CN106254115A - A kind of cross-domain mapping method of virtual data center - Google Patents

A kind of cross-domain mapping method of virtual data center Download PDF

Info

Publication number
CN106254115A
CN106254115A CN201610640554.7A CN201610640554A CN106254115A CN 106254115 A CN106254115 A CN 106254115A CN 201610640554 A CN201610640554 A CN 201610640554A CN 106254115 A CN106254115 A CN 106254115A
Authority
CN
China
Prior art keywords
virtual
virtual machine
link
resource
overbar
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN201610640554.7A
Other languages
Chinese (zh)
Other versions
CN106254115B (en
Inventor
汤露阳
廖丹
谭智
赵东成
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.)
Guangdong Institute Of Electronic And Information Engineering University Of Electronic Science And Technology Of China
Original Assignee
Guangdong Platinum Information Technology Co Ltd
Institute of Electronic and Information Engineering of Dongguan UESTC
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 Guangdong Platinum Information Technology Co Ltd, Institute of Electronic and Information Engineering of Dongguan UESTC filed Critical Guangdong Platinum Information Technology Co Ltd
Priority to CN201610640554.7A priority Critical patent/CN106254115B/en
Publication of CN106254115A publication Critical patent/CN106254115A/en
Application granted granted Critical
Publication of CN106254115B publication Critical patent/CN106254115B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/08Configuration management of networks or network elements
    • H04L41/0893Assignment of logical groups to network elements
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/08Configuration management of networks or network elements
    • H04L41/0896Bandwidth or capacity management, i.e. automatically increasing or decreasing capacities
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00Network arrangements or protocols for supporting network services or applications
    • H04L67/01Protocols
    • H04L67/10Protocols in which an application is distributed across nodes in the network
    • H04L67/1001Protocols in which an application is distributed across nodes in the network for accessing one among a plurality of replicated servers
    • H04L67/1004Server selection for load balancing
    • H04L67/1008Server selection for load balancing based on parameters of servers, e.g. available memory or workload
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00Network arrangements or protocols for supporting network services or applications
    • H04L67/01Protocols
    • H04L67/10Protocols in which an application is distributed across nodes in the network
    • H04L67/1001Protocols in which an application is distributed across nodes in the network for accessing one among a plurality of replicated servers
    • H04L67/1031Controlling of the operation of servers by a load balancer, e.g. adding or removing servers that serve requests

Abstract

The invention discloses a kind of cross-domain mapping method of virtual data center, virtual data center is mapped to distributed multiple data centers of bottom, and minimize total mapping cost and energy efficient, in order to adapt to the motility that VDC maps, one simple and common solution is to start anew to remap VDC, and VDC maps dynamically is to be mapped in bottom physical infrastructure by the virtual resource newly increased, physical resource is distributed for the dummy node newly increased or virtual link, the virtual resource reduced is deleted from physical infrastructure simultaneously, discharge its physical resource taken.

Description

A kind of cross-domain mapping method of virtual data center
Technical field
The invention belongs to network communication technology field, more specifically, relate to the cross-domain of a kind of virtual data center and reflect Shooting method.
Background technology
In recent years, cloud computing provided a kind of effective model and allows many tenants to share cloud resource.Traditionally, as Amazon There is provided resource Deng cloud computing supplier with the form of virtual machine, and have ignored the network service operated on these virtual machines and want Ask.This results in a lot in the problem about network performance, safety and manageability.Limit, recently to solve this Some research approaches propose this concept of virtual data center, advocate the form with virtual data center rather than the shape of virtual machine Formula provides resource.The most both it had been provided that the server with guarantee was also provided that the guaranteed Internet resources of tool.Only compare tradition Thering is provided the mode of virtual machine, hire out resource with the form of virtual data center VDC (Virtual Data Center), service carries The isolation of application program better performance and service quality (Quality of Servive, QoS) can be supplied to for business.At cloud meter In calculation, infrastructure provider has physical infrastructure (such as, typical data center), is responsible for the thing of each data center Reason resource is divided into virtual resource (such as, virtual machine), and then these virtual resources are leased to service provider services provides Business.Service provider by by application deployment in the infrastructure that infrastructure provider provides, right rear line carries For corresponding service.Service provider can be abstract for VDC for the request of physical resource.Data center's physical resource virtual Changing makes the management of the physical resource to data center become flexible, efficient, and is conducive to improving server and network money The utilization rate in source.
Hire out resource with the form of VDC and bring a problem, it is simply that the most optimally by server and data central network Network resource distributes to multiple virtual data center, and maximizes total revenue, minimizes the energy resource consumption total amount of data center.For One effective resource management scheme of design of VDC is a challenging problem.Present invention research to be that VDC is cross-domain reflect Penetrate.It should be noted that one NP-hard problem of VDC cross-domain mapping.But, make problem get worse with, basis sets Executing provider s can be that service provider s provides more motility, by allowing service provider s to carry according to according to them Demand for service extends up or reduces downwards the scale of VDC.Such as, service provider may wish to increase the capacity of VDC Adapt to the quick growth of demand for services.The mode increasing VDC capacity can be the resource requirement by increasing VM or VL, it is possible to To be VM or VL new by increase.On the other hand, service provider can also be at the scale joint of service minimizing of idle period VDC The about hiring cost of physical resource.The scale reducing the VDC mapped can cut operating costs.Particularly, VDC integrates and calculates Method, it is intended to maximize the resource service condition opening machine, allows to close idle machines simultaneously.
Therefore, the present invention have studied the VDC cross-domain mapping problem that resource requirement dynamically changes, and uses virtual machine (vm) migration dynamic Adjustresources configures, to meet the probability of VDC resource requirement fluctuation.
Existing method is primarily directed to the dynamic mapping of virtual network, and the method for the dynamic mapping of virtual network has Dynamic virtual network mapping algorithm (Dynamic Virtual Network Mapping Algorithm, DVNMA), DVNMA Algorithm proposes a kind of framework and algorithm reconfigured and map VN, it is achieved that minimize total reconfiguring and becoming of mapping VN This.
Cost between territory when DVNMA algorithm does not accounts for cross-domain mapping, it only considered and reconfigures inside data center With map the situation of VN, and it does not accounts for energy-conservation problem.Additionally, DVNMA algorithm does not also account for there is multiple VN simultaneously Request change and system total reconfigure cost.Due to DVNMA algorithm research is that VN mapping maps similar with VDC, but Also have some different.Such as, the VM belonging to identical VN in VNE can not map on the same server, and VDC belongs in mapping The VM of identical VDC then can be mapped in together.
Summary of the invention
It is an object of the invention to overcome the deficiencies in the prior art, it is provided that the cross-domain mapping side of a kind of virtual data center Method, can also update the request Added User while virtual data center carries out cross-domain mapping, uses virtual machine (vm) migration to move State adjustresources configures, to meet the probability of VDC resource requirement fluctuation.
For achieving the above object, the cross-domain mapping method of a kind of virtual data center of the present invention, it is characterised in that bag Include following steps:
(1), use the initial cross-domain algorithm of virtual data center that virtual data center carries out cross-domain mapping:
(1.1) in virtual data center, randomly choose a virtual machine carry out cross-domain mapping;
(1.2), in the unmapped virtual machine of virtual data center, select to be connected and communication bandwidth with maps virtual machine Maximum virtual machine maps;
(1.3), step (1.2) is repeated, until having mapped all of virtual machine;
(1.4), estimation virtual machine maps cost Cos taddn
Cost a d d n Σ v ∈ V 2 - V 1 Σ v ‾ ∈ V ‾ r e q ( v ) * h v v ‾ * cos t ( v ‾ ) - - - ( 1 )
Wherein, V2-V1Represent the dynamic change of virtual machine;Represent the set of server;V represents a virtual machine;Table Show a station server;Req (v) represents the resource requirement of virtual machine v;Representing current virtual machine mapping relations, whether v is mapped in ServerOn,Or 0, wherein, 1 represents that v is mapped in serverOn, 0 represents that v is not mapped inOn;Represent ServerResource unit costs;
(1.5), the mapping cost Cos t of estimation original virtual data center internal virtual linkaddl
Cost a d d l = Σ e ∈ ( E 2 - E 1 ) Σ e ‾ ∈ M E ( e ) ϵ c a p ( e ‾ ) + δ * b ( e ) * cos ( e ‾ ) * s 1 e - - - ( 2 )
Wherein, δ be one close to 0 minimum normal number;E2-E1Represent the dynamic change of virtual link;ε is used to control The parameter of load balancing importance;Represent a physical link;E represents a virtual link;B (e) represents virtual link e's Resource request;Represent physical linkResource unit costs;cos tbRepresentation unit bandwidth expense;Represent thing Reason linkResource capacity;MEE () represents the physical pathway that virtual link e maps;Or 0,1 represents that virtual link e maps In territory, 0 represents that virtual link e is not mapped in territory;
(1.6), between estimation territory, virtual link maps cost Cos taddb
Cost a d d b = Σ e ∈ ( E 2 - E 1 ) s 2 e * b ( e ) * cos t b - - - ( 3 )
Wherein,Or 0,1 represents that virtual link e is mapped between territory, and 0 represents that virtual link e is not mapped between territory;
(2) user received when, virtual data center being carried out cross-domain mapping increases request newly and integrates;
(2.1), the request of resource of deleting is received;
(2.1.1), reduce the resource of designated virtual machine in request, the part physical resource that release reduces, change this void The resource requirement of plan machine;
(2.1.2), the resource of virtual link corresponding to designated virtual machine is reduced, the part band width in physical that release reduces, Change the bandwidth demand of this virtual link;
(2.2), the request of assembly of deleting is received;
(2.2.1), designated virtual machine in removal request, discharge the physical resource on the physical host that this virtual machine takies, And from virtual data center, delete this virtual machine;
(2.2.2), delete the virtual link that designated virtual machine is corresponding, discharge in the physical pathway that this virtual link takies Bandwidth, deletes this virtual link from virtual data center;
(2.3) request of increase node resource, is received;
(2.3.1), the virtual machine increasing resource will be needed to put in queue, by the virtual machine in queue according to increasing resource Amount descending;
If (2.3.2) server needing the virtual machine increasing resource to be mapped has enough surplus resources capacity, then Directly increase resources of virtual machine;Otherwise, by by the virtual machine (vm) migration on server to other servers, completing to increase void The request of plan machine resource;
(2.4) request of increase link circuit resource, is received;
For needing to increase the virtual link of bandwidth resources, if in the case of physical link bandwidth resources abundance, directly Connect the bandwidth increasing virtual link;If the resource of physical link is inadequate, then the shortest k routing algorithm is used to be torn open by virtual link Divide and be mapped on a plurality of bottom path;
(2.5) request of increase virtual machine, is received;
(2.5.1), will need increase new virtual machine put in queue, by the virtual machine in queue according to map Amount of bandwidth descending between virtual machine;
(2.5.2), according to descending order, the virtual machine newly increased is mapped;
(2.6), the request that packet adjusts is received;
(2.6.1), delete link in the territory needing to delete, and carry out the node needed in the packet migrated migrating between territory With remap;
(2.6.2), the bandwidth of link in territory is reduced, and to needing the node in the packet migrated to carry out migrating between territory and weighing Map;
(2.6.3), the bandwidth of link in territory is increased, and to needing the node in the packet migrated to carry out migrating between territory and weighing Map;
(2.6.4), estimation virtual machine remaps cost;
C o s t r e m n = Σ v ∈ V Σ v ‾ ∈ V ‾ req t ( v ) v * h v v ‾ * cos t ( v ‾ ) - Σ v ∈ V Σ v ‾ ∈ V ‾ req t - 1 ( v ) * h v v ‾ * cos t ( v ‾ ) - - - ( 4 )
Wherein, reqtV () represents the resource requirement of t virtual machine v, reqt-1V () represents the money of t-1 moment virtual machine v Source demand.
(2.6.5), estimated data's central interior virtual link remaps cost;
Cost r e m l = Σ e ∈ ( E ) Σ e ‾ ∈ M E ( e ) ϵ c a p ( e ‾ ) + δ b t ( e ) * cos t ( e ‾ ) * s 1 e - Σ e ∈ ( E ) Σ e ‾ ∈ M E ( e ) ϵ c a p ( e ‾ ) + δ b t - 1 ( e ) * cos t ( e ‾ ) * s 1 e - - - ( 5 )
Wherein, btE () represents the resource request of t virtual link e, bt-1E () represents the money of t-1 moment virtual link e Source is asked, and E represents the set of virtual link.
(2.6.6), between estimation territory, virtual link remaps cost;
cost r e m b = Σ e ∈ ( E 2 - E 1 ) s 2 e * b t ( e ) * cos t b - Σ e ∈ ( E 2 - E 1 ) s 2 e * b t - 1 ( e ) * cos t b - - - ( 6 )
(2.7) request that virtual data center is integrated, is received;
(2.7.1), the Servers-all that travels through in each frame, by the utilization rate of all of server according to from small to large Sequence, and is not above virtual machine (vm) migration that is minimum for utilization rate and that be less than on the server of lower limit to utilization rate maximum successively On the server of limit, until utilization rate all being closed less than the server of lower limit, or closing server cannot be continued;
(2.7.2), judge whether to exist between each server a plurality of virtual link, if it does not exist, then do not carry out whole Close;If there is, it may be judged whether leading between the two server can be reduced by the virtual machine on two servers of exchange Letter bandwidth, if it is then integrate, does not integrates.
The goal of the invention of the present invention is achieved in that
The cross-domain mapping method of a kind of virtual data center of the present invention, is mapped to the distributed of bottom by virtual data center Multiple data centers, and minimize total mapping cost and energy efficient, in order to adapt to the motility that VDC maps, one Simple and common solution is to start anew to remap VDC, and dynamic VDC mapping is to be reflected by the virtual resource newly increased It is mapped in bottom physical infrastructure, distributes physical resource for the dummy node newly increased or virtual link, simultaneously by minimizing Virtual resource is deleted from physical infrastructure, discharges its physical resource taken.
Meanwhile, the cross-domain mapping method of a kind of virtual data center of the present invention also has the advantages that
(1) applied widely.Traditional dynamic mapping is mostly for virtual network requests.This method can be suitably used for void Intend data center requests.Therefore, compared with traditional algorithm, the scope of application of this method is wider.
(2) total mapping cost is little.The dynamic virtual data center mapping algorithm proposed due to the present invention, by using Migrating between territory and reduce cost between territory, by mapping cost in migrating minimizing territory in using territory, thus it is little to reduce total mapping cost, Therefore, compared with traditional algorithm, the total mapping cost of this method is less.
(3) energy consumption is low.The dynamic virtual data center mapping algorithm proposed due to the present invention, migrates whole by VM in territory Close VDC and reduce the number opening physical host, thus reduce energy resource consumption, therefore compared with traditional algorithm, the energy of this method Consume lower.
Accompanying drawing explanation
Fig. 1 is the cross-domain mapping method flow diagram of a kind of virtual data center of the present invention;
Schematic diagram is migrated in territory when Fig. 2 is to increase node resource;
Fig. 3 is migration schematic diagram in territory during packet adjustment;
Fig. 4 is that virtual data center integrates schematic diagram.
Detailed description of the invention
Below in conjunction with the accompanying drawings the detailed description of the invention of the present invention is described, in order to those skilled in the art is preferably Understand the present invention.Requiring particular attention is that, in the following description, when known function and design detailed description perhaps When can desalinate the main contents of the present invention, these are described in and will be left in the basket here.
Embodiment
Fig. 1 is the cross-domain mapping method flow diagram of a kind of virtual data center of the present invention.
In the present embodiment, as it is shown in figure 1, the cross-domain mapping method of a kind of virtual data center of the present invention, including following Step:
(1), use the initial cross-domain algorithm of virtual data center that virtual data center carries out cross-domain mapping:
(1.1) in virtual data center, randomly choose a virtual machine carry out cross-domain mapping;
(1.2), in the unmapped virtual machine of virtual data center, select to be connected and communication bandwidth with maps virtual machine Maximum virtual machine maps;
(1.3), step (1.2) is repeated, until having mapped all of virtual machine;
(1.4), estimation virtual machine maps cost Cos taddn
Cost a d d n Σ v ∈ V 2 - V 1 Σ v ‾ ∈ V ‾ r e q ( v ) * h v v ‾ * cos t ( v ‾ ) - - - ( 1 )
Wherein, V2-V1Represent the dynamic change of virtual machine;Represent the set of server;V represents a virtual machine;Table Show a station server;Req (v) represents the resource requirement of virtual machine v;Representing current virtual machine mapping relations, whether v is mapped in ServerOn;Represent serverResource unit costs;
(1.5), the mapping cost Cos t of estimation original virtual data center internal virtual linkaddl
Cost a d d l = Σ e ∈ ( E 2 - E 1 ) Σ e ‾ ∈ M E ( e ) ϵ c a p ( e ‾ ) + δ * b ( e ) * cos t ( e ‾ ) * s 1 e - - - ( 2 )
Wherein, δ be one close to 0 minimum normal number;E2-E1Represent the dynamic change of virtual link;ε is used to control The parameter of load balancing importance;Represent a physical link;E represents a virtual link;B (e) represents virtual link e's Resource request;Represent physical linkResource unit costs;cos tbRepresentation unit bandwidth expense;Represent thing Reason linkResource capacity;MEE () represents the physical pathway that virtual link e maps;Or 0,1 represents that virtual link e maps In territory, 0 represents that virtual link e is not mapped in territory;
(1.6), between estimation territory, virtual link maps cost Cos taddb
C o s t a d d b = Σ e ∈ ( E 2 - E 1 ) s 2 e * b ( e ) * cos t b - - - ( 3 )
Wherein,Or 0,1 represents that virtual link e is mapped between territory, and 0 represents that virtual link e is not mapped between territory;
(2) user received when, virtual data center being carried out cross-domain mapping increases request newly and integrates;
(2.1), the request of resource of deleting is received;
(2.1.1), reduce the resource of designated virtual machine in request, the part physical resource that release reduces, change this void The resource requirement of plan machine;
(2.1.2), the resource of virtual link corresponding to designated virtual machine is reduced, the part band width in physical that release reduces, Change the bandwidth demand of this virtual link;
(2.2), the request of assembly of deleting is received;
(2.2.1), designated virtual machine in removal request, discharge the physical resource on the physical host that this virtual machine takies, And from virtual data center, delete this virtual machine;
(2.2.2), delete the virtual link that designated virtual machine is corresponding, discharge in the physical pathway that this virtual link takies Bandwidth, deletes this virtual link from virtual data center;
(2.3) request of increase node resource, is received;
(2.3.1), the virtual machine increasing resource will be needed to put in queue, by the virtual machine in queue according to increasing resource Amount descending;
If (2.3.2) server needing the virtual machine increasing resource to be mapped has enough surplus resources capacity, then Directly increase resources of virtual machine;Otherwise, by by the virtual machine (vm) migration on server to other servers, completing to increase void The request of plan machine resource;
In the present embodiment, in the algorithm increasing dummy node resource and increase dummy node, virtual machine (vm) migration is used Complete, by using virtual machine (vm) migration to enable to increasing during dummy node resource and dummy node link in territory Map cost minimization.As in figure 2 it is shown, the example migrated in a VM territory.When the resource requirement of dummy node a increases, and carrying a Physical server F in surplus resources off-capacity, now need to map a and link by moving to one in VM territory Map the server of cost minimization.The request of the resource of dummy node a is completed to increase by a being moved to server E.
(2.4) request of increase link circuit resource, is received;
For needing to increase the virtual link of bandwidth resources, if in the case of physical link bandwidth resources abundance, directly Connect the bandwidth increasing virtual link;If the resource of physical link is inadequate, then the shortest k routing algorithm is used to be torn open by virtual link Divide and be mapped on a plurality of bottom path;
(2.5) request of increase virtual machine, is received;
(2.5.1), will need increase new virtual machine put in queue, by the virtual machine in queue according to map Amount of bandwidth descending between virtual machine;
(2.5.2), according to descending order, the virtual machine newly increased is mapped;
(2.6), the request that packet adjusts is received;
(2.6.1), delete link in the territory needing to delete, and carry out the node needed in the packet migrated migrating between territory With remap;
(2.6.2), the bandwidth of link in territory is reduced, and to needing the node in the packet migrated to carry out migrating between territory and weighing Map;
In the present embodiment, as it is shown on figure 3, cause moving when a virtual link vl bandwidth demand in territory reduces Transposition is whole, divides all VM belonging to same VDC in the dc of this link place.It is divided into three if not VM being divided into two groups Group, group1, group2 and group3.If group1 moves to dc1, group2 and moves to dc2, group3 at former dc not Dynamic, and two end points of vl are respectively in group1 and group3.Total link band between so group group2 and group3 The wide impact not receiving the minimizing of link vl bandwidth, chain total between group2 and group3 before and after namely vl bandwidth reduces Road bandwidth does not change.Now, the link total bandwidth being mapped in network between territory between group2 and dc2 does not the most change.By In link vl bandwidth reduce before, between the territory between VM and dc2 in group2 the network bandwidth less than group2 and group3 it Between total bandwidth.After link vl bandwidth reduces, in group2, between territory between VM and dc2, the network bandwidth is also less than group2 And the total bandwidth between group3.Therefore the VM in group2 need not move in dc2, thus, the VM in group2 is permissible It is merged into a group that need not migrate with the VM in group3.
As it is shown on figure 3, when the bandwidth of link dc increases to 5 from 2, by being moved from former data center network by dummy node d Move on to the data center network at c place, reduce the bandwidth consumption of network between territory, thus reduce cost between territory.
(2.6.3), the bandwidth of link in territory is increased, and to needing the node in the packet migrated to carry out migrating between territory and weighing Map;
(2.6.4), estimation virtual machine remaps cost;
C o s t r e m n = Σ v ∈ V Σ v ‾ ∈ V ‾ req t ( v ) v * h v v ‾ * cos t ( v ‾ ) - Σ v ∈ V Σ v ‾ ∈ V ‾ req t - 1 ( v ) * h v v ‾ * cos t ( v ‾ ) - - - ( 4 )
Wherein, reqtV () represents the resource requirement of t virtual machine v, reqt-1V () represents the money of t-1 moment virtual machine v Source demand.
(2.6.5), estimated data's central interior virtual link remaps cost;
Cost r e m l = Σ e ∈ ( E ) Σ e ‾ ∈ M E ( e ) ϵ c a p ( e ‾ ) + δ b t ( e ) * cos t ( e ‾ ) * s 1 e - Σ e ∈ ( E ) Σ e ‾ ∈ M E ( e ) ϵ c a p ( e ‾ ) + δ b t - 1 ( e ) * cos t ( e ‾ ) * s 1 e - - - ( 5 )
Wherein, btE () represents the resource request of t virtual link e, bt-1E () represents the money of t-1 moment virtual link e Source is asked, and E represents the set of virtual link.
(2.6.6), between estimation territory, virtual link remaps cost;
C o s t r e m b = Σ e ∈ ( E 2 - E 1 ) s 2 e * b t ( e ) * cos t b - Σ e ∈ ( E 2 - E 1 ) s 2 e * b t - 1 ( e ) * cos t b - - - ( 6 )
(2.7) request that virtual data center is integrated, is received;
Along with VDC request scale reduce or VDC exceedes effect duration and leaves system, the utilization rate of substantial amounts of physical node Can become relatively low, also can produce resource fragmentation.In this case, it is desirable to dynamically integrate the service that VDC makes part open Device can cut out.The purpose periodically integrated is the quantity of the server in order to minimize the unlatching that VDC uses.
As shown in Figure 4, server a and server b is in same frame, and server c and server d is in another frame On.Therefore, when carrying out the integration of VDC local, the VM on server a and b can only be integrated, or by server c and d VM integrates.
Below the concrete step integrated is illustrated:
(2.7.1), the Servers-all that travels through in each frame, by the utilization rate of all of server according to from small to large Sequence, and is not above virtual machine (vm) migration that is minimum for utilization rate and that be less than on the server of lower limit to utilization rate maximum successively On the server of limit, until utilization rate all being closed less than the server of lower limit, or closing server cannot be continued;
(2.7.2), judge whether to exist between each server a plurality of virtual link, if it does not exist, then do not carry out whole Close;If there is, it may be judged whether leading between the two server can be reduced by the virtual machine on two servers of exchange Letter bandwidth, if it is then integrate, does not integrates.
Although detailed description of the invention illustrative to the present invention is described above, in order to the technology of the art Personnel understand the present invention, the common skill it should be apparent that the invention is not restricted to the scope of detailed description of the invention, to the art From the point of view of art personnel, as long as various change limits and in the spirit and scope of the present invention that determine in appended claim, these Change is apparent from, and all utilize the innovation and creation of present inventive concept all at the row of protection.

Claims (2)

1. the cross-domain mapping method of a virtual data center, it is characterised in that comprise the following steps:
(1), use the initial cross-domain algorithm of virtual data center that virtual data center carries out cross-domain mapping:
(1.1) in virtual data center, randomly choose a virtual machine carry out cross-domain mapping;
(1.2), in the unmapped virtual machine of virtual data center, select to be connected with maps virtual machine and communication bandwidth maximum Virtual machine map;
(1.3), step (1.2) is repeated, until having mapped all of virtual machine;
(1.4), estimation virtual machine maps cost Costaddn
C o s t a d d n = Σ v ∈ V 2 - V 1 Σ v ‾ ∈ V ‾ r e q ( v ) * h v v ‾ * cos t ( v ‾ ) - - - ( 1 )
Wherein, V2-V1Represent the dynamic change of virtual machine;Represent the set of server;V represents a virtual machine;Represent one Server;Req (v) represents the resource requirement of virtual machine v;Representing current virtual machine mapping relations, whether v is mapped in serverOn;Represent serverResource unit costs;
(1.5), the mapping cost Cost of estimation original virtual data center internal virtual linkaddl
C o s t a d d l = Σ e ∈ ( E 2 - E 1 ) Σ e ‾ ∈ M E ( e ) ϵ c a p ( e ‾ ) + δ * b ( e ) * cos t ( e ‾ ) * s 1 e - - - ( 2 )
Wherein, δ be one close to 0 minimum positive normal number;E2-E1Represent the dynamic change of virtual link;ε is used to control to bear Carry the parameter of equilibrium importance;Represent a physical link;E represents a virtual link;B (e) represents the money of virtual link e Source is asked;Represent physical linkResource unit costs;costbRepresentation unit bandwidth expense;Represent physics LinkResource capacity;MEE () represents the physical pathway that virtual link e maps;Or 0,1 represents that virtual link e is mapped in In territory, 0 represents that virtual link e is not mapped in territory;
(1.6), between estimation territory, virtual link maps cost Costaddb
C o s t a d d b = Σ e ∈ ( E 2 - E 1 ) s 2 e * b ( e ) * cos t b - - - ( 3 )
Wherein,Or 0,1 represents that virtual link e is mapped between territory, and 0 represents that virtual link e is not mapped between territory;
(2) user received when, virtual data center being carried out cross-domain mapping increases request newly and integrates;
(2.1), the request of resource of deleting is received;
(2.1.1), reduce the resource of designated virtual machine in request, the part physical resource that release reduces, change this virtual machine Resource requirement;
(2.1.2), reduce the resource of virtual link corresponding to designated virtual machine, the part band width in physical that release reduces, change The bandwidth demand of this virtual link;
(2.2), the request of assembly of deleting is received;
(2.2.1), designated virtual machine in removal request, discharge the physical resource on the physical host that this virtual machine takies, and from Virtual data center is deleted this virtual machine;
(2.2.2), delete the virtual link that designated virtual machine is corresponding, discharge the band in the physical pathway that this virtual link takies Width, deletes this virtual link from virtual data center;
(2.3) request of increase node resource, is received;
(2.3.1), the virtual machine increasing resource will be needed to put in queue, by the virtual machine in queue according to increasing stock number fall Sequence arranges;
If (2.3.2) server needing the virtual machine increasing resource to be mapped has enough surplus resources capacity, the most directly Increase resources of virtual machine;Otherwise, by by the virtual machine (vm) migration on server to other servers, completing to increase virtual machine The request of resource;
(2.4) request of increase link circuit resource, is received;
For needing to increase the virtual link of bandwidth resources, if in the case of physical link bandwidth resources abundance, directly increased Add the bandwidth of virtual link;If the resource of physical link is inadequate, then uses the shortest k routing algorithm to be split by virtual link and reflect It is mapped on a plurality of bottom path;
(2.5) request of increase virtual machine, is received;
(2.5.1), will need increase new virtual machine put in queue, by the virtual machine in queue according to maps virtual Amount of bandwidth descending between machine;
(2.5.2), according to descending order, the virtual machine newly increased is mapped;
(2.6), the request that packet adjusts is received;
(2.6.1), link in the territory needing to delete is deleted, and to needing the node in the packet migrated to carry out migrating between territory and weighing Map;
(2.6.2), the width of link in territory is reduced, and to needing the node in the packet migrated to carry out migrating between territory and remapping;
(2.6.3), increase the bandwidth of link in territory, and carry out the node needed in the packet migrated migrating and replay between territory Penetrate;
(2.6.4), estimation virtual machine remaps cost;
C o s t r e m n = Σ v ∈ V Σ v ‾ ∈ V ‾ req t ( v ) * h v v ‾ * cos t ( v ‾ ) - Σ v ∈ V Σ v ‾ ∈ V ‾ req t - 1 ( v ) * h v v ‾ * cos t ( v ‾ ) - - - ( 4 )
Wherein, reqtV () represents the resource requirement of t virtual machine v, reqt-1V () represents that the resource of t-1 moment virtual machine v needs Ask.
(2.6.5), estimated data's central interior virtual link remaps cost;
C o s t r e m l = Σ e ∈ ( E ) Σ e ‾ ∈ M E ( e ) ϵ c a p ( e ‾ ) + δ b t ( e ) * cos t ( e ‾ ) * s 1 e - Σ e ∈ ( E ) Σ e ‾ ∈ M E ( e ) ϵ c a p ( e ‾ ) + δ b t - 1 ( e ) * cos t ( e ‾ ) * s 1 e - - - ( 5 )
Wherein, btE () represents the resource request of t virtual link e, bt-1E () represents that the resource of t-1 moment virtual link e please Asking, E represents the set of virtual link.
(2.6.6), between estimation territory, virtual link remaps cost;
C o s t r e m b = Σ e ∈ ( E 2 - E 1 ) s 2 e * b t ( e ) * cos t b - Σ e ∈ ( E 2 - E 1 ) s 2 e * b t - 1 ( e ) * cos t b - - - ( 6 )
(2.7) request that virtual data center is integrated, is received;
(2.7.1), the Servers-all that travels through in each frame, by the utilization rate of all of server according to arranging from small to large Sequence, to utilization rate maximum and is not above the upper limit by virtual machine (vm) migration that is minimum for utilization rate and that be less than on the server of lower limit successively Server on, until utilization rate all being closed less than the server of lower limit, or closing server cannot be continued;
(2.7.2), judge whether to exist between each server a plurality of virtual link, if it does not exist, then do not integrate;As Fruit exists, it may be judged whether can reduce the communication band between the two server by the virtual machine on two servers of exchange Width, if it is then integrate, does not integrates.
The cross-domain mapping method of a kind of virtual data center the most according to claim 1, it is characterised in that described Or 0, wherein, 1 represents that v is mapped in serverOn, 0 represents that v is not mapped inOn.
CN201610640554.7A 2016-08-05 2016-08-05 A kind of cross-domain mapping method of virtual data center Active CN106254115B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201610640554.7A CN106254115B (en) 2016-08-05 2016-08-05 A kind of cross-domain mapping method of virtual data center

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201610640554.7A CN106254115B (en) 2016-08-05 2016-08-05 A kind of cross-domain mapping method of virtual data center

Publications (2)

Publication Number Publication Date
CN106254115A true CN106254115A (en) 2016-12-21
CN106254115B CN106254115B (en) 2019-03-15

Family

ID=58078976

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201610640554.7A Active CN106254115B (en) 2016-08-05 2016-08-05 A kind of cross-domain mapping method of virtual data center

Country Status (1)

Country Link
CN (1) CN106254115B (en)

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20110066566A1 (en) * 2009-09-16 2011-03-17 International Business Machines Corporation Conceptual representation of business processes for cross-domain mapping
CN103391233A (en) * 2013-07-31 2013-11-13 清华大学 Mapping method of cross-domain virtual network
CN104468803A (en) * 2014-12-12 2015-03-25 华为技术有限公司 Virtual data center resource mapping method and equipment
CN104901999A (en) * 2015-03-31 2015-09-09 电子科技大学 Method for reliability perceptive cross-domain mapping of virtual data center

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20110066566A1 (en) * 2009-09-16 2011-03-17 International Business Machines Corporation Conceptual representation of business processes for cross-domain mapping
CN103391233A (en) * 2013-07-31 2013-11-13 清华大学 Mapping method of cross-domain virtual network
CN104468803A (en) * 2014-12-12 2015-03-25 华为技术有限公司 Virtual data center resource mapping method and equipment
CN104901999A (en) * 2015-03-31 2015-09-09 电子科技大学 Method for reliability perceptive cross-domain mapping of virtual data center

Also Published As

Publication number Publication date
CN106254115B (en) 2019-03-15

Similar Documents

Publication Publication Date Title
EP3530037B1 (en) System and method for network slice management in a management plane
Sun et al. Live migration for multiple correlated virtual machines in cloud-based data centers
CN105610715B (en) A kind of cloud data center multi-dummy machine migration scheduling method of planning based on SDN
CN104270421B (en) A kind of multi-tenant cloud platform method for scheduling task for supporting Bandwidth guaranteed
CN104935628B (en) A kind of method that multiple associated virtual machines are migrated between multiple data centers
CN103294521A (en) Method for reducing communication loads and energy consumption of data center
CN103812748B (en) Mapping method of survivable virtual network
Tsai et al. Two-tier multi-tenancy scaling and load balancing
CN102664814A (en) Grey-prediction-based adaptive dynamic resource allocation method for virtual network
CN103401939A (en) Load balancing method adopting mixing scheduling strategy
CN105391651B (en) Virtual optical network multi-layer resource convergence method and system
CN103560967A (en) Virtual data center mapping method sensing service requirement
CN109525426A (en) A kind of operation control system and method for the open MEC platform based on FV
CN107277832A (en) A kind of mapping method of wireless dummy network
CN108880888A (en) A kind of SDN network method for predicting based on deep learning
CN109379281A (en) A kind of traffic scheduling method and system based on time window
CN106161603B (en) A kind of method of networking, equipment and framework
EP3278592A1 (en) Technique for extending capacities of a radio access network
CN112073237B (en) Large-scale target network construction method in cloud edge architecture
Liu et al. Sn-vne: A virtual network embedding algorithm for satellite networks
Li et al. Dynamic SDN controller association mechanism based on flow characteristics
CN103595610A (en) Destroy-resistant mapping method for uncertainty resource demand multicast virtual network
CN110535673A (en) Construction method, equipment and the storage medium of virtual subnet
CN105138391B (en) The multitasking virtual machine distribution method of cloud system justice is distributed towards wide area
CN108923961B (en) Multi-entry network service function chain optimization method

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
CP03 Change of name, title or address
CP03 Change of name, title or address

Address after: 523000, No. two, No. 17, Songshan hi tech Industrial Development Zone, Guangdong, Dongguan

Co-patentee after: Guangdong platinum Information Technology Co., Ltd.

Patentee after: Guangdong Institute of electronic and information engineering, University of Electronic Science and technology of China

Address before: 523808, No. two, No. 17, Songshan hi tech Industrial Development Zone, Guangdong, Dongguan

Co-patentee before: Guangdong platinum Information Technology Co., Ltd.

Patentee before: Institute of Electronic and Information Engineering In Dongguan, UESTC

PP01 Preservation of patent right
PP01 Preservation of patent right

Effective date of registration: 20220524

Granted publication date: 20190315