtual
Machine is not scheduled, the out-of-work time；(2) reduce the gross migration time, i.e. move to be migrated virtual machine in purpose from starting
The time of normal work is started on physical host；(3) reduce the transport overhead migrated network, i.e. migrate transmission data total amount.
At present, the main method of real-time migration of virtual machine has front duplication (Pre-copy) and rear duplication (Post-copy).Its
It is replicated in after migrating processes starts before in, periodically virutal machine memory is copied to purpose physical host.It is implemented as: first
First replicate virtual machine full memory；In each cycle the most later, the internal memory changed in replicating the previous cycle；Entering
Go after some replicative cycles, virtual machine is being migrated source hang-up；Finally replicate remaining internal memory and other virtual machine states；
After all states (internal memory, processor etc.) of virtual machine are copied to destination, virtual machine restarts.
The dirty page produced in the previous cycle due to each replicative cycle copies to destination, and volume of transmitted data is big；
For internal memory writes more application, the problem that volume of transmitted data is big is particularly acute.
Summary of the invention
The embodiment of the present invention provide a kind of Data Migration methods, devices and systems, utilize spatial locality principle and time
Between principle of locality, reduce the transmission of dirty data, thus reduce volume of transmitted data.
The embodiment of the present invention provides a kind of method of Data Migration, including:
Calculating the rewriting rate of each storage band, described storage band obtains by needing the memory space segmentation carrying out Data Migration
Arrive；
The dirty data in the storage band that rewriting rate is low is preferentially sent to destination host；
Send after state reaches preset requirement at dirty data and terminate Data Migration.
The embodiment of the present invention additionally provides the device of a kind of Data Migration, including:
Rewriting rate computing unit, for calculating the rewriting rate of each storage band, described storage band is carried out data by needs
The memory space segmentation migrated obtains；
Transmitting element, for preferentially sending the dirty data in the storage band that rewriting rate is low to destination host；
Migrate control unit, terminate Data Migration for sending after state reaches preset requirement at dirty data.
The embodiment of the present invention additionally provides the system of a kind of Data Migration, including: source host, destination host, wherein, described
The device of any one Data Migration that source host provides for the embodiment of the present invention.
The technical scheme that the embodiment of the present invention provides, will need the memory space segmentation carrying out Data Migration to obtain by using
To storage band；Preferentially send the dirty data technological means in the storage band that rewriting rate is low to destination host, thus utilize sky
Between principle of locality and temporal locality principle, i.e. during the execution of application, the use domain of the existence of memory space is concentrated
Characteristic, and the higher memory space of the rate that rewrites has the biggest probability to be again modified, and storage higher for rewriting rate is empty
Between dirty data postpone send, utilize write merging absorb more internal memory write operation, decrease the transmission of dirty data, thus reduce
Volume of transmitted data.
Accompanying drawing explanation
In order to be illustrated more clearly that the technical scheme of the embodiment of the present invention, required use in embodiment being described below
Accompanying drawing be briefly described, it should be apparent that, below describe in accompanying drawing be only some embodiments of the present invention, for this
From the point of view of the those of ordinary skill of field, on the premise of not paying creative work, it is also possible to obtain other according to these accompanying drawings
Accompanying drawing.
Fig. 1 is embodiment of the present invention method flow schematic diagram；
Fig. 2 is embodiment of the present invention method flow schematic diagram；
Fig. 3 is embodiment of the present invention method flow schematic diagram；
Fig. 4 is embodiment of the present invention method flow schematic diagram；
Fig. 5 is embodiment of the present invention method flow schematic diagram；
Fig. 6 is embodiment of the present invention apparatus structure schematic diagram；
Fig. 7 is embodiment of the present invention apparatus structure schematic diagram；
Fig. 8 is embodiment of the present invention apparatus structure schematic diagram；
Fig. 9 is embodiment of the present invention system structure schematic diagram.
Detailed description of the invention
Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is carried out clear, complete
Describe, it is clear that described embodiment is only a part of embodiment of the present invention rather than whole embodiments wholely.Based on
Embodiment in the present invention, it is every other that those of ordinary skill in the art are obtained under not making creative work premise
Embodiment, broadly falls into the scope of protection of the invention.
A kind of method of Data Migration, as it is shown in figure 1, in source host and destination host running, if need into
Row Data Migration, then source host can perform following steps:
101: calculate the rewriting rate of each storage band, above-mentioned storage band divides by needing the memory space carrying out Data Migration
Section obtains；
Above-mentioned storage band can be stored in internal memory, it is also possible to is stored in such as other the storage medium such as hard disk, for tool
The storage medium embodiment of the present invention of body not limits.If storage band is internal memory band, then internal memory band is carried out by needs
The memory sections of Data Migration obtains.Subsequent embodiment all illustrates as a example by internal memory band.
102: preferentially send the dirty data in the storage band that rewriting rate is low to destination host；
If storage band is internal memory band；
The most above-mentioned 102 are: preferentially send the dirty data in the internal memory band that rewriting rate is low to destination host；
One of given below 102 implements example: according to the rewriting rate of internal memory band, high by rewriting rate to internal memory band
Low order sorts；By the result of above-mentioned sequence, start to send internal memory to destination host successively from the internal memory band of minimum rewriting rate
Dirty data in band.Certainly, adopt and realize 102 also possible in other ways, this embodiment of the present invention is not limited.
It addition, the dirty data in above-mentioned internal memory band can represent with the dirty page form in internal memory band, and
Subsequent embodiment all will illustrate as example, certainly adopt and represent the reality not affecting the embodiment of the present invention in other ways
Existing, this embodiment of the present invention is not limited.
103: send after state reaches preset requirement at dirty data and terminate Data Migration.
If storage band is internal memory band；Then 103 are: send after state reaches preset requirement at dirty data and terminate data
Migrate.
In above-mentioned 103, dirty data transmission state reaches preset requirement and includes: the cycle times that dirty data sends reaches first
Setting value and existing dirty data less than the dirty data in the second setting value or full memory bar migrate complete in.If certainly set
The preset requirement putting other nor affects on the realization of the embodiment of the present invention, not limits this embodiment of the present invention.
Further, if using network transmission dirty data, transmission dirty data may be limited in bandwidth, is the most also
Improving the utilization rate of the network bandwidth, said method 102 and 103 can be:
Circulation performs: in acquisition current time period, source host is to the available bandwidth of destination host；Preferential to destination host
Send the dirty data in the storage band that rewriting rate is low, and calculate the dirty number sent to destination host within the above-mentioned time cycle
According to amount；If having sent dirty data amount to meet or exceed available bandwidth, then stop sending dirty to destination host in this time cycle
Data；
Send after state reaches preset requirement at dirty data and stop the flow process that above-mentioned circulation performs, terminate Data Migration.
The executive agent of said method is the main frame in Data Migration source, the technical scheme that the embodiment of the present invention provides, and passes through
Use and the memory space segmentation carrying out Data Migration will be needed to obtain storing band；Preferentially send rewriting rate to destination host low
Dirty data technological means in storage band, thus utilize spatial locality principle and temporal locality principle, i.e. in application
During execution, memory space is used the characteristic that domain of the existence is concentrated, and the higher memory space of the rate that rewrites has the most general
Rate will be modified again, postpones to send by the dirty data of memory space higher for rewriting rate, and utilization is write merging and absorbed more
Internal memory write operation, decreases the transmission of dirty data, thus decreases volume of transmitted data, decreases opening of the network bandwidth simultaneously
Pin.
Present invention below embodiment gives a more specifically implementation, and it is real that the present embodiment realizes virutal machine memory
Time migrate, as in figure 2 it is shown, include:
201: source virtual machine internal memory is divided into the internal memory band of some equal lengths.
More specifically, the pseudo-physical memory (Pseudo-physical of virtual machine (i.e. source virtual machine) can be will be migrated
Memory) it is divided into some continuous and equal-sized internal memory band (MemoryStripe).As it is shown on figure 3, the signal of left side internal memory
For the internal memory before not dividing, box indicating page, wherein black box represents dirty page；After middle internal memory is illustrated as being divided
Internal memory band can be with numbered 1～5.
More specifically, the data structure that can be defined as follows, as the memory element of memory bar information:
struct stripe{
unsigned long start_page；
unsigned long page_num；
int num；
unsigned long dirty_page_num；
double rate_mbyte；
}；
The number of the page of internal memory band can be according to having determined since the size of virtual machine storage allocation, additionally page
Can be set to as far as possible 8 the most several, so can allow the sub-memory bits chart of each internal memory band is integral words joint, it is simple to
Calculate and management.Such as: to being assigned with the virtual machine of 256MB memory size, internal memory band could be arranged to 64 memory pages.
202: calculate the rewriting rate of all internal memory bands.
Calculate the method for rewriting rate of internal memory band and may is that to calculate the current dirty internal memory number of pages of internal memory band dirty with transmission
The business of the cycle time of page；What rewriting rate embodied is virtual machine instant ratio (that is: the unit of writing page in internal memory band
In time, the quantity of the dirty page produced in internal memory band), if each memory bar carrying space is identical, directly calculate dirty internal memory
This embodiment of the present invention is not limited by number of pages as rewriting rate is also possible.
203: according to the rewriting rate of the internal memory band calculated in 202, internal memory band is sorted in ascending order；Such as right part in Fig. 3
Internal memory is illustrated, for the result after sequence.Certainly this step can also be to sort in descending order, to this embodiment of the present invention not limit
Fixed.
204: estimate the physical host available network bandwidth to destination physical host of source, i.e. current from source physics
Main frame sends the network packet available network bandwidth size to destination physical host.Start in the cycle of internal memory real-time migration
After can be immediately performed this step, it may be assumed that this step can start any time before 206 in the cycle of internal memory real-time migration
Perform, do not affect the realization of the embodiment of the present invention.
In the present embodiment, the physical host of source sends dirty data to the physical host of destination, uses periodically transmission
Mode is carried out；A confirmable available network bandwidth can be had within the time in each cycle.
205: by the ranking results in 203, send the dirty page in internal memory band successively to the physical host of destination
(in Fig. 3 according to: 5, the order of 2,1,3,4 sends), and calculate the data volume of dirty page transmitted.Specifically can be by
The dirty data number of pages of transmission internal memory band is accumulative to be added.
206: by the available bandwidth estimated in 204 compared with the data volume of the dirty page transmitted, if transmitted
Data volume be not above estimating available bandwidth, then continue the dirty page in copy transmissions subsequent memory band, otherwise stop
The transmission of the dirty page in this cycle, re-executes step 204 to 206 at next cycle, until dirty page end of transmission
Enter 207；
Being understandable that in 206, the condition of entrance 207 can also is that other condition, such as: dirty data sends
Cycle times reaches the first setting value and existing dirty data is less than the second setting value, and this enters the condition of 207 can be according to moving
That moves dirty data needs setting, not limits this embodiment of the present invention.
207: terminate internal memory real-time migration.
The technical scheme that the embodiment of the present invention provides, will need the memory space segmentation carrying out Data Migration to obtain by using
To storage band；Preferentially send the dirty data technological means in the storage band that rewriting rate is low to destination host, thus utilize sky
Between principle of locality and temporal locality principle, i.e. during the execution of application, the use domain of the existence of memory space is concentrated
Characteristic, and the higher memory space of the rate that rewrites has the biggest probability to be again modified, and storage higher for rewriting rate is empty
Between dirty data postpone send, utilize write merging absorb more internal memory write operation, decrease the transmission of dirty data, thus reduce
Volume of transmitted data, decreases the expense of the network bandwidth simultaneously.
Will be given below in embodiment corresponding with Fig. 2, only carry out ordered propagation virutal machine memory containing dirty pages, do not carry out bandwidth control
The method flow of system, as shown in Figure 4, including:
401: start and carry out internal memory real-time migration flow process to target virtual machine；
402: start virtual machine shadow page table and log in the pattern of dirty page (Log_Dirty), start Log_Dirty pattern
Dirty page bitmap or system bitmap table can be got；Dirty page bitmap or system bitmap table comprise: each memory bar
Dirty internal memory number of pages in band and specifically which page are the information of dirty page；
403: determine whether that last is taken turns, can be specifically to judge whether indicator is 1, this indicator can be pre-
If a value for control send internal memory round；If it is, enter 404, otherwise enter 405；
404: terminate internal memory real-time migration；
405: judge whether that meeting last takes turns condition；This condition is to terminate the condition of real-time migration of virtual machine, such as: exceed
20 take turns internal memory replicates and transmission containing dirty pages face is less than 100 pages etc., can be arranged as required to.The embodiment of the present invention is specifically set
Not limit.If so, enter 406, otherwise enter 407；
406: arranging indicator is 1；
407: replicate the dirty page bitmap table clearing systems bitmap table obtained in transmission cycle next time；Subsequently into
403。
Above-mentioned Fig. 4 flow process can be completed by following programmed instruction hardware:
MEMORY PROPAGATION ALGORITHM
LAST_ITER←0
START MEMORY PROPAGATION
STARTVM SHADOW MEMORY LOG DIRTY MODE
DO
BEGIN
ORDERED MEMORY STRIPE PROPAGATION
IF LAST_ITER=1
STOP MEMORY PROPAGATION
BREAK
ELSE IF
IF SATISFY LAST ITERATION
LAST_ITER=1
END IF
SCAN DIRTY MEMORY AND CLEAN MEMORY BITMAP
END IF
END WHILE(1)
Will be given below in embodiment corresponding with Fig. 2, only carry out bandwidth control, do not carry out ordered propagation virutal machine memory dirty
The method flow of page, as it is shown in figure 5, include:
501: calculate the rewriting rate of internal memory band；
502: utilize fast row's algorithm to internal memory belt row sequence；
503: judge that dirty page has been sent, if it is, enter 504, otherwise enter 505；Or, sentence
Whether the disconnected accumulative dirty page sent has exceeded the network bandwidth；The most then wait that the next one sends the cycle of dirty interior page, then
Enter 505, be otherwise directly entered 505；Available network bandwidth is estimated and can be used self-evaluating method or be loaded into path (Pathload)
Method is estimated.
504: terminate internal memory real-time migration；
505: according to ranking results successively from the beginning of the internal memory band that rewriting rate is low, send dirty page to destination host；
Subsequently into 503.
Above-mentioned Fig. 5 flow process can be completed by following programmed instruction hardware:
ORDERED PROPAGATION ALGORITHM
ANB←AVAILABLE NETWORK BANDWIDTH
ACC_RR←0
CALCULATE MEMORY STRIPE REWRITE RATES
QUICK SORT MEMORY STRIPE
WHILE (MEMORY STRIPE NOT COMPLETE OR ACC_RR ＜ ANB)
BEGIN
COPY AND TRANSFER DIRTY PAGE IN MEMORY STRIPE
ACC_RR=ACC_RR+MEMORY STRIPE REWRITING RATE
END
SAVE DIRTY MEMORY PAGE BITMAP
BEGIN NEXT MEMORY PROPAGATION ITERATION
The technical scheme that the embodiment of the present invention provides, will need the memory space segmentation carrying out Data Migration to obtain by using
To storage band；Preferentially send the dirty data technological means in the storage band that rewriting rate is low to destination host, thus utilize sky
Between principle of locality and temporal locality principle, i.e. during the execution of application, the use domain of the existence of memory space is concentrated
Characteristic, and the higher memory space of the rate that rewrites has the biggest probability to be again modified, and storage higher for rewriting rate is empty
Between dirty data postpone send, utilize write merging absorb more internal memory write operation, decrease the transmission of dirty data, thus reduce
Volume of transmitted data, decreases the expense of the network bandwidth simultaneously.
The embodiment of the present invention additionally provides the device of a kind of Data Migration, as shown in Figure 6, and including:
Rewriting rate computing unit 601, for calculating the rewriting rate of each storage band, above-mentioned storage band is by needs number
Obtain according to the memory space segmentation migrated；
The specific implementation that memory space segmentation obtains storing band is referred to, in embodiment of the method the 201 of Fig. 2
The explanation of part.
Transmitting element 602, for preferentially sending the dirty data in the storage band that rewriting rate is low to destination host；
Migrate control unit 603, terminate Data Migration for sending after state reaches preset requirement at dirty data.
Specifically, above-mentioned rewriting rate computing unit 601, for calculating the rewriting rate of each internal memory band, above-mentioned internal memory band
Obtained by needing the memory sections carrying out Data Migration；
Above-mentioned transmitting element 602, for preferentially sending the dirty data in the internal memory band that rewriting rate is low to destination host.
Further, as it is shown in fig. 7, said apparatus also includes:
Bandwidth acquiring unit 701, in obtaining current time period, source host is to the available bandwidth of destination host；This reality
Execute the physical host of source in example and send dirty data to the physical host of destination, use the mode periodically sent to carry out；?
A confirmable available network bandwidth can be had in the time in each cycle.
Data volume computing unit 702, for calculating the dirty data amount sent within the above-mentioned time cycle to destination host；
Data send control unit 703, if meeting or exceeding available bandwidth, then in this cycle for sending dirty data amount
Stop sending dirty data to destination host；
Above-mentioned migration control unit 603, stops above-mentioned circulation after reaching preset requirement in dirty data transmission state and holds
The flow process of row, terminates Data Migration.
Alternatively, as shown in Figure 8, transmitting element 602 includes:
Sequence subelement 6021, for the rewriting rate according to internal memory band, is arranged by rewriting rate sequence internal memory band
Sequence；Sequence herein may is that being sorted in ascending order, it is also possible to is to sort in descending order, not limits this embodiment of the present invention.
Send subelement 6022, for by the result of above-mentioned sequence, from the internal memory band of minimum rewriting rate start successively to
Destination host sends the dirty data in internal memory band.
Specifically, above-mentioned migration control unit 603, if for dirty data send cycle times reach the first setting value and
Existing dirty data is less than the second setting value, or dirty data all migrates complete, then terminate Data Migration.Terminate data to move herein
The condition moved can according to migrate dirty data need set, this embodiment of the present invention is not limited.
The embodiment of the present invention additionally provides the system of a kind of Data Migration, as it is shown in figure 9, include: source host 901, target
Main frame 902, the device of any one Data Migration that above-mentioned source host 901 provides for the embodiment of the present invention.
The technical scheme that the embodiment of the present invention provides, will need the memory space segmentation carrying out Data Migration to obtain by using
To storage band；Preferentially send the dirty data technological means in the storage band that rewriting rate is low to destination host, thus utilize sky
Between principle of locality and temporal locality principle, i.e. during the execution of application, the use domain of the existence of memory space is concentrated
Characteristic, and the higher memory space of the rate that rewrites has the biggest probability to be again modified, and storage higher for rewriting rate is empty
Between dirty data postpone send, utilize write merging absorb more internal memory write operation, decrease the transmission of dirty data, thus reduce
Volume of transmitted data, decreases the expense of the network bandwidth simultaneously.
One of ordinary skill in the art will appreciate that all or part of step realizing in above-described embodiment method is permissible
Instructing relevant hardware by program to complete, program can be stored in a kind of computer-readable recording medium, mentioned above
Storage medium can be read only memory, disk or CD etc..
The methods, devices and systems of a kind of Data Migration provided the embodiment of the present invention above have carried out detailed Jie
Continuing, principle and the embodiment of the present invention are set forth by specific case used herein, and the explanation of above example is only
It is the method and core concept thereof being adapted to assist in and understanding the present invention；Simultaneously for one of ordinary skill in the art, according to this
Bright thought, the most all will change, and to sum up, this specification content should not be construed as
Limitation of the present invention.
Claims (10)
1. the method for a Data Migration, it is characterised in that including:
Calculating the rewriting rate of each storage band, described storage band obtains by needing the memory space segmentation carrying out Data Migration；
Circulation performs: in acquisition current time period, source host is to the available bandwidth of destination host, preferential to described destination host
Send the dirty data in the storage band that rewriting rate is low, and calculate and send out to described destination host in described current time period
The dirty data amount sent, if having sent dirty data amount to meet or exceed described available bandwidth, then stops to described in this time cycle
Destination host sends dirty data；
Send at dirty data and after state reaches preset requirement, stop the flow process that described circulation performs, terminate Data Migration；
Wherein, described storage band is internal memory band, and described internal memory band obtains by needing the memory sections carrying out Data Migration；
Wherein, the rewriting rate of described calculating each storage band includes: calculate the current dirty internal memory number of pages of each described internal memory band with
Send the business of the cycle time of dirty page, obtain the rewriting rate of each described internal memory band.
Method the most according to claim 1, it is characterised in that preferentially send in the storage band that rewriting rate is low to destination host
Dirty data include: preferentially send the dirty data in the internal memory band that rewriting rate is low to destination host.
Method the most according to claim 2, it is characterised in that preferentially send in the internal memory band that rewriting rate is low to destination host
Dirty data, including:
Rewriting rate according to internal memory band, sorts by rewriting rate sequence to internal memory band；
By the result of described sequence, start to send internal memory band to destination host successively from the internal memory band of minimum rewriting rate
Dirty data.
4. according to method described in Claims 2 or 3, it is characterised in that the dirty data in described internal memory band includes: internal memory band
In dirty page.
5. according to method described in Claims 2 or 3, it is characterised in that the dirty data of described internal memory band sends state and reaches pre-
If requiring to include:
The cycle times that dirty data sends reaches the first setting value and existing dirty data is less than the second setting value, or full memory
Dirty data in band migrates complete.
6. the device of a Data Migration, it is characterised in that including:
Rewriting rate computing unit, for calculating the rewriting rate of each storage band, described storage band is carried out Data Migration by needs
Memory space segmentation obtain；
Transmitting element, for preferentially sending the dirty data in the storage band that rewriting rate is low to destination host；
Migrate control unit, terminate Data Migration for sending after state reaches preset requirement at dirty data；
Bandwidth acquiring unit, in obtaining current time period, source host is to the available bandwidth of destination host；
Data volume computing unit, for calculating the dirty data amount sent in described current time period to destination host；
Data send control unit, if meeting or exceeding available bandwidth for sending dirty data amount, then stop in this time cycle
Only send dirty data to destination host；
Described bandwidth acquiring unit, data volume computing unit and data send control unit each function and perform for circulation；
Described migration control unit, stops, for sending at dirty data, the stream that described circulation performs after state reaches preset requirement
Journey, terminates Data Migration；
Wherein, described storage band is internal memory band, and described internal memory band obtains by needing the memory sections carrying out Data Migration；
Wherein, described rewriting rate computing unit specifically for: calculate the current dirty internal memory number of pages of each described internal memory band and transmission
The business of the cycle time of dirty page, obtains the rewriting rate of each described internal memory band.
The most according to claim 6, device, it is characterised in that described rewriting rate computing unit, be used for calculating each internal memory band
Rewriting rate；
Described transmitting element, for preferentially sending the dirty data in the internal memory band that rewriting rate is low to destination host.
Device the most according to claim 7, it is characterised in that transmitting element includes:
Sequence subelement, for the rewriting rate according to internal memory band, sorts by rewriting rate sequence to internal memory band；
Send subelement, for by the result of described sequence, starting successively to destination host from the internal memory band of minimum rewriting rate
Send the dirty data in internal memory band.
9. according to device described in claim 7 or 8, it is characterised in that
Described migration control unit, if the cycle times being used for dirty data transmission reaches the first setting value and existing dirty data is less than
Dirty data in second setting value, or full memory band migrates complete, then terminate Data Migration.
10. a system for Data Migration, including: source host, destination host, it is characterised in that described source host is that right is wanted
Seek the device described in 6 to 9 any one.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201010243657.2A CN102136993B (en) | 2010-07-29 | 2010-07-29 | A kind of methods, devices and systems of Data Migration |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201010243657.2A CN102136993B (en) | 2010-07-29 | 2010-07-29 | A kind of methods, devices and systems of Data Migration |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102136993A CN102136993A (en) | 2011-07-27 |
| CN102136993B true CN102136993B (en) | 2016-12-21 |
Family
ID=44296653
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201010243657.2A Active CN102136993B (en) | 2010-07-29 | 2010-07-29 | A kind of methods, devices and systems of Data Migration |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102136993B (en) |
Families Citing this family (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2013105217A1 (en) * | 2012-01-10 | 2013-07-18 | 富士通株式会社 | Virtual machine management program, method and device |
| JP5995071B2 (en) * | 2012-09-19 | 2016-09-21 | 学校法人 中央大学 | Memory controller, data storage device, and memory control method |
| CN103473368A (en) * | 2013-09-27 | 2013-12-25 | 中国科学院信息工程研究所 | Virtual machine real-time migration method and system based on counting rank ordering |
| US9436751B1 (en) | 2013-12-18 | 2016-09-06 | Google Inc. | System and method for live migration of guest |
| CN104346210B (en) * | 2014-09-04 | 2017-12-29 | 中国船舶重工集团公司第七0九研究所 | The online moving method of virtual machine and system based on temporal and spatial correlations |
| CN105468457B (en) * | 2015-11-24 | 2019-04-09 | 无锡江南计算技术研究所 | A kind of parallel system local migration fault-tolerance approach based on difference identification |
| CN105893114A (en) * | 2016-04-05 | 2016-08-24 | 浪潮电子信息产业股份有限公司 | Virtual machine migration method, target host machine and source host machine |
| CN106534140A (en) * | 2016-11-25 | 2017-03-22 | 西安烽火电子科技有限责任公司 | Transmission system and method of SIP message |
| TWI624757B (en) | 2017-05-24 | 2018-05-21 | 財團法人工業技術研究院 | Data processing method, data processing system, and computer program product |
| CN107368438A (en) * | 2017-09-13 | 2017-11-21 | 郑州云海信息技术有限公司 | A kind of dirty data detection method and system |
| CN111343250A (en) * | 2020-02-16 | 2020-06-26 | 西安奥卡云数据科技有限公司 | Reliable remote data protection method based on network bearing capacity |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1696902A (en) * | 2004-05-11 | 2005-11-16 | 国际商业机器公司 | System, method and program to migrate a virtual machine |
| CN101101562A (en) * | 2007-07-10 | 2008-01-09 | 北京大学 | Dummy machine external storage on-line migration method |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9594582B2 (en) * | 2009-01-05 | 2017-03-14 | International Business Machines Corporation | Detection and management of dynamic migration of virtual environments |
-
2010
- 2010-07-29 CN CN201010243657.2A patent/CN102136993B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1696902A (en) * | 2004-05-11 | 2005-11-16 | 国际商业机器公司 | System, method and program to migrate a virtual machine |
| CN101101562A (en) * | 2007-07-10 | 2008-01-09 | 北京大学 | Dummy machine external storage on-line migration method |
Non-Patent Citations (3)
| Title |
|---|
| Xen环境下实时迁移结构和算法研究;阮敏;《中国优秀硕士学位论文全文数据库 信息科技辑(月刊)》;20090915(第9期);摘要，第36页倒数第1段，第39页第1-2段，图4.4 * |
| 广域网络存储虚拟化技术研究;邓劲生;《中国优秀博硕士学位论文全文数据库(博士) 信息科技辑(月刊)》;20061015(第10期);第46-47页 * |
| 虚拟化Xen及其实时迁移技术研究;孙昱;《中国优秀硕士学位论文全文数据库 信息科技辑(月刊)》;20080715(第7期);第39-54页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102136993A (en) | 2011-07-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102136993B (en) | A kind of methods, devices and systems of Data Migration | |
| CN103098014B (en) | Storage system | |
| Li et al. | NUMA-aware shared-memory collective communication for MPI | |
| Tiwari et al. | Active Flash: Towards {Energy-Efficient},{In-Situ} Data Analytics on {Extreme-Scale} Machines | |
| Xu et al. | Managing performance overhead of virtual machines in cloud computing: A survey, state of the art, and future directions | |
| CN104123171B (en) | Virtual machine migrating method and system based on NUMA architecture | |
| CN113424144A (en) | Computing data storage system | |
| CN106201659B (en) | A kind of method and host of live migration of virtual machine | |
| Svärd et al. | Principles and performance characteristics of algorithms for live VM migration | |
| CN103955399A (en) | Migrating method and device for virtual machine, as well as physical host | |
| CN109684074A (en) | Physical machine resource allocation methods and terminal device | |
| CN105095094A (en) | Memory management method and equipment | |
| Sakai et al. | High-performance and highly reliable file system for the K computer | |
| CN107408018A (en) | For adapting to the mechanism of refuse collection resource allocation in solid-state drive | |
| US20210329354A1 (en) | Telemetry collection technologies | |
| CN104268003A (en) | Memory state migration method applicable to dynamic migration of virtual machine | |
| Xu et al. | Rethink the storage of virtual machine images in clouds | |
| Li et al. | Efficient live virtual machine migration for memory write-intensive workloads | |
| CN116450053B (en) | Data storage method, device, system, electronic equipment and storage medium | |
| CN113723443A (en) | Distributed training method and system for large visual model | |
| CN101729421B (en) | Storage method and device based on time division multiplex | |
| CN110990133A (en) | Edge computing service migration method and device, electronic equipment and medium | |
| CN112306628B (en) | Virtual network function resource management system based on multi-core server | |
| US20190102091A1 (en) | Modifying accesibility based on memory access patterns | |
| CN111400241B (en) | Data reconstruction method and device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant |