CN104158675B - Calculate node dispositions method, processing node, controller and system - Google Patents

Calculate node dispositions method, processing node, controller and system Download PDF

Info

Publication number
CN104158675B
CN104158675B CN201310174927.2A CN201310174927A CN104158675B CN 104158675 B CN104158675 B CN 104158675B CN 201310174927 A CN201310174927 A CN 201310174927A CN 104158675 B CN104158675 B CN 104158675B
Authority
CN
China
Prior art keywords
information
calculate node
disposed
node
deployment
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
CN201310174927.2A
Other languages
Chinese (zh)
Other versions
CN104158675A (en
Inventor
李峰
周洪
殷悦
陈国海
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
Huawei 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 Huawei Technologies Co Ltd filed Critical Huawei Technologies Co Ltd
Priority to CN201310174927.2A priority Critical patent/CN104158675B/en
Priority to PCT/CN2014/076828 priority patent/WO2014183574A1/en
Publication of CN104158675A publication Critical patent/CN104158675A/en
Application granted granted Critical
Publication of CN104158675B publication Critical patent/CN104158675B/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
    • G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
    • G06F9/5061—Partitioning or combining of resources
    • G06F9/5072—Grid computing
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/50—Network service management, e.g. ensuring proper service fulfilment according to agreements
    • H04L41/508—Network service management, e.g. ensuring proper service fulfilment according to agreements based on type of value added network service under agreement
    • H04L41/5096—Network service management, e.g. ensuring proper service fulfilment according to agreements based on type of value added network service under agreement wherein the managed service relates to distributed or central networked applications

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Software Systems (AREA)
  • Theoretical Computer Science (AREA)
  • Mathematical Physics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Data Exchanges In Wide-Area Networks (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

A kind of calculate node dispositions method of offer of the embodiment of the present invention, processing node, controller and system.This method includes：It receives the deployment recommendation request message that processing node is sent and deployment scheme is determined according to link information and/or flow information；The deployment recommendation response message for including deployment scheme is sent to processing node, so that processing node disposes calculate node to be disposed according to deployment scheme.Calculate node dispositions method provided in an embodiment of the present invention, controller, for new calculate node, can provide the deployment scheme of preferable deployed position according to the link information between each data center in operation system, flow information between each calculate node etc.；For having disposed calculate node in operation system, the deployment scheme for redeploying position can be provided, so that processing node can be disposed the calculate node to be disposed according to deployment scheme, to improve the communication quality between calculate node, reduce the communication flows between data center.

Description

Calculate node dispositions method, processing node, controller and system
Technical field
The present embodiments relate to the communications field more particularly to a kind of calculate node dispositions method, processing node, controllers And system.
Background technology
The appearance of virtual technology provides more modes for service deployment, as largely used virtual machine in cloud computing (Virtual Machine, hereinafter referred to as VM) is provided when content supplier (Content Provider, hereinafter referred to as CP) Inner capacities is bigger, when needing more computing resources, can apply to data center (Data Center, hereinafter referred to as DC) Dispose more virtual machines；When the portfolio of CP declines, it is possible to reduce the quantity of virtual machine.
In the prior art, when disposing virtual machine, by judging the surplus resources of server in DC, such as memory, center Whether processing unit (Central Processing Unit, hereinafter referred to as CPU) meets calculating demand, determines whether to need Virtual machine is disposed on that server.If asking to be CPU number to be two for example, receiving deploying virtual machine, inside save as 1024M, disk read-write number are 50 times, then a server is selected from resource pool, judge the remaining money for the server chosen Source：The number of CPU whether at least two, memory whether be at least 1024M, disk read-write number whether be at least 50 times, if One of which is undesirable, then the server is marked not meet the requirement of this selection, again new to selecting in resource pool Server.
However, above-mentioned virtual machine deployment method is selected from resource pool every time when needing to dispose multiple virtual machines One server, judges whether the surplus resources of the server meet demand and then dispose virtual machine, each virtual machine possibility portion Administration is on the server of different data centers.If there are a large amount of network communications between virtual machine, the virtual machine of communication is participated in Positioned at different data centers, then the network flow between data center can be caused to increase, data congestion, communication quality is poor.
Invention content
The embodiment of the present invention provides a kind of calculate node dispositions method, processing node, controller and system, controller according to Link information, flow information etc. treat deployment calculate node and provide rational deployment scheme so that processing node can be according to deployment Scheme disposes the calculate node to be disposed to reduce the communication flows between data center, improves communication quality.
The first aspect, the embodiment of the present invention provide a kind of calculate node dispositions method, including：
The deployment recommendation request message that processing node is sent is received, the deployment recommendation request message, which carries, waits for that deployment calculates The description information of node；
According to link information and/or flow information, deployment scheme is determined, wherein the link information includes the processing Link information between each data center that node is managed, and/or, each data center for being managed of processing node with not Belong to the link information between each data center that the processing node is managed；The flow information waits for that deployment calculates described in being Flow information between node and the relevant calculate node of calculate node to be disposed, wherein wait for that deployment calculates section with described The relevant calculate node of point is the calculate node for having with the calculate node to be disposed communication requirement；
The deployment recommendation response message for including the deployment scheme is sent to the processing node.
Described in also being carried in the possible realization method of in the first aspect the first, in the deployment recommendation request message The deployment request information of calculate node to be disposed；
It is described to determine deployment scheme according to link information and/or flow information, including：
According to the link information and/or flow information, the deployment scheme for meeting the deployment request information is determined.
In conjunction with the first possible realization method of the first aspect, second of possible realization side in the first aspect In formula, the deployment request information includes：
Relative position information, the calculate node to be disposed between the calculate node to be disposed are saved with calculating has been disposed Relative position information between point, described waits for deployment calculating at the QoS requirement information between calculate node dispose Node communicates always with QoS requirement information, the calculate node to be disposed disposed between calculate node across data center A kind of information in traffic requirement information or combinations thereof.
In conjunction with the first or second of possible realization method of the first aspect, the first aspect, in the first aspect The third possible realization method in, the description information includes the identification information of calculate node to be disposed, described waits disposing Tenant identification information belonging to the quantity information of calculate node or the calculate node to be disposed.
In conjunction with the first aspect, the first aspect the first, second or the third possible realization method, first In 4th kind of possible realization method of a aspect, the deployment recommendation request message for receiving processing node and sending, the portion Administration suggests that request message carries the description information of calculate node to be disposed, including：
By acting on behalf of the deployment recommendation request message for receiving the processing node and sending；
It is described to send the deployment recommendation response message for including the deployment scheme to the processing node, so that the place Node is managed according to calculate node to be disposed described in deployment scheme deployment, including：
The deployment recommendation response message for including the deployment scheme is sent by described act on behalf of to the processing node, so that The processing node is obtained according to calculate node to be disposed described in deployment scheme deployment.
In conjunction with the first aspect, the first aspect the first to any middle realization in the 4th kind of possible realization method Mode, in the 5th kind of possible realization method in the first aspect, the calculate node to be disposed includes newly-increased calculate node Or calculate node is disposed.
The second aspect, the embodiment of the present invention provide a kind of calculate node dispositions method, including：
The deployment information request message that processing node is sent is received, the deployment information request message, which carries, waits for that deployment calculates The description information of node；
Obtain link information and/or flow information, wherein the link information include the processing node managed it is each Link information between data center, and/or, each data center and be not belonging to the processing section that the processing node is managed Link information between each data center that point is managed, the flow information are that the calculate node to be disposed waits for portion with described Affix one's name to the flow information between the relevant calculate node of calculate node, wherein the relevant calculate node of calculate node to be disposed To there is the calculate node of communication requirement with calculate node to be disposed；
The deployment information response message comprising the link information and/or flow information is sent to the processing node.
In the possible realization method of in the second aspect the first, the description information includes calculate node to be disposed Identification information, the quantity information of the calculate node to be disposed or the tenant identification information belonging to the calculate node to be disposed.
In conjunction with the possible realization method of the first of the second aspect or the second aspect, the second of the second aspect In the possible realization method of kind, the deployment information request message for receiving processing node and sending, the deployment information request disappears Breath carries the description information of calculate node to be disposed, including：
By acting on behalf of the deployment information request message for receiving the processing node and sending, the deployment information request message is taken The description information of band calculate node to be disposed；
It is described to disappear to deployment information response of the processing node transmission comprising the link information and/or flow information Breath, so that the processing node determines deployment scheme according to the link information and/or flow information, including：
The deployment information comprising the link information and/or flow information is sent by described act on behalf of to the processing node Response message, so that the processing node determines deployment scheme according to the link information and/or flow information.
In conjunction with the first or second of possible realization method of the second aspect, the second aspect, in the second aspect The third possible realization method in, the calculate node to be disposed includes newly-increased calculate node or has disposed calculate node.
In terms of third, the embodiment of the present invention provides a kind of calculate node dispositions method, including：
Deployment recommendation request message is sent to controller, the deployment recommendation request message carries calculate node to be disposed Description information, so that the controller determines deployment scheme according to link information and/or flow information, wherein the link Information includes the link information between each data center that the processing node is managed, and/or, the processing node is managed Each data center and the link information that is not belonging between each data center for being managed of processing node；The flow information For the flow information between the calculate node to be disposed and the relevant calculate node of calculate node to be disposed, wherein with The relevant calculate node of calculate node to be disposed is the calculate node for having with the calculate node to be disposed communication requirement；
Receive the deployment recommendation response message for including the deployment scheme that the controller is sent.
In the first possible realization method in terms of third, the deployment recommendation sent to the controller is asked The deployment request information for seeking calculate node to be disposed described in also being carried in message, so that the controller is believed according to the link Breath and/or flow information determine the deployment scheme for meeting the deployment request information.
In conjunction with the first possible realization method in terms of third, second of possible realization side in terms of third In formula, the deployment request information includes：Relative position information between the calculate node to be disposed described waits for that deployment calculates Node and the QoS requirement disposed between the relative position information between calculate node, the calculate node to be disposed are believed Breath, described wait for that deployment calculates and saves calculate node dispose and the QoS requirement information disposed between calculate node A kind of information or combinations thereof in across the data center communication total flow require information of point.
The first in terms of in conjunction with third, in terms of third or second of possible realization method, in terms of third The third possible realization method in, the description information includes the identification information of calculate node to be disposed, described waits disposing Tenant identification information belonging to the quantity information of calculate node or the calculate node to be disposed.
In conjunction in terms of third, third aspect the first, second or the third possible realization method, in third It is described to send deployment recommendation request message to controller in 4th kind of possible realization method of a aspect, including：Pass through agency Deployment recommendation request message is sent to controller；
The deployment recommendation response message for including the deployment scheme for receiving the controller and sending, including：Pass through The agency receives the deployment recommendation response message for including the deployment scheme that the controller is sent.
In conjunction with it is in terms of third, third aspect the first, second or the third possible realization method, the In 4th kind of possible realization method of three aspects, the calculate node to be disposed includes newly-increased calculate node or has disposed meter Operator node.
4th aspect, the embodiment of the present invention provide a kind of calculate node dispositions method, including：
Deployment information request message is sent to controller, the deployment information request message carries calculate node to be disposed Description information, so that the controller obtains link information and/or flow information, wherein the link information includes described Link information between each data center that processing node is managed, and/or, each data center that the processing node is managed With the link information being not belonging between each data center that the processing node is managed, the flow information waits disposing to be described Flow information between calculate node and the relevant calculate node of calculate node to be disposed, wherein wait for deployment meter with described The relevant calculate node of operator node is the calculate node for having with the calculate node to be disposed communication requirement；
Receive the deployment information response message comprising the link information and/or flow information that the controller is sent；
Deployment scheme is determined according to the link information and/or flow information.
In the first possible realization method of the 4th aspect, the description information includes calculate node to be disposed Identification information, the quantity information of the calculate node to be disposed or the tenant identification information belonging to the calculate node to be disposed.
In conjunction with the first possible realization method of the 4th aspect or the 4th aspect, at second of the 4th aspect It is described to send deployment information request message to controller in possible realization method, including：By acting on behalf of to controller sending part Affix one's name to information request message；
The response message comprising the link information and/or flow information for receiving the controller and sending, including； The deployment information response comprising the link information and/or flow information that the controller is sent is received by the agency to disappear Breath.
In conjunction with the first or second of possible realization method of the 4th aspect, the 4th aspect, at the 4th aspect The third possible realization method in, the calculate node to be disposed includes newly-increased calculate node or has disposed calculate node.
5th aspect, the embodiment of the present invention provide a kind of controller, including：
Receiving module, the deployment recommendation request message sent for receiving processing node, the deployment recommendation request message Carry the description information of calculate node to be disposed；
Determining module, for according to link information and/or flow information, determining deployment scheme, wherein the link information Link information between each data center managed including the processing node, and/or, the processing node is managed each Data center and the link information being not belonging between each data center that the processing node is managed；The flow information is institute State the flow information between calculate node to be disposed and the relevant calculate node of calculate node to be disposed, wherein with it is described The relevant calculate node of calculate node to be disposed is the calculate node for having with the calculate node to be disposed communication requirement；
Sending module, for sending the deployment recommendation response message for including the deployment scheme to the processing node.
In the first possible realization method of the 5th aspect, the deployment recommendation that the receiving module receives is asked The deployment request information of calculate node to be disposed described in also being carried in message；
The determining module is additionally operable to according to the link information and/or flow information, and determination meets the deployment request The deployment scheme of information.
In conjunction with the first possible realization method of the 5th aspect, in second of possible realization side of the 5th aspect In formula, the deployment request information includes：
Relative position information, the calculate node to be disposed between the calculate node to be disposed are saved with calculating has been disposed Relative position information between point, described waits for deployment calculating at the QoS requirement information between calculate node dispose Node communicates always with QoS requirement information, the calculate node to be disposed disposed between calculate node across data center A kind of information in traffic requirement information or combinations thereof.
In conjunction with the first or second of possible realization method of the 5th aspect, the 5th aspect, at the 5th aspect The third possible realization method in, the description information includes the identification information of calculate node to be disposed, described waits disposing Tenant identification information belonging to the quantity information of calculate node or the calculate node to be disposed.
In conjunction with the 5th aspect, the 5th aspect the first, second or the third possible realization method, the 5th In 4th kind of possible realization method of a aspect, the receiving module is additionally operable to receive the processing node transmission by acting on behalf of Deployment recommendation request message；
The sending module is additionally operable to send the portion for including the deployment scheme by described act on behalf of to the processing node Response message is suggested by administration.
In conjunction with the 5th aspect, the 5th aspect the first, second, the third or the 4th possible realization method, In the 5th kind of possible realization method of the 5th aspect, the calculate node to be disposed includes increasing calculate node or portion newly Affix one's name to calculate node.
6th aspect, the embodiment of the present invention provide a kind of controller, including：
Receiving module, the deployment information request message sent for receiving processing node, the deployment information request message Carry the description information of calculate node to be disposed；
Acquisition module, for obtaining link information and/or flow information, wherein the link information includes the processing Link information between each data center that node is managed, and/or, each data center for being managed of processing node with not Belong to the link information between each data center that the processing node is managed, the flow information waits for that deployment calculates to be described Flow information between node and the relevant calculate node of calculate node to be disposed, wherein the calculate node to be disposed Relevant calculate node is the calculate node for having with calculate node to be disposed communication requirement；
Sending module, for sending the deployment letter comprising the link information and/or flow information to the processing node Cease response message.
In the first possible realization method of the 6th aspect, the description information includes calculate node to be disposed Identification information, the quantity information of the calculate node to be disposed or the tenant identification information belonging to the calculate node to be disposed.
In conjunction with the first possible realization method of the 6th aspect or the 6th aspect, at second of the 6th aspect In possible realization method, the receiving module is additionally operable to by acting on behalf of the deployment information request for receiving the processing node and sending Message, the deployment information request message carry the description information of calculate node to be disposed；
The sending module be additionally operable to by it is described act on behalf of sent to the processing node comprising the link information and/or The deployment information response message of flow information, so that the processing node is determined according to the link information and/or flow information Deployment scheme.
In conjunction with the first or second of possible realization method of the 6th aspect, the 6th aspect, at the 6th aspect The third possible realization method in, the calculate node to be disposed includes newly-increased calculate node or has disposed calculate node.
7th aspect, the embodiment of the present invention provide a kind of processing node, including：
Sending module, for sending deployment recommendation request message to controller, the deployment recommendation request message carrying waits for The description information for disposing calculate node, so that the controller is according to link information and/or flow information, determining section management side Case, wherein the link information includes the link information between each data center that the processing node is managed, and/or, institute State each data center and the link being not belonging between each data center that the processing node is managed that processing node is managed Information；The flow information is described between calculate node to be disposed and the relevant calculate node of calculate node to be disposed Flow information, wherein with the relevant calculate node of calculate node to be disposed be with the calculate node to be disposed to have communication The calculate node of demand；
Receiving module, the deployment recommendation response message for including the deployment scheme sent for receiving the controller.
In the first possible realization method of the 7th aspect, the sending module is additionally operable to send out to the controller The deployment recommendation request message for sending the deployment request information of calculate node to be disposed described in carrying, so that the controller According to link information and/or flow information, the deployment scheme for meeting the deployment request information is determined.
In conjunction with the first possible realization method of the 7th aspect, in second of possible realization side of the 7th aspect In formula, the deployment request information includes：
Relative position information, the calculate node to be disposed between the calculate node to be disposed are saved with calculating has been disposed Relative position information between point, described waits for deployment calculating at the QoS requirement information between calculate node dispose Node communicates always with QoS requirement information, the calculate node to be disposed disposed between calculate node across data center A kind of information in traffic requirement information or combinations thereof.
In conjunction with the first or second of possible realization method of the 7th aspect, the 7th aspect, at the 7th aspect The third possible realization method in, the description information includes the identification information of calculate node to be disposed, described waits disposing Tenant identification information belonging to the quantity information of calculate node or the calculate node to be disposed.
In conjunction with the 7th aspect, the 7th aspect the first, second or the third possible realization method, the 7th In 4th kind of possible realization method of a aspect, the sending module is additionally operable to send deployment recommendation to controller by agency Request message；
The receiving module is additionally operable to through agency's reception controller transmission include the deployment scheme Deployment recommendation response message.
In conjunction with the 7th aspect, the 7th aspect the first, second, the third or the 4th kind of possible realization side Formula, the 7th aspect the 5th kind of possible realization method in, calculate node dispose include increase newly calculate node or Calculate node is disposed.
8th aspect, a kind of processing node of the embodiment of the present invention, including：
Sending module, for sending deployment information request message to controller, the deployment information request message carrying waits for The description information of calculate node is disposed, so that the controller obtains link information and/or flow information, wherein the chain Road information includes the link information between each data center that the processing node is managed, and/or, the processing node is managed Each data center of reason and the link information being not belonging between each data center that the processing node is managed, the flow letter Breath is the flow information between calculate node to be disposed and the relevant calculate node of calculate node to be disposed, wherein It is the calculate node for having with the calculate node to be disposed communication requirement with the relevant calculate node of calculate node to be disposed；
Receiving module, the deployment comprising the link information and/or flow information sent for receiving the controller Info response message；
Determining module, for determining deployment scheme according to the link information and/or flow information.
In the first possible realization method of the 8th aspect, the description information includes calculate node to be disposed Node identification information, the quantity information of the calculate node to be disposed, the tenant identification letter belonging to the calculate node to be disposed Breath, alternatively, the characteristic information belonging to the calculate node to be disposed.
In conjunction with the first possible realization method of the 8th aspect or the 8th aspect, at second of the 8th aspect In possible realization method, the sending module is additionally operable to send deployment information request message to controller by agency；
The receiving module be additionally operable to by the agency receive that the controller sends comprising the link information and/ Or the deployment information response message of flow information.
In conjunction with the first or second of possible realization method of the 8th aspect, the 8th aspect, at the 8th aspect The third possible realization method in, the calculate node to be disposed includes newly-increased calculate node or has disposed calculate node.
9th aspect, the embodiment of the present invention provide a kind of operation system, including the control described in the 5th aspect as above Processing node described in device and the 7th aspect as above.
Tenth aspect, the embodiment of the present invention provide a kind of operation system, including the control described in the 6th aspect as above Processing node described in device and the 8th aspect as above.
Calculate node management method, processing node, controller and system provided in an embodiment of the present invention, controller is according to industry Link information or calculate node to be disposed in business system between each data center that calculate node to be disposed can dispose can portions Link information between each data center that the data center of administration and calculate node to be disposed can not dispose calculates section with deployment is waited for Flow information etc. between the relevant each calculate node of point, for new calculate node, it may be determined that go out preferable deployed position Deployment scheme so that processing node can be disposed new calculate node according to deployment scheme；For portion in operation system The calculate node of administration can provide the deployment scheme for redeploying position so that processing node can according to deployment scheme to this It disposes calculate node and carries out position adjustment, the communication flows between data center is changed into the flow in data center, to It improves the communication quality between calculate node, reduce the communication flows between data center.
Description of the drawings
In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, to embodiment or will show below There is attached drawing needed in technology description to be briefly described, it should be apparent that, the accompanying drawings in the following description is this hair Some bright embodiments for those of ordinary skill in the art without having to pay creative labor, can be with Obtain other attached drawings according to these attached drawings.
The first service system structure schematic diagram that Fig. 1 is applicable in for calculate node dispositions method of the present invention；
Fig. 2 is the flow chart of calculate node dispositions method embodiment one of the present invention；
Fig. 3 is the flow chart of calculate node dispositions method embodiment two of the present invention；
Fig. 4 is the flow chart of calculate node dispositions method embodiment three of the present invention；
Fig. 5 is the flow chart of calculate node dispositions method example IV of the present invention；
The second service system structure schematic diagram that Fig. 6 is applicable in for calculate node dispositions method of the present invention；
Fig. 7 is the structural schematic diagram of controller embodiment one of the present invention；
Fig. 8 is the structural schematic diagram of controller embodiment two of the present invention；
Fig. 9 is the structural schematic diagram of present invention processing node embodiment one；
Figure 10 is the structural schematic diagram of present invention processing node embodiment two；
Figure 11 is the structural schematic diagram of controller embodiment three of the present invention；
Figure 12 is the structural schematic diagram of controller example IV of the present invention；
Figure 13 is the structural schematic diagram of present invention processing node embodiment three；
Figure 14 is the structural schematic diagram of present invention processing node embodiment four.
Specific implementation mode
In order to make the object, technical scheme and advantages of the embodiment of the invention clearer, below in conjunction with the embodiment of the present invention In attached drawing, technical scheme in the embodiment of the invention is clearly and completely described, it is clear that described embodiment is A part of the embodiment of the present invention, instead of all the embodiments.Based on the embodiments of the present invention, those of ordinary skill in the art The every other embodiment obtained without creative efforts, shall fall within the protection scope of the present invention.
The first service system structure schematic diagram that Fig. 1 is applicable in for calculate node dispositions method of the present invention.Fig. 1 is please referred to, In the present embodiment, processing node administration data center DC_A, DC_B, DC_C, as shown in solid black lines in figure, data center DC_D Be not belonging to processing node administration, and there is communication between data center DC_A, DC_B, DC_C, data center DC_B and DC_D it Between there is communication, can dispose or deploy calculate node (not shown) in each data center.Wherein, it counts Operator node refer to it is various with computing capability, the function module of data processing can be carried out, such as virtual machine (Virtual Machine, hereinafter referred to as VM), calculate container (Linux Container, hereinafter referred to as LXC) and physical server etc. by soft Part or hard-wired function module, wherein it is, for example, to provide virtual execution on operating system level for process to calculate container Environment, CPU distribution times with special ratios, input and output (Input Output, hereinafter referred to as IO), limitation can make The container of memory size etc.
Handling node can be by hardware or software realization, for example, it may include hard-wired calculate node management Center can be third party application, the tenant etc. of software realization, system that the present invention is not limited thereto.When processing node Specially calculate node administrative center when, be also responsible for the management of each calculate node in operation system, as shown in phantom in FIG., The calculate node of each data center can be managed collectively, or administrative section calculate node according to demand.In addition, also may be used It is targetedly divided into sub- administrative center so that node will be handled, such as is divided into Virtual Machine Manager subcenter, Container Management Heart etc., Virtual Machine Manager subcenter are responsible for the management of virtual machine, such as virtual machine establishment, startup, delete, freeze and migrate； Container Management subcenter is responsible for calculating the management of container, such as calculates the establishment of container, startup, deletes, freezes and migrate.Specifically , processing node by with the server of each data center interact to complete calculate node management.
As shown in fig. 1 by dash-dotted lines, controller can be based on hypertext transfer protocol (Hyper Text with processing node Transport Protocol, hereinafter referred to as HTTP) or transmission control protocol (Transmission Control Protocol, hereinafter referred to as TCP) it is communicated；Alternatively, User Datagram Protocol (User Datagram can also be based on Protocol, hereinafter referred to as UDP) it is communicated.Controller receives the portion that processing node is sent by the connection with processing node Request message is suggested by administration, and the deployment scheme of calculate node to be disposed is determined according to link information and/or flow information, will be disposed Scheme carrying is sent to processing node in deployment recommendation response message；Alternatively, controller is connect by the connection with processing node The deployment information request message that processing node is sent is received, obtains link information and/or flow information, and the link got is believed Breath and/or flow information carrying are sent to processing node in deployment information response message so that processing node is determined to wait for portion Affix one's name to the deployment scheme of calculate node.Wherein, link information includes：Link letter between each data center that processing node is managed Breath, and/or, each data center and be not belonging between each data center that the processing node is managed that processing node is managed Link information；Flow information is the flow letter between calculate node to be disposed and the relevant calculate node of calculate node to be disposed Breath, wherein with the relevant calculate node of calculate node dispose to there is with calculate node to be disposed the calculate node of communication requirement. Fig. 1 is please referred to, the data center that processing node is managed is DC_A, DC_B, DC_C, then controller is according to DC_A, DC_B, DC_C Between link information determine deployment scheme；Or it can also be true according to the link information between DC_A, DC_B, DC_C and DC_D Determine deployment scheme.In addition, other than the link information between considering data center, it is also contemplated that the stream between calculate node to be disposed Information is measured so that it is determined that deployment scheme.For example, for increasing calculate node newly, it can be according to link information, flow information or deployment Require information etc. determines preferable deployed position；For having disposed calculate node in operation system, according between calculate node Link information or deployment request information between flow information, data center etc. provide the suggestion of the position redeployed, alternatively, giving Go out to have disposed the flow list or traffic matrix of the flow information of calculate node.
Wherein, link information refers to the link-state information between operation system Zhong Ge data centers, includes total band of link Width, idle bandwidth, time delay, shake, packet loss etc..In general, the link between data center is two-way, and two-way state can It can be different.In addition, link information is not unalterable, with time change, can by controller or other Real-time, event triggering property or periodic acquisition is carried out using to link information, collected link information is stored To formation link information database in database.Table 1 is to believe into line link between Tu1Zhong data centers DC_A, DC_B, DC_C The link information table for ceasing acquisition and obtaining.By taking the first row as an example, L_AB indicates total bandwidths of the DC_A to DC_B transmission datas when, surplus The information such as remaining bandwidth, time delay, shake, packet loss, as shown in Table 1, L_AB and L_BA are different.
Table 1
Link identification Total bandwidth Remaining bandwidth Time delay Shake Packet loss
L_AB 10G 5G 50ms 20ms 20ppm
L_AC 10G 3G 50ms 10ms 10ppm
L_BA 10G 2G 50ms 40ms 10ppm
L_BC 1G 200M 50ms 20ms 200ppm
L_CA 1G 200M 50ms 10ms 50ppm
L_CB 1G 200M 50ms 40ms 10ppm
The flow information of calculate node refers to the flow status information communicated between calculate node, including certain time period The information such as total communication flows, average bandwidth, time burst length, burst bandwidth.In general, the flow letter between calculate node Breath is not unalterable, changes with the variation of time, can be by controller or other application to flow information Real-time, event triggering property or periodic acquisition is carried out, collected flow information is stored in database and is formed Flow information database.Using calculate node as virtual machine instance, Fig. 1 is please referred to, it is assumed that two have been disposed in data center DC_A Two virtual machines VM_3 and VM_4 are disposed in virtual machine VM_1 and VM_2, DC_B, VM_1, VM_2, VM_3, VM_4 are mutual There are communication, certain time period carries out flow information acquisition between VM_1, VM_2, VM_3, VM_4 and can be obtained each calculating section The flow status information communicated between point.Flow information table of the table 2 between VM_1, VM_2, VM_3, VM_4.Wherein, with T12 For, T12 can indicate total communication flows, average bandwidth, time burst length, burst bandwidth between VM_1 and VM_2 etc. Information.
Table 2
It should be noted that explained in detail the present invention for handling node and controller and independently disposing in Fig. 1 It states.However, the present invention is not limited thereto, system will be located in actual application scenarios by what software or hardware mode were realized Reason node and controller can also be deployed on the same server in operation system, alternatively, being deployed in different clothes respectively It is engaged on device.
Fig. 2 is the flow chart of calculate node dispositions method embodiment one of the present invention.The present embodiment is applicable to operation system In disposed calculate node and be unsatisfactory for business demand and need to dispose to have disposed in the scene or operation system of new calculate node Calculate node needs to adjust the scene of deployed position, and the present embodiment carries out the embodiment of the present invention using controller as executive agent detailed It is thin to illustrate.Specifically, the present embodiment includes the following steps：
101, the deployment recommendation request message that processing node is sent is received, deployment recommendation request message, which carries, waits for that deployment calculates The description information of node.
Service operation initial stage, the calculate node that content supplier (Content Provider, hereinafter referred to as CP) needs Quantity is few, and still, the quantity of the continuous development reckoned with business, the calculate node that CP needs is also corresponding more and more, It needs to dispose new calculate node in certain data centers in operation system；Alternatively, due to handle same business, Multiple calculate nodes of deployment are deployed in different data centers, according to demand, need the calculating section disposed to those It clicks through line position and sets adjustment.At this point, processing node sends deployment recommendation request message to controller, controller receives the carrying and waits for portion Affix one's name to the description information of calculate node.
Wherein, description information include the identification information of calculate node to be disposed, calculate node to be disposed quantity, wait disposing Tenant identification etc. belonging to calculate node, wherein tenant identification is, for example, the title of tenant or identification corresponding with title Number (IDentity, hereinafter referred to as ID), as Tencent, Baidu, 3456123,7890123 (3456123,7890123 be Tencent and Baidu's corresponding Digital ID under controller) etc., using tenant identification as wait dispose calculate node description information when, table Show all calculate nodes under the tenant；The mark of one or more calculate nodes, such as { VM_a } indicate a virtual machine a, { VM_a, VM_b, VM_c } indicates one group of virtual machine a, b, c；Alternatively, it is also possible to which certain one kind can be identified for that calculate node to be disposed Identification information of the characteristic information as calculate node to be disposed, as Internet protocol (Internet Protocol, it is simple below Claim IP) address information, network segment address information or medium access control (media access control, detect MAC below) Location information, IP address range information etc., wherein 192.168.3.X/24 indicates under Subnet Identification 192.168.3.X, interconnection All calculate nodes that 24 are 192.168.1 before the address fidonetFido (Internet Protocol, hereinafter referred to as IP).
102, according to link information and/or flow information, deployment scheme is determined, wherein link information includes processing node Link information between each data center managed, and/or, each data center for being managed of processing node be not belonging to handle Link information between each data center that node is managed；Flow information is calculate node to be disposed and calculate node to be disposed Flow information between relevant calculate node.Wherein, the relevant calculate node of calculate node to be disposed is to be calculated with deployment is waited for Node has the calculate node of communication requirement.
After receiving the deployment recommendation request message that processing node is sent, controller is managed each according to processing node Each data center and be not belonging to each number that processing node is managed that link information, processing node between data center are managed According at least one of the flow information between the link information between center and the relevant calculate node of calculate node to be disposed Information determines the deployment scheme of calculate node to be disposed.For example, for new calculate node, it may be determined that go out preferable deployment Position；For the calculate node disposed in operation system, weight is provided according to flow information, the link information etc. between calculate node The suggestion of new deployed position.
Further, the deployment request information of calculate node to be disposed can also be carried in deployment recommendation request message, controlled Device processed determines the deployment scheme for meeting deployment request information according to link information and/or flow information.If specifically, business system Any calculate node is not disposed in each data center in system, can only consider to handle in each data that node is administered at this time Link information in the heart determines the deployment scheme for meeting deployment request information according to link information.Deployment request information can be Treat between deployment calculate node or calculate node to be disposed with disposed deployment relative position between calculate node, communicated matter The constraints that amount requires or across the data center communication total flow of calculate node to be disposed requires.
103, the deployment recommendation response message for including deployment scheme is sent to processing node.
When controller is to be sent to processing node comprising deployment scheme after disposing calculate node and determining deployment scheme Deployment recommendation response message so that processing node disposes calculate node to be disposed according to deployment scheme.For example, processing node can be with According to deployment recommendation response message, calculate node to be disposed is deployed in the number that link tape remaining is roomy, time delay is small, packet loss is low According to center；Alternatively, there are the calculate nodes of a large amount of network communications to be deployed in the same data center by two.Deployment scheme can be with It is that controller is wanted according to the flow information between link information or calculate node to be disposed and relative calculate node, deployment At least one of the information such as information information is asked to determine；Disappear alternatively, processing node can not also adopt deployment recommendation request Breath.
Calculate node dispositions method provided in an embodiment of the present invention, controller are managed according to processing node in operation system Each data center between link information or each data center for being managed of processing node and be not belonging to handle node and managed Each data center between link information and the flow information etc. between the relevant each calculate node of calculate node to be disposed, it is right In new calculate node, it may be determined that go out the suggestion of preferable deployed position；For the calculate node disposed in operation system, The suggestion for redeploying position can be provided, by the way that there will be the calculate nodes of a large amount of network communications to be deployed in the same data In the heart, the communication flows between data center is changed into the flow in data center, it is logical between calculate node to improve Believe quality, reduce the communication flows between data center.
Fig. 3 is the flow chart of calculate node dispositions method embodiment two of the present invention.The present embodiment is to execute master with controller The embodiment of the present invention is described in detail in body.Specifically, the present embodiment includes the following steps：
201, the deployment information request message that processing node is sent is received, deployment information request message, which carries, waits for that deployment calculates The description information of node.
202, link information and/or flow information are obtained, wherein link information includes each data for handling node and being managed Link information between center, and/or, each data center for being managed of processing node and being not belonging to handle node managed it is each Link information between data center, flow information are calculate node to be disposed and the relevant calculate node of calculate node to be disposed Between flow information, wherein with the relevant calculate node of calculate node to be disposed be to communicate with calculate node dispose need The calculate node asked.
Controller receives the deployment information request message that processing node is sent, and obtains each data that processing node is managed Link information between center, and/or, each data center and be not belonging to what processing node was managed that pending node is managed Link information between each data center, flow information are that calculate node to be disposed is saved with the relevant calculating of calculate node to be disposed Flow information between point, for example, the flow list for needing to carry out the calculate node of position adjustment disposed can be provided Or traffic matrix.Controller can immediately obtain relevant after receiving the deployment information request message that processing node is sent Link information or flow information；It is waited for alternatively, being obtained after traffic requests message can also be carried only in deployment recommendation request message The flow information etc. of calculate node is disposed, is meet demand, the specific format of specific deployment information request message can be designed, System that the present invention is not limited thereto.
203, the deployment information response message comprising link information and/or flow information is sent to processing node.
It is in place of the present embodiment and the difference of above-described embodiment one, in the present embodiment, controller is not such as embodiment one It determines to be then forwarded to processing node after deployment scheme in step 102, but by the link information got and/or flow information Response message be transmitted directly to processing node, by processing node voluntarily according to those information determining portion management side cases.Specifically, place Reason node can be according at least one of link information, flow information or deployment request information for determining according to demand letter Breath determines deployment scheme.
Calculate node dispositions method provided in an embodiment of the present invention, controller obtain link information, calculate section with deployment is waited for Flow information etc. between the relevant each calculate node of point is simultaneously sent to processing node so that processing node in operation system The calculate node of deployment can provide the suggestion for redeploying position, by the way that there will be the calculate node portions of a large amount of network communications Communication flows between data center is changed into the flow in data center, to improve by administration in the same data center The communication flows between communication quality, reduction data center between calculate node.
Fig. 4 is the flow chart of calculate node dispositions method embodiment three of the present invention.The present embodiment is applicable to operation system In the calculate node disposed be unsatisfactory for business demand and need to dispose in the scene or operation system of new calculate node portion The calculate node of administration needs to adjust the scene of deployed position, and the present embodiment is executive agent to the embodiment of the present invention to handle node It is described in detail.Specifically, the present embodiment includes the following steps：
301, deployment recommendation request message is sent to controller, deployment recommendation request message carries calculate node to be disposed Description information, so that controller determines deployment scheme according to link information and/or flow information, wherein link information packet Include the link information between each data center for being managed of processing node, and/or, each data center that processing node is managed with The link information being not belonging between each data center that processing node is managed；Flow information is calculate node to be disposed and waits for portion Affix one's name to the flow information between the relevant calculate node of calculate node, wherein be with the relevant calculate node of calculate node to be disposed There is the calculate node of communication requirement with calculate node to be disposed.
The description as described in link information and flow information can be found in embodiment illustrated in fig. 2 in the present embodiment, no longer superfluous herein It states.
302, the deployment recommendation response message comprising deployment scheme that controller is sent is received.
After controller receives the deployment recommendation request message that processing node is sent and determines deployment scheme, saved to processing Point sends the deployment recommendation response message for including deployment scheme, and processing node receives the deployment recommendation response message.
After the deployment recommendation response message for receiving controller transmission, processing node can be selected according to the deployment recommendation The deployment scheme that response message includes completes to wait for that deployment calculates by being interacted to the server for the data center administered The deployment of node, for example, the establishment of new calculate node, startup, the deletion for the calculate node disposed, freezing and migrating； Alternatively, processing node can not also adopt the deployment scheme of deployment recommendation response message return.
Calculate node dispositions method provided in an embodiment of the present invention, for new calculate node, processing node is according to control The suggestion for the preferable deployed position that device provides carries out the establishment etc. of calculate node, for the calculating section disposed in operation system Point, the suggestion for redeploying position provided according to controller carry out position adjustment, for example, freezing some or certain having disposed Calculate node, these calculate nodes are re-created in other data centers, so that there will be the meters of a large amount of network communications Operator node is deployed in the same data center, and the communication flows between data center is changed into the flow in data center, Or multiple data centers that link-quality is relatively good, to improve the communication quality between calculate node, reduce data center Between communication flows.
Fig. 5 is the flow chart of calculate node dispositions method example IV of the present invention.The present embodiment is to execute to handle node The embodiment of the present invention is described in detail in main body.Specifically, the present embodiment includes the following steps：
401, deployment information request message is sent to controller, deployment information request message carries calculate node to be disposed Description information, so that controller obtains link information and/or flow information, wherein link information includes that processing node is managed Link information between each data center of reason, and/or, each data center and be not belonging to processing node that processing node is managed Link information between each data center managed, flow information are that calculate node to be disposed is related to calculate node to be disposed Calculate node between flow information, wherein with the relevant calculate node of calculate node to be disposed be and wait for deployment calculate section Point has the calculate node of communication requirement.
The description as described in link information and flow information can be found in embodiment illustrated in fig. 2 in the present embodiment, no longer superfluous herein It states.
402, the deployment information response message comprising link information and/or flow information that controller is sent is received.
Controller can immediately obtain relevant link after receiving the deployment information request message that processing node is sent Information or flow information；It waits disposing alternatively, obtaining after traffic requests message can also be carried only in deployment recommendation request message The flow information etc. of calculate node is meet demand, can design the specific format of specific deployment information request message, this hair It is bright to be not limited thereto system.It handles node and receives the deployment information comprising link information and/or flow information that controller is sent Response message.
403, deployment scheme is determined according to link information and/or flow information.
Processing node voluntarily determines deployment scheme according to link information and/or flow information.Specifically, processing node can be with Portion is determined according at least one of link information, flow information or the deployment request information determined according to demand information Management side case.
Calculate node dispositions method provided in an embodiment of the present invention, processing node are calculated according in operation system with deployment is waited for The link information between relevant each data center of data center where node and the relevant each meter of calculate node to be disposed Flow information etc. between operator node determines the suggestion for redeploying position for the calculate node disposed in operation system, Position adjustment is carried out, for example, freezing some or certain calculate nodes disposed, it is several to re-create this in other data centers A calculate node, to which there will be the calculate nodes of a large amount of network communications to be deployed in the same data center, by data center Between communication flows be changed into flow in data center or the relatively good multiple data centers of link-quality, to carry The communication flows between communication quality, reduction data center between high calculate node.
Above-described embodiment one, embodiment two, embodiment three and example IV are respectively to execute from controller and processing node The angle of main body is expounded the present invention, wherein embodiment one is determined to wait disposing meter with embodiment three by controller The deployment scheme of operator node, in embodiment two and example IV, processing node receive obtained according to controller and send with After the relevant link information of deployment calculate node and flow information, the deployment of calculate node to be disposed is determined by processing node Scheme.In the following, by different embodiments come the present invention will be described in detail.
The second service system structure schematic diagram that Fig. 6 is applicable in for calculate node dispositions method of the present invention.Fig. 6 is please referred to, In the present embodiment, be in operation system, handle for node administration 3 data centers DC_A, DC_B, DC_C to the present invention into What row elaborated, calculate node can be disposed or deployed on the server of each data center (not to be shown in figure Go out).In addition, being not shown in figure processing node and controller.
In calculate node dispositions method embodiment five of the present invention, the deployment recommendation that processing node is sent to controller is asked Message carries the description information of calculate node to be disposed, and needs to create two new calculating sections for example, this describes instruction in message Point.Table 3 acquires the first link information obtained between Tu6Zhong data centers DC_A, DC_B, DC_C into downlink message Table.
Table 3
Link identification Total bandwidth Remaining bandwidth Time delay Shake Packet loss
L_AB 10G 5G 50ms 20ms 20ppm
L_AC 10G 5G 50ms 10ms 10ppm
L_BC 10G 2G 50ms 40ms 10ppm
L_BA 10G 5G 50ms 20ms 20ppm
L_CA 10G 5G 50ms 10ms 50ppm
L_CB 10G 200M 50ms 40ms 10ppm
Controller receives the instruction that processing node is sent and needs to create the deployment recommendation request of two new calculate nodes After message, based on the link information stored in link information database at this time, for example, link information table shown in table 3, traversal The table, in the maximum data of summation for determining the outbound remaining bandwidth and inbound link remaining bandwidth of each data center The heart, using the data center determined as the data center for disposing new calculate node.Specifically, table 3 is please referred to, DC_A's Outbound remaining bandwidth is 10G (L_AB 5G, L_AC 5G), and inbound link remaining bandwidth is 10G (L_BA 5G, L_CA For 5G), total surplus bandwidth is 20G；The outbound remaining bandwidth of DC_B is 7G (L_BA 5G, L_BC 2G), inbound link Remaining bandwidth is 5.2G (L_AB 5G, L_CB 200M), and total surplus bandwidth is 12.2G；The outbound remaining bandwidth of DC_C For 5.2G (L_CA 5G, L_CB 200M), inbound link remaining bandwidth is 7G (L_AC 5G, L_BC 2G, total surplus bandwidth For 7G).Thus, the total surplus bandwidth of DC_A is maximum, total surplus band is roomy, in general, can more meet calculate node across The communicating requirement of data center improves communication quality.Therefore, this 2 calculate nodes are disposed in controller suggestion in DC_A, i.e., really The deployment scheme of 2 calculate nodes can be disposed by being scheduled on DC_A.
In calculate node dispositions method embodiment six of the present invention, the deployment recommendation that processing node is sent to controller is asked Message also carries the deployment request information of calculate node to be disposed to limit other than carrying the description information of calculate node to be disposed The relative position information between calculate node to be disposed is made, needs to create two new calculating for example, this describes instruction in message Node, deployment request information indicate that two new calculate nodes are in different data centers.Table 4 is Tu6Zhong data centers The second link information table for acquiring and obtaining into downlink message between DC_A, DC_B, DC_C.
Table 4
Link identification Total bandwidth Remaining bandwidth Time delay Shake Packet loss
L_AB 10G 5G 50ms 20ms 20ppm
L_AC 10G 5G 50ms 10ms 10ppm
L_BC 10G 2G 50ms 40ms 10ppm
L_BA 10G 5G 50ms 20ms 20ppm
L_CA 1G 200M 50ms 10ms 50ppm
L_CB 1G 200M 50ms 40ms 10ppm
Controller receives the instruction that processing node is sent and needs to create two new calculate nodes and this two calculating sections After point needs the deployment recommendation request message in different data centers, based on the chain stored in link information database at this time Road information, for example, link information table shown in table 4, controller traverse the table, determine that the link between data center is remaining The maximum Liang Ge data centers of bandwidth, using the Liang Ge data centers as the data center for disposing new calculate node.Specifically, Table 4 is please referred to, the link remaining bandwidth between DC_A and DC_B is 10G (L_AB 5G, L_BA 5G)；Between DC_A and DC_C Link remaining bandwidth be 5.2G (L_AC 5G, L_CA 200M)；Link remaining bandwidth between DC_B and DC_C is 2.2G (L_BC 2G, L_CB 200M).Thus, the link remaining bandwidth between DC_A and DC_B is maximum.Therefore, controller is built View respectively disposes a calculate node in DC_A and DC_B, that is, determines respectively to dispose calculate node in DC_A and DC_B Deployment scheme.At this point, due to each calculate node deployed position there are two, can determine according to demand some count The prioritized deployment data center of operator node, it is proposed that dispose the meter in the relatively high deployment data center of the priority level determined Remaining calculate node is deployed in other data centers by operator node；If alternatively, without priority requirement, it is recommended that by two Calculate node is random to be deployed in DC_A and DC_B.
Based on above-described embodiment six, optionally, deployment request information can also be to the communication matter between new calculate node Amount requires to be indicated, for example, carrying the description information and deployment request of calculate node to be disposed in deployment recommendation request message Information describes instruction in message and needs to create two new calculate nodes, and deployment request information indicates this two new calculating sections Point is in different data centers, and the shake communicated between two new calculate nodes is less than 30ms and is more than 15ms.Table 5 is The third link information table for acquiring and obtaining into downlink message between Tu6Zhong data centers DC_A, DC_B, DC_C.
Table 5
Link identification Total bandwidth Remaining bandwidth Time delay Shake Packet loss
L_AB 10G 5G 50ms 20ms 20ppm
L_AC 10G 3G 50ms 10ms 10ppm
L_BC 10G 2G 50ms 40ms 10ppm
L_BA 10G 5G 50ms 20ms 20ppm
L_CA 1G 200M 50ms 10ms 50ppm
L_CB 1G 200M 50ms 40ms 10ppm
Controller receives the instruction that processing node is sent and needs to create two new calculate nodes, two calculate nodes It needs to be less than the portion that 30ms is more than 15ms in the shake communicated between different data centers and two new calculate nodes After request message is suggested by administration, based on the link information stored in link information database at this time, for example, link shown in table 5 is believed Cease table, controller traverses the table, determines the Liang Ge data centers of shake meet demand communicated between data center, by this two A data center is as the data center for disposing new calculate node.Specifically, please referring to table 5, communicated between DC_A and DC_B Be dithered as 20ms；What is communicated between DC_A and DC_C is dithered as 10ms；What is communicated between DC_B and DC_C is dithered as 40ms.By This can be obtained, and the shake communicated between only DC_A and DC_B meets the condition in deployment request information.Therefore, controller suggestion exists A calculate node is respectively disposed in DC_A and DC_B, that is, determines the deployment for respectively disposing a calculate node in DC_A and DC_B Scheme.
Based on above-described embodiment six, optionally, the calculating that deployment request information can also be between new calculate node, new Node and the QoS requirement disposed between calculate node are indicated.It is waited for for example, being carried in deployment recommendation request message The description information and deployment request information for disposing calculate node describe instruction in message and need to create two new calculate nodes VM1 and VM2, the instruction of deployment request information VM1 and VM2 are in different data centers, VM1 with disposed the portions calculate node VMX Affix one's name to different data centers and VM1 with disposed calculate node VMX between the shake that communicates be less than 15ms.Fig. 6 is please referred to, VMX is deployed in DC_A in advance.Table 6 between Tu6Zhong data centers DC_A, DC_B, DC_C into downlink message acquire and The 4th link information table obtained.
Table 6
Link identification Total bandwidth Remaining bandwidth Time delay Shake Packet loss
L_AB 10G 5G 50ms 20ms 20ppm
L_AC 10G 3G 50ms 10ms 20ppm
L_BC 10G 2G 50ms 40ms 10ppm
L_BA 10G 5G 50ms 20ms 20ppm
L_CA 1G 200M 50ms 10ms 50ppm
L_CB 1G 200M 50ms 40ms 10ppm
Controller receives the instruction that processing node is sent and needs to create two new calculate node VM1 and VM2, and deployment is wanted Information instruction VM1 is asked to be in different data centers from VM2, VM1 is deployed in from calculate node VMX has been disposed in different data The heart and VM1 with disposed deployment recommendation request message of the shake less than 15ms that is communicated between calculate node VMX after, based at this time The link information stored in link information database, for example, link information table shown in table 6, controller traverse the table, determine The Liang Ge data centers for going out the shake meet demand communicated between data center, using the Liang Ge data centers meter new as deployment The data center of operator node.Specifically, please referring to table 6, what is communicated between DC_A and DC_B is dithered as 20ms；DC_A and DC_C it Between communicate be dithered as 10ms；What is communicated between DC_B and DC_C is dithered as 40ms.Thus, between only DC_A and DC_B The shake of communication meets the condition in deployment request information, and VMX is deployed in DC_A in advance.Therefore, controller suggestion is in DC_ Calculate node VM1 is disposed in C, and calculate node VM2 is disposed in DC_A or DC_B, that is, determines to dispose calculate node in DC_C VM1 disposes the deployment scheme of calculate node VM2 in DC_A or DC_B.
Based on above-described embodiment six, optionally, calculate node to be disposed be operation system in dispose calculate node, such as belong to In at least one calculate node of a certain subnet, the deployment recommendation request message that processing node is sent to controller, which carries, to be waited disposing The description information of calculate node, quantity of the description information in addition to including calculate node to be disposed can also include Subnet Identification；Portion It can also be to indicate that the calculate node of the subnet is minimum across the communication flows of data center to affix one's name to require information.For example, calculate node For virtual machine, Subnet Identification 192.168.10.X/24 shares 4 calculate nodes：VM_1, VM_2, VM_3 and VM_4 are (in figure It is not shown) belong to the subnet, current deployed position is that the first two is in DC_C, latter two is in DC_A.Table 7A is in Fig. 6 The the first communication stream scale for carrying out communication flows acquisition between VM_1, VM_2, VM_3 and VM_4 and obtaining.In order to briefly describe See, is only here the communication flows listed between each calculate node, is calculated with MB, in other feasible embodiments, The communication stream scale can also include the information such as average bandwidth, time burst length, burst bandwidth between each calculate node.
Table 7A
Controller receive processing node send instruction Subnet Identification be 192.168.10.X/24 under each virtual machine across After the deployment recommendation request message of the communication total flow of data center 2 virtual machines of minimum and each data center deployment, control Device divides the deployed position of each virtual machine under the subnet, and the requirement of division is to be divided into 4 virtual machines to belong to The data center DC_C and DC_A of subnet 192.168.10.X/24, two virtual machines of each data center's division, then possible stroke Divide result as shown in table 7B, table 7B is that division result illustrates table in table 7A.
Table 7B
For dividing a line of serial number 1 in table 7B, which indicates that VM_1 and VM_2 are divided into DC_C by controller, VM_1 and VM_2 is located at the same data center, the communication flows of VM_1 to VM_2 and the communication flows phase of VM_2 to VM_1 at this time It when the inner stream flow in DC_C, therefore can be not counted between data center, i.e. communication flows between DC_C and DC_A, such as In table 7A shown in the dash area of the upper left corner；Similarly, VM_3 and VM_4 is divided into DC_A, and VM_3 and VM_4 is located at same at this time In a data, the communication flows of VM_3 to VM_4 and the communication flows of VM_4 to VM_3 are equivalent to the inner stream flow of DC_A, therefore It can be not counted between data center, i.e. communication flows between DC_C and DC_A, as shown in lower right corner dash area in table 7A. Therefore, if not considered according to table 7A with dividing division methods deployment VM_1, VM_2, VM_3 and VM_4 shown in serial number 1 The upper left corner and the flow inside the data center of lower right corner part, then the communication flows summation between DC_A and DC_C is 240MB.
Similarly, it can show that the communication flows between the DC_A and DC_C under the other modes for dividing serial number 2~6 divide is total With specifically, as shown in table 7B.
Controller traversal list 7B has found to be in same number shown in serial number 6 as divided serial number 3 or dividing as VM_2 and VM_3 When according to center, the communication flows summation between DC_A and DC_C is minimum, is 160MB；It was found that when VM_2 and VM_3 is in not With data center when, communication flows summation between DC_A and DC_C is 240MB, bigger.Therefore, controller suggestion Deployment is re-started to VM_1, VM_2, VM_3 and VM_4, VM_2 and VM_3 are deployed in DC_A, VM_1 and VM_4 is disposed In DC_C, alternatively, VM_2 and VM_3 are deployed in DC_C, VM_1 and VM_4 are deployed in DC_A.
Based on above-described embodiment six, optionally, calculate node to be disposed is to have disposed calculate node in operation system, is such as belonged to In at least one calculate node of a certain tenant, the deployment recommendation request message that processing node is sent to controller, which carries, to be waited disposing The description information of calculate node, quantity of the description information in addition to including calculate node to be disposed can also include tenant identification；Portion It can also be to indicate that the calculate node of the subnet is minimum across the communication flows of data center and each data center to affix one's name to require information Dispose 2 calculate nodes.For example, calculate node is virtual machine, tenant identification CP_1234 shares 6 calculate nodes：VM_1、 VM_2, VM_3, VM_4, VM_5 and VM_6 (not shown) belong to the tenant, and current deployed position is that VM_1, VM_2 are in DC_C, VM_3, VM_4 are in DC_A, VM_5 and VM_6 and are in DC_B.Table 8A is VM_1, VM_2, VM_3, VM_4, VM_5 in Fig. 6 The the second communication stream scale for carrying out communication flows acquisition between VM_6 and obtaining.Only it is row here for the sake of briefly describing Go out the communication flows between each calculate node, has been calculated with MB, in other feasible embodiments, the communication stream scale May include the information such as average bandwidth, time burst length, burst bandwidth between each calculate node.
Table 8A
The instruction tenant identification that controller receives processing node transmission is each virtual machine under CP_1234 across data center After communication flows is minimum and the deployment recommendation request message of each 2 virtual machines of data center deployment, controller is to the tenant Under the deployed position of each virtual machine divided, the requirement of division is to be divided into 6 virtual machines to belong to tenant CP_1234 Data center DC_C, DC_A and DC_B in, each data center divides two virtual machines, then possible division result such as table 8B Shown, table 8B is that division result illustrates table in table 8A.
Table 8B
For dividing a line of serial number 3 in table 8B, which indicates that VM_1 and VM_4 are divided into DC_C by controller, VM_1 and VM_4 is located at the same data center, the communication flows of VM_1 to VM_4 and the communication flows phase of VM_4 to VM_1 at this time When the inner stream flow in DC_C, therefore the communication flows that can be not counted between data center A；Similarly, VM_2 and VM_3 quilts It is divided into data DC_A, the communication flows of VM_2 and VM_3 and the communication flows of VM_3 to VM_2 are equivalent in DC_A at this time Inner stream flow, therefore, the communication flows that can be not counted between data center；Similarly, VM_5 and VM_6 is divided into data DC_ In B, the communication flows of VM_5 and VM_6 and the communication flows of VM_6 to VM_5 are equivalent to the inner stream flow in DC_B at this time, because This, the communication flows that can be not counted between data center, specifically, leading between being not counted in data center under this kind of dividing mode In the 8A of the part of letter flow amount such as table shown in dash area.Therefore, if with divide need for division methods shown in 3 dispose VM_1, VM_2, VM_3, VM_4, VM_5 and VM_6 do not consider dash area shown in table 8A then according to table 8A, then DC_A, DC_B with Communication flows summation between DC_C is 480MB.
Similarly, it can show that the communication flows between other divide each data center under division methods shown in serial number is total With specifically, as shown in table 8B.It should be noted that under different calculate node numbers and different divisions require, divide Mode is also different, and is not enumerated herein.In actual scene, can select according to demand calculate node number and Dividing mode.
Controller traversal list 8B has found to be in same when VM_1 and VM_4 is in the same data center, VM_2 and VM_3 When a data center, VM_5 and VM_6 are in the same data center, the communication flows summation between each data center is 480MB, it is smaller.Therefore, controller suggestion re-starts deployment to VM_1, VM_2, VM_3, VM_4, VM_5 and VM_6, will VM_1 and VM_4 is deployed in the same data center, VM_2 and VM_3 is deployed in the same data center, by VM_5 and VM_ 6 are deployed in the same data center.Specifically, such as dash area in table 8B, that is, divide division methods shown in serial number 3 or 10.
In calculate node dispositions method embodiment seven of the present invention, for having disposed calculate node in operation system, handle The deployment recommendation request message that node is sent to controller is also carried and is waited for other than carrying the description information of calculate node to be disposed The traffic requests information of calculate node is disposed, the flow status for having disposed calculate node is known in request.For example, having with calculate node Body is virtual machine instance, please refers to Fig. 6, and 9 virtual machines are shared in operation system, and VM_1, VM_2, VM_3 are deployed in DC_A, VM_4, VM_5, VM_6 are deployed in DC_B, and VM_7, VM_8, VM_9 are deployed in DC_C, are not shown in figure specific virtual machine. Table 9A in Fig. 4 between VM_1, VM_2, VM_3, VM_4, VM_5, VM_6, VM_7, VM_8, VM_9 carry out communication flows acquisition and The third communication flowmeter obtained.Only it is the communication stream listed between each calculate node here for the sake of briefly describing Amount, is calculated with MB, and in other feasible embodiments, which can also include the average band between each calculate node The information such as width, time burst length, burst bandwidth.
Table 9A
If what controller received the carrying of deployment recommendation request message is the traffic requests message of VM1, traversal list 9A is carried Other virtual machines that there is communication between VM1 are taken out, then extract corresponding flow, the communication flows letter extracted Breath such as table 9B shows that table 9B is the result signal table for carrying out communication flows information extraction in table 9A to VM.
Table 9B
In the case of the row where VM_2, when indicating that VM_1 is communicated with VM_2, the communication flows of VM_1 to VM_2 goes out Flow is 100MB；The communication flows of VM_2 to VM_1, i.e. inbound traffics are 90MB.Controller is in the communication flows letter for extracting VM1 Breath, processing node is sent to by the communication flows information, so that processing node can voluntarily determine VM_ according to the flow information 1, the deployment scheme of VM_2, VM_3, VM_4, VM_5, VM_6, VM_7, VM_8, VM_9.Specifically, controller can be by VM1's Mark and table 9B shown in content carrying processing node is sent in deployment recommendation response message, each communication can also be calculated It is then forwarded to processing node after the communication flows summation of object, for example, by taking communication object is VM_2 as an example, outflow 100MB, If flow is 90MB, then communication flows summation is 190MB, the deployment recommendation response message that controller is sent to calculate node management It is 190MB to carry the mark of VM_1 and the communication flows summation of VM_1 and VM_2.In actual scene, controller can be with The flow information of the communication flows information or section communication object of VM1 and all communication objects is sent to processing section by selection Point.
It is processing node and controller point-to-point communication, processing node is directly to controller in above-mentioned each embodiment Deployment recommendation request message is sent, and receives the deployment recommendation response message of controller transmission；Controller directly receives processing section The deployment recommendation request message that point is sent directly sends deployment recommendation response message after determining deployment scheme to processing node. However, in actual scene, processing node can also be communicated indirectly with controller.For example, in calculate node portion of the present invention In management side formula embodiment eight, controller can be determined by acting on behalf of the deployment recommendation request message for receiving processing node and sending out Deployment recommendation response message is sent to processing node by agency after deployment scheme；Similarly, processing node can also pass through agency Deployment recommendation request message is sent to controller；The deployment recommendation response message that controller is sent is received by agency.Due to generation Reason may be implemented the conversion of message format and the conversion of content and handle the decoupling between node and controller, and limitation is small, uses Range is more extensive.Therefore, in actual application scenarios, processing node can be selected directly logical with controller according to demand Letter, alternatively, by agency to indirect communication between selection processing node and controller.
In addition, in above-mentioned each embodiment, link information is that link information includes each data for handling node and being managed Link information between center, however, link information may be to handle each data center for being managed of node and be not belonging to locate Link information between each data center that reason node is managed.For example, Fig. 1 is please referred to, if desired by new calculate node portion Administration is in DC_B or DC_C, and new calculate node to be disposed is communicated with the calculate node presence in DC_D, then needs to consider DC_ Link information between B and DC_C, DC_B and DC_D, DC_C and DC_D.
Fig. 7 is the structural schematic diagram of controller embodiment one of the present invention.Controller 100 provided in this embodiment, it can be achieved that Each step of the method applied to controller in any embodiment of the present invention, details are not described herein again for specific implementation process.Tool Body, controller 100 provided in this embodiment may include：
Receiving module 11, the deployment recommendation request message sent for receiving processing node, deployment recommendation request message are taken The description information of band calculate node to be disposed；
Determining module 12, for according to link information and/or flow information, determining deployment scheme, wherein link information packet Include the link information between each data center for being managed of processing node, and/or, each data center that processing node is managed with The link information being not belonging between each data center that processing node is managed；Flow information is calculate node to be disposed and waits for portion Affix one's name to the flow information between the relevant calculate node of calculate node, wherein be with the relevant calculate node of calculate node to be disposed There is the calculate node of communication requirement with calculate node to be disposed；
Sending module 13, for sending the deployment recommendation response message for including deployment scheme to processing node.
Controller provided in an embodiment of the present invention, according to each data that can be disposed with calculate node to be disposed in operation system What the data center that link information or calculate node to be disposed between center can be disposed can not dispose with calculate node to be disposed The flow information etc. between the relevant each calculate node of link information and calculate node to be disposed between each data center, for New calculate node, it may be determined that go out the suggestion of preferable deployed position；It, can for the calculate node disposed in operation system To provide the suggestion for redeploying position, by the way that there will be the calculate nodes of a large amount of network communications to be deployed in the same data center In, the communication flows between data center is changed into the flow in data center, to improve the communication between calculate node Communication flows between quality, reduction data center.
Further, the portion of calculate node to be disposed also is carried in the deployment recommendation request message that receiving module 11 receives Affix one's name to require information；
Determining module 12 is additionally operable to, according to link information and/or flow information, determine the deployment for meeting deployment request information Scheme.
Further, deployment request information includes：Relative position information between calculate node to be disposed waits for that deployment calculates Node and disposed QoS requirement information between the relative position information between calculate node, calculate node to be disposed, Calculate node to be disposed and QoS requirement information between calculate node, calculate node to be disposed have been disposed across data center Communicate a kind of information or combinations thereof in total flow require information.
Further, description information include the identification information of calculate node to be disposed, calculate node to be disposed quantity letter Cease or calculate node to be disposed belonging to tenant identification information.
Further, receiving module 11 is additionally operable to by acting on behalf of the deployment recommendation request message for receiving processing node and sending；
Sending module 13 is additionally operable to send the deployment recommendation response message for including deployment scheme by acting on behalf of to processing node.
Further, calculate node to be disposed includes newly-increased calculate node or has disposed calculate node.
Fig. 8 is the structural schematic diagram of controller embodiment two of the present invention.Controller 200 provided in this embodiment, it can be achieved that The each step for the method applied to controller that any embodiment of the present invention provides, details are not described herein again for specific implementation process. Specifically, controller 200 provided in this embodiment may include：
Receiving module 21, the deployment information request message sent for receiving processing node, deployment information request message are taken The description information of band calculate node to be disposed；
Acquisition module 22, for obtaining link information and/or flow information, wherein link information includes processing node institute Link information between each data center of management, and/or, each data center and be not belonging to processing section that processing node is managed Link information between each data center that point is managed, flow information are calculate node to be disposed and calculate node phase to be disposed Flow information between the calculate node of pass, wherein the relevant calculate node of calculate node to be disposed is to calculate section with deployment is waited for Point has the calculate node of communication requirement；
Sending module 23, for sending the deployment information response comprising link information and/or flow information to processing node Message.
Controller provided in an embodiment of the present invention, by obtaining link information and the relevant each meter of calculate node to be disposed Flow information between operator node etc. is simultaneously sent to processing node so that processing node is for the calculating section disposed in operation system Point can provide the suggestion for redeploying position, by the way that there will be the calculate nodes of a large amount of network communications to be deployed in same number According in center, the communication flows between data center is changed into the flow in data center, to improve between calculate node Communication quality, reduce data center between communication flows.
Further, description information include the identification information of calculate node to be disposed, calculate node to be disposed quantity letter Cease or calculate node to be disposed belonging to tenant identification information..
Further, receiving module 21 is additionally operable to by acting on behalf of the deployment information request message for receiving processing node and sending, Deployment information request message carries the description information of calculate node to be disposed；
Sending module 23 is additionally operable to send the deployment comprising link information and/or flow information to processing node by acting on behalf of Info response message, so that processing node determines deployment scheme according to link information and/or flow information.
Further, calculate node to be disposed includes newly-increased calculate node or has disposed calculate node.
Fig. 9 is the structural schematic diagram of present invention processing node embodiment one.Processing node 300 provided in this embodiment, can The each step for realizing the method for being applied to processing node that any embodiment of the present invention provides, implements process herein no longer It repeats.Specifically, processing node 300 provided in this embodiment may include：
Sending module 31, for sending deployment recommendation request message to controller, the carrying of deployment recommendation request message waits for portion The description information for affixing one's name to calculate node, so that controller determines deployment scheme according to link information and/or flow information, wherein Link information includes the link information handled between each data center that node is managed, and/or, processing node is managed each Data center and the link information being not belonging between each data center that processing node is managed；Flow information is to wait for that deployment calculates Flow information between node and the relevant calculate node of calculate node to be disposed, wherein relevant with calculate node to be disposed Calculate node is the calculate node for having with calculate node to be disposed communication requirement；
Receiving module 32, the deployment recommendation response message comprising deployment scheme for receiving controller transmission.
Processing node provided in an embodiment of the present invention, for new calculate node, processing node is provided according to controller The suggestion of preferable deployed position, carries out the establishment etc. of calculate node, for the calculate node disposed in operation system, according to control What device processed provided redeploys the suggestion of position, position adjustment is carried out, for example, freezing some or certain calculating sections disposed Point re-creates these calculate nodes in other data centers, to which there will be the calculate node portions of a large amount of network communications Administration is in the same data center, the flow or link that the communication flows between data center are changed into data center The preferably multiple data centers of mass ratio, it is logical between the communication quality between calculate node, reduction data center to improve Letter flow amount.
Further, sending module 31 is additionally operable to send the deployment request information for carrying calculate node to be disposed to controller Deployment recommendation request message so that controller, according to link information and/or flow information, determination meets deployment request information Deployment scheme.
Further, deployment request information includes：Relative position information between calculate node to be disposed waits for that deployment calculates Node and disposed QoS requirement information between the relative position information between calculate node, calculate node to be disposed, Calculate node to be disposed and QoS requirement information between calculate node, calculate node to be disposed have been disposed across data center Communicate a kind of information or combinations thereof in total flow require information.
Further, description information include the identification information of calculate node to be disposed, calculate node to be disposed quantity letter Cease or calculate node to be disposed belonging to tenant identification information..
Further, sending module 31 is additionally operable to send deployment recommendation request message to controller by agency；
Receiving module 32 is additionally operable to receive the deployment recommendation comprising deployment scheme that controller is sent by agency and respond to disappear Breath.
Further, calculate node to be disposed includes newly-increased calculate node or has disposed calculate node.
Figure 10 is the structural schematic diagram of present invention processing node embodiment two.Processing node 400 provided in this embodiment, can The each step for realizing the method for being applied to processing node that any embodiment of the present invention provides, implements process herein no longer It repeats.Specifically, processing node 400 provided in this embodiment may include：
Sending module 41, for sending deployment information request message to controller, the carrying of deployment information request message waits for portion The description information of calculate node is affixed one's name to, so that controller obtains link information and/or flow information, wherein link information includes Link information between each data center for being managed of processing node, and/or, each data center that processing node is managed with not Belong to the link information between each data center for being managed of processing node, flow information is calculate node to be disposed and wait disposing Flow information between the relevant calculate node of calculate node, wherein with the relevant calculate node of calculate node to be disposed be with Calculate node to be disposed has the calculate node of communication requirement；
Receiving module 42, the deployment information comprising link information and/or flow information for receiving controller transmission are rung Answer message；
Determining module 43, for determining deployment scheme according to link information and/or flow information.
Processing node provided in an embodiment of the present invention, according in the data in operation system and where calculate node to be disposed Flow letter between link information between the relevant each data center of the heart and the relevant each calculate node of calculate node to be disposed Breath etc. determines the suggestion for redeploying position for the calculate node disposed in operation system, carries out position adjustment, example Such as, freeze some or certain calculate nodes disposed, these calculate nodes are re-created in other data centers, to There will be the calculate nodes of a large amount of network communications to be deployed in the same data center, and the communication flows between data center is turned Become flow in data center or the relatively good multiple data centers of link-quality, to improve between calculate node Communication flows between communication quality, reduction data center.
Further, description information include the identification information of calculate node to be disposed, calculate node to be disposed quantity letter Cease or calculate node to be disposed belonging to tenant identification information
Further, sending module 41 is additionally operable to send deployment information request message to controller by agency；
Receiving module 42 is additionally operable to receive the portion comprising link information and/or flow information that controller is sent by agency Affix one's name to info response message.
Figure 11 is the structural schematic diagram of controller embodiment three of the present invention.As shown in figure 11, control provided in this embodiment Device 500 includes processor 51 and memory 52.Controller 500 can also include transmitter 53, receiver 54.It transmitter 53 and connects Receiving device 54 can be connected with processor 51.Wherein, the storage of memory 52 executes instruction, when controller 500 is run, processor 51 It is communicated between memory 52, processor 51 calls executing instruction in memory 52, implements for executing method shown in Fig. 2 Example, implementing principle and technical effect are similar, and details are not described herein again.
Figure 12 is the structural schematic diagram of controller example IV of the present invention.As shown in figure 12, control provided in this embodiment Device 600 includes processor 61 and memory 62.Controller 600 can also include transmitter 63, receiver 64.It transmitter 63 and connects Receiving device 64 can be connected with processor 61.Wherein, the storage of memory 62 executes instruction, when controller 600 is run, processor 61 It is communicated between memory 62, processor 61 calls executing instruction in memory 62, implements for executing method shown in Fig. 3 Example, implementing principle and technical effect are similar, and details are not described herein again.
Figure 13 is the structural schematic diagram of present invention processing node embodiment three.As shown in figure 13, place provided in this embodiment It includes processor 71 and memory 72 to manage node 700.It can also includes transmitter 73, receiver 74 to handle node 700.Transmitter 73 can be connected with receiver 74 with processor 71.Wherein, the storage of memory 72 executes instruction, when handling the operation of node 700, It is communicated between processor 71 and memory 72, processor 71 calls executing instruction in memory 72, for executing side shown in Fig. 4 Method embodiment, implementing principle and technical effect are similar, and details are not described herein again.
Figure 14 is the structural schematic diagram of present invention processing node embodiment four.As shown in figure 14, place provided in this embodiment It includes processor 81 and memory 82 to manage node 800.It can also includes transmitter 83, receiver 84 to handle node 800.Transmitter 83 can be connected with receiver 84 with processor 81.Wherein, the storage of memory 82 executes instruction, when handling the operation of node 800, It is communicated between processor 81 and memory 82, processor 81 calls executing instruction in memory 82, for executing side shown in Fig. 5 Method embodiment, implementing principle and technical effect are similar, and details are not described herein again.
Embodiment based on the above method and device, the present invention also provides a kind of operation system, it may include such as Fig. 7 or figure Controller shown in 11 and handle node as shown in Fig. 9 or Figure 13；Or, it may include the controller as shown in Fig. 8 or Figure 12 And node is handled as shown in Figure 10 or Figure 14, concrete operating principle please refers to above method embodiment, and details are not described herein again.
In several embodiments provided herein, it should be understood that disclosed system, device and method can be with It realizes by another way.For example, the apparatus embodiments described above are merely exemplary, for example, the unit It divides, only a kind of division of logic function, formula that in actual implementation, there may be another division manner, such as multiple units or component It can be combined or can be integrated into another system, or some features can be ignored or not executed.Another point, it is shown or The mutual coupling, direct-coupling or communication connection discussed can be the indirect coupling by some interfaces, device or unit It closes or communicates to connect, can be electrical, machinery or other forms.
The unit illustrated as separating component may or may not be physically separated, aobvious as unit The component shown may or may not be physical unit, you can be located at a place, or may be distributed over multiple In network element.Some or all of unit therein can be selected according to the actual needs to realize the mesh of this embodiment scheme 's.
Finally it should be noted that：The above embodiments are only used to illustrate the technical solution of the present invention., rather than its limitations；To the greatest extent Present invention has been described in detail with reference to the aforementioned embodiments for pipe, it will be understood by those of ordinary skill in the art that：Its according to So can with technical scheme described in the above embodiments is modified, either to which part or all technical features into Row equivalent replacement；And these modifications or replacements, various embodiments of the present invention technology that it does not separate the essence of the corresponding technical solution The range of scheme.

Claims (42)

1. a kind of calculate node dispositions method, which is characterized in that including：
The deployment recommendation request message that processing node is sent is received, the deployment recommendation request message carries calculate node to be disposed Description information；
According to link information and flow information, deployment scheme is determined, wherein the link information includes that the processing node is managed Link-state information between each data center of reason, and/or, each data center for being managed of processing node be not belonging to Link-state information between each data center that the processing node is managed；The flow information waits for that deployment calculates described in being Flow information between node and the relevant calculate node of calculate node to be disposed, wherein wait for that deployment calculates section with described The relevant calculate node of point is the calculate node for having with the calculate node to be disposed communication requirement；
The deployment recommendation response message for including the deployment scheme is sent to the processing node.
2. according to the method described in claim 1, it is characterized in that, waiting for portion described in also being carried in the deployment recommendation request message Affix one's name to the deployment request information of calculate node；
It is described that deployment scheme is determined according to link information and flow information, including：
According to the link information and flow information, the deployment scheme for meeting the deployment request information is determined.
3. according to the method described in claim 2, it is characterized in that, the deployment request information includes：
Relative position information, the calculate node to be disposed between the calculate node to be disposed and disposed calculate node it Between relative position information, the QoS requirement information between the calculate node to be disposed, the calculate node to be disposed With QoS requirement information, the calculate node to be disposed disposed between calculate node total flow is communicated across data center A kind of information in require information or combinations thereof.
4. according to claims 1 to 3 any one of them method, which is characterized in that the description information includes waiting for that deployment calculates The identification information of node, the quantity information of the calculate node to be disposed or the tenant identification belonging to the calculate node to be disposed Information.
5. according to claims 1 to 3 any one of them method, which is characterized in that the deployment for receiving processing node and sending It is recommended that request message, including：
By acting on behalf of the deployment recommendation request message for receiving the processing node and sending；
It is described to send the deployment recommendation response message for including the deployment scheme to the processing node, so that the processing saves It puts according to calculate node to be disposed described in deployment scheme deployment, including：
The deployment recommendation response message for including the deployment scheme is sent by described act on behalf of to the processing node, so that described Node is handled according to calculate node to be disposed described in deployment scheme deployment.
6. according to claims 1 to 3 any one of them method, which is characterized in that the calculate node to be disposed includes newly-increased Calculate node has disposed calculate node.
7. a kind of calculate node dispositions method, which is characterized in that including：
The deployment information request message that processing node is sent is received, the deployment information request message carries calculate node to be disposed Description information；
Obtain link information and flow information, wherein the link information includes in each data that the processing node is managed Link-state information between the heart, and/or, each data center and be not belonging to the processing node that the processing node is managed Link-state information between each data center managed, the flow information are that the calculate node to be disposed is waited for described Dispose the flow information between the relevant calculate node of calculate node, wherein the relevant calculating section of calculate node to be disposed Point is the calculate node for having with calculate node to be disposed communication requirement；
The deployment information response message for including the link information and flow information is sent to the processing node, so that the place Reason node determines deployment scheme according to the link information and flow information.
8. the method according to the description of claim 7 is characterized in that the description information includes the mark of calculate node to be disposed Information, the quantity information of the calculate node to be disposed or the tenant identification information belonging to the calculate node to be disposed.
9. method according to claim 7 or 8, which is characterized in that the deployment information for receiving processing node transmission is asked Message, the deployment information request message is asked to carry the description information of calculate node to be disposed, including：
By acting on behalf of the deployment information request message for receiving the processing node and sending, the deployment information request message carrying waits for Dispose the description information of calculate node；
It is described to send the deployment information response message for including the link information and flow information to the processing node, so that institute It states processing node and deployment scheme is determined according to the link information and flow information, including：
Disappeared by described act on behalf of to deployment information response of the processing node transmission comprising the link information and flow information Breath, so that the processing node determines deployment scheme according to the link information and flow information.
10. method according to claim 7 or 8, which is characterized in that the calculate node to be disposed includes that newly-increased calculate is saved Point has disposed calculate node.
11. a kind of calculate node dispositions method, which is characterized in that including：
Deployment recommendation request message is sent to controller, the deployment recommendation request message carries the description of calculate node to be disposed Information, so that the controller determines deployment scheme according to link information and flow information, wherein the link information packet The link-state information between each data center that processing node is managed is included, and/or, each number that the processing node is managed According to center and the link-state information being not belonging between each data center that the processing node is managed；The flow information is Flow information between the calculate node to be disposed and the relevant calculate node of calculate node to be disposed, wherein with institute It is the calculate node for having with the calculate node to be disposed communication requirement to state the relevant calculate node of calculate node to be disposed；
Receive the deployment recommendation response message for including the deployment scheme that the controller is sent.
12. according to the method for claim 11, which is characterized in that the deployment recommendation request sent to the controller The deployment request information of calculate node to be disposed described in also being carried in message, so that the controller is according to the link information And flow information, determine the deployment scheme for meeting the deployment request information.
13. according to the method for claim 12, which is characterized in that the deployment request information includes：
Relative position information, the calculate node to be disposed between the calculate node to be disposed and disposed calculate node it Between relative position information, the QoS requirement information between the calculate node to be disposed, the calculate node to be disposed With QoS requirement information, the calculate node to be disposed disposed between calculate node total flow is communicated across data center A kind of information in require information or combinations thereof.
14. according to claim 11~13 any one of them method, which is characterized in that the description information includes waiting for deployment meter The identification information of operator node, the quantity information of the calculate node to be disposed or tenant's mark belonging to the calculate node to be disposed Know information.
15. according to claim 11~13 any one of them method, which is characterized in that described to send deployment recommendation to controller Request message, including：
By agency deployment recommendation request message is sent to controller；
The deployment recommendation response message for including the deployment scheme for receiving the controller and sending, including：
The deployment recommendation response message for including the deployment scheme that the controller is sent is received by the agency.
16. according to claim 11~13 any one of them method, which is characterized in that the calculate node to be disposed includes new Increase calculate node or disposes calculate node.
17. a kind of calculate node dispositions method, which is characterized in that including：
Deployment information request message is sent to controller, the deployment information request message carries the description of calculate node to be disposed Information, so that the controller obtains link information and flow information, wherein the link information includes that processing node is managed Link-state information between each data center of reason, and/or, each data center for being managed of processing node be not belonging to Link-state information between each data center that the processing node is managed, the flow information wait for that deployment calculates to be described Flow information between node and the relevant calculate node of calculate node to be disposed, wherein wait for that deployment calculates section with described The relevant calculate node of point is the calculate node for having with the calculate node to be disposed communication requirement；
Receive the deployment information response message for including the link information and flow information that the controller is sent；
Deployment scheme is determined according to the link information and flow information.
18. according to the method for claim 17, which is characterized in that the description information includes the mark of calculate node to be disposed Know information, the quantity information of the calculate node to be disposed or the tenant identification information belonging to the calculate node to be disposed.
19. the method according to claim 17 or 18, which is characterized in that described to disappear to controller transmission deployment information request Breath, including：
By agency deployment information request message is sent to controller；
The response message for including the link information and flow information for receiving the controller and sending, including；
The response message for including the link information and flow information that the controller is sent is received by the agency.
20. the method according to claim 17 or 18, which is characterized in that the calculate node to be disposed includes newly-increased calculates Node has disposed calculate node.
21. a kind of controller, which is characterized in that including：
Receiving module, the deployment recommendation request message sent for receiving processing node, the deployment recommendation request message carry The description information of calculate node to be disposed；
Determining module, for according to link information and flow information, determining deployment scheme, wherein the link information includes institute The link-state information between each data center that processing node is managed is stated, and/or, each number that the processing node is managed According to center and the link-state information being not belonging between each data center that the processing node is managed；The flow information is Flow information between the calculate node to be disposed and the relevant calculate node of calculate node to be disposed, wherein with institute It is the calculate node for having with the calculate node to be disposed communication requirement to state the relevant calculate node of calculate node to be disposed；
Sending module, for sending the deployment recommendation response message for including the deployment scheme to the processing node.
22. controller according to claim 21, which is characterized in that the deployment recommendation request that the receiving module receives The deployment request information of calculate node to be disposed described in also being carried in message；
The determining module is additionally operable to, according to the link information and flow information, determine the portion for meeting the deployment request information Management side case.
23. controller according to claim 22, which is characterized in that the deployment request information includes：
Relative position information, the calculate node to be disposed between the calculate node to be disposed and disposed calculate node it Between relative position information, the QoS requirement information between the calculate node to be disposed, the calculate node to be disposed With QoS requirement information, the calculate node to be disposed disposed between calculate node total flow is communicated across data center A kind of information in require information or combinations thereof.
24. according to claim 21~23 any one of them controller, which is characterized in that the description information includes waiting disposing The identification information of calculate node, the quantity information of the calculate node to be disposed or the tenant belonging to the calculate node to be disposed Identification information.
25. according to claim 21~23 any one of them controller, which is characterized in that the receiving module is additionally operable to pass through Agency receives the deployment recommendation request message that the processing node is sent；
The sending module is additionally operable to build by described act on behalf of to the deployment of the processing node transmission comprising the deployment scheme Discuss response message.
26. according to claim 21~23 any one of them controller, which is characterized in that the calculate node to be disposed includes Newly-increased calculate node has disposed calculate node.
27. a kind of controller, which is characterized in that including：
Receiving module, the deployment information request message sent for receiving processing node, the deployment information request message carry The description information of calculate node to be disposed；
Acquisition module, for obtaining link information and flow information, wherein the link information includes that the processing node is managed Link-state information between each data center of reason, and/or, each data center for being managed of processing node be not belonging to Link-state information between each data center that the processing node is managed, the flow information wait for that deployment calculates to be described Flow information between node and the relevant calculate node of calculate node to be disposed, wherein the calculate node to be disposed Relevant calculate node is the calculate node for having with calculate node to be disposed communication requirement；
Sending module disappears for sending the deployment information response comprising the link information and flow information to the processing node Breath.
28. controller according to claim 27, which is characterized in that the description information includes calculate node to be disposed Identification information, the quantity information of the calculate node to be disposed or the tenant identification information belonging to the calculate node to be disposed.
29. the controller according to claim 27 or 28, which is characterized in that the receiving module is additionally operable to connect by agency The deployment information request message that the processing node is sent is received, the deployment information request message carries calculate node to be disposed Description information；
The sending module is additionally operable to send comprising the link information and flow letter by described act on behalf of to the processing node The deployment information response message of breath, so that the processing node determines deployment scheme according to the link information and flow information.
30. the controller according to claim 27 or 28, which is characterized in that the calculate node to be disposed includes newly-increased meter Operator node has disposed calculate node.
31. a kind of processing node, which is characterized in that including：
Sending module, for sending deployment recommendation request message to controller, the deployment recommendation request message, which carries, to be waited disposing The description information of calculate node, so that the controller determines deployment scheme according to link information and flow information, wherein The link information includes the link-state information handled between each data center that node is managed, and/or, the processing section Each data center and the link-state information being not belonging between each data center that the processing node is managed that point is managed； The flow information is the flow between calculate node to be disposed and the relevant calculate node of calculate node to be disposed Information, wherein with the relevant calculate node of calculate node to be disposed be with the calculate node to be disposed to have communication requirement Calculate node；
Receiving module, the deployment recommendation response message for including the deployment scheme sent for receiving the controller.
32. processing node according to claim 31, which is characterized in that the sending module is additionally operable to the controller The deployment recommendation request message for sending the deployment request information of calculate node to be disposed described in carrying, so that the control Device determines the deployment scheme for meeting the deployment request information according to link information and flow information.
33. processing node according to claim 32, which is characterized in that the deployment request information includes：
Relative position information, the calculate node to be disposed between the calculate node to be disposed and disposed calculate node it Between relative position information, the QoS requirement information between the calculate node to be disposed, the calculate node to be disposed With QoS requirement information, the calculate node to be disposed disposed between calculate node total flow is communicated across data center A kind of information in require information or combinations thereof.
34. handling node according to claim 31~33 any one of them, which is characterized in that the description information includes waiting for portion Affix one's name to the identification information of calculate node, the quantity information of the calculate node to be disposed or the rent belonging to the calculate node to be disposed Family identification information.
35. handling node according to claim 31~33 any one of them, which is characterized in that the sending module is additionally operable to lead to It crosses agency and sends deployment recommendation request message to controller；
The receiving module is additionally operable to receive the deployment for including the deployment scheme that the controller is sent by the agency It is recommended that response message.
36. handling node according to claim 31~33 any one of them, which is characterized in that the calculate node packet to be disposed It includes newly-increased calculate node or has disposed calculate node.
37. a kind of processing node, which is characterized in that including：
Sending module, for sending deployment information request message to controller, the deployment information request message, which carries, to be waited disposing The description information of calculate node, so that the controller obtains link information and flow information, wherein the link information packet The link-state information between each data center that processing node is managed is included, and/or, each number that the processing node is managed According to center and the link-state information being not belonging between each data center that the processing node is managed, the flow information is Flow information between the calculate node to be disposed and the relevant calculate node of calculate node to be disposed, wherein with institute It is the calculate node for having with the calculate node to be disposed communication requirement to state the relevant calculate node of calculate node to be disposed；
Receiving module, the deployment information response comprising the link information and flow information sent for receiving the controller Message；
Determining module, for determining deployment scheme according to the link information and flow information.
38. according to the processing node described in claim 37, which is characterized in that the description information includes calculate node to be disposed Identification information, the calculate node to be disposed quantity information or belonging to the calculate node to be disposed tenant identification letter Breath.
39. the processing node according to claim 37 or 38, which is characterized in that the sending module is additionally operable to pass through agency Deployment information request message is sent to controller；
The receiving module is additionally operable to through agency's reception controller transmission include the link information and flow The deployment information response message of information.
40. the processing node according to claim 37 or 38, which is characterized in that the calculate node to be disposed includes newly-increased Calculate node has disposed calculate node.
41. a kind of operation system, which is characterized in that including such as claim 21~26 any one of them controller and as weighed Profit requires 31~36 any one of them to handle node.
42. a kind of operation system, which is characterized in that including claim 27~30 any one of them controller and such as right It is required that 37~40 any one of them handle node.
CN201310174927.2A 2013-05-13 2013-05-13 Calculate node dispositions method, processing node, controller and system Active CN104158675B (en)

Priority Applications (2)

Application Number Priority Date Filing Date Title
CN201310174927.2A CN104158675B (en) 2013-05-13 2013-05-13 Calculate node dispositions method, processing node, controller and system
PCT/CN2014/076828 WO2014183574A1 (en) 2013-05-13 2014-05-06 Computer node deployment method, processing node, controller, and system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201310174927.2A CN104158675B (en) 2013-05-13 2013-05-13 Calculate node dispositions method, processing node, controller and system

Publications (2)

Publication Number Publication Date
CN104158675A CN104158675A (en) 2014-11-19
CN104158675B true CN104158675B (en) 2018-09-11

Family

ID=51884088

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201310174927.2A Active CN104158675B (en) 2013-05-13 2013-05-13 Calculate node dispositions method, processing node, controller and system

Country Status (2)

Country Link
CN (1) CN104158675B (en)
WO (1) WO2014183574A1 (en)

Families Citing this family (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105677454B (en) * 2014-11-20 2019-08-27 华为技术有限公司 The integration method of computing resource, device and system
CN105656662B (en) * 2014-12-08 2019-02-12 华为技术有限公司 A kind of Fault Locating Method and device
CN105791073B (en) * 2014-12-23 2019-07-19 华为技术有限公司 The method and apparatus of service deployment in a kind of virtualization network
US10333851B2 (en) 2016-10-18 2019-06-25 Huawei Technologies Co., Ltd. Systems and methods for customizing layer-2 protocol
CN110275756B (en) * 2018-03-13 2023-04-18 华为技术有限公司 Method and device for deploying virtualized network element
CN109889370B (en) * 2019-01-10 2021-12-21 中国移动通信集团海南有限公司 Network equipment position determining method and device and computer readable storage medium
CN113344152A (en) * 2021-04-30 2021-09-03 华中农业大学 System and method for intelligently detecting and uploading full-chain production information of dairy products

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102082692A (en) * 2011-01-24 2011-06-01 华为技术有限公司 Method and equipment for migrating virtual machines based on network data flow direction, and cluster system
CN102112981A (en) * 2008-07-31 2011-06-29 思科技术公司 Distribution of virtual machines in a communication network
CN103023799A (en) * 2011-09-27 2013-04-03 日电（中国）有限公司 Central control unit and virtual machine migration method used for virtual machine migration

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8145760B2 (en) * 2006-07-24 2012-03-27 Northwestern University Methods and systems for automatic inference and adaptation of virtualized computing environments

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102112981A (en) * 2008-07-31 2011-06-29 思科技术公司 Distribution of virtual machines in a communication network
CN102082692A (en) * 2011-01-24 2011-06-01 华为技术有限公司 Method and equipment for migrating virtual machines based on network data flow direction, and cluster system
CN103023799A (en) * 2011-09-27 2013-04-03 日电（中国）有限公司 Central control unit and virtual machine migration method used for virtual machine migration

Also Published As

Publication number Publication date
WO2014183574A1 (en) 2014-11-20
CN104158675A (en) 2014-11-19

Similar Documents

Publication Publication Date Title
CN104158675B (en) Calculate node dispositions method, processing node, controller and system
CN109618002B (en) Micro-service gateway optimization method, device and storage medium
CN105207873B (en) A kind of message processing method and device
CN105554065B (en) Handle method, converting unit and the applying unit of message
US9231869B2 (en) Ensuring predictable and quantifiable networking performance
CN108370341B (en) Resource allocation method, virtual network function manager and network element management system
CN108989372A (en) Method, registration center and the equipment of service discovery
EP3703316B1 (en) Frame aggregation in a wireless network
US20220086044A1 (en) Analyzing and configuring workload distribution in slice-based networks to optimize network performance
AU2020450956B2 (en) Tenant-driven dynamic resource allocation for virtual network functions
CN108702316A (en) A kind of resource allocation methods and device of VNF
CN111432247B (en) Traffic scheduling method, traffic scheduling device, server and storage medium
CN108337275A (en) Task distribution method, device and equipment for Distributor
EP3574674B1 (en) A network monitoring entity, method, system and computer program for a communication network implementing network slices
CN106302230B (en) A kind of data transmission method and device
CN109818867A (en) A kind of configuration method and device
CN106375486B (en) A kind of system of cloud platform software definition DHCP service
CN109639796A (en) A kind of implementation of load balancing, device, equipment and readable storage medium storing program for executing
CN102158406B (en) Intelligent routing method for computer network links
CN108259221A (en) Flow control methods, device, system, storage medium and processor
CN107809495A (en) Address management method and device
CN101860454A (en) Network performance data processing method and device thereof
CN108040118A (en) Visit capacity statistical method, device and storage medium
CN108039986A (en) Monitor the method, apparatus and storage medium of web application performances
CN108833570A (en) A kind of cluster-based storage and balanced transmission system based on cloud storage

Legal Events

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