CN109617811B - Rapid migration method for mobile application in cloud network - Google Patents

Rapid migration method for mobile application in cloud network Download PDF

Info

Publication number
CN109617811B
CN109617811B CN201910084329.3A CN201910084329A CN109617811B CN 109617811 B CN109617811 B CN 109617811B CN 201910084329 A CN201910084329 A CN 201910084329A CN 109617811 B CN109617811 B CN 109617811B
Authority
CN
China
Prior art keywords
migration
path
service function
request
network
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
Application number
CN201910084329.3A
Other languages
Chinese (zh)
Other versions
CN109617811A (en
Inventor
孙健
周润
孙罡
虞红芳
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.)
University of Electronic Science and Technology of China
Original Assignee
University of Electronic Science and Technology of China
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 University of Electronic Science and Technology of China filed Critical University of Electronic Science and Technology of China
Priority to CN201910084329.3A priority Critical patent/CN109617811B/en
Publication of CN109617811A publication Critical patent/CN109617811A/en
Application granted granted Critical
Publication of CN109617811B publication Critical patent/CN109617811B/en
Expired - Fee Related legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L45/00Routing or path finding of packets in data switching networks
    • H04L45/12Shortest path evaluation
    • H04L45/125Shortest path evaluation based on throughput or bandwidth
    • 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
    • H04L41/50Network service management, e.g. ensuring proper service fulfilment according to agreements
    • H04L41/5041Network service management, e.g. ensuring proper service fulfilment according to agreements characterised by the time relationship between creation and deployment of a service
    • H04L41/5054Automatic deployment of services triggered by the service manager, e.g. service implementation by automatic configuration of network components
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00Network arrangements or protocols for supporting network services or applications
    • H04L67/50Network services
    • H04L67/56Provisioning of proxy services
    • H04L67/563Data redirection of data network streams

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Mobile Radio Communication Systems (AREA)
  • Data Exchanges In Wide-Area Networks (AREA)

Abstract

The invention discloses a rapid migration method in mobile application in a cloud network, which aims to improve the service quality of users and realize seamless migration of services, shortens the migration time of SFC to the greatest extent, and comprehensively considers the costs of migration operation overhead, resource consumption and the like in order to ensure the success and rationality of service migration.

Description

Rapid migration method for mobile application in cloud network
Technical Field
The invention belongs to the technical field of data migration, and particularly relates to a rapid migration method of mobile applications in a cloud network.
Background
With the rapid development of wireless communication technology and the rapid popularization of various mobile devices, the number of mobile service requests is increasing dramatically, and more users want to be able to access their service requests flexibly and conveniently at any time and any place. Virtualization of mobile network functions and application of mobile cloud computing further facilitate this, but today with an unprecedented proliferation of mobile applications, operators must reconsider management schemes for mobile services, especially for long distance and/or high speed moving users (such as vehicles and sub-high speed trains). When a user moves, the SFC needs to change a service path in time according to a time-varying access position of the user to ensure the continuation of the service, which provides requirements and challenges for the NFV-based SFC to be capable of dynamically changing and having self-adaptability.
Network function virtualization lays the foundation for virtualization of service functions and execution thereof on virtual machines. Any Service request may be represented by a Service Function Chain (SFC), which is a set of Virtual Network Functions (VNFs) that perform functions according to a given order. The operation of VNFs requires the instantiation of VNF instances (VNFI), which are typically software modules executing on VMs.
Similar to mobile telephony, people are more intolerant of sudden interruptions in the call than the inability to reach the initial call. Service interruption not only reduces user experience and makes users feel happy and complain, but also may interrupt important file transmission, service request, business negotiation and cooperative transaction, and miss important cooperative opportunity or cause huge economic loss. More and more users therefore want to access and continue their requests conveniently and quickly, whenever and wherever they move or are stationary. That is, when the user moves, the SFC needs to change the service path in time according to the time-varying access position of the user, and readjust the deployment policy of the SFC to ensure the continuation of the service, which provides requirements and challenges for the NFV-based SFC to be capable of dynamically changing and having self-adaptability. The SFC migration technology can solve the problem to a certain extent. Unfortunately, most of the current research on SFC migration focuses on how to implement migration, and also only stays at the research level of static SFC, and ignores the fundamental purpose of migration: the online service is quickly recovered, and the sudden interruption of the service is avoided. To avoid the ongoing service interruption when the user moves outside the service area of the original ap server, the SFC of the user should be migrated quickly to ensure the service continuity when the user moves and achieve seamless service migration (i.e. without interrupting the ongoing service).
In order to process load balance of underlying network nodes and implement migration of VNFs, an SFC deployment method is developed to respond to constantly changing working loads and integrate VNFIs on servers as few as possible, so that energy consumption is reduced; this approach only considers offline migration of static SFCs. The static SFC deployment scheme makes the path between the user and the VNFs suboptimal, which not only causes unnecessary bandwidth consumption, but also reduces the user experience, and has no practical application value. In addition, a migration policy of VNFIs is proposed, which considers investment loss due to QoS degradation during migration and loss of information to users due to information loss occurring during migration, with the goal of minimizing energy consumption and revenue loss due to QoS degradation; the method only tries to transfer VNFIs to the servers as few as possible, reduces energy consumption, does not consider transfer requirements brought by online movement of users at all, cannot meet the requirements in the present day when the movement requests are rapidly increased, and therefore has no great research significance. The scholars also provide how to implement SFC deployment and adjustment in a dynamic environment, and build an algorithm to re-deploy the user SFCs in the server to balance resource consumption and operation overhead, so as to jointly optimize the deployment of the new user SFCs and the adjustment of the user SFCs in the service center; although this method considers the mobility problem of the online users, the system will all redeploy and adjust all SFCs in the network whenever there is a new user access request or a new user mobility request, unfortunately this method is done at the expense of huge time complexity, it is not compensated in real application, and the robustness of the method is very poor, when the network state at the moment is not enough to support the access request of all new users at the current moment and the migration request of users in service, the whole algorithm system will break down.
Disclosure of Invention
Aiming at the defects in the prior art, the rapid migration method for the mobile application in the cloud network provided by the invention solves the problems of large bandwidth consumption, long time consumption and complex operation in the traditional migration method.
In order to achieve the purpose of the invention, the invention adopts the technical scheme that: a method for rapidly migrating mobile applications in a cloud network comprises the following steps: a method for rapidly migrating mobile applications in a cloud network comprises the following steps:
s1, checking the access state of the user request in the cloud network in real time;
s2 service function request SR in access state when user requestsmWhen the position of the mobile terminal is changed, the service function request SR is acquiredmIs moved source point smAnd destination point t after movementmAn access location;
s3 underlying network N of cloud networkGWherein the deployment point of the last virtual network function VNF in the user request SFC is determined
Figure GDA0002370807510000037
And the destination point t after movingmShortest migration path therebetween
Figure GDA0002370807510000031
And proceeds to step S4;
s4, judging the shortest migration path
Figure GDA0002370807510000032
Whether the access is available;
if yes, go to step S5;
if not, go to step S7;
s5, converting the shortest transfer path
Figure GDA0002370807510000033
Selected as service function request SRmCorresponding optimal migration pathmStep S6;
s6, calculating the shortest migration path
Figure GDA0002370807510000034
Corresponding bandwidth consumption is set, a time-bandwidth balance parameter k is set, and judgment is carried out
Figure GDA0002370807510000035
Whether the result is true or not;
wherein the bandwidth consumption comprises the shortest migration path
Figure GDA0002370807510000036
Bandwidth consumption of the corresponding entire migration path
Figure GDA0002370807510000041
And shortest migration path
Figure GDA0002370807510000042
Bandwidth consumption of corresponding end migration path
Figure GDA0002370807510000043
If yes, go to step S8;
if not, go to step S7;
s7 underlying network N of cloud networkGIn determining a service function request SRmMiddle shifted source point smTo the moved destination point tmShortest migration path of
Figure GDA0002370807510000044
And proceeds to step S9;
s8, forming sub-paths according to the selected components
Figure GDA0002370807510000045
Determining a service function request SRmCorresponding optimal migration pathmAnd proceeds to step S13;
s9, judging the shortest migration path
Figure GDA0002370807510000046
Whether the access is available;
if yes, go to step S10;
if not, go to step S11;
s10, determining the shortest migration path
Figure GDA0002370807510000047
Total migration overhead in (1)
Figure GDA0002370807510000048
And shortest migration path
Figure GDA0002370807510000049
Total migration overhead in (1)
Figure GDA00023708075100000410
Judgment of
Figure GDA00023708075100000411
Whether the result is true or not;
if yes, return to step S8;
if not, go to step S12;
s11, detecting the user request SFC in service in the cloud network, revoking the expired user request SFC, and returning the physical resources
Figure GDA00023708075100000412
And returns to step S1;
s12, converting the shortest transfer path
Figure GDA00023708075100000413
Request SR as a service functionmCorresponding optimal migration pathmAnd proceeds to step S13;
s13, determining the optimal migration pathmFulfilling a current service function request SRmThe fast migration of the mobile application is realized.
The invention has the beneficial effects that:
(1) the migration is efficient and rapid. Aiming at the uncertainty and flexibility of user position movement, the P1-MP migration scheme provided by the method can realize the rapid online migration of the SFC, firstly, the user experience is considered, the migration time overhead is minimized, and the seamless connection of the service request is realized.
(2) The utilization rate of network resources is high. The method considers the dynamic deployment and migration of the online SFCs, has respective online survival time for service requests of the SFCs which arrive at an irregular time, and cancels the service requests when the service requests arrive at the time, and simultaneously releases occupied resources, thereby continuing to use the SFCs in the future.
(3) The migration success rate is high. The invention provides a characteristic that a P2-MP migration scheme complements the P1-MP migration scheme to specially optimize migration time, comprehensively considers the bearing capacity of a physical network and the limitation of resources, and when the P1-MP migration scheme cannot return an effective migration path or the bandwidth overhead of the obtained path is too large, in order to ensure the success of migration, the P1-MP migration scheme searches a reasonable migration path with the lowest cost for SFC.
Drawings
Fig. 1 is a flowchart of an implementation of a fast migration method for a cloud network mobile application according to the present invention.
FIG. 2 is a diagram illustrating the determination of the shortest migration path in the present invention
Figure GDA0002370807510000051
The method implements a flow diagram.
FIG. 3 is a diagram illustrating the determination of the shortest migration path in the present invention
Figure GDA0002370807510000052
The method implements a flow diagram.
Fig. 4 is a schematic diagram of two situations of a user request in service when the location of the user is changed in the present invention.
Detailed Description
The following description of the embodiments of the present invention is provided to facilitate the understanding of the present invention by those skilled in the art, but it should be understood that the present invention is not limited to the scope of the embodiments, and it will be apparent to those skilled in the art that various changes may be made without departing from the spirit and scope of the invention as defined and defined in the appended claims, and all matters produced by the invention using the inventive concept are protected.
As shown in fig. 1, a method for fast migrating a mobile application in a cloud network includes the following steps:
s1, checking the access state of the user request in the cloud network in real time;
s2 service function request SR in access state when user requestsmWhen the position of the mobile terminal is changed, the service function request SR is acquiredmIs moved source point smAnd destination point t after movementmAn access location;
s3 underlying network N of cloud networkGWherein the deployment point of the last virtual network function VNF in the user request SFC is determined
Figure GDA00023708075100000613
And the destination point t after movingmShortest migration path therebetween
Figure GDA0002370807510000061
And proceeds to step S4;
s4, judging the shortest migration path
Figure GDA0002370807510000062
Whether the access is available;
if yes, go to step S5;
if not, go to step S7;
s5, converting the shortest transfer path
Figure GDA0002370807510000063
Selected as service function request SRmCorresponding optimal migration pathmStep S6;
s6, calculating the shortest migration path
Figure GDA0002370807510000064
Corresponding bandwidth consumption is set, a time-bandwidth balance parameter k is set, and judgment is carried out
Figure GDA0002370807510000065
Whether the result is true or not;
wherein the bandwidth consumption comprises the shortest migration path
Figure GDA0002370807510000066
Bandwidth consumption of the corresponding entire migration path
Figure GDA0002370807510000067
And shortest migration path
Figure GDA0002370807510000068
Bandwidth consumption of corresponding end migration path
Figure GDA0002370807510000069
If yes, go to step S8;
if not, go to step S7;
s7 underlying network N of cloud networkGIn determining a service function request SRmMiddle shifted source point smTo the moved destination point tmShortest migration path of
Figure GDA00023708075100000610
And proceeds to step S9;
s8, forming sub-paths according to the selected components
Figure GDA00023708075100000611
Determining a service function request SRmCorresponding optimal migration pathmAnd proceeds to step S13;
s9, judging the shortest migration path
Figure GDA00023708075100000612
Whether the access is available;
if yes, go to step S10;
if not, go to step S11;
s10, determining the shortest migration path
Figure GDA0002370807510000071
Total migration overhead in (1)
Figure GDA0002370807510000072
And shortest migration path
Figure GDA0002370807510000073
Total migration overhead in (1)
Figure GDA0002370807510000074
Judgment of
Figure GDA0002370807510000075
Whether the result is true or not;
if yes, return to step S8;
if not, go to step S12;
s11, detecting the user request SFC in service in the cloud network, revoking the expired user request SFC, and returning the physical resources
Figure GDA0002370807510000076
And returns to step S1;
s12, converting the shortest transfer path
Figure GDA0002370807510000077
Request SR as a service functionmCorresponding optimal migration pathmAnd proceeds to step S13;
s13, determining the optimal migration pathmFulfilling a current service function request SRmThe fast migration of the mobile application is realized.
User service request SR ═ (S)F,LF,RFS, t) denotes a user request (SFC), each SFC consisting of a series of Virtual Network Functions (VNFs) with a certain connection order, each VNF representing a network function (e.g. intrusion detection System IDS, Firewall, etc.)
The service function request SR in the above step S2mIs SRm=(SF,LF,RF,sm,tm);
Wherein S isFRequesting a set of several virtual network functions VNFs in the SFC for a user, and SF={f1,f2,...f|SF|-SF | is the number of virtual network functions VNF;
LFfor connecting virtual links between two adjacent virtual network functions VNFs, and LF={l1,l2,...l|LF|H, | LF | is the number of links for a user to request SFC;
RFthe total amount of resources required to request the SFC for the entire user;
smthe source point after the user moves;
tmafter moving for the userA destination point;
the underlying network is composed of various physical devices (e.g., servers and routers) and connecting links (e.g., optical fibers), each with corresponding resource attributes, so we use an undirected graph NGIndicates an underlying network (physical network), and therefore the underlying network N in step S3GComprises the following steps:
NG={VG,EG,RG}
wherein, VGIs an underlying network NGA set of nodes in (1); vG={v1,v2,...v|VG|H, | VG | is underlying network NGThe number of nodes in;
EGis a VGIs an underlying network NGSet of links in (1), EG={E1,E2,...E|EG|}, | EG | is underlying network NGThe number of links in (1);
RGis an underlying network NGThe total amount of resources that can be provided.
For ease of description and understanding, we set:
1. one physical node may deploy different types of VNFs;
2. one VNF can only be deployed in the same physical node;
3. multiple VNFs in the same SFC are unavailable to one physical node;
4. VNFs of different SFCs allow deployment on the same physical node if the resources are sufficient.
In the method of the invention, an optimal migration path can be selected for the SFC according to service requests of different users and current network use conditions, firstly, the method of the invention continuously checks the access state of the current user request, judges whether a new user request is accessed or whether the position of the user in service is changed, if the two conditions occur simultaneously, in order to ensure the user service in service to continue, the SFC which is rapidly migrated in the service user is preferentially selected, and then the request of the new user is deployed, otherwise, the processing is carried out according to the sequence of the appearing requests, therefore, the service function request SR in the access state of the user request ismWhen the position is changed, as shown in fig. 2, step S3 specifically includes:
s31, obtaining service function request SRmLast virtual network function point f in (1)lastTo the moved destination point tmBandwidth requirement of
Figure GDA0002370807510000081
S32, traversing underlying network NGIn each link, the underlying network N is deletedGBandwidth of the remaining resources
Figure GDA0002370807510000082
Less than bandwidth requirement
Figure GDA0002370807510000083
Corresponding link etTo obtain the underlying network NGSubfigure N ofSG
Wherein e ist∈EG;NSG=NG-etTo obtain NSGThe purpose of the method is to reduce the search space of the shortest path and save time;
s33, in sub-figure NSGWherein the deployment point of the last virtual network function VNF in the user request SFC is determined
Figure GDA00023708075100000919
And the destination point t after movingmShortest migration path therebetween
Figure GDA0002370807510000091
Wherein, subfigure NSGEach link in (b) satisfies the last virtual network function point flastTo the moved destination point tmBandwidth requirement of
Figure GDA0002370807510000092
The step S4 is specifically:
determining the shortest migration path
Figure GDA0002370807510000093
Corresponding total migration overhead
Figure GDA0002370807510000094
Whether the result is true or not;
if yes, the shortest migration path
Figure GDA0002370807510000095
If not, go to step S7;
if not, the shortest migration path
Figure GDA0002370807510000096
If yes, go to step S5;
wherein the total migration overhead
Figure GDA0002370807510000097
The calculation formula of (2) is as follows:
Figure GDA0002370807510000098
in the formula (I), the compound is shown in the specification,
Figure GDA0002370807510000099
is the shortest migration path
Figure GDA00023708075100000910
Corresponding bandwidth consumption;
Figure GDA00023708075100000911
is the shortest migration path
Figure GDA00023708075100000912
Corresponding seek time consumption.
The bandwidth consumption of the entire migration path in the above step S6
Figure GDA00023708075100000913
The calculation formula of (2) is as follows:
Figure GDA00023708075100000914
in the formula (I), the compound is shown in the specification,
Figure GDA00023708075100000915
requesting an SR for a service functionmBandwidth consumption before migration;
Figure GDA00023708075100000916
requesting an SR for a service functionmBandwidth consumption of the end path prior to migration;
Figure GDA00023708075100000917
to make the shortest migration path
Figure GDA00023708075100000918
As the optimal migration pathmService function request SR in migration processmThe bandwidth consumption of (2).
As shown in fig. 3, the step S7 specifically includes:
s71 traversing service function request SRmAnd determining a set S of virtual network functions VNFFVirtual network function point f in (1)iWhether it is an intermediate function point;
wherein f isi∈SF
If yes, go to step S72;
if not, go to step S74;
s72, traversing the current underlying network NGEach node in (2) determining the underlying network NGNode v inmAnd its corresponding shortest path, supbpath, and go to step S73;
wherein v ism∈VGNode vmIs an underlying network NGA contracted migration destination point v at middle distancejNearest and satisfying virtual network function point fiA node of the resource demand of (c);
s73, connecting the node vmAs a virtual network function point fiAnd node v is migrated tomAnd the corresponding shortest path supbpath is added to the shortest migration path as an intermediate path
Figure GDA0002370807510000101
In the get service function request SRmCorresponding shortest migration path
Figure GDA0002370807510000102
Proceeding to step S9;
s74, traversing underlying network NGDetermining a passing deployment point
Figure GDA0002370807510000108
The shortest path of (1), and proceeds to step S75;
wherein the deployment point
Figure GDA0002370807510000109
Satisfy the last virtual network function point flastThe resource requirements of (1);
s75, point to be deployed
Figure GDA00023708075100001010
As the last virtual network function point flastMigrate the sink point and deploy the point
Figure GDA00023708075100001011
And the corresponding shortest path supbpath is added to the shortest migration path as an intermediate path
Figure GDA0002370807510000103
In the get service function request SRmCorresponding shortest migration path
Figure GDA0002370807510000104
The process advances to step S9.
The optimum transition in the above step S8Path moving pathmComprises the following steps:
Figure GDA0002370807510000105
in the formula, pathbeforeRequesting SR for performing a service functionmA deployment path before migration;
Figure GDA0002370807510000106
requesting an SR for a service functionmA deployment path of the end path prior to migration;
Figure GDA0002370807510000107
requesting a deployment point of a last virtual network function, VNF, in the SFC for a user
Figure GDA00023708075100001012
And the destination point t after movingmThe migration path therebetween.
The step S9 is specifically:
determining the shortest migration path
Figure GDA0002370807510000111
Corresponding total migration overhead
Figure GDA0002370807510000112
Whether the result is true or not;
if yes, the shortest migration path
Figure GDA0002370807510000113
If not, go to step S11;
if not, the shortest migration path
Figure GDA0002370807510000114
If yes, go to step S10;
wherein the total migration overhead
Figure GDA0002370807510000115
The calculation formula of (2) is as follows:
Figure GDA0002370807510000116
in the formula (I), the compound is shown in the specification,
Figure GDA0002370807510000117
is the shortest migration path
Figure GDA0002370807510000118
Corresponding bandwidth consumption;
Figure GDA0002370807510000119
is the shortest migration path
Figure GDA00023708075100001110
Corresponding migration operation consumption;
Figure GDA00023708075100001111
is the shortest migration path
Figure GDA00023708075100001112
Corresponding seek time consumption;
Figure GDA00023708075100001113
is the shortest migration path
Figure GDA00023708075100001114
Corresponding time consumption.
In the above step S13, when the service function request SR is mademBest migration path ofmIs the shortest migration path
Figure GDA00023708075100001115
Service function request SR at corresponding migration pathmThe migration method specifically comprises the following steps:
a1, traversing the optimal migration pathmAnd issuing a service function request SRmFormer deployment path pbeforeThe deployment situation of each link in the network;
determine the optimal migration pathmAnd issuing a service function request SRmFormer deployment path pbeforeWhether there is the same virtual link l innCorresponding to the same link et
Wherein ln∈LF,et∈EG
If yes, go to step A2;
if not, go to step A3;
a2, directly according to the optimal migration pathmDirect service function request SRmMigration of (2);
the resource release and deduction are not repeated in the migration process;
a3, from underlying network NGMiddle deduction pathmIn (1)
Figure GDA00023708075100001116
Corresponding bandwidth resource is released, and the user sends out service function request SRmFront side
Figure GDA00023708075100001117
Occupied bandwidth resource and then according to the optimal migration pathmDirect service function request SRmMigration of (2);
when the service function requests the SRmBest migration path ofmIs the shortest migration path
Figure GDA0002370807510000121
When, the service function requests the SRmThe migration method specifically comprises the following steps:
b1 traversing the optimal migration path
Figure GDA0002370807510000122
And issuing a service function request SRmFormer deployment path pbeforeDeployment of each node and link in the networkDetermining the optimal migration path
Figure GDA0002370807510000123
And issuing a service function request SRmFormer deployment path pbeforeWhether there is the same virtual network function point fiCorresponding to the same node vmOr with the same virtual link lnCorresponding to the same link etThe case (1);
if yes, go to step B2;
if not, go to step B3;
b2, directly according to the optimal migration path
Figure GDA0002370807510000124
Direct service function request SRmMigration of (2);
the resource release and deduction are not repeated in the migration process;
b3, slave underlay network NGMiddle deduction
Figure GDA0002370807510000125
Corresponding node and bandwidth resource
Figure GDA0002370807510000126
And releases the user to send out the service function request SRmFront pathbeforeOccupied node and its corresponding bandwidth resource, then according to optimum transfer path
Figure GDA0002370807510000127
Direct service function request SRmMigration of (2);
wherein the bandwidth resources
Figure GDA0002370807510000128
Comprises the following steps:
Figure GDA0002370807510000129
in the formula (I), the compound is shown in the specification,
Figure GDA00023708075100001210
path for optimal migration pathmIs the shortest migration path
Figure GDA00023708075100001211
Service function request SR at corresponding migration pathmBandwidth resources consumed during migration;
the above step S10;
returned physical resources
Figure GDA00023708075100001212
Comprises the following steps:
Figure GDA00023708075100001213
in the formula (I), the compound is shown in the specification,
Figure GDA00023708075100001214
is an underlying network NGThe remaining physical resources;
Figure GDA00023708075100001215
requesting physical resources occupied by the SFCs for all users that have not migrated;
Figure GDA00023708075100001216
as the optimal migration pathmIs the shortest migration path
Figure GDA00023708075100001217
When the corresponding migration path is used, all users request physical resources occupied by the SFCs;
Figure GDA0002370807510000131
as the optimal migration pathmIs the shortest migration path
Figure GDA0002370807510000132
And when the corresponding migration path is used, all the users request the physical resources occupied by the SFCs.
In order to better perform quick migration on user requests, two migration path determining methods in the method of the invention can be divided into a P1-MP migration scheme and a P2-MP migration scheme, wherein the P1-MP migration scheme is used for determining the shortest migration path
Figure GDA0002370807510000133
Further determine the optimal migration pathmThe P2-MP migration scheme is to determine the shortest migration path
Figure GDA0002370807510000134
Further determine the optimal migration pathmThe method of (1);
the P1-MP migration scheme is a rapid migration method, and aims to put the user experience at the head and shorten the SR to the greatest extentmThe migration time of the SFC is shortened, and the rapid and efficient migration of the SFC is realized; the P1-MP migration scheme is rapidly embodied in the following points:
1. at the beginning of the way-finding, the P1-MP migration scheme first targets the underlying network NGThe remaining bandwidth resources in (1) are judged, and the bandwidth resources which do not meet the requirement f are removedlast→tmLink e of bandwidth requirementtObtaining subgraph NSGThe search space of the way-finding algorithm is reduced, and the way-finding time is greatly saved;
2. P1-MP migration scheme directly searches deployment point of last VNF of SFC
Figure GDA0002370807510000138
And the destination point t after movingmMigration paths without changing the deployment path before the SFC;
3. VNF deployment is not needed, so that the resource use condition of the node is not considered in the routing process, and only bandwidth resource constraint is considered;
4. migration of VNF does not need to be considered, so that release of original deployment node resources and deduction of new migration node resources are carried out differently, and only f needs to be addedlast→tmMigrating to a new path
Figure GDA0002370807510000135
Namely, the migration time is shortened to a great extent.
Since the P1-MP migration scheme does not change all deployments before the last VNF of the SFC, when f is searchedlast→tmIs transferred to
Figure GDA0002370807510000136
While, the whole SRmIs also determined as
Figure GDA0002370807510000137
Thus, the whole SR is omittedmSearching the optimal migration path, releasing the node/bandwidth resource in the original deployment path and deducting the node/bandwidth resource in the new migration path, and only considering the last section of the virtual link flast→tmIs being migrated.
Of more interest to the P1-MP migration scheme is minimizing SRmWhen the user position moves, the P1-MP migration scheme only needs to migrate the access point of the user, and then is SRmS inm→f1Or flast→tmFinding nearest migration paths
Figure GDA0002370807510000141
The deployment position of each VNF in the SFC is not changed, so that the P1-MP migration scheme is suitable for SFCs with important service requests or higher delay requirements;
for the jth SFC:
the migration overhead of the P1-MP migration scheme is:
Figure GDA0002370807510000142
in the formula (I), the compound is shown in the specification,
Figure GDA0002370807510000143
and
Figure GDA0002370807510000144
the total migration overhead, bandwidth consumption and seek time consumption of the P1-MP migration scheme, respectively. For the underlying network, the occupation of the node resources by the SFC is not changed by migration, so that the cost of the node resources is not considered when the migration cost is calculated.
The bandwidth consumption of the P1-MP migration scheme is:
Figure GDA0002370807510000145
wherein, s → t ∈ EG,flast→t∈EG,flast→tm∈EG,
Figure GDA0002370807510000146
Is a link etBandwidth consumption of (d);
the path-finding time overhead of the P1-MP migration scheme is:
Figure GDA0002370807510000147
Figure GDA0002370807510000148
deployment point for user requesting last virtual network function VNF in SFC in P1-MP migration scheme
Figure GDA0002370807510000149
And the destination point t after movingmOverhead of inter-path seek time;
the total resource consumption of the P1-MP migration scheme is:
Figure GDA0002370807510000151
wherein, s → t, flast→t,
Figure GDA0002370807510000152
Figure GDA0002370807510000153
For the j-th SFC, node resources occupied by VNFs are more in the migration path in the P1-MP migration scheme,
Figure GDA0002370807510000154
is a VNF fiNode resource occupancy.
The P2-MP migration scheme determines the optimal migration path on the premise that the P2-MP migration scheme is unreachable, and the P2-MP migration scheme considers the SRmAccording to the overall situation of the current underlying network NGThe medium resource usage is SRmFinding a migration path with optimal resources, and if the position of a user is enlarged and changed, not using the SRmThe overall migration will cause the SFC mapping path to be too long and the network resource to be occupied too high, thereby lengthening the transmission delay between the source node and the destination node and increasing the network load. For this reason, the migration scheme of P2-MP aims to satisfy the mobile requirement of the user and comprehensively considers SRmThe occupation situation of the bottom layer resource is SRmSearching a minimum migration path of comprehensive overhead (including migration cost and transmission cost) to ensure the continuability and reasonability of the service after the user position is changed, which is mainly embodied in the following points:
1. due to the limitation of bandwidth resources, NGMay have none in
Figure GDA0002370807510000155
The reachable path of (1), at which point the P2-MP migration scheme needs to be invoked;
2. a time and bandwidth balance parameter k is set in the P1-MP migration scheme, if the fast recovery of the P1-MP migration scheme must sacrifice more bandwidth, consider NGThe resources of (1) are tense, and a P2-MP migration scheme can be called;
3. in the face of the SFC after the position change, the migration scheme of P2-MP is according to the current underlying network NGIs SRmRe-searching for a piece sm→tmShortest path, and therefore minimal bottom layer resource occupation, SRmMigration ofThe matching degree of the path and the current network state is highest;
the P2-MP migration scheme is complementary to the P1-MP migration scheme, and the P2-MP migration scheme comprehensively considers the resource overhead, migration operation overhead, path-finding time overhead and migration time overhead of a migration path and migrates the whole SFC. It is obvious that the P2-MP migration scheme aims to realize the trade-off between user service quality and resource consumption, and reduce SR while reducing service delay as much as possiblemOccupation of network resources. Therefore, the P2-MP migration scheme is more suitable for the case where the user has less strict requirement on the delay and the network resources are more tight.
For the jth SFC:
the total migration overhead of the P2-MP migration scheme is:
Figure GDA0002370807510000161
in the formula (I), the compound is shown in the specification,
Figure GDA0002370807510000162
fi-1→fi∈SRm,
Figure GDA0002370807510000163
Figure GDA0002370807510000164
is the shortest migration path
Figure GDA0002370807510000165
Corresponding bandwidth consumption;
Figure GDA0002370807510000166
is the shortest migration path
Figure GDA0002370807510000167
Corresponding migration operation consumption;
Figure GDA0002370807510000168
is the shortest migration path
Figure GDA0002370807510000169
Corresponding seek time consumption;
Figure GDA00023708075100001610
is the shortest migration path
Figure GDA00023708075100001611
Corresponding migration time consumption.
The bandwidth requirements of the P2-MP migration scheme are:
Figure GDA00023708075100001612
in the formula (I), the compound is shown in the specification,
Figure GDA00023708075100001613
the migration operation overhead of the P2-MP migration scheme is as follows:
Figure GDA0002370807510000171
wherein the content of the first and second substances,
Figure GDA0002370807510000172
migration operation overhead for a single VNF;
the path-finding time overhead of the P2-MP migration scheme is:
Figure GDA0002370807510000173
in the formula (f)i-1→fi∈SRm,
Figure GDA00023708075100001710
Figure GDA0002370807510000174
And
Figure GDA0002370807510000175
migration protocol for P2-MP for SFCs, respectivelym→tm,fi-1→fiThe seek time overhead of;
the migration time overhead of the P2-MP migration scheme is:
Figure GDA0002370807510000176
in the formula (I), the compound is shown in the specification,
Figure GDA0002370807510000177
total resource consumption of the P2-MP migration scheme:
Figure GDA0002370807510000178
in the formula (I), the compound is shown in the specification,
Figure GDA0002370807510000179
the relationship between the P1-MP migration scheme and the P2-MP migration scheme is shown as follows:
Figure GDA0002370807510000181
when the system detects the user position movement, the method of the invention gives priority to the P1-MP migration scheme, and only when the P1-MP migration scheme can not find an effective path or migration overhead
Figure GDA0002370807510000182
Only if so, the P2-MP migration scheme is enabled, whereas for the P1-MP migration scheme: involving no migration of VNF, it
Figure GDA0002370807510000183
Are all zero; however, for the P2-MP migration scheme:
Figure GDA0002370807510000184
and the whole link sm→tmMust have a path-finding time overhead greater than sm→f1Or flast→tmThe seek time overhead of; it is obvious that the bandwidth consumption of the P2-MP migration scheme is definitely less than that of the P1-MP migration scheme, because for dynamically changing network resources, the P2-MP migration scheme is SR according to the use condition of the network at the current momentmThe shortest migration path sought.
In one embodiment of the invention, the method for SR is also providedmWhen migration is carried out, a calculation method for the resource occupied by the underlying network is adopted;
for the underlying network, there may be SFCs of 3 users at a certain time, that is, when no location movement occurs, service requests for migration implemented by using the P1-MP migration scheme and the P2-MP migration scheme, and their occupation of physical network resources is:
Figure GDA0002370807510000185
in the formula of Unm+Up1+Up2=Unm+Um=US,
Figure GDA0002370807510000191
Respectively, no migration of SFC, use of p1MP migration scheme, Using p2-resource consumption of MP migration scheme; u shapenm,Up1,Up2,UmRespectively, no migration, use p1MP migration scheme, Using p2-MP migration scheme, number of user requests for which migration takes place;
the total node resource consumption in the underlying network is as follows:
Figure GDA0002370807510000192
in the formula (I), the compound is shown in the specification,
Figure GDA0002370807510000193
p is used without migration of SFC1MP migration scheme, Using p2-node resource requirements of the MP migration scheme;
the total resource link consumption in the underlying network is:
Figure GDA0002370807510000194
in the formula (I), the compound is shown in the specification,
Figure GDA0002370807510000195
p is used without migration of SFC1MP migration scheme, Using p2-bandwidth resource requirements of MP migration scheme.
In an embodiment of the present invention, there are also provided an optimization objective and its constraint conditions when migrating by using the method of the present invention:
among these, the optimization objectives (total overhead for all request migrations) are:
Figure GDA0002370807510000201
the total resource constraint conditions of the optimization target are as follows:
Figure GDA0002370807510000202
in the formula (I), the compound is shown in the specification,
Figure GDA0002370807510000203
is binary number and represents whether the service request SR is successfully deployed in the underlying network
The node resource constraint of the optimization objective is as follows:
Figure GDA0002370807510000204
wherein, i is not equal to j,
Figure GDA0002370807510000205
one VNF point can be deployed on only one physical node; namely, it is
Figure GDA0002370807510000206
During the deployment of the same SFC request, at most one function on one service function chain can be deployed at one physical point n (to avoid the occurrence of a ring), and a server that has been used as a deployment node (forwarding node) can no longer serve as a forwarding node (deployment node) of the same SFC, that is, the same SFC request is deployed at a point n
Figure GDA0002370807510000207
In one embodiment of the invention, an example of fast migration of a computing cloud network mobile application is provided: in a computing cloud network, mobility and uncertainty of mobile users may cause significant degradation of network performance and even interruption of ongoing service requests, and thus, it is difficult to guarantee continuity of service. Service migration has great potential in addressing these issues, determining when, where, and how to migrate these services after user migration and demand changes. In the face of service requests of user movement in service, two solutions of a P1-MP migration scheme and a P2-MP migration scheme are provided, aiming at preferentially considering minimization of service migration delay, improving user service quality (QoS), preventing interruption of ongoing service, and meanwhile comprehensively considering reasonable utilization of operation overhead and underlying network resources of SFC migration within the acceptable delay range of users, so that high efficiency, reasonableness, practicability and optimality of SFC migration are jointly realized. As shown in fig. 4, service requests SFC1, SFC2, SFC3 for 3 users are shown.
At first, the access points of their source and destination nodes are access points respectively
Figure GDA0002370807510000211
Access point 3, access point
Figure GDA0002370807510000212
Access point 4, access point
Figure GDA0002370807510000213
An access point 4; the corresponding deployment path is Access Point 2 → Server
Figure GDA0002370807510000214
Server
3 → access point 3, access point 7 → server
Figure GDA0002370807510000215
Server 5 → server 4 → access point 4, access point 1 → server
Figure GDA0002370807510000216
Server
3 → server 4 → access point 4; after that, the user positions of SFC2 and SFC3 move (SFC1 does not change), the access point of SFC2 source point s2 becomes access point 2, the access point of SFC3 destination point t3 becomes access point 6, and the broken line indicates that the user requests the change of the access point after moving, and therefore, the deployment path corresponding to SR needs to be changed.
In order to quickly restore user service in the face of the online mobile problem of the user, firstly, a P1-MP migration scheme is adopted to find a migration path for an ongoing SFC, the original deployment point of a VNF is not changed by the P1-MP migration scheme, and only the shortest path is found as an SR (scheduling request) between a new access point of the SFC and the nearest placement point of the VNFmMigration path of sub-path to shorten SRmThe seek time of (1). Of course, the target migration path must satisfy the bandwidth resource constraint and the currently corresponding time-bandwidth balance parameter k requirement. Like SFC2, the access point is changed from access point 7 to access point 2, and the P1-MP migration scheme simply changes the mapping sub-path between access point 7 → server 7 to access point
Figure GDA0002370807510000217
The server 7 (the server 2 is only used as a forwarding point here), and the VNF is not migrated, so that the VNF migration operation overhead and migration time overhead are avoided, and the service recovery time is greatly reduced;
when the P1-MP migration scheme can not return to effective migration path or needs a large amount of bandwidth resources to doAt the cost, we resort to the P2-MP migration scheme. Complementary to the P1-MP migration scheme, the P2-MP migration scheme comprehensively considers the total cost of migration time overhead, operation overhead and bandwidth resource overhead and is SRmAnd finding an optimal migration path. E.g., SFC3, the access point changes from access point 4 to access point 6 due to network topology and resource constraints, p2MP will be the whole SRmAnd re-finding the migration path. At this moment, the SFC3 migrates to the path access point 1 → server 2 → server 7 → server 6 → access point 6, where the deployment points of VNF4 and VNF6 have both migrated, which brings about a certain migration time consumption and operation overhead, but the new path of migration is the network matching SR at the current timemThe shortest path of (3), which occupies the least bandwidth resources. Conversely, if we still use the P1-MP migration scheme at this point, a large amount of resource consumption will result between the server 4 → server 6 paths.
In one embodiment of the present invention, a method for implementing and deploying fast migration of a mobile application in a cloud network is provided: the network operator can deploy the rapid migration method in the mobile cloud computing provided by the invention on a control layer in a control router of the SDN, and the SDN control router can schedule a control management function carried by the SDN control router to collect information of the whole underlying network, and acquire the conditions of all nodes and link resources in the network and the connection topology conditions between the nodes. The router can acquire the topology of the whole network and corresponding resource information by the centralized control mode. When the user position in service changes, the invention firstly uses the P1-MP migration scheme to carry out rapid migration on the SFC changed by the source and destination nodes, thereby improving the service quality of the user and ensuring the fund profit and income of operators. Meanwhile, the P2-MP migration scheme considers the specific use condition of the underlying network resources, comprehensively considers the migration overhead of the service, the network resource consumption and the migration time overhead, and searches the optimal path with the lowest cost in the range acceptable by the user.
The invention has the beneficial effects that:
(1) the migration is efficient and rapid. Aiming at the uncertainty and flexibility of user position movement, the P1-MP migration scheme provided by the method can realize the rapid online migration of the SFC, firstly, the user experience is considered, the migration time overhead is minimized, and the seamless connection of the service request is realized.
(2) The utilization rate of network resources is high. The method considers the dynamic deployment and migration of the online SFCs, has respective online survival time for service requests of the SFCs which arrive at an irregular time, and cancels the service requests when the service requests arrive at the time, and simultaneously releases occupied resources, thereby continuing to use the SFCs in the future.
(3) The migration success rate is high. The P2-MP migration scheme complements the characteristic that the P1-MP migration scheme specifically optimizes the migration time, comprehensively considers the bearing capacity of a physical network and the limitation of resources, and when the P1-MP migration scheme cannot return an effective migration path or the bandwidth overhead of the obtained path is too large, in order to ensure the migration success, the P1-MP migration scheme searches a reasonable migration path with the lowest cost for the SFC.

Claims (10)

1. A method for rapidly migrating mobile applications in a cloud network is characterized by comprising the following steps:
s1, checking the access state of the user request in the cloud network in real time;
s2 service function request SR in access state when user requestsmWhen the position of the mobile terminal is changed, the service function request SR is acquiredmIs moved source point smAnd destination point t after movementmAn access location;
s3 underlying network N of cloud networkGWherein a deployment point V of a last virtual network function VNF in a user request SFC is determinedi flastAnd the destination point t after movingmShortest migration path therebetween
Figure FDA0002424634790000011
And proceeds to step S4;
s4, judgment
Figure FDA0002424634790000012
Whether the access is available;
if yes, go to step S5;
if not, go to step S7;
s5, mixing
Figure FDA0002424634790000013
Selected as service function request SRmCorresponding optimal migration pathmStep S6;
s6, calculating
Figure FDA0002424634790000014
Corresponding bandwidth consumption is set, a time-bandwidth balance parameter k is set, and judgment is carried out
Figure FDA0002424634790000015
Whether the result is true or not;
wherein the bandwidth consumption comprises
Figure FDA0002424634790000016
Bandwidth consumption of the corresponding entire migration path
Figure FDA0002424634790000017
And
Figure FDA0002424634790000018
bandwidth consumption of corresponding end migration path
Figure FDA0002424634790000019
If yes, go to step S8;
if not, go to step S7;
s7 underlying network N of cloud networkGIn determining a service function request SRmMiddle shifted source point smTo the moved destination point tmShortest migration path therebetween
Figure FDA00024246347900000110
And proceeds to step S9;
s8, according to the selection
Figure FDA00024246347900000111
Determining a service function request SRmCorresponding optimal migration pathmAnd proceeds to step S13;
s9, judgment
Figure FDA0002424634790000021
Whether the access is available;
if yes, go to step S10;
if not, go to step S11;
s10, determining
Figure FDA0002424634790000022
Total migration overhead in (1)
Figure FDA0002424634790000023
And
Figure FDA0002424634790000024
total migration overhead in (1)
Figure FDA0002424634790000025
Judgment of
Figure FDA0002424634790000026
Whether the result is true or not;
if yes, return to step S8;
if not, go to step S12;
s11, detecting the user request SFC in service in the cloud network, revoking the expired user request SFC, and returning the physical resources
Figure FDA0002424634790000027
And returns to step S1;
s12, mixing
Figure FDA0002424634790000028
Request SR as a service functionmCorresponding optimal migration pathmAnd proceeds to step S13;
s13, determining the optimal migration pathmFulfilling a current service function request SRmThe fast migration of the mobile application is realized.
2. The method for fast migrating the mobile application in the cloud network according to claim 1, wherein the service function request SR in step S2mIs SRm=(SF,LF,RF,sm,tm);
Wherein S isFRequesting a set of several virtual network functions VNFs in the SFC for a user, and SF={f1,f2,...f|SF|-SF | is the number of virtual network functions VNF;
LFfor connecting virtual links between two adjacent virtual network functions VNFs, and LF={l1,l2,...l|LF|H, | LF | is the number of links for a user to request SFC;
RFthe total amount of resources required to request the SFC for the entire user;
smthe source point after the user moves;
tmthe destination point is the destination point of the user after moving;
the underlying network N in step S3GComprises the following steps:
NG={VG,EG,RG}
wherein, VGIs an underlying network NGA set of nodes in (1); vG={v1,v2,...v|VG|H, | VG | is underlying network NGThe number of nodes in;
EGis a VGIs an underlying network NGSet of links in (1), EG={E1,E2,...E|EG|Is | EG |Underlying network NGThe number of links in (1);
RGis an underlying network NGThe total amount of resources that can be provided.
3. The method for rapidly migrating the mobile application in the cloud network according to claim 2, wherein the step S3 specifically includes:
s31, obtaining service function request SRmLast virtual network function point f in (1)lastTo the moved destination point tmBandwidth requirement of
Figure FDA0002424634790000031
S32, traversing underlying network NGIn each link, the underlying network N is deletedGBandwidth of the remaining resources
Figure FDA0002424634790000032
Less than bandwidth requirement
Figure FDA0002424634790000033
Corresponding link etTo obtain the underlying network NGSubfigure N ofSG
Wherein e ist∈EG
S33, in sub-figure NSGWherein a deployment point V of a last virtual network function VNF in a user request SFC is determinedi flastAnd the destination point t after movingmIn between
Figure FDA0002424634790000034
4. The method for rapidly migrating the mobile application in the cloud network according to claim 2, wherein the step S4 specifically includes:
judgment of
Figure FDA0002424634790000035
Corresponding total migration overhead
Figure FDA0002424634790000036
Whether the result is true or not;
if so, then
Figure FDA0002424634790000037
If not, go to step S7;
if not, then
Figure FDA0002424634790000038
If yes, go to step S5;
wherein the total migration overhead
Figure FDA0002424634790000039
The calculation formula of (2) is as follows:
Figure FDA00024246347900000310
in the formula (I), the compound is shown in the specification,
Figure FDA00024246347900000311
is composed of
Figure FDA00024246347900000312
Corresponding bandwidth consumption;
Figure FDA0002424634790000041
is composed of
Figure FDA0002424634790000042
Corresponding seek time consumption.
5. The method for fast migrating a mobile application in a cloud network according to claim 2, wherein the bandwidth consumption of the whole migration path in the step S6
Figure FDA0002424634790000043
The calculation formula of (2) is as follows:
Figure FDA0002424634790000044
in the formula (I), the compound is shown in the specification,
Figure FDA0002424634790000045
requesting an SR for a service functionmBandwidth consumption before migration;
Figure FDA0002424634790000046
requesting an SR for a service functionmBandwidth consumption of the end path prior to migration;
Figure FDA0002424634790000047
to be composed of
Figure FDA0002424634790000048
As the optimal migration pathmService function request SR in migration processmThe bandwidth consumption of (2).
6. The method for rapidly migrating the mobile application in the cloud network according to claim 2, wherein the step S7 specifically includes:
s71 traversing service function request SRmAnd determining a set S of virtual network functions VNFFVirtual network function point f in (1)iWhether it is an intermediate function point;
wherein f isi∈SF
If yes, go to step S72;
if not, go to step S74;
s72, traversing the current underlying network NGEach node in (2) determining the underlying network NGNode v inmAnd its corresponding shortest path, supbpath,and proceeds to step S73;
wherein v ism∈VGNode vmIs an underlying network NGA contracted migration destination point v at middle distancejNearest and satisfying virtual network function point fiA node of the resource demand of (c);
s73, connecting the node vmAs a virtual network function point fiAnd node v is migrated tomAnd its corresponding shortest path, supbpath, as an intermediate path to join
Figure FDA0002424634790000049
In the get service function request SRmCorresponding to
Figure FDA00024246347900000410
Proceeding to step S9;
s74, traversing underlying network NGDetermines a passing deployment point Vi flastThe shortest path of (1), and proceeds to step S75;
wherein, the deployment point Vi flastSatisfy the last virtual network function point flastThe resource requirements of (1);
s75, deploying the point Vi flastAs the last virtual network function point flastAnd will deploy point Vi flastAnd its corresponding shortest path, supbpath, as an intermediate path to join
Figure FDA0002424634790000051
In the get service function request SRmCorresponding to
Figure FDA0002424634790000052
The process advances to step S9.
7. The method for fast migrating a mobile application in a cloud network according to claim 2, wherein the optimal migration path in step S8mComprises the following steps:
Figure FDA0002424634790000053
in the formula, pathbeforeRequesting SR for performing a service functionmA deployment path before migration;
Figure FDA0002424634790000054
requesting an SR for a service functionmA deployment path of the end path prior to migration.
8. The method for rapidly migrating a mobile application in a cloud network according to claim 1, wherein the step S9 specifically includes:
judgment of
Figure FDA0002424634790000055
Corresponding total migration overhead
Figure FDA0002424634790000056
Whether the result is true or not;
if so, then
Figure FDA0002424634790000057
If not, go to step S11;
if not, then
Figure FDA0002424634790000058
If yes, go to step S10;
wherein the total migration overhead
Figure FDA0002424634790000059
The calculation formula of (2) is as follows:
Figure FDA00024246347900000510
in the formula (I), the compound is shown in the specification,
Figure FDA00024246347900000511
is composed of
Figure FDA00024246347900000512
Corresponding bandwidth consumption;
Figure FDA00024246347900000513
is composed of
Figure FDA00024246347900000514
Corresponding migration operation consumption;
Figure FDA00024246347900000515
is composed of
Figure FDA00024246347900000516
Corresponding seek time consumption;
Figure FDA00024246347900000517
is composed of
Figure FDA00024246347900000518
Corresponding migration time consumption.
9. The method for fast migrating a mobile application in a cloud network according to claim 2, wherein in step S13, when the service function requests SR, the method further comprisesmBest migration path ofmIs composed of
Figure FDA00024246347900000519
Service function request SR at corresponding migration pathmThe migration method specifically comprises the following steps:
a1, traversing the optimal migration pathmAnd issuing a service function request SRmFormer deployment path pbeforePart of each link inDeploying a situation;
determine the optimal migration pathmAnd issuing a service function request SRmFormer deployment path pbeforeWhether there is the same virtual link l innCorresponding to the same link et
Wherein ln∈LF,et∈EG
If yes, go to step A2;
if not, go to step A3;
a2, directly according to the optimal migration pathmDirect service function request SRmMigration of (2);
a3, from underlying network NGMiddle deduction pathmIn (1)
Figure FDA0002424634790000061
Corresponding bandwidth resource is released, and the user sends out service function request SRmFront side
Figure FDA0002424634790000062
Occupied bandwidth resource and then according to the optimal migration pathmDirect service function request SRmMigration of (2);
when the service function requests the SRmBest migration path ofmIs composed of
Figure FDA0002424634790000063
When, the service function requests the SRmThe migration method specifically comprises the following steps:
b1 traversing the optimal migration path
Figure FDA0002424634790000064
And issuing a service function request SRmFormer deployment path pbeforeThe deployment situation of each node and each link in the network is used for judging the optimal migration path
Figure FDA0002424634790000065
And send out service functionsRequest SRmFormer deployment path pbeforeWhether there is the same virtual network function point fiCorresponding to the same node vmOr with the same virtual link lnCorresponding to the same link etThe case (1);
if yes, go to step B2;
if not, go to step B3;
b2, directly according to the optimal migration path
Figure FDA0002424634790000066
Direct service function request SRmMigration of (2);
b3, slave underlay network NGMiddle deduction
Figure FDA0002424634790000067
Corresponding node and bandwidth resource, and releasing service function request SR sent by usermFront pathbeforeOccupied node and its corresponding bandwidth resource, then according to the optimum migration path
Figure FDA0002424634790000071
Direct service function request SRmIs being migrated.
10. The method for fast migrating the mobile application in the cloud network according to claim 2, wherein the physical resource returned in step S11
Figure FDA0002424634790000072
Comprises the following steps:
Figure FDA0002424634790000073
in the formula (I), the compound is shown in the specification,
Figure FDA0002424634790000074
is an underlying network NGThe remaining physical resources;
Figure FDA0002424634790000075
requesting physical resources occupied by the SFCs for all users that have not migrated;
Figure FDA0002424634790000076
as the optimal migration pathmIs composed of
Figure FDA0002424634790000077
When the corresponding migration path is used, all users request physical resources occupied by the SFCs;
Figure FDA0002424634790000078
as the optimal migration pathmIs composed of
Figure FDA0002424634790000079
And when the corresponding migration path is used, all the users request the physical resources occupied by the SFCs.
CN201910084329.3A 2019-01-29 2019-01-29 Rapid migration method for mobile application in cloud network Expired - Fee Related CN109617811B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201910084329.3A CN109617811B (en) 2019-01-29 2019-01-29 Rapid migration method for mobile application in cloud network

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201910084329.3A CN109617811B (en) 2019-01-29 2019-01-29 Rapid migration method for mobile application in cloud network

Publications (2)

Publication Number Publication Date
CN109617811A CN109617811A (en) 2019-04-12
CN109617811B true CN109617811B (en) 2020-06-30

Family

ID=66020983

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201910084329.3A Expired - Fee Related CN109617811B (en) 2019-01-29 2019-01-29 Rapid migration method for mobile application in cloud network

Country Status (1)

Country Link
CN (1) CN109617811B (en)

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112737854B (en) * 2020-12-30 2023-06-09 国网河南省电力公司信息通信公司 Service chain migration method and device based on energy consumption and service quality
CN112910960A (en) * 2021-01-15 2021-06-04 郑州轻工业大学 Virtual network online migration method and device with time delay, resource and energy consumption perception
CN113038520B (en) * 2021-03-31 2022-05-10 广东电网有限责任公司电力调度控制中心 Topology-aware virtual network service fault recovery method and system
CN114124796A (en) * 2021-10-26 2022-03-01 清华大学 Service chain deployment processing method and device, computer storage medium and terminal

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107562510A (en) * 2016-06-30 2018-01-09 华为技术有限公司 The management method and management equipment of a kind of application example
CN109104313A (en) * 2018-08-20 2018-12-28 电子科技大学 A kind of SFC dynamic deployment method with flow consciousness with energy perception
CN109117321A (en) * 2018-07-27 2019-01-01 山东师范大学 A kind of full link application moving method of cloud platform based on label figure

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107332913B (en) * 2017-07-04 2020-03-27 电子科技大学 Optimized deployment method of service function chain in 5G mobile network
CN109257399B (en) * 2017-07-13 2023-01-24 上海擎感智能科技有限公司 Cloud platform application program management method, management platform and storage medium
CN108076156B (en) * 2017-12-27 2020-09-08 北京航空航天大学 Mixed cloud system based on Chinese cloud product

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107562510A (en) * 2016-06-30 2018-01-09 华为技术有限公司 The management method and management equipment of a kind of application example
CN109117321A (en) * 2018-07-27 2019-01-01 山东师范大学 A kind of full link application moving method of cloud platform based on label figure
CN109104313A (en) * 2018-08-20 2018-12-28 电子科技大学 A kind of SFC dynamic deployment method with flow consciousness with energy perception

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
Bandwidth Provisioning for Virtual Machine;Uttam Mandal;《IEEE TRANSACTIONS ON CLOUD COMPUTING》;20160323;全文 *
自组织网络环境下基于遗传蚁群算法的计算迁移技术研究;卢占翔;《中国优秀硕士学位论文全文数据库 信息科技辑》;20180315;全文 *

Also Published As

Publication number Publication date
CN109617811A (en) 2019-04-12

Similar Documents

Publication Publication Date Title
CN109617811B (en) Rapid migration method for mobile application in cloud network
WO2022021176A1 (en) Cloud-edge collaborative network resource smooth migration and restructuring method and system
EP3300316B1 (en) Deterministic controller-based path query
EP3466122B1 (en) Mobile edge computing
CN109525426B (en) FV-based service control system and method for open MEC platform
Bi et al. Intelligent quality of service aware traffic forwarding for software-defined networking/open shortest path first hybrid industrial internet
CN111626352A (en) Adaptive energy consumption optimal vehicle clustering method based on fuzzy C-means
WO2023024219A1 (en) Joint optimization method and system for delay and spectrum occupancy in cloud-edge collaborative network
CN109379230B (en) Service function chain deployment method based on breadth-first search
CN101521616A (en) Method and system for neighbor migration in Border Gateway Protocol (BGP) distributed system
Yi et al. Design and implementation of network-aware VNF migration mechanism
Priyadarsini et al. An energy-efficient load distribution framework for SDN controllers
CN113076177B (en) Dynamic migration method of virtual machine in edge computing environment
US11824765B2 (en) Fast redirect of traffic when pods fail
WO2023019604A1 (en) Minimum network energy consumption optimization method and system based on traffic grooming
Somula et al. POUPR: properly utilizing user-provided recourses for energy saving in mobile cloud computing
CN101272347B (en) Load sharing method and device
CN106992906A (en) The method of adjustment and system of a kind of access rate
CN114513449A (en) Method and system for optimizing routing in domain
CN112615750A (en) Backbone network flow granularity adjusting system and method based on SDN
CN105657054B (en) A kind of content center network caching method based on K mean algorithms
CN108923961B (en) Multi-entry network service function chain optimization method
Rizou et al. Providing qos guarantees in large-scale operator networks
Fekih et al. Proactive content caching strategy with router reassignment in content centric networks based SDN
Fekih et al. SDN-based replication management framework for CCN networks

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for 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: 20200630

Termination date: 20210129

CF01 Termination of patent right due to non-payment of annual fee