CN105159780B - The high availability virtual network mapping method and device of oriented multilayer time cloud application - Google Patents
The high availability virtual network mapping method and device of oriented multilayer time cloud application Download PDFInfo
- Publication number
- CN105159780B CN105159780B CN201510520914.5A CN201510520914A CN105159780B CN 105159780 B CN105159780 B CN 105159780B CN 201510520914 A CN201510520914 A CN 201510520914A CN 105159780 B CN105159780 B CN 105159780B
- Authority
- CN
- China
- Prior art keywords
- layer
- link
- vector
- feasible
- request
- 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.)
- Expired - Fee Related
Links
Abstract
The present invention relates to a kind of high availability virtual network mapping method and devices of oriented multilayer time cloud application, belong to multi-level cloud application technical field.The method of the present invention obtains the multi-level application request of user's proposition first;Then the virtual machine quantity descending sort that the request proposed to user is needed according to each layer;Next the sequence of the virtual machine quantity needed according to each layer from high to low is each Layer assignment resources of virtual machine, specific as follows:The feasible vector S of all allocation plans is calculated the virtual machine quantity of this layer request needs first according to the tree topology under current cloud platformeWith minimum feasible valueAnd corresponding object vector PeWith minimum target valueSecondly, optimal distributing scheme of each of the links to this layer is obtained by recursion method;Finally export the optimal distributing scheme of the multi-level application request of the user.The prior art is compared, Resource Allocation Formula of the invention takes into account resource utilization and availability, while ensureing availability, improves the efficiency of whole system.
Description
Technical field
The present invention relates to a kind of mapping method of virtual network and device, more particularly to the height of a kind of oriented multilayer time cloud application
Availability mapping method of virtual network and device belong to multi-level cloud application technical field, particularly belong to virtual network availability
With the fields such as resource utilization.
Background technology
In recent years, more and more service providers moved to their multi-level applications among public cloud network.
In traditional cloud network environment, cloud provider possesses physical equipment, virtual network resource can be provided to service provider, every
A service provider obtains after an individual virtual network resource, you can certain service is provided for each user.Virtually
Internet startup disk realizes mapping of the virtual network to actual physical networks.In order to realize that better application service quality, service carry
Would generally there are the specific demand of some Resource Guarantees (such as resources of virtual machine, bandwidth resources), but some to virtual network for quotient
The failure of physical network (such as interchanger, link) often makes a big impact to the availability of service.Some existing researchs carry
Go out the method for ensureing data center network availability, for example designs a kind of novel topological structure with network redundancy, but
It is still a thing for perplexing many people that virtual network resource availability caused by being data center's failure, which reduces,.Therefore, it takes
Provider be engaged in for improving virtual network availability and ensureing that the two targets of virtual resource have prodigious demand.
It is operated on identical equipment simultaneously now with multiple virtual networks, so it is necessary to a kind of virtual net of design
Network is embedded in mechanism, meets the needs of service provider while realizing the efficient utilization of resources.But it due to high efficiency and can be used
Property itself contradicts, so realizing that the two targets are extremely difficult simultaneously.The present invention proposes the virtual net of high availability
Network is embedded in frame, while improving resource utilization the availability with system.In addition, the present invention can measure the utilization of resources respectively
Rate expense and availability expense are solved using the two expenses as optimization problem.
In multi-level cloud application, the utilization rate and availability of resource are two contradictory targets itself, while considering this
Two targets, and the balance for obtaining the two targets has very important significance for multi-level cloud application, but regrettably
Currently lack the effective solution method of associated row regarding to the issue above.
Invention content
The purpose of the present invention is can not take into account resource utilization and availability the two contradictory mesh to solve the prior art
Mark carries out the problem of virtual network mapping, proposes a kind of high availability virtual network mapping method of oriented multilayer time cloud application,
This method can effectively improve the availability and resource utilization of virtual network, meanwhile, the present invention has flexibility height, availability
By force, available and efficient feature.
Idea of the invention is that being asked according to user, the resource utilization calculated separately under each Resource Allocation Formula is opened
Pin and availability expense obtain resource allocation when the sum of expense minimum, optimal resource allocation scheme as of the present invention.
The purpose of the present invention is what is be achieved through the following technical solutions:
A kind of high availability virtual network mapping method of oriented multilayer time cloud application, includes the following steps:
1. service provider (user) proposes the request T={ T with k layers of application to cloud provideri(1≤i≤k),
WhereinNiIndicate the quantity of the virtual machine using i-th layer of needs;I-th layer of expression is upper each virtual
The bandwidth that machine intralayer communication needs;Indicate the bandwidth that i-th layer of upper each virtual machine interlayer communication needs;
2. cloud provider obtains the tree topology G=(V, E) under current cloud platform, wherein V indicates Servers-all
Set, E indicates the set of all links, and root is the root link in E;And according in request T using the virtual of every layer of demand
Machine quantity NiDescending arrangement is carried out, and result is put into array T';
3. for each layer of request T in T'i, calculate the virtual machine feasible vector S of each of the links subtreeeIt can with minimum
Row valueFeasible vector SeExpression it is as shown in table 1:
1 feasible vector S of tablee
x | 0 | 1 | …… | Ni |
y | 0/1 | 0/1 | …… | 0/1 |
Indicate the Resource Allocation Formula F of link e subtrees is that is, when link e subtrees distribute x resources of virtual machine
No feasible (y=1 indicates feasible, otherwise infeasible);For variable x, if meeting x≤rv, then corresponding y values are 1, rvIndicate clothes
The quantity of business device v empty slots;Minimum feasible valueWhen indicating y=1, the minimum value of x;
4. according to feasible vector SeIn corresponding virtual resource allocation scheme F calculate the corresponding object vector P of link eeMost
Small object value
Pe=C (F)+α A (F) (1)
Wherein, C (F) indicates the sum of the bandwidth cost of data center's structure all links at allocation plan F, passes through following formula
It calculates:
Wherein, weIndicate the weight that link e occupies, ceIndicate the capacity of link e, beIndicate the link at allocation plan F
Bandwidth demand is calculate by the following formula:
Wherein,Indicate the virtual machine quantity of i-th layer contained in the subtree being connected with G links e application,Indicate G
The virtual machine quantity for i-th layer of application that the middle remainder for removing the subtree that is connected with link e contains,
It indicates to set the required bandwidth of intralayer communication with the link e subtrees being connected and residue,It indicates
With the link e subtrees being connected and the remaining tree required bandwidth of interlayer communication;Wherein, the bandwidth for distributing to link e is not more than chain
The remaining bandwidth capacity of road e, i.e.,Re indicates the residual capacity of link e, if be> re, illustrate link e's
Remaining bandwidth is insufficient for the demand of user under allocation plan F, then gives up the program;
A (F) indicates availability expense of data center's structure at allocation plan F, is calculate by the following formula:
Wherein, J indicates the server set that may be failed in data center, pjp,jJ (, jj ∈ JJ) and it indicates each to take in J
The possibility of business device j failures;
θij=∑v∈Vfivzvj/Ni (5)
θijWhen indicating that j fails in J, the i-th layer of resources of virtual machine quantity distributed in j occupies the ratio of i-th layer of family resource requirement
Example, fivIndicate the quantity of the virtual resource of the i-th Layer assignment in server v to application, zvjIndicate whether server v is j,
Therefore zvjValue is 0 or 1;As v=j, zvj=1;As v ≠ j, zvj=0;α is the parameter of availability expense A (F), is carried by cloud
It is provided for quotient;Minimum target valueFor object vector PeMinimum value;
5. cloud provider is according to feasible vector SeWith minimum feasible valueAnd object vector PeWith minimum target value
I-th layer of optimal virtual resource allocation scheme is obtained by traversal method;
Preferably, the traversal method is recursion method alloc (r, e, n), this method is obtained to i-th layer of application distribution
Virtual resource quantity, and with user request NiIt is compared, if the two is equal, returns to true, illustrate that there are i-th layer
The allocation plan of application;If the two is unequal, false is returned, illustrates that there is no the allocation plans of i-th layer of application;Wherein,
Parameter r indicates that user's request, parameter e indicate that link, parameter n indicate that user asks the virtual resource quantity of distribution;Also, it will be with
Original execution links of the link root as method alloc, i.e., by Ti,root,NiIn substitution method alloc (r, e, n);alloc
The workflow of (r, e, n) is as follows:
(1) if level (e)=0 (level indicates levels of the link e residing for data center, when level (e)=0, chain
Road e is connected directly with server), the virtual resource quantity of distribution is
It indicates to work as object vector PeIt is minimizedWhen corresponding l values, judge doneVM in feasible vector SeWhether legal (x=is worked as
When doneVM, whether y is 1), if illegal, to return to -1;If legal, doneVM values are returned, i.e., are applied to i-th layer
Quantity allotted is the resources of virtual machine of l in the server v that link e is connected, and is mathematically represented as fiv=l;
(2) if level (e) ≠ 0, doneVM ← 0d, on set eVseMt (ee) t as -1 layer of mp link e subtree level (e)
Link set, link e' ∈ set (e), to each of the links e' in set (e) according toValue carries out ascending order arrangement, calculates empty
Quasi- resource quantity available base,For each of the links e' in set (e), calculateWith tmp=alloc (r, e', ne'), tmp is the numerical value that recurrence executes that method alloc (steps 5) obtain;
If tmp ≠ -1, base and doneVM carries out following assignment:With doneVM ← doneVM+
tmp;If tmp=-1, alloc return to -1 value;After each of the links e' operations in having executed set (e), judge
DoneVM is in feasible vector SeWhether legal (i.e. as x=doneVM, whether y is 1), if illegal, to return to -1;If
It is legal, then return to doneVM values;
6. executing the step after 5, you can obtain the Resource Allocation Formula { f that virtual network is applied at i-th layeriv, v ∈ V,
I is fixed, and executes the request in the 2nd step in T' using other layers successively, obtains other layers of Resource Allocation Formula { fjv},1≤j
≤ k, j ≠ i, v ∈ V are final to obtain virtual network insertion solution F={ fiv, 1≤i≤k, v ∈ V, fivIt is represented to application
I-th layer of virtual resource distributed in server v quantity.
A kind of high availability virtual network mapping device of oriented multilayer time cloud application, including sequentially connected user request
Receiving module, request computing module, resource distribution module and resource allocation output module;User's request receiving module is for receiving
User asks T={ Ti(1≤i≤k), and export user request to request computing module;Computing module is asked to be used for basis
User asks the tree topology under T and current cloud platform to calculate separately the feasible vector of link and corresponding object vector,
And export user's request and result of calculation to resource distribution module:Feasible vector, object vector;Resource distribution module is for calling
Alloc methods are asked according to user, feasible vector and object vector be the certain resources of virtual machine of each Layer assignment for applying, and
Allocation plan is output to resource allocation output module;Resource allocation output module is carried for exporting Resource Allocation Formula to cloud
For quotient.
Preferably, the request computing module is further by feasible vector computing module, object vector computing module, band
Wide overhead computational module and availability overhead computational module composition, feasible vector computing module and object vector computing module are direct
It is connected, object vector computing module is connected with bandwidth cost computing module and availability overhead computational module respectively;It is described feasible
Vector calculation module is for enumerating all possible Resource Allocation Formula under each of the links subtree;Bandwidth cost computing module is used for
Calculate the link bandwidth utilization rate (i.e. bandwidth cost) under certain Resource Allocation Formula;Availability overhead computational module is based on
Calculate influence (i.e. availability expense) of the server failure to the overall situation under certain Resource Allocation Formula;Object vector computing module is used
In the sum of bandwidth cost and availability expense under certain Resource Allocation Formula.
Advantageous effect
Under the prior art, many researchs both for raising data center resource utilization rate (i.e. bandwidth cost) or can be used
Property expense, not by the combined research of the two, otherwise this causes available data center resource utilization low or available
Property is low, and the two is difficult to be optimal counterbalance effect simultaneously.And a kind of High Availabitity of oriented multilayer proposed by the present invention time cloud application
Property mapping method of virtual network, improves the availability and resource utilization of virtual network.Meanwhile the invention also provides one
The method of new quantization multilayer virtual network availability expense, bandwidth and availability expense are solved as an optimization problem
Certainly;In addition, since this optimization problem is a np hard problem, so the present invention devises a heuritic approach, Neng Gou
This is solved the problems, such as in polynomial time, reduces time complexity.The present invention can solve all tree topology knots of data center
The resource allocation problem of structure has the characteristics that flexibility is high;Method proposed by the present invention can be deployed at once data center it
On, have the characteristics that availability is strong;Meanwhile Resource Allocation Formula of the invention takes into account resource utilization and availability, ensures
While availability, the efficiency of whole system is improved.
Description of the drawings
Fig. 1 is the virtual resource allocation schematic diagram of low availability, low bandwidth overhead;
Fig. 2 is the virtual resource allocation schematic diagram of high availability, high bandwidth expense;
Fig. 3 is in k layers of application layer, interlayer communication bandwidth demand schematic diagram;
Fig. 4 is the initial primary topology schematic diagram of data center;
Fig. 5 is the topological structure schematic diagram of data center after resource allocation;
Fig. 6 is the method for the present invention flow diagram;
Fig. 7 is the method for the present invention apparatus structure schematic diagram.
Specific implementation mode
To make the target of the present invention, technical solution and advantage are more explicit, below in conjunction with attached drawing to the present invention's
Embodiment is described in detail.The present embodiment is with the technical scheme is that guidance carries out actual practice veritification, simultaneously
Detailed embodiment and specific operating process are given, but protection scope of the present invention is not only limited in following implementation
Example.
Fig. 1 and Fig. 2 indicates the virtual resource point of low availability, low bandwidth overhead and high availability, high bandwidth expense respectively
With schematic diagram.The structure of two figures is identical, respectively containing there are one core router (Cor), two convergence routers (Agg), four
A server, each server most multipotency put four virtual machines, and core router connects two convergence routers, convergence router
Connect two servers.If the low availability of Fig. 1 is embodied in the failure of server 1, whole system can paralyse state;Its
The communication that low bandwidth overhead embodies in the server 1 between four virtual machines can be realized in server internal, it is not necessary to it is other
Virtual machine in server is communicated, and the bandwidth entirely communicated in this way is maintained at lower degree.The high availability of Fig. 2 is embodied in
If there are one failures, the other three server still can work, be unlikely to total system and paralyse in four servers;
Four virtual machines that its high bandwidth expense is embodied in distribution are distributed in four different servers, the communication between virtual machine across
Different servers has been got over, there is higher bandwidth cost.
Fig. 3 indicates every interior from level to level, interlayer communication bandwidth demand of k layers of application.By taking first layer as an example, intralayer communication passes through
Virtual switch (Intra-tier VS) carries out in layer, and the bandwidth needed isInterlayer communication passes through interlayer virtual switch
(Cross-tier VS) is carried out, and the bandwidth needed is
Fig. 4 indicates the initial primary topology of data center, including a convergence router (Agg), two servers, each
Server can accommodate four virtual machines, and each of the links capacity is 1Gbps.Convergence router connects two servers, server 1,2
The probability of failure is respectively 0.01,0.02.
The implementation process of the method for the present invention is illustrated by taking data center's structure shown in Fig. 4 as an example below:First, if
It is 200 to determine step 4 Parameters in Formula α, and weight is respectively 1.
It is illustrated in figure 6 the method for the present invention flow diagram, is included the following steps:
The P 1. user files a request, is shown in Table 2:
2 user of table asks P
Parameter<1,0,150>Indicate that user asks first layer to need 1 virtual resource, in layer, interlayer communication bandwidth demand
Difference 0,150Mbps;Parameter<3,0,200>Indicate that user asks the second layer to need 3 virtual resources, in layer, interlayer communication band
Wide demand difference 0,200Mbps.
2. the virtual resource demand descending of pair each layer of request arrange to obtain T '=<3,1>.
3. for the 2nd layer, the feasible vector S of each of the links is calculated1、S2, as shown in Table 3, 4:
3 link l of table1Feasible vector S1
x | 0 | 1 | 2 | 3 |
y | 1 | 1 | 1 | 1 |
At this point, minimum feasible value
4 link l of table2Feasible vector S2
x | 0 | 1 | 2 | 3 |
y | 1 | 1 | 1 | 1 |
At this point, minimum feasible value
4. the corresponding object vector of feasible vector of above-mentioned each link is:
5 object vector P of table1
x | 0 | 1 | 2 | 3 |
p | 16/9 | 12/5 | 26/15 | 4 |
At this point, minimum feasible value P1 min=26/15.
6 object vector P of table2
x | 0 | 1 | 2 | 3 |
P | 4 | 26/15 | 12/5 | 16/9 |
At this point, minimum feasible value
5. by T2, root, 3 are updated in alloc methods, due to level (root) ≠ 0, set (root)={ l1,l2,
According toAscending order arranges to obtain { l1,l2, base=3- (0+0)=3, for each of the links l in set (root)1,
l2,
Execute following steps:(1)By T2,l1, 3 are updated in alloc methods, due to
level(l1)=0, so tmp=2, so the virtual machine quantity distributed in server 1 is 2, because of tmp ≠ -1, base
=3- (2-0)=1, doneVM=0+2=2.(2)By T2,l2, 1 is updated in alloc methods,
Due to level (l2)=0, thus tmp=1, so the virtual machine quantity distributed in server 2 is 1, because of tmp ≠ -1,
Base=1- (1-0)=0, doneVM=2+1=3, finally, alloc (Ti,root,Ni) return result be doneVM=3, this
When meet alloc (T2,root,N2)=N2, so the program is feasible.In summary, for applying the 2nd layer, allocation plan
For:2 resources of virtual machine are distributed in the server 1, and 1 resources of virtual machine, i.e. f are distributed in server 221=2, f22=1.
6. for the 1st layer, the feasible vector S of each of the links is calculated1、S2, as shown in table 7,8:
7 link l of table1Feasible vector S1
x | 0 | 1 |
y | 1 | 1 |
At this point, minimum feasible value
8 link l of table2Feasible vector S2
x | 0 | 1 |
y | 1 | 1 |
At this point, minimum feasible value
The corresponding object vector of feasible vector of above-mentioned each link is:
9 object vector P of table1
x | 0 | 1 |
p | 89/20 | 49/20 |
At this point, minimum feasible value P1 min=49/20.
10 object vector P of table2
x | 0 | 1 |
P | 49/20 | 89/20 |
At this point, minimum feasible value
By T1, root, 1 is updated in alloc methods, due to level (root) ≠ 0, set (root)={ l1,l2, it presses
According toAscending order arranges to obtain { l1,l2, base=1- (0+0)=1, for each of the links l in set (root)1,l2,
Execute following steps:(1)By T2,l1, 1 is updated in alloc methods, due to level (l1)=
0, so tmp=1, so the virtual machine quantity distributed in server 1 is 1, because tmp ≠ -1, base=1- (1-0)=
0, doneVM=0+1=1.(2)By T2,l2, 0 is updated in alloc methods, due to level
(l2)=0, so tmp=0, so the virtual machine quantity distributed in server 2 is 0, because of tmp ≠ -1, base=0-
(0-0)=0, doneVM=1+0=1, finally, alloc (Ti,root,Ni) return result be doneVM=1, meet at this time
alloc(T1,root,N1)=N1, so the program is feasible.In summary, for applying the 1st layer, allocation plan is:
1 resources of virtual machine is distributed in server 1, and 0 resources of virtual machine, i.e. f are distributed in server 211=1.
7. so allocation plan is:Server 1, which is given, applies 1 virtual machine of the first Layer assignment, to using the second Layer assignment 2
Virtual machine, server 2, which is given, applies 1 virtual machine of the second Layer assignment, is expressed as F={ f21=2, f22=1, f11=1 }.Such as Fig. 5 institutes
Show.
The high availability virtual network for being illustrated in figure 7 a kind of oriented multilayer time cloud application made based on the above method is reflected
Injection device structural schematic diagram, as can be seen from Figure, apparatus of the present invention include sequentially connected user's request receiving module, request
Computing module, resource distribution module and resource allocation output module;User's request receiving module asks T=for receiving user
{Ti(1≤i≤k), and export user request to request computing module;Ask computing module be used for according to user ask T and
Tree topology under current cloud platform calculates separately the feasible vector of link and corresponding object vector, and to resource allocation
Module exports user's request and result of calculation:Feasible vector, object vector;Resource distribution module is for calling alloc method roots
Be the certain resources of virtual machine of each Layer assignment for applying according to user's request, feasible vector and object vector, and by allocation plan
It is output to resource allocation output module;Resource allocation output module gives cloud provider for exporting Resource Allocation Formula.
Preferably, the request computing module is further by feasible vector computing module, object vector computing module, band
Wide overhead computational module and availability overhead computational module composition, feasible vector computing module and object vector computing module are direct
It is connected, object vector computing module is connected with bandwidth cost computing module and availability overhead computational module respectively;It is described feasible
Vector calculation module is for enumerating all possible Resource Allocation Formula under each of the links subtree;Bandwidth cost computing module is used for
Calculate the link bandwidth utilization rate (i.e. bandwidth cost) under certain Resource Allocation Formula;Availability overhead computational module is based on
Calculate influence (i.e. availability expense) of the server failure to the overall situation under certain Resource Allocation Formula;Object vector computing module is used
In the sum of bandwidth cost and availability expense under certain Resource Allocation Formula.
The bandwidth cost solution of link proposed by the present invention:Pass through bandwidth allocation in link
Ratio, and assign link certain weight, really reflect the link bandwidth expense in data center's tree topology;It proposes
Availability expense solution:By the way that virtual resource is put down by the ratio that server failure is influenced
Side, can better discriminate between the influence brought by server failure, be consistent with actual conditions.
Above-described specific descriptions have all carried out the purpose, technical scheme and advantage benefit of invention further detailed
Describe in detail it is bright, it should be understood that the above is only a specific embodiment of the present invention, the guarantor being not intended to limit the present invention
Range is protected, all within the spirits and principles of the present invention, any modification, equivalent substitution, improvement and etc. done should be included in this
Within the protection domain of invention.
Claims (3)
1. a kind of high availability virtual network mapping method of oriented multilayer time cloud application, it is characterised in that:Include the following steps:
(1) service provider proposes the request T={ T with k layers of application to cloud provideri, wherein 1≤i≤k,NiIndicate the quantity of the virtual machine using i-th layer of needs;Indicate i-th layer of upper each virtual machine layer
The bandwidth that interior communication needs;Indicate the bandwidth that i-th layer of upper each virtual machine interlayer communication needs;
(2) cloud provider obtains the tree topology G=(V, E) under current cloud platform, and wherein V indicates the collection of Servers-all
It closes, E indicates the set of all links, this Rigen's link is root;And according to the virtual machine quantity for applying every layer of demand in request T
NiDescending arrangement is carried out, and result is put into array T';
(3) for each layer of request T in T'i, calculate the virtual machine feasible vector S of each of the links subtreeeWith minimum feasible valueFeasible vector SeIndicate the Resource Allocation Formula F of link e subtrees, i.e., when in link e subtrees x resources of virtual machine of distribution
When, if it is feasible:Y=1 indicates feasible, otherwise infeasible;For variable x, if meeting x≤rv, then corresponding y values are 1, rvTable
Show the quantity of server v empty slots;Minimum feasible valueWhen indicating y=1, the minimum value of x;
(4) according to feasible vector SeIn corresponding virtual resource allocation scheme F calculate the corresponding object vector P of link eeAnd minimum
Desired value
Pe=C (F)+α A (F) (1)
Wherein, C (F) indicates the sum of the bandwidth cost of data center's structure all links at allocation plan F, passes through following formula meter
It calculates:
Wherein, weIndicate the weight that link e occupies, ceIndicate the capacity of link e, beIndicate the bandwidth of the link at allocation plan F
Demand is calculate by the following formula:
Wherein,Indicate the virtual machine quantity of i-th layer contained in the subtree being connected with G links e application,It indicates to remove in G
The virtual machine quantity of i-th layer of application that the remainder of subtree that goes to be connected with link e contains,It indicates
The required bandwidth of intralayer communication is set with the link e subtrees being connected and residue,Expression and chain
Subtree connected road e and the remaining tree required bandwidth of interlayer communication;Wherein, the bandwidth of link e is distributed to no more than link e's
Remaining bandwidth capacity, i.e. be≤re,reThe residual capacity for indicating link e, if be> re, illustrate the residue of link e
Bandwidth is insufficient for the demand of user under allocation plan F, then gives up the program;
A (F) indicates availability expense of data center's structure at allocation plan F, is calculate by the following formula:
Wherein, J indicates the server set that may be failed in data center, pj(j ∈ J) indicates each server j failures in J
Possibility;
θij=∑v∈Vfivzvj/Ni (5)
θijWhen indicating that j fails in J, the i-th layer of resources of virtual machine quantity distributed in j occupies the ratio of the i-th layer of resource requirement in family, fiv
Indicate the quantity of the virtual resource of the i-th Layer assignment in server v to application, zvjIndicate whether server v is j, therefore zvjIt takes
Value is 0 or 1;As v=j, zvj=1;As v ≠ j, zvj=0;
α is the parameter of availability expense A (F), is provided by cloud provider;Minimum target valueFor object vector PeMinimum
Value;
(5) cloud provider is according to i-th layer of feasible vector SeWith minimum feasible valueAnd object vector PeWith minimum target valueI-th layer of optimal virtual resource allocation scheme is obtained by traversal method;
(6) it executes the step after 5, you can obtain the Resource Allocation Formula { f that virtual network is applied at i-th layeriv, v ∈ V, i is solid
It is fixed, and the request in the 2nd step in T' using other layers is executed successively, obtain other layers of Resource Allocation Formula { fjv},1≤j≤
K, j ≠ i, v ∈ V, it is final to obtain virtual network insertion solution F={ fiv, 1≤i≤k, v ∈ V, fivIt is represented to application
The quantity of i-th layer of virtual resource distributed in server v;
Step (5) described traversal method is recursion method alloc (r, e, n), and this method is obtained to i-th layer using the virtual of distribution
Resource quantity, and the N asked with useriCompared, if the two is equal, return to true, illustrate there are i-th layer apply
Allocation plan;If the two is unequal, false is returned, illustrates that there is no the allocation plans of i-th layer of application;Wherein, parameter r
Indicate that user's request, parameter e indicate that link, parameter n indicate that user asks the virtual resource quantity of distribution;It implements process
It is as follows:
(5.1) if level (e)=0, wherein level indicates layers of the link e residing for data center
Secondary, when level (e)=0, link e is connected directly with server, and the virtual resource quantity of distribution isIt indicates to work as object vector PeIt is minimizedWhen corresponding l values, sentence
Disconnected doneVM is in feasible vector SeWhether legal, i.e., as x=doneVM, whether y is 1, if illegal, returns to -1;Such as
Fruit is legal, then returns to doneVM values, i.e., apply the virtual machine that quantity allotted is l in the server v that link e is connected to i-th layer
Resource is mathematically represented as fiv=l;
(5.2) if level (e) ≠ 0, if set (e) is the link set of -1 layer of link e subtree level (e), link e' ∈ set
(e), to each of the links e' in set (e) according toValue carries out ascending order arrangement, calculates virtual resource quantity available base,For each of the links e' in set (e), calculateAnd tmp=
alloc(r,e',ne'), tmp is that recurrence execution method alloc (steps 5) obtain numerical value;If tmp ≠ -1, base and doneVM
Carry out following assignment:With doneVM ← doneVM+tmp;If tmp=-1, alloc are returned
Return -1 value;After each of the links e' operations in having executed set (e), judge doneVM in feasible vector SeIt is whether legal, i.e.,
As x=doneVM, whether y is 1, if illegal, returns to -1;If legal, doneVM values are returned.
2. a kind of face of the high availability virtual network mapping method based on a kind of oriented multilayer described in claim 1 time cloud application
To the high availability virtual network mapping device of multi-level cloud application, it is characterised in that:Sequentially connected user asks to receive mould
Block, request computing module, resource distribution module and resource allocation output module;User's request receiving module is asked for receiving user
Seek T={ Ti, wherein 1≤i≤k, and export user request to request computing module;Computing module is asked to be used for according to user
Tree topology under request T and current cloud platform calculates separately the feasible vector of link and corresponding object vector, and to
Resource distribution module exports user's request and result of calculation:Feasible vector, object vector;Resource distribution module is for calling
Alloc methods are asked according to user, feasible vector and object vector be the certain resources of virtual machine of each Layer assignment for applying, and
Allocation plan is output to resource allocation output module;Resource allocation output module is carried for exporting Resource Allocation Formula to cloud
For quotient;The request computing module is used for using claim 1 step (1)-(5) realization, and the alloc methods are wanted using right
1 step (5) is asked to realize.
3. a kind of high availability virtual network mapping device of oriented multilayer according to claim 2 time cloud application, special
Sign is:The request computing module is further calculated by feasible vector computing module, object vector computing module, bandwidth cost
Module and availability overhead computational module composition, feasible vector computing module and object vector computing module are connected directly, target
Vector calculation module is connected with bandwidth cost computing module and availability overhead computational module respectively;The feasible vector calculates mould
Block is for enumerating all possible Resource Allocation Formula under each of the links subtree;Bandwidth cost computing module is for calculating at certain
Link bandwidth utilization rate under Resource Allocation Formula, i.e. bandwidth cost;Availability overhead computational module is for calculating in certain money
Server failure is to global influence under the allocation plan of source, you can with property expense;Object vector computing module is used in certain money
The sum of bandwidth cost and availability expense under the allocation plan of source.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510520914.5A CN105159780B (en) | 2015-08-21 | 2015-08-21 | The high availability virtual network mapping method and device of oriented multilayer time cloud application |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510520914.5A CN105159780B (en) | 2015-08-21 | 2015-08-21 | The high availability virtual network mapping method and device of oriented multilayer time cloud application |
Publications (2)
Publication Number | Publication Date |
---|---|
CN105159780A CN105159780A (en) | 2015-12-16 |
CN105159780B true CN105159780B (en) | 2018-07-20 |
Family
ID=54800643
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201510520914.5A Expired - Fee Related CN105159780B (en) | 2015-08-21 | 2015-08-21 | The high availability virtual network mapping method and device of oriented multilayer time cloud application |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN105159780B (en) |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102223281A (en) * | 2011-06-24 | 2011-10-19 | 清华大学 | Method for remapping resource demand dynamic change based on virtual network |
CN102904794A (en) * | 2012-09-27 | 2013-01-30 | 北京邮电大学 | Method and device for mapping virtual network |
CN103218175A (en) * | 2013-04-01 | 2013-07-24 | 无锡成电科大科技发展有限公司 | Multi-tenant cloud storage platform access control system |
CN103856563A (en) * | 2014-03-06 | 2014-06-11 | 南京理工大学 | Cloud computing network resource distribution method based on service quality requirement drive |
CN104144135A (en) * | 2014-07-25 | 2014-11-12 | 电子科技大学 | Resource distribution method and survivability resource distribution method used for multicast virtual network |
CN104270421A (en) * | 2014-09-12 | 2015-01-07 | 北京理工大学 | Multi-user cloud platform task scheduling method supporting bandwidth guarantee |
Family Cites Families (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US8671407B2 (en) * | 2011-07-06 | 2014-03-11 | Microsoft Corporation | Offering network performance guarantees in multi-tenant datacenters |
CN102331948A (en) * | 2011-09-01 | 2012-01-25 | 杭州湾云计算技术有限公司 | Resource state-based virtual machine structure adjustment method and adjustment system |
US9071609B2 (en) * | 2012-10-08 | 2015-06-30 | Google Technology Holdings LLC | Methods and apparatus for performing dynamic load balancing of processing resources |
CN103067524A (en) * | 2013-01-18 | 2013-04-24 | 浪潮电子信息产业股份有限公司 | Ant colony optimization computing resource distribution method based on cloud computing environment |
EP2770431B1 (en) * | 2013-02-25 | 2017-01-04 | Telefonica S.A. | System and method to trigger cross-layer optimizations in a network |
CN104202263B (en) * | 2014-09-12 | 2017-03-29 | 北京理工大学 | A kind of multi-tenant data midbandwidth resource fairness distribution method |
-
2015
- 2015-08-21 CN CN201510520914.5A patent/CN105159780B/en not_active Expired - Fee Related
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102223281A (en) * | 2011-06-24 | 2011-10-19 | 清华大学 | Method for remapping resource demand dynamic change based on virtual network |
CN102904794A (en) * | 2012-09-27 | 2013-01-30 | 北京邮电大学 | Method and device for mapping virtual network |
CN103218175A (en) * | 2013-04-01 | 2013-07-24 | 无锡成电科大科技发展有限公司 | Multi-tenant cloud storage platform access control system |
CN103856563A (en) * | 2014-03-06 | 2014-06-11 | 南京理工大学 | Cloud computing network resource distribution method based on service quality requirement drive |
CN104144135A (en) * | 2014-07-25 | 2014-11-12 | 电子科技大学 | Resource distribution method and survivability resource distribution method used for multicast virtual network |
CN104270421A (en) * | 2014-09-12 | 2015-01-07 | 北京理工大学 | Multi-user cloud platform task scheduling method supporting bandwidth guarantee |
Also Published As
Publication number | Publication date |
---|---|
CN105159780A (en) | 2015-12-16 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN105103506B (en) | For the method and system for the non-homogeneous bandwidth request allocation bandwidth in system for cloud computing | |
CN104270421B (en) | A kind of multi-tenant cloud platform method for scheduling task for supporting Bandwidth guaranteed | |
CN104683488B (en) | Streaming computing system and its dispatching method and device | |
CN102098686A (en) | 'Three-layer and one-experience' evaluation model for mobile communication network optimization | |
CN102780574B (en) | The collocation method of the office data of service-oriented, device and check method, device | |
CN111106999A (en) | IP-optical network communication service joint distribution method and device | |
CN105872079A (en) | Chain balancing method based on domain name system (DNS) | |
CN108984125A (en) | A kind of method, apparatus and storage medium of resource allocation | |
CN108062243A (en) | Generation method, task executing method and the device of executive plan | |
DE202014010920U1 (en) | Traffic design for large-scale data center networks | |
CN111162865A (en) | Virtual optical network mapping method for sensing fragments in space division multiplexing elastic optical network | |
CN109995580A (en) | VN mapping method based on GA_PSO hybrid algorithm in 5G network slice | |
Zhang et al. | SLA aware cost efficient virtual machines placement in cloud computing | |
CN104778528A (en) | Method and system for obtaining smart city evaluation indexes by utilizing big data | |
CN110519090A (en) | A kind of accelerator card distribution method, system and the associated component of FPGA cloud platform | |
CN103384251A (en) | Multi-class safety service combinable safety network construction method and device | |
CN111541628A (en) | Power communication network service resource allocation method and related device | |
CN111092827B (en) | Power communication network resource allocation method and device | |
CN114186750A (en) | Routing planning method and device for hierarchical requirements of power business of smart power grid | |
CN105159780B (en) | The high availability virtual network mapping method and device of oriented multilayer time cloud application | |
CN105072049B (en) | The resource allocation methods and device of the multi-level flexible application in data-oriented center | |
CN108711074A (en) | Business sorting technique, device, server and readable storage medium storing program for executing | |
CN110677306A (en) | Network topology replica server configuration method and device, storage medium and terminal | |
CN107360480B (en) | A kind of ONU grouping planing method based on load balancing | |
CN107528742A (en) | A kind of virtual machine deployment method of facing cloud data center network optimization |
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 | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20180720 Termination date: 20210821 |
|
CF01 | Termination of patent right due to non-payment of annual fee |