CN103577122B - Implementation method that distribution application system migrates between platform and device - Google Patents

Implementation method that distribution application system migrates between platform and device Download PDF

Info

Publication number
CN103577122B
CN103577122B CN201310547871.0A CN201310547871A CN103577122B CN 103577122 B CN103577122 B CN 103577122B CN 201310547871 A CN201310547871 A CN 201310547871A CN 103577122 B CN103577122 B CN 103577122B
Authority
CN
China
Prior art keywords
node
platform
target platform
data
target
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
Application number
CN201310547871.0A
Other languages
Chinese (zh)
Other versions
CN103577122A (en
Inventor
蔡毅
池金燕
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.)
Huawei Technologies Co Ltd
Original Assignee
Hangzhou Huawei Digital Technologies Co Ltd
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 Hangzhou Huawei Digital Technologies Co Ltd filed Critical Hangzhou Huawei Digital Technologies Co Ltd
Priority to CN201310547871.0A priority Critical patent/CN103577122B/en
Publication of CN103577122A publication Critical patent/CN103577122A/en
Priority to PCT/CN2014/086106 priority patent/WO2015067093A1/en
Application granted granted Critical
Publication of CN103577122B publication Critical patent/CN103577122B/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
    • 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
    • G06F9/48—Program initiating; Program switching, e.g. by interrupt
    • G06F9/4806—Task transfer initiation or dispatching
    • G06F9/4843—Task transfer initiation or dispatching by program, e.g. task dispatcher, supervisor, operating system
    • G06F9/485—Task life-cycle, e.g. stopping, restarting, resuming execution
    • G06F9/4856—Task life-cycle, e.g. stopping, restarting, resuming execution resumption being on a different machine, e.g. task migration, virtual machine migration

Landscapes

  • Engineering & Computer Science (AREA)
  • Software Systems (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Information Transfer Between Computers (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

The invention discloses implementation method and device that a kind of distribution application system migrates between platform, method includes: the number n of the node required for acquisition target platform；Utilize total data capacity size of all nodes of source platform, determine the size of the data space of each node in n node of target platform；The operating system of the hardware configuration target platform according to n node of target platform；The distribution application system of the distribution application system configuration target platform according to source platform；After the operating system having configured target platform and distribution application system, according to the size of the data space of each node of target platform, the application data on source platform are moved on n node of target platform respectively.By the way, the present invention can support the migration from traditional X-ray 86 IA frame serverPC to ARM IA frame serverPC, supports that N number of node of distribution application system migrates scene to M node.

Description

Implementation method that distribution application system migrates between platform and device
Technical field
The present invention relates to field of computer technology, particularly relate to a kind of distribution application system flat The implementation method migrated between platform and device.
Background technology
In current IT market, traditional server hardware provider has been difficult to affect client's choosing Type, emerge more dependence total solution and the consultancy service promotion market application of new product, with Shi Cuisheng added value and profit, Client application migration service is exactly that one of them is crucial.Should Ensure the smooth migration of Client application system, Customer Transition risk to be reduced and workload, ability Reduce client's total cost of ownership TCO, the satisfaction of Win Clients and use wish.
IBM System services migrating technology comprises the following steps: application system is at source platform running Time middle, find the system service that application system is directly relied on；By the system service on source platform it Between dependence model and the system service that directly relied on of the application system of described discovery, obtain Obtain the system service that application system is indirectly relied on；Find what application system directly or indirectly relied on The configuration of system service；The system service directly or indirectly relied on according to application system and configuration thereof Mapping relations between source platform and target platform, by application system institute on source platform directly and Connect the system service of dependence and configuration be mapped as target platform, application system institute directly and The system service indirectly relied on and configuration thereof；It is used for target platform, application system institute directly by described The system service and the configuration thereof that connect and indirectly rely on are arranged on target platform, thus realize application system The system environments that system is relied on migration from source platform to target platform.
Present inventor finds in long-term research and development, and above-mentioned prior art is not supported from tradition X86-based server is to the migration of ARM IA frame serverPC；Distribution application system cannot be supported N number of node M node is migrated scene.
Summary of the invention
The technical problem that present invention mainly solves is to provide a kind of distribution application system between platform The implementation method migrated and device, it would be preferable to support from traditional X-ray 86 IA frame serverPC to ARM framework The migration of server, supports that N number of node of distribution application system migrates scene to M node.
First aspect, the present invention provides the realization side that a kind of distribution application system migrates between platform Method, including: the number n of the node required for acquisition target platform, so that described target platform The hardware of n node can the business of all nodes of load source platform；Described source platform is utilized to own Total data capacity size of node, determines the number of each node in n node of described target platform According to the size of memory space, wherein, total data capacity size etc. of all nodes of described source platform Size in the total data space of n node of described target platform；According to described target platform The operating system of target platform described in the hardware configuration of n node, so that described target platform Operating system and the Hardware match of described n node of target platform；Distribution according to described source platform Formula application system configures the distribution application system of described target platform, so that described target platform The application of distribution application system of applied environment and described source platform of distribution application system Difference between environment is in preset range；Configuring the operating system of good described target platform and dividing After cloth application system, according to the size of the data space of each node of described target platform, Application data on described source platform are moved on n node of described target platform respectively.
In the first possible implementation of first aspect, described source platform node is X86 frame Structure server or virtual machine, described target platform node is advanced Reduced Instruction Set Computer equipment ARM IA frame serverPC or virtual machine.
In the implementation that the second of first aspect is possible, required for described acquisition target platform The step of number n of node, including: by one node of described source platform, two nodes ..., M-1 node, service bearer amount P of m node actual measurement1、P2、…、Pm-1、PmDivided by reason Service bearer amount Q of opinion1、Q2、…、Qm-1、Qm, it is thus achieved that the cluster that described source platform is corresponding expands Hold coefficient C1、C2、…、Cm-1、Cm, wherein, Cm=Pm/Qm, define described source platform Node number is m, and described m is the natural number being not zero；According to the cluster that described source platform is corresponding Dilation coefficient C1、C2、…、Cm-1、Cm, it is thus achieved that described cluster dilation coefficient C and described node Functional relation C(x between number x)；According to described cluster dilation coefficient and described node Functional relation C(x between number), hardware performance S of described source platform single node, described source Combination property utilization rate δ of platform single node, described target platform single node hardware performance T with And described target platform performance amount of resources reserved ρ, by applied business load balancing algorithm S* δ * Cm=T*(1-ρ) * C(x), it is thus achieved that the number n of the node required for described target platform.
In the third possible implementation of first aspect, described utilize described source platform own Total data capacity size of node, determines the number of each node in n node of described target platform According to the step of the size of memory space, including: each node in n node of described target platform The size of data space is removed equal to total data capacity size of all nodes of described source platform Business with described n.
In the 4th kind of possible implementation of first aspect, described every according to described target platform Application data on described source platform are moved to by the size of the data space of individual node respectively Step on n node of described target platform, comprises determining that all nodes on described source platform Total original application data and total back-up application data；According to each node of described target platform The size of data space, by total original application data of nodes all on described source platform Move to respectively on n node of described target platform with total back-up application data.
Second invention, the present invention provides the realization dress that a kind of distribution application system migrates between platform Putting, described device includes: obtains module, determine module, the first configuration module, the second configuration mould Block and Data Migration module；Described acquisition module is for obtaining node required for target platform Number n, so that the hardware of described n node of target platform can all nodes of load source platform Business；Described determine that module for obtaining the node required for target platform in described acquisition module Number n after, utilize total data capacity size of all nodes of described source platform, determine described The size of the data space of each node in n node of target platform, wherein, described source is put down The data that total data capacity size of all nodes of platform is equal to n node of described target platform total are deposited The size in storage space；Hard for according to described n node of target platform of described first configuration module Part configures the operating system of described target platform, so that the operating system of described target platform and institute State the Hardware match of n node of target platform；Described second configuration module is for putting down according to described source The distribution application system of platform configures the distribution application system of described target platform, so that described The applied environment of the distribution application system of target platform and the Distributed Application system of described source platform Difference between the applied environment of system is in preset range；Described Data Migration module is described first Configuration module and the second configuration module have been respectively configured operating system and the distribution of described target platform After formula application system, according to the size of the data space of each node of described target platform, will Application data on described source platform move on n node of described target platform respectively.
In the first possible implementation of second aspect, described source platform node is X86 frame Structure server or virtual machine, described target platform node is advanced Reduced Instruction Set Computer equipment ARM IA frame serverPC or virtual machine.
In the implementation that the second of second aspect is possible, described acquisition module includes: first Obtain unit, the second acquisition unit and the 3rd obtains unit；Described first acquisition unit is used for will One node of described source platform, two node ..., m-1 node, the industry of m node actual measurement Business bearing capacity P1、P2、…、Pm-1、PmDivided by theoretical service bearer amount Q1、Q2、…、Qm-1、 Qm, it is thus achieved that the cluster dilation coefficient C that described source platform is corresponding1、C2、…、Cm-1、Cm, wherein, Cm=Pm/Qm, the node number defining described source platform is m, and described m is the nature being not zero Number；Described second obtains unit for obtaining, according to described first, the described source platform pair that unit obtains The cluster dilation coefficient C answered1、C2、…、Cm-1、Cm, it is thus achieved that described cluster dilation coefficient C with Functional relation C(x between described node number x)；Described 3rd obtains unit for basis Described second obtains the letter between described cluster dilation coefficient and the described node number that unit obtains Number relational expressions C(x), hardware performance S of described source platform single node, described source platform single node Combination property utilization rate δ, hardware performance T of described target platform single node and described target Platform property amount of resources reserved ρ, by applied business load balancing algorithm S* δ * Cm=T*(1-ρ) * C(x), it is thus achieved that the number n of the node required for described target platform.
In the third possible implementation of second aspect, described determine that module is specifically for institute State the total data capacity size utilizing all nodes of described source platform, determine described target platform n The size of the data space of each node in individual node, wherein, described target platform n joint In point, the size of the data space of each node is equal to the total of all nodes of described source platform Data capacity size is divided by the business of described n.
In the 4th kind of possible implementation of second aspect, described Data Migration module includes: Determine unit and data migration unit；Described determine that unit is for determining all joints on described source platform Total original application data of point and total back-up application data；Described data migration unit is used for pressing According to the size of the data space of each node of described target platform, determine that unit determines by described Described source platform on total original application data of all nodes divide with total back-up application data Do not move on n node of described target platform.
The invention has the beneficial effects as follows: be different from the situation of prior art, the present invention obtains target and puts down The number n of the node required for platform；Utilize total data capacity size of all nodes of source platform, Determine the size of the data space of each node in n node of target platform；Put down according to target The operating system of the hardware configuration target platform of n node of platform；Distributed Application according to source platform The distribution application system of system configuration target platform；Deposit according to the data of each node of target platform Application data on source platform are moved to n node of target platform by the size in storage space respectively On.In this way, can support from traditional X-ray 86 IA frame serverPC to ARM infrastructure services The migration of device；The N number of node that can also support distribution application system migrates field to M node Scape.
Accompanying drawing explanation
Fig. 1 is implementation method one embodiment that distribution application system of the present invention migrates between platform Flow chart；
Fig. 2 is the migration effect of the implementation method that distribution application system of the present invention migrates between platform Schematic diagram；
Fig. 3 is another embodiment party of implementation method that distribution application system of the present invention migrates between platform The flow chart of formula；
Fig. 4 is that the implementation method that distribution application system of the present invention migrates between platform is divided at Hadoop Application schematic diagram in cloth application system；
Fig. 5 be distribution application system of the present invention migrate between platform realize device one embodiment Structural representation；
Fig. 6 be distribution application system of the present invention migrate between platform realize another embodiment party of device The structural representation of formula；
Fig. 7 be distribution application system of the present invention migrate between platform realize the another embodiment party of device The structural representation of formula.
Detailed description of the invention
The present invention is described in detail with embodiment below in conjunction with the accompanying drawings.
It is the implementation method that distribution application system of the present invention migrates between platform refering to Fig. 1, Fig. 1 The flow chart of one embodiment, including:
Step S101: the number n of the node required for acquisition target platform, so that target platform The hardware of n node can the business of all nodes of load source platform.
The node of source platform is the node of distribution application system platform before migration, such as The frame of X86-based or blade server, or virtual machine etc..X86-based server refers to use The processor sets of x86 instruction set, it uses complicated order framework computer (CISC, Complex Instruction Set Computer) processor of instruction set.
Target platform node refers to the node of distribution application system platform after migration, such as ARM IA frame serverPC or virtual machine, ARM IA frame serverPC is referred to as advanced reduced instruction in the past Collection machine (Advanced RISC Machine), is 32 bit reduced instruction set computers (RISC) Central processing unit (processor) framework, it is widely used in many embedded systems (embedded) in design.
In order to by the application Data Migration on all for source platform nodes to target platform, need to ensure The hardware of all nodes of target platform can the business of all nodes of load source platform.Therefore, root According to hardware configuration and the service conditions of all node loads of source platform of each node of source platform, Determine the number n of node required for target platform.
Source platform hardware configuration information include the central processor CPU of node hardware, internal memory, network, The information such as hard disk.The mode of the hardware configuration information obtaining each node of source platform can be manually Mode, it is also possible to by automatic mode, wherein may is that migration work by automatic mode Have and set up by safety shell protocol SSH or Telnet TELNET or alternate manner The communication connection of node each with source platform and target platform, such that it is able to obtain each node of source platform Hardware configuration information.Each node of source platform can also be obtained of course through this automatic mode Other information of other information and target platform.Wherein, Migration tools operates in terminal, with The each node of source platform, each node of target platform are in consolidated network, can intercommunication.
Such as, source platform is N number of node, and target platform needs M node just can put down with source N number of node of platform is corresponding.
Step S102: utilize total data capacity size of all nodes of source platform, determines that target is put down The size of the data space of each node in n node of platform, wherein, all nodes of source platform Total data capacity size equal to the size of the total data space of n node of target platform.
The size of the data space of each node of target platform refers to that each node can be received Data capacity size；Total data capacity size of all nodes of source platform is equal to target platform n The size of the data space that individual node is total, if the configuration of n node of target platform is complete The same, then total data capacity size of all nodes of source platform averagely arrives target platform n On node, it is the size of the data space of each node in n node of target platform；As Really the configuration of n node of target platform is not duplicate, then can be according to concrete configuration feelings Condition is the size of each node reasonable distribution data space of n node of target platform.
Step S103: according to the operating system of the hardware configuration target platform of n node of target platform, So that the operating system of target platform and the Hardware match of n node of target platform.
The principle matched according to operating system and hardware, according to the hardware of n node of target platform I.e. can configure the operating system of target platform.
In actual applications, the principle matched according to operating system and hardware, can make in advance Operating system OS Mirroring Mapping table, provides the corresponding relation between operating system and hardware configuration. Such as, give x86 framework pass corresponding from the os release of the different linux kernel of ARM framework System.
Step S104: configuring the distributed of target platform according to the distribution application system of source platform should By system, so that the distribution of the applied environment of the distribution application system of target platform and source platform Difference between the applied environment of formula application system is in preset range.
The distribution application system of source platform include application software install file, application software patch, Application configuration parameter of each node etc..
Distribution application system moves to target platform between source platform, and perfect condition is distributed answering There is not any change with systematic difference environment, but in actual applications, this be to be difficult to accomplish, And also there is no need to allow the applied environment of distribution application system of target platform and source platform Difference is not had, as long as this difference can receive between the applied environment of distribution application system Scope is all possible.Therefore, target platform is configured according to the distribution application system of source platform Distribution application system, if the applied environment of the distribution application system of target platform and source platform Distribution application system applied environment between difference in preset range.
Step S105: after the operating system having configured target platform and distribution application system, press According to the size of the data space of each node of target platform, the application data on source platform are divided Do not move on n node of target platform.
After the operating system having configured target platform and distribution application system, according to target platform Application data on source platform can be migrated by the size of the data space of each node respectively On n node of target platform.
It is the implementation method that distribution application system of the present invention migrates between platform refering to Fig. 2, Fig. 2 Migration effect schematic diagram.Migration tools 11 operates in terminal, with source server (source platform) 12 each nodes, each node of destination server (target platform) 13 are in consolidated network, can Intercommunication.Operation system is Distributed Application, and data are distributed on each node, between node There is the relation of initial data and Backup Data, source server 12 node can be X86-based Server or virtual machine, destination server 13 node is the server of ARM framework or virtual Machine.Distribution application system has N number of node server, Migration tools 11 to lead on source platform 12 Spend the collection of the configuration of the N number of node hardware to source platform 12, and the joining of target platform 13 node Confidence ceases, and cooks up the cluster scale that target platform 13 is M ARM IA frame serverPC node； According further to total data capacity size of all nodes of source platform 12, carry out according to M node The division of the size of the data space of each node, cooks up each node of target platform 13 The size of data space, is configuring operating system and the distribution application system of target platform After, finally complete business datum and migrate and synchronize, it is achieved distribution application system is from N number of X86 The bulk migration of IA frame serverPC node to M ARM IA frame serverPC node.
Embodiment of the present invention obtains the number n of the node required for target platform；Utilize source platform Total data capacity size of all nodes, determines the number of each node in n node of target platform Size according to memory space；The operation of the hardware configuration target platform according to n node of target platform System；The distribution application system of the distribution application system configuration target platform according to source platform； According to the size of the data space of each node of target platform, by the application data on source platform Move to respectively on n node of target platform.In this way, can support from tradition X86-based server is to the migration of ARM IA frame serverPC；Distributed Application system can also be supported N number of node of system migrates scene to M node.
It is the implementation method that distribution application system of the present invention migrates between platform refering to Fig. 3, Fig. 3 The embodiment of the flow chart of another embodiment, present embodiment and Fig. 1 is essentially identical, identical Part refers to Fig. 1 and corresponding explanatory note, and difference refers to following content:
Step S201: the number n of the node required for acquisition target platform, so that target platform The hardware of n node can the business of all nodes of load source platform.
The node of source platform is the node of distribution application system platform before migration, such as The frame of X86-based or blade server, or virtual machine etc..X86-based server refers to use The processor sets of x86 instruction set, it uses complicated order framework computer (CISC, Complex Instruction Set Computer) processor of instruction set.
Target platform node refers to the node of distribution application system platform after migration, such as ARM IA frame serverPC or virtual machine, ARM IA frame serverPC is referred to as advanced reduced instruction in the past Collection machine (Advanced RISC Machine), is 32 bit reduced instruction set computers (RISC) Central processing unit (processor) framework, it is widely used in many embedded systems (embedded) in design.
In order to by the application Data Migration on all for source platform nodes to target platform, need to ensure The hardware of all nodes of target platform can the business of all nodes of load source platform.Therefore, root According to hardware configuration and the service conditions of all node loads of source platform of each node of source platform, Determine the number n of node required for target platform.
Wherein, step S201 includes three sub-steps, respectively:
Sub-step S201a: by one node of source platform, two node ..., m-1 node, m Service bearer amount P of individual node actual measurement1、P2、…、Pm-1、PmDivided by theoretical service bearer amount Q1、Q2、…、Qm-1、Qm, it is thus achieved that the cluster dilation coefficient C that source platform is corresponding1、C2、…、 Cm-1、Cm, wherein, Cm=Pm/Qm, the node number of definition source platform is that m, m are The natural number of zero.
Sub-step S201b: according to the cluster dilation coefficient C that source platform is corresponding1、C2、…、Cm-1、 Cm, it is thus achieved that the functional relation C(x between cluster dilation coefficient C and node number x).
Sub-step S201c: according to the functional relation C between cluster dilation coefficient and node number (x), hardware performance S of source platform single node, combination property utilization rate δ of source platform single node, Hardware performance T of target platform single node and target platform performance amount of resources reserved ρ, by answering With business load balance algorithm S* δ * Cm=T*(1-ρ) * C(x), it is thus achieved that required for target platform The number n of node.
Step S202: utilize total data capacity size of all nodes of source platform, determines that target is put down The size of the data space of each node in n node of platform, wherein, all nodes of source platform Total data capacity size equal to the size of the total data space of n node of target platform.
The size of the data space of each node of target platform refers to that each node can be received Data capacity size；Total data capacity size of all nodes of source platform is equal to target platform n The size of the data space that individual node is total, if the configuration of n node of target platform is complete The same, then total data capacity size of all nodes of source platform averagely arrives target platform n On node, it is the size of the data space of each node in n node of target platform；As Really the configuration of n node of target platform is not duplicate, then can be according to concrete configuration feelings Condition is the size of each node reasonable distribution data space of n node of target platform.
Wherein, if the configuration of n node of target platform is duplicate, step S202 has Body is: utilizes total data capacity size of all nodes of source platform, determines target platform n joint The size of the data space of each node in point, each node in n node of target platform The size of data space is equal to total data capacity size of all nodes of source platform divided by n's Business.
Step S203: according to the operating system of the hardware configuration target platform of n node of target platform, So that the operating system of target platform and the Hardware match of n node of target platform.
The principle matched according to operating system and hardware, according to the hardware of n node of target platform I.e. can configure the operating system of target platform.
In actual applications, the principle matched according to operating system and hardware, can make in advance Operating system OS Mirroring Mapping table, provides the corresponding relation between operating system and hardware configuration. Such as, give x86 framework pass corresponding from the os release of the different linux kernel of ARM framework System.
Step S204: configuring the distributed of target platform according to the distribution application system of source platform should By system, so that the distribution of the applied environment of the distribution application system of target platform and source platform Difference between the applied environment of formula application system is in preset range.
The distribution application system of source platform include application software install file, application software patch, Application configuration parameter of each node etc..
Distribution application system moves to target platform between source platform, and perfect condition is distributed answering There is not any change with systematic difference environment, but in actual applications, this be to be difficult to accomplish, And also there is no need to allow the applied environment of distribution application system of target platform and source platform Difference is not had, as long as this difference can receive between the applied environment of distribution application system Scope is all possible.Therefore, target platform is configured according to the distribution application system of source platform Distribution application system, if the applied environment of the distribution application system of target platform and source platform Distribution application system applied environment between difference in preset range.
Step S205: after the operating system having configured target platform and distribution application system, press According to the size of the data space of each node of target platform, the application data on source platform are divided Do not move on n node of target platform.
After the operating system having configured target platform and distribution application system, according to target platform Application data on source platform can be migrated by the size of the data space of each node respectively On n node of target platform.
Wherein, step S205 includes two sub-steps, and content is as follows:
Sub-step S205a: determine that on source platform, total original application data of all nodes are with total Back-up application data.
On source platform, what some node was deposited is original application data, and some node is deposited The back-up application data of original application data, it is determined that on source platform, all nodes is total original After application data and total back-up application data, can be by total original application Data Migration to target On the node of platform some, by the joint of total back-up application Data Migration to other some On point so that the pass between original application data and back-up application data tie up to migrate after still Will be apparent from understanding.
Sub-step S205b: according to the size of the data space of each node of target platform, will On source platform, total original application data of all nodes migrate respectively with total back-up application data On n node of target platform.
Embodiment of the present invention obtains the number n of the node required for target platform；Utilize source platform Total data capacity size of all nodes, determines the number of each node in n node of target platform Size according to memory space；The operation of the hardware configuration target platform according to n node of target platform System；The distribution application system of the distribution application system configuration target platform according to source platform； According to the size of the data space of each node of target platform, by the application data on source platform Move to respectively on n node of target platform.In this way, can support from tradition X86-based server is to the migration of ARM IA frame serverPC；Distributed Application system can also be supported N number of node of system migrates scene to M node.
Below to move to from the Hadoop distribution application system on X86 blade server Illustrating as a example by ARM IA frame serverPC node, concrete application schematic diagram is as shown in Figure 4.
By technical scheme, Hadoop application system can be completed from X86 blade service Device node cluster moves on ARM server node cluster, and source platform has N number of node, migrates Object includes the operating system of each node, Hadoop application system, Hadoop application data, Bulk migration step approximately as:
(1) Migration tools client and the Hadoop each node of application system source platform and target platform Each node is in same LAN, installation and deployment Migration tools in Migration tools client, fortune Row Migration tools.
(2) IP, root account of input source platform each node OS and connection on Migration tools Mode；It is simultaneously entered management IP and the management account of target platform node, supports that batch imports, logical Cross SSH/TELNET agreement to set up and source platform and the communication connection of each node of target platform.
(3) by Migration tools without agency's (Agent) acquisition mode, each node of source platform is collected Hardware configuration information, hardware performance data, application data set conjunction, application system configuration information etc. Deng；Hardware configuration information includes CPU model and data, internal memory model and quantity, Disk model With quantity, Raid(redundant arrays of inexpensive disks) relation, network interface card model and quantity, bond(binding) Relation；Hardware performance data includes the cpu busy percentage of each node, memory usage etc., and respectively The data message etc. of individual node traffic bearing capacity；Data acquisition system includes the data capacity of each node, each Deposit on individual node is original application data or back-up application data；OS configuration includes OS version Basis, system mend, host name, file system, business IP, basic software bag, patch, trust The information such as main frame, security strategy；Application system configuration information include hadoop software version, patch, The information such as application configuration parameter.
(3) Migration tools is by one node of source platform, two node ..., m-1 node, m Service bearer amount P of individual node actual measurement1、P2、…、Pm-1、PmDivided by theoretical service bearer amount Q1、Q2、…、Qm-1、Qm, it is thus achieved that the cluster dilation coefficient C that source platform is corresponding1、C2、…、 Cm-1、Cm, wherein, Cm=Pm/Qm, the node number of definition source platform is m；According to source platform Corresponding cluster dilation coefficient C1、C2、…、Cm-1、Cm, it is thus achieved that cluster dilation coefficient C and joint Functional relation C(x between some number x)；According between cluster dilation coefficient and node number Functional relation C(x), hardware performance S of source platform single node, source platform single node combine Close performance utilization rate δ, hardware performance T of target platform single node and target platform performance resource Pre-allowance ρ, by applied business load balancing algorithm S* δ * Cm=T*(1-ρ) * C(x), obtain The number n of node required for target platform.
(4) Migration tools utilizes total data capacity size of all nodes of source platform, determines mesh The size of the data space of each node in mark n node of platform.Specifically: by source platform The load of Hadoop applied business stops, and interrupts service, stops data and updates.According to source platform node Between the original application data deposited and back-up application data, by the Hadoop data of N number of node Carry out centralized integration respectively according to original application data and back-up application data, it is thus achieved that total original should By size and the size of total back-up application data of data, then for the target platform after migrating Number of nodes is repartitioned, and calculating target platform, to need how many nodes to deposit total original Application data and how many nodes deposit total back-up application data, and each node can be deposited The size of data space.
(5) by the hardware configuration information analysis to each node of target platform, generate target and put down The OS of M node of platform installs and configuration script, and the intervention combining expert adjusts, and completes target The OS environmental structure of platform and OS configuration.
(6) by the application system configuration information of the N number of node of source platform, target platform M is generated The Distributed Application software of individual node is installed and configuration script, and the intervention combining expert adjusts, and joins Put the distribution application system of target platform so that the distribution application system of target platform should By the difference between the applied environment of environment and the distribution application system of source platform at preset range In.
(7) according to the size of data space of each node of target platform, by source platform Total original application data of all nodes and total back-up application data move to target respectively and put down On n node of platform, the data syn-chronization completing M node of target platform migrates.
(8) IP, domain name, DNS and the letter on the node of M target platform of expert's involved switching The relation of appointing, adjusts and confirms the Hadoop configuration parameter after migrating, starting target platform Hadoop Application, and service switchover load, completion system bulk migration.
Refering to Fig. 5, Fig. 5 be distribution application system of the present invention migrate between platform realize device The structural representation of one embodiment, this device includes: obtain module 101, determine module 102, First configuration module 103, second configures module 104 and Data Migration module 105.
It should be noted that the device of present embodiment can perform the step in Fig. 1 and Fig. 3.
Obtain module 101 and be used for obtaining the number n of the node required for target platform, so that mesh The hardware of mark n node of platform can the business of all nodes of load source platform.
The node of source platform is the node of distribution application system platform before migration, such as The frame of X86-based or blade server, or virtual machine etc..X86-based server refers to use The processor sets of x86 instruction set, it uses the processor of complicated order framework computer instruction set.
Target platform node refers to the node of distribution application system platform after migration, such as ARM IA frame serverPC or virtual machine, ARM IA frame serverPC is referred to as advanced reduced instruction in the past Collection machine, is 32 bit reduced instruction set computer CPU architecture, and it is widely used in In many Embedded System Design.
In order to by the application Data Migration on all for source platform nodes to target platform, need to ensure The hardware of all nodes of target platform can the business of all nodes of load source platform.Therefore, root According to hardware configuration and the service conditions of all node loads of source platform of each node of source platform, Determine the number n of node required for target platform.
Source platform hardware configuration information include the central processor CPU of node hardware, internal memory, network, The information such as hard disk.The mode of the hardware configuration information obtaining each node of source platform can be manually Mode, it is also possible to by automatic mode, wherein may is that migration work by automatic mode Have and set up by safety shell protocol SSH or Telnet TELNET or alternate manner The communication connection of node each with source platform and target platform, such that it is able to obtain each node of source platform Hardware configuration information.Each node of source platform can also be obtained of course through this automatic mode Other information of other information and target platform.Wherein, Migration tools operates in terminal, with The each node of source platform, each node of target platform are in consolidated network, can intercommunication.
Determine that module 102 for obtaining the number of the node required for target platform in acquisition module 101 After mesh n, utilize total data capacity size of all nodes of source platform, determine target platform n The size of the data space of each node in node, wherein, all nodes of source platform total The size of the data space that data capacity size is total equal to n node of target platform.
The size of the data space of each node of target platform refers to that each node can be received Data capacity size；Total data capacity size of all nodes of source platform is equal to target platform n The size of the data space that individual node is total, if the configuration of n node of target platform is complete The same, then total data capacity size of all nodes of source platform averagely arrives target platform n On node, it is the size of the data space of each node in n node of target platform；As Really the configuration of n node of target platform is not duplicate, then can be according to concrete configuration feelings Condition is the size of each node reasonable distribution data space of n node of target platform.
First configuration module 103 is for the hardware configuration target platform according to n node of target platform Operating system so that the operating system of target platform and the hardware of n node of target platform Join.
The principle matched according to operating system and hardware, according to the hardware of n node of target platform I.e. can configure the operating system of target platform.
In actual applications, the principle matched according to operating system and hardware, can make in advance Operating system OS Mirroring Mapping table, provides the corresponding relation between operating system and hardware configuration. Such as, give x86 framework pass corresponding from the os release of the different linux kernel of ARM framework System.
Second configuration module 104 configures target platform for the distribution application system according to source platform Distribution application system so that the applied environment of the distribution application system of target platform and source Difference between the applied environment of the distribution application system of platform is in preset range.
The distribution application system of source platform include application software install file, application software patch, Application configuration parameter of each node etc..
Distribution application system moves to target platform between source platform, and perfect condition is distributed answering There is not any change with systematic difference environment, but in actual applications, this be to be difficult to accomplish, And also there is no need to allow the applied environment of distribution application system of target platform and source platform Difference is not had, as long as this difference can receive between the applied environment of distribution application system Scope is all possible.Therefore, target platform is configured according to the distribution application system of source platform Distribution application system, if the applied environment of the distribution application system of target platform and source platform Distribution application system applied environment between difference in preset range.
Data Migration module 105 is joined respectively in the first configuration module 103 and the second configuration module 104 After having put operating system and the distribution application system of target platform, according to determining that module 102 determines The size of data space of each node of target platform, the application data on source platform are divided Do not move on n node of target platform.
After the operating system having configured target platform and distribution application system, according to target platform Application data on source platform can be migrated by the size of the data space of each node respectively On n node of target platform.
Embodiment of the present invention obtains the number n of the node required for target platform；Utilize source platform Total data capacity size of all nodes, determines the number of each node in n node of target platform Size according to memory space；The operation of the hardware configuration target platform according to n node of target platform System；The distribution application system of the distribution application system configuration target platform according to source platform； According to the size of the data space of each node of target platform, by the application data on source platform Move to respectively on n node of target platform.In this way, can support from tradition X86-based server is to the migration of ARM IA frame serverPC；Distributed Application system can also be supported N number of node of system migrates scene to M node.
Refering to Fig. 6, Fig. 6 be distribution application system of the present invention migrate between platform realize device The embodiment of the structural representation of another embodiment, present embodiment and Fig. 5 is essentially identical, Something in common refers to Fig. 5 and corresponding explanatory note, and difference refers to following content.
This device includes: obtain module 201, determine module 202, first configure module 203, the Two configuration module 204 and Data Migration modules 205.
It should be noted that the device of present embodiment can perform the step in Fig. 3.
Obtain module 201 and be used for obtaining the number n of the node required for target platform, so that mesh The hardware of mark n node of platform can the business of all nodes of load source platform.
The node of source platform is the node of distribution application system platform before migration, such as The frame of X86-based or blade server, or virtual machine etc..X86-based server refers to use The processor sets of x86 instruction set, it uses the processor of complicated order framework computer instruction set.
Target platform node refers to the node of distribution application system platform after migration, such as ARM IA frame serverPC or virtual machine, ARM IA frame serverPC is referred to as advanced reduced instruction in the past Collection machine, is 32 bit reduced instruction set computer CPU architecture, and it is widely used in In many Embedded System Design.
In order to by the application Data Migration on all for source platform nodes to target platform, need to ensure The hardware of all nodes of target platform can the business of all nodes of load source platform.Therefore, root According to hardware configuration and the service conditions of all node loads of source platform of each node of source platform, Determine the number n of node required for target platform.
Wherein, it is thus achieved that module 201 includes: first obtains unit 2011, second obtains unit 2012 And the 3rd obtain unit 2013.
First obtain unit 2011 for by one node of source platform, two nodes ..., m-1 Node, service bearer amount P of m node actual measurement1、P2、…、Pm-1、PmDivided by theoretical industry Business bearing capacity Q1、Q2、…、Qm-1、Qm, it is thus achieved that the cluster dilation coefficient C that source platform is corresponding1、 C2、…、Cm-1、Cm, wherein, Cm=Pm/Qm, the node number of definition source platform is m, m For the natural number being not zero.
Second obtains unit 2012 is used for the collection that the source platform according to the first acquisition unit acquisition is corresponding Group dilation coefficient C1、C2、…、Cm-1、Cm, it is thus achieved that cluster dilation coefficient C and node number x Between functional relation C(x).
3rd obtain unit 2013 for according to second obtain cluster dilation coefficient that unit obtains with Functional relation C(x between node number), hardware performance S of source platform single node, source put down Combination property utilization rate δ of platform single node, hardware performance T of target platform single node and target Platform property amount of resources reserved ρ, by applied business load balancing algorithm S* δ * Cm=T*(1-ρ) * C(x), it is thus achieved that the number n of the node required for target platform.
Determine that module 202 for obtaining the number of the node required for target platform in acquisition module 201 After mesh n, utilize total data capacity size of all nodes of source platform, determine target platform n The size of the data space of each node in node, wherein, all nodes of source platform total The size of the data space that data capacity size is total equal to n node of target platform.
The size of the data space of each node of target platform refers to that each node can be received Data capacity size；Total data capacity size of all nodes of source platform is equal to target platform n The size of the data space that individual node is total, if the configuration of n node of target platform is complete The same, then total data capacity size of all nodes of source platform averagely arrives target platform n On node, it is the size of the data space of each node in n node of target platform；As Really the configuration of n node of target platform is not duplicate, then can be according to concrete configuration feelings Condition is the size of each node reasonable distribution data space of n node of target platform.
Wherein, if the configuration of n node of target platform is duplicate, then determine module 202, specifically for utilizing total data capacity size of all nodes of source platform, determine target platform The size of the data space of each node in n node, wherein, n node of target platform In the size of data space of each node hold equal to the total data of all nodes of source platform Amount size is divided by the business of n.
First configuration module 203 is for the hardware configuration target platform according to n node of target platform Operating system so that the operating system of target platform and the hardware of n node of target platform Join.
The principle matched according to operating system and hardware, according to the hardware of n node of target platform I.e. can configure the operating system of target platform.
In actual applications, the principle matched according to operating system and hardware, can make in advance Operating system OS Mirroring Mapping table, provides the corresponding relation between operating system and hardware configuration. Such as, give x86 framework pass corresponding from the os release of the different linux kernel of ARM framework System.
Second configuration module 204 configures target platform for the distribution application system according to source platform Distribution application system so that the applied environment of the distribution application system of target platform and source Difference between the applied environment of the distribution application system of platform is in preset range.
The distribution application system of source platform include application software install file, application software patch, Application configuration parameter of each node etc..
Distribution application system moves to target platform between source platform, and perfect condition is distributed answering There is not any change with systematic difference environment, but in actual applications, this be to be difficult to accomplish, And also there is no need to allow the applied environment of distribution application system of target platform and source platform Difference is not had, as long as this difference can receive between the applied environment of distribution application system Scope is all possible.Therefore, target platform is configured according to the distribution application system of source platform Distribution application system, if the applied environment of the distribution application system of target platform and source platform Distribution application system applied environment between difference in preset range.
Data Migration module 205 is joined respectively in the first configuration module 203 and the second configuration module 204 After having put operating system and the distribution application system of target platform, according to determining that module 202 determines The size of data space of each node of target platform, the application data on source platform are divided Do not move on n node of target platform.
After the operating system having configured target platform and distribution application system, according to target platform Application data on source platform can be migrated by the size of the data space of each node respectively On n node of target platform.
Wherein, Data Migration module 205 comprises determining that unit 2051 and data migration unit 2052.
Determine unit 2051 for determine on source platform total original application data of all nodes with Total back-up application data.
On source platform, what some node was deposited is original application data, and some node is deposited The back-up application data of original application data, it is determined that on source platform, all nodes is total original After application data and total back-up application data, can be by total original application Data Migration to target On the node of platform some, by the joint of total back-up application Data Migration to other some On point so that the pass between original application data and back-up application data tie up to migrate after still Will be apparent from understanding.
Data migration unit 2052 is for the data space according to each node of target platform Size, will determine total original application data of all nodes on the source platform that unit 2051 determines Move to respectively on n node of target platform with total back-up application data.
Embodiment of the present invention obtains the number n of the node required for target platform；Utilize source platform Total data capacity size of all nodes, determines the number of each node in n node of target platform Size according to memory space；The operation of the hardware configuration target platform according to n node of target platform System；The distribution application system of the distribution application system configuration target platform according to source platform； According to the size of the data space of each node of target platform, by the application data on source platform Move to respectively on n node of target platform.In this way, can support from tradition X86-based server is to the migration of ARM IA frame serverPC；Distributed Application system can also be supported N number of node of system migrates scene to M node.
Refering to Fig. 7, Fig. 7 be distribution application system of the present invention migrate between platform realize device The structural representation of another embodiment, this device includes: processor 71 and described processor The memorizer 72 of 71 couplings, migration engine 73 and data/address bus 74, wherein, memorizer 72 And migration engine 73 is connected with processor 71 by data/address bus 74.
Processor 72 is for obtaining the number n of the node required for target platform, so that described mesh The hardware of mark n node of platform can the business of all nodes of load source platform；Described source is utilized to put down Total data capacity size of all nodes of platform, determines each joint in n node of described target platform The size of the data space of point, wherein, total data capacity of all nodes of described source platform The size of the data space that size is total equal to n node of described target platform；According to described mesh The hardware controls migration engine 73 of mark n node of platform configures the operating system of described target platform, So that the operating system of described target platform and the Hardware match of described n node of target platform； Distribution application system according to described source platform controls migration engine 73 and configures described target platform Distribution application system so that the applied environment of the distribution application system of described target platform And the difference between the applied environment of the distribution application system of described source platform is in preset range； After the operating system and distribution application system of the good described target platform of configuration, according to described target The size of the data space of each node of platform, controls migration engine 73 by described source platform On application data move to respectively on n node of described target platform.
Wherein, described source platform node is X86-based server or virtual machine, described target platform Node is advanced Reduced Instruction Set Computer device A RM IA frame serverPC or virtual machine.
Wherein, described processor 72 be additionally operable to by one node of described source platform, two nodes ..., M-1 node, service bearer amount P of m node actual measurement1、P2、…、Pm-1、PmDivided by reason Service bearer amount Q of opinion1、Q2、…、Qm-1、Qm, it is thus achieved that the cluster that described source platform is corresponding expands Hold coefficient C1、C2、…、Cm-1、Cm, wherein, Cm=Pm/Qm, define described source platform Node number is m, and described m is the natural number being not zero；According to the cluster that described source platform is corresponding Dilation coefficient C1、C2、…、Cm-1、Cm, it is thus achieved that described cluster dilation coefficient C and described node Functional relation C(x between number x)；According to described cluster dilation coefficient and described node Functional relation C(x between number), hardware performance S of described source platform single node, described source Combination property utilization rate δ of platform single node, described target platform single node hardware performance T with And described target platform performance amount of resources reserved ρ, by applied business load balancing algorithm S* δ * Cm=T*(1-ρ) * C(x), it is thus achieved that the number n of the node required for described target platform.
Processor 72 is additionally operable to utilize total data capacity size of all nodes of described source platform, Determine the size of the data space of each node in n node of described target platform, wherein, In n node of described target platform, the size of the data space of each node is put down equal to described source Total data capacity size of all nodes of platform is divided by the business of described n.
Processor 72 is additionally operable to determine total original application data of all nodes on described source platform With total back-up application data；Data space big according to each node of described target platform Little, control migration engine 73 by total original application data of nodes all on described source platform with Total back-up application data move on n node of described target platform respectively.
Embodiment of the present invention obtains the number n of the node required for target platform；Utilize source platform Total data capacity size of all nodes, determines the number of each node in n node of target platform Size according to memory space；The operation of the hardware configuration target platform according to n node of target platform System；The distribution application system of the distribution application system configuration target platform according to source platform； According to the size of the data space of each node of target platform, by the application data on source platform Move to respectively on n node of target platform.In this way, can support from tradition X86-based server is to the migration of ARM IA frame serverPC；Distributed Application system can also be supported N number of node of system migrates scene to M node.
In several embodiments provided by the present invention, it should be understood that disclosed system, Apparatus and method, can realize by another way.Such as, device described above is implemented Mode is only schematically, such as, described module or the division of unit, it is only a kind of logic Function divides, and actual can have other dividing mode when realizing, and the most multiple unit or assembly can To combine or to be desirably integrated into another system, or some features can be ignored, or does not performs. Another point, shown or discussed coupling each other or direct-coupling or communication connection can be By INDIRECT COUPLING or the communication connection of some interfaces, device or unit, can be electrical, machinery Or other form.
The described unit illustrated as separating component can be or may not be physically separate , the parts shown as unit can be or may not be physical location, i.e. may be located at One place, or can also be distributed on multiple NE.Can select according to the actual needs Select some or all of unit therein to realize the purpose of present embodiment scheme.
It addition, each functional unit in each embodiment of the present invention can be integrated in a process In unit, it is also possible to be that unit is individually physically present, it is also possible to two or more unit Integrated in a unit.Above-mentioned integrated unit both can realize to use the form of hardware, it is possible to To use the form of SFU software functional unit to realize.
If described integrated unit realizes and as independent product using the form of SFU software functional unit When product are sold or use, can be stored in a computer read/write memory medium.Based on so Understanding, part that prior art is contributed by technical scheme the most in other words or Completely or partially can embodying with the form of software product of this technical scheme of person, this computer Software product is stored in a storage medium, including some instructions with so that a computer sets Standby (can be personal computer, server, or the network equipment etc.) or processor (processor) Perform all or part of step of method described in each embodiment of the present invention.And aforesaid storage is situated between Matter includes: USB flash disk, portable hard drive, read only memory (ROM, Read-Only Memory), Random access memory (RAM, Random Access Memory), magnetic disc or CD etc. are each Plant the medium that can store program code.
The foregoing is only embodiments of the present invention, not thereby limit the patent model of the present invention Enclosing, every equivalent structure utilizing description of the invention and accompanying drawing content to be made or equivalence flow process become Change, or be directly or indirectly used in other relevant technical fields, be the most in like manner included in the present invention's In scope of patent protection.

Claims (10)

1. the implementation method that a distribution application system migrates between platform, it is characterised in that Including:
The number n of the node required for acquisition target platform, so that described target platform n The hardware of node can the business of all nodes of load source platform；
Utilize total data capacity size of all nodes of described source platform, determine that described target is put down The size of the data space of each node in n node of platform, wherein, described source platform institute The data having total data capacity size of node to be equal to n node of described target platform total are deposited The size in storage space；
The operation system of target platform described in hardware configuration according to described n node of target platform System, so that the operating system of described target platform and the hardware of described n node of target platform Coupling；
Distribution application system according to described source platform configures the distributed of described target platform Application system, so that the applied environment of the distribution application system of described target platform is with described Difference between the applied environment of the distribution application system of source platform is in preset range；
After the operating system and distribution application system of the good described target platform of configuration, according to institute State the size of the data space of each node of target platform, by the application on described source platform Data move on n node of described target platform respectively.
Method the most according to claim 1, it is characterised in that described source platform node is X86-based server or virtual machine, described target platform node is advanced reduced instruction set computer meter Calculate machine equipment ARM IA frame serverPC or virtual machine.
Method the most according to claim 1, it is characterised in that described acquisition target platform The step of the number n of required node, including:
By one node of described source platform, two node ..., m-1 node, m nodes Service bearer amount P of actual measurement1、P2、…、Pm-1、PmDivided by theoretical service bearer amount Q1、 Q2、…、Qm-1、Qm, it is thus achieved that the cluster dilation coefficient C that described source platform is corresponding1、C2、…、 Cm-1、Cm, wherein, Cm=Pm/Qm, the node number defining described source platform is m, described M is the natural number being not zero；
According to the cluster dilation coefficient C that described source platform is corresponding1、C2、…、Cm-1、Cm, obtain Obtain functional relation C (x) between described cluster dilation coefficient C and described node number x；
According to functional relation C (x) between described cluster dilation coefficient and described node number, Hardware performance S of described source platform single node, the combination property of described source platform single node utilize Rate δ, hardware performance T of described target platform single node and described target platform performance resource Pre-allowance ρ, by applied business load balancing algorithm S × δ × Cm=T × (1-ρ) × C (x), Obtain the number n of node required for described target platform.
Method the most according to claim 1, it is characterised in that described utilize described source put down Total data capacity size of all nodes of platform, determines in n node of described target platform each The step of the size of the data space of node, including: in n node of described target platform The size of the data space of each node is equal to total number of all nodes of described source platform According to amount of capacity divided by the business of described n.
Method the most according to claim 1, it is characterised in that described according to described target Application data on described source platform are divided by the size of the data space of each node of platform Do not move to the step on n node of described target platform, including:
Determine that on described source platform, total original application data of all nodes should with total backup Use data；
According to the size of the data space of each node of described target platform, described source is put down On platform, total original application data of all nodes move to respectively with total back-up application data On n node of described target platform.
6. what a distribution application system migrated between platform realizes device, it is characterised in that Described device includes: obtains module, determine module, the first configuration module, the second configuration module And Data Migration module；
Described acquisition module is used for obtaining the number n of the node required for target platform, so that The hardware of described n node of target platform can the business of all nodes of load source platform；
Described determine that module for obtaining the node required for target platform in described acquisition module Number n after, utilize total data capacity size of all nodes of described source platform, determine institute State the size of the data space of each node in n node of target platform, wherein, described Total data capacity size of all nodes of source platform is total equal to n node of described target platform The size of data space；
Described first configuration module is for the hardware configuration according to described n node of target platform The operating system of described target platform, so that the operating system of described target platform and described mesh The Hardware match of mark n node of platform；
Described second configuration module configures for the distribution application system according to described source platform The distribution application system of described target platform, so that the Distributed Application of described target platform Between the applied environment of the distribution application system of systematic difference environment and described source platform Difference is in preset range；
Described Data Migration module is joined respectively in described first configuration module and the second configuration module After having put operating system and the distribution application system of described target platform, put down according to described target The size of the data space of each node of platform, by the application data on described source platform respectively Move on n node of described target platform.
Device the most according to claim 6, it is characterised in that described source platform node is X86-based server or virtual machine, described target platform node is advanced reduced instruction set computer meter Calculate machine equipment ARM IA frame serverPC or virtual machine.
Device the most according to claim 6, it is characterised in that described acquisition module includes: First obtains unit, the second acquisition unit and the 3rd acquisition unit；
Described first obtain unit for by one node of described source platform, two nodes ..., M-1 node, service bearer amount P of m node actual measurement1、P2、…、Pm-1、PmDivided by Theoretical service bearer amount Q1、Q2、…、Qm-1、Qm, it is thus achieved that the collection that described source platform is corresponding Group dilation coefficient C1、C2、…、Cm-1、Cm, wherein, Cm=Pm/Qm, define described source and put down The node number of platform is m, and described m is the natural number being not zero；
Described second obtains unit puts down for the described source obtaining unit acquisition according to described first The cluster dilation coefficient C that platform is corresponding1、C2、…、Cm-1、Cm, it is thus achieved that described cluster dilatation system Functional relation C (x) between number C and described node number x；
Described 3rd obtains unit for obtaining, according to described second, the described cluster that unit obtains Functional relation C (x) between dilation coefficient and described node number, described source platform single-unit Hardware performance S of point, combination property utilization rate δ of described source platform single node, described target Hardware performance T of platform single node and described target platform performance amount of resources reserved ρ, pass through Applied business load balancing algorithm S × δ × Cm=T × (1-ρ) × C (x), it is thus achieved that described target is put down The number n of the node required for platform.
Device the most according to claim 6, it is characterised in that described determine that module is concrete For the described total data capacity size utilizing all nodes of described source platform, determine described mesh The size of the data space of each node in mark n node of platform, wherein, described target In n node of platform, the size of the data space of each node is equal to described source platform institute There is total data capacity size business divided by described n of node.
Device the most according to claim 6, it is characterised in that described Data Migration mould Block comprises determining that unit and data migration unit；
Described determine that unit is for determining total original application of all nodes on described source platform Data and total back-up application data；
Described data migration unit stores for the data according to each node of described target platform The size in space, determines the total former of all nodes on the described source platform that unit determines by described The application data that begin and total back-up application data move to n joint of described target platform respectively On point.
CN201310547871.0A 2013-11-06 2013-11-06 Implementation method that distribution application system migrates between platform and device Active CN103577122B (en)

Priority Applications (2)

Application Number Priority Date Filing Date Title
CN201310547871.0A CN103577122B (en) 2013-11-06 2013-11-06 Implementation method that distribution application system migrates between platform and device
PCT/CN2014/086106 WO2015067093A1 (en) 2013-11-06 2014-09-09 Method and device for effecting migration of distributed application system between platforms

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201310547871.0A CN103577122B (en) 2013-11-06 2013-11-06 Implementation method that distribution application system migrates between platform and device

Publications (2)

Publication Number Publication Date
CN103577122A CN103577122A (en) 2014-02-12
CN103577122B true CN103577122B (en) 2016-08-17

Family

ID=50048988

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201310547871.0A Active CN103577122B (en) 2013-11-06 2013-11-06 Implementation method that distribution application system migrates between platform and device

Country Status (2)

Country Link
CN (1) CN103577122B (en)
WO (1) WO2015067093A1 (en)

Families Citing this family (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103577122B (en) * 2013-11-06 2016-08-17 杭州华为数字技术有限公司 Implementation method that distribution application system migrates between platform and device
CN104935618B (en) * 2014-03-19 2018-01-19 福建福昕软件开发股份有限公司 A kind of clustered deploy(ment) method
CN104468759B (en) * 2014-11-27 2018-06-01 中国联合网络通信集团有限公司 The method and apparatus that application migration is realized in PaaS platform
JP6548540B2 (en) * 2014-12-16 2019-07-24 キヤノン株式会社 Management system and control method of management system
CN107145309B (en) * 2017-05-04 2020-05-12 北京数盾信息科技有限公司 Distributed storage system based on ARM architecture
CN111435939B (en) * 2019-01-14 2023-05-05 百度在线网络技术（北京）有限公司 Method and device for dividing storage space of node
CN111459416B (en) * 2020-04-24 2021-02-23 杭州网银互联科技股份有限公司 Distributed storage-based thermal migration system and migration method thereof

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102103544A (en) * 2009-12-16 2011-06-22 腾讯科技（深圳）有限公司 Method and device for realizing distributed cache
CN103369042A (en) * 2013-07-10 2013-10-23 中国人民解放军国防科学技术大学 Data processing method and data processing device

Family Cites Families (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6775673B2 (en) * 2001-12-19 2004-08-10 Hewlett-Packard Development Company, L.P. Logical volume-level migration in a partition-based distributed file system
JP2005115438A (en) * 2003-10-03 2005-04-28 Mitsubishi Electric Corp Data management apparatus
US20090150461A1 (en) * 2007-12-07 2009-06-11 Brocade Communications Systems, Inc. Simplified snapshots in a distributed file system
JP5038163B2 (en) * 2008-01-07 2012-10-03 キヤノン株式会社 Distribution apparatus, system, information processing method, and program
US9047136B2 (en) * 2010-06-11 2015-06-02 Oracle International Corporation Method and system for migrating the state of a virtual cluster
WO2014032262A1 (en) * 2012-08-30 2014-03-06 华为技术有限公司 Implementation method and system for migrating application among platforms<0}
CN103577122B (en) * 2013-11-06 2016-08-17 杭州华为数字技术有限公司 Implementation method that distribution application system migrates between platform and device

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102103544A (en) * 2009-12-16 2011-06-22 腾讯科技（深圳）有限公司 Method and device for realizing distributed cache
CN103369042A (en) * 2013-07-10 2013-10-23 中国人民解放军国防科学技术大学 Data processing method and data processing device

Also Published As

Publication number Publication date
WO2015067093A1 (en) 2015-05-14
CN103577122A (en) 2014-02-12

Similar Documents

Publication Publication Date Title
CN103577122B (en) Implementation method that distribution application system migrates between platform and device
CN103765372B (en) It is configured to the object storage system of input/output operations
US11681668B2 (en) System and method for developing and implementing a migration plan for migrating a file system
CN107707393B (en) Multi-active system based on Openstack O version characteristics
CN107391629B (en) Method, system, server and computer storage medium for data migration between clusters
CN102110071B (en) Virtual machine cluster system and implementation method thereof
US11146626B2 (en) Cloud computing environment with replication system configured to reduce latency of data read access
WO2017092017A1 (en) Method, apparatus and system for migrating virtual machine backup information
CN110362381A (en) HDFS cluster High Availabitity dispositions method, system, equipment and storage medium
US10942759B2 (en) Seamless virtual standard switch to virtual distributed switch migration for hyper-converged infrastructure
US20150100826A1 (en) Fault domains on modern hardware
US20130036213A1 (en) Virtual private clouds
Tsai et al. Two-tier multi-tenancy scaling and load balancing
US20210109890A1 (en) System and method for planning and configuring a file system migration
CN106168884A (en) Access the computer system of object storage system
CN104572274A (en) Cross-cloud-node migration system and cross-cloud-node migration method
US20180189124A1 (en) Rebuilding the namespace in a hierarchical union mounted file system
CN103414767A (en) Method and device for deploying application software on cloud computing platform
CN104954458A (en) Configuration method of cloud operating system based on Loongson platform
CN105095103A (en) Storage device management method and device used for cloud environment
CN204650507U (en) Virtual resources integration platform in a kind of system for cloud computing
CN116089011A (en) Method and device for creating mirror warehouse, storage medium and electronic equipment
Vaquero et al. Deploying large-scale datasets on-demand in the cloud: treats and tricks on data distribution
JP2016206952A (en) Migration support device and migration support method
CN203301532U (en) Cloud desktop system

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
C14 Grant of patent or utility model
GR01 Patent grant
TR01 Transfer of patent right
TR01 Transfer of patent right

Effective date of registration: 20200417

Address after: 518129 Bantian HUAWEI headquarters office building, Longgang District, Guangdong, Shenzhen

Patentee after: HUAWEI TECHNOLOGIES Co.,Ltd.

Address before: 301, A building, room 3, building 301, foreshore Road, No. 310052, Binjiang District, Zhejiang, Hangzhou

Patentee before: Huawei Technologies Co.,Ltd.