CN110190987B - Virtual network function reliability deployment method based on backup income and remapping - Google Patents

Virtual network function reliability deployment method based on backup income and remapping Download PDF

Info

Publication number
CN110190987B
CN110190987B CN201910380700.0A CN201910380700A CN110190987B CN 110190987 B CN110190987 B CN 110190987B CN 201910380700 A CN201910380700 A CN 201910380700A CN 110190987 B CN110190987 B CN 110190987B
Authority
CN
China
Prior art keywords
reliability
backup
virtual network
service
network function
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN201910380700.0A
Other languages
Chinese (zh)
Other versions
CN110190987A (en
Inventor
朱晓荣
冯爽
付康平
王一忠
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.)
Nanjing University of Posts and Telecommunications
Original Assignee
Nanjing University of Posts and Telecommunications
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 Nanjing University of Posts and Telecommunications filed Critical Nanjing University of Posts and Telecommunications
Priority to CN201910380700.0A priority Critical patent/CN110190987B/en
Publication of CN110190987A publication Critical patent/CN110190987A/en
Application granted granted Critical
Publication of CN110190987B publication Critical patent/CN110190987B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L12/00—Data switching networks
    • H04L12/28—Data switching networks characterised by path configuration, e.g. LAN [Local Area Networks] or WAN [Wide Area Networks]
    • H04L12/46—Interconnection of networks
    • H04L12/4641—Virtual LANs, VLANs, e.g. virtual private networks [VPN]
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/06—Management of faults, events, alarms or notifications
    • H04L41/0654—Management of faults, events, alarms or notifications using network fault recovery
    • H04L41/0663—Performing the actions predefined by failover planning, e.g. switching to standby network elements

Abstract

The invention discloses a virtual network function reliability deployment method based on backup income and remapping. The invention firstly introduces a double-sharing backup mechanism, effectively improves the reliability of the service function chain, simultaneously reduces the occupation of the backup to the resource as much as possible, thereby improving the resource utilization rate of the system, and then provides a reasonable and efficient virtual network function remapping method aiming at the virtual network function which is not subjected to the reliability backup in the service function chain. The invention provides a method for solving the problems that the virtual network function in the service function chain is backed up and restored after a fault occurs and the reliability of the service function chain is reduced due to the change of the reliability of the virtual network function after remapping.

Description

Virtual network function reliability deployment method based on backup income and remapping
Technical Field
The invention relates to a virtual network function reliability deployment method based on a backup income index and remapping, and belongs to the technical field of networks.
Background
The 5G is a new generation mobile communication system developed for the demand of mobile communication after 2020. With the overall arrival of 5G networks, network virtualization technology becomes one of the key technologies for network service model transition, and is the basis of dynamic and flexible deployment of network services. The network function virtualization is realized by converting the network function software on the original special hardware into a virtual network function deployed on a general server, so that the operation cost is reduced, the flexibility of the network function is improved, and meanwhile, new problems are caused, such as the problem that the reliability of the network service is influenced by the network function, and the condition that the network service is invalid and even the network is broken down due to the network function failure also occur at times.
In the NFV network, network services are implemented by a service function chain SFC, and currently, research on service function chain deployment is mostly focused on acceptance rate of network service requests, utilization rate of system resources, and the like, and research on reliability of virtual network functions is little and few. At present, for improving the reliability of the virtual network function, a common method is to allocate redundant resources to perform virtual network function backup when performing virtual network function mapping; when the virtual network function fails, the original request is directly transferred to a backup resource, so that the continuity and high reliability of the service are guaranteed. However, the redundant backup increases reliability, sacrifices occupation of system resources by the backup, and fails to solve the problem of reduced reliability of service of a service function chain in which the virtual network function is located after the virtual network function is failed to recover. How to backup the service function chain can effectively improve the reliability of the service function chain to improve the effectiveness of the service function chain as much as possible is a difficult problem to be solved.
The method for reliably deploying the virtual network function based on the backup profit and the remapping, which is provided by the invention, not only introduces double shared backups, but also optimizes the reliable deployment by using the backup profit and the weight coefficient to reduce the occupation of system resources as much as possible, and solves the problem of recovering the virtual network function when the service function chain fails by implementing the remapping method on the virtual network function in the service function chain. The method has great practical significance for optimizing the deployment of the reliability of the virtual network function.
Disclosure of Invention
The technical problem is as follows: the invention aims to model the reliability problem of the virtual network function during the deployment, improve the traditional redundancy backup method, provide a virtual network function reliability deployment method based on backup income and remapping, introduce a double-sharing backup mechanism, add a backup income strategy, effectively improve the reliability of a service function chain, reduce the occupation of backup on resources as much as possible and improve the resource utilization rate of a system; a reasonable and efficient virtual network function remapping method is provided for virtual network functions which are not subjected to reliable backup in a service function chain; the method is firstly provided for solving the problem that the reliability of the service function chain is reduced due to the change of the reliability of the virtual network function after the backup recovery and the remapping of the virtual network function in the service function chain are carried out.
The technical scheme is as follows: the invention relates to a virtual network function reliability deployment method based on backup income and remapping, which comprises the following steps:
step 1, based on a double sharing backup mechanism of backup income, providing a backup income index virtual network function selection method with backup, and considering the reliability of a service function chain and the occupation effectiveness of system resources;
step 2, for the virtual network function mapping problem of the backup node, a solution based on a simulated annealing method is provided;
and step 3, providing a reliability compensation strategy aiming at the problem that the reliability of the service function chain is reduced after the virtual network function is subjected to backup recovery or remapping.
Wherein:
the concrete content of the step 1 is as follows:
double sharing equivalent model: (as shown in FIG. 8)
After the double sharing mechanism backup is carried out, the virtual network function vnf of the backupaCan be regarded as two virtual network functions vnfb、vnfc(ii) a Since two shares are assigned to the virtual network function vnfaSo that the two virtual network functions cannot function simultaneously,
only the virtual network function vnf is considered under the modeliReliability of (2):
ri'＝ri+(1-ri)(1-rj)rb (8)
the increment is as follows:
Δri'＝ri+(1-ri)(1-rj)rb-ri＝(1-ri)(1-rj)rb (9)
compared with the original exclusive backup protection, the reliability of the double shared backup is reduced as follows:
Δri-Δri'＝ra(1-ri)-(1-ri)(1-rj)rb (10)
wherein: r represents virtual network function reliability. Virtual network function vnf before backupiHas a reliability of riVirtual network function vnfjHas a reliability of rj. Backup nodeFor vnfiAnd vnfjRespectively in ra、rb。
To clarify the effect of backup, assume ra＝rb(ii) a The average reliability of the general service is substituted into 99.9%, the reliability of the double-sharing backup is reduced to 0.0000999% compared with the exclusive backup, even for the carrier-class service, the reduction of the reliability can be ignored, but the backup cost is reduced to almost half of the original backup cost;
in order to reduce the complexity of the Model, a Double Shared Equivalent Model (DSEM) is provided to use the original backed-up virtual network function vnfiVnf with its doubly shared backupbViewed as a whole-a network function fiAs a new virtual network function vnfi', with a reliability of
ri'＝ri+(1-ri)(1-rj)rb (11)
Virtual network function vnfjVnf with its doubly shared backupcViewed as a whole-a network function fjAs a new virtual network function vnf'jWith a reliability of
rj'＝rj+(1-ri)(1-rj)rc (12)
Wherein the virtual network function vnfcHas a reliability of rc；
At this time, vnf is set for the virtual network functioni', the virtual network function vnf will be backed up by inspecting the occupation of the resourceaThe resource occupation of (a) is taken as a special resource occupation and is not distributed to a virtual network function vnfi、vnfjThereby reducing subsequent network functions fjComplexity of resource allocation problem during backup;
network model
The NFV network G ═ N, E, where N denotes a set of nodes and E denotes a set of links between nodes; multiple virtual network function instances can be placed on each node, and each node n has own exclusive computing resourceCnStorage resource MnAnd network resource BnVirtual network functions vnf placed on nodesiRequiring a node to allocate its own dedicated computing resource ciStorage resource miAnd network resource biAnd the sum of various resources occupied by the virtual network function deployed on the node must not exceed the resources owned by the node;
defining a service request in the NFV network, wherein the service request comprises a service type, a service reliability requirement and service strength; the service request type in the NFV network is the service function chain request siService function chain set denoted S, service function chain SiIs a set NF of network functions ordered by a certain logiciThe components are mixed; virtual network function set NFiThe virtual network function in (1) is denoted as fijEach network function fijCorresponding to one or more virtual network functions vnfijWherein i denotes that the virtual network function belongs to a service function chain siJ indicates that the virtual network function is the jth virtual network function that constitutes a service function chain, with a reliability of rijDefine a Boolean variable aijnWhen vnfijWhen placed on node n, aijn1 is ═ 1; when vnfijWhen not placed on node n, aijn0; the type of service request determines the service function chain siOf type (D), on which all network functions are denoted FiThe strength of the service request determines the size of the resources required by the virtual network functions that make up the service function chain; the reliability requirements of the service determine the service function chain siHas a reliability requirement of Ri；
After introducing the redundant backup of the virtual network function, the virtual network function vnfijThe network model of (2) needs to be changed to some extent; assuming deployment of a single virtual network function vnfijThe occupied computing, storage and network resources are respectively cij、mijAnd bijWith a reliability of rijFor virtual network functions vnfijDeploying redundant backups on node n
Figure BDA0002053277310000041
Need to be distributed
Figure BDA0002053277310000042
The required computing, storage and network resources are respectively ci、mi、biWhile for the network function fjIn other words, the reliability of the virtual network function vnf is improved to a certain extent, and in order to reduce the complexity of the model, the occupied resources are still regarded as the virtual network function vnfijOccupied resources, virtual network function backup
Figure BDA0002053277310000043
Network function f is not counted in occupied resourcesjOccupied resources, but including its service function chain siOccupied backup overhead;
redundant backups can increase the reliability of a virtual network function, thus defining a network function fijBack-up profit g (f)ij) Comprises the following steps:
Figure BDA0002053277310000044
the backup revenue for the service function chain is G(s)i) Comprises the following steps:
Figure BDA0002053277310000045
the virtual network function backup occupies a large amount of resources while increasing reliability, and the resources occupied by the backup cannot bring any benefit when the virtual network function is not failed, so that the resource occupation of the backup is reduced as much as possible while the reliability of the service function chain meets the service requirement, and the backup benefit of the service function chain is maximized;
optimizing the target:
Figure BDA0002053277310000051
constraint conditions are as follows:
Figure BDA0002053277310000052
Figure BDA0002053277310000053
Figure BDA0002053277310000054
Figure BDA0002053277310000055
Figure BDA0002053277310000056
the objective function (14) represents that maximizing backup of service function chains is an optimization objective for revenue text; the constraint (15) represents the reliability r of the service function chainSFCThe reliability requirement R of the service needs to be meti(ii) a The constraint conditions (16) to (18) represent that the sum of the occupation of all virtual network functions on the node computing, storing and network resources cannot exceed the resource allocated by the node; the constraint (19) indicates that the virtual network function must be deployed to one node and can only be deployed to one node.
The specific content of the step 2 is as follows:
dual-sharing reliability deployment method based on backup benefits
The network service has corresponding reliability requirements on the service function chain, when the service function chain cannot meet the service reliability requirements, the reliability of a single network function is improved by performing redundant backup on the virtual network function forming the service function chain, so that the reliability of the service function chain meets the service requirements; how to backup can effectively improve the reliability of the service function chain so as to improve the effectiveness of the service function chain as much as possible, which is a difficult problem to be solved;
in a reliability system, in the system of which the overall performance depends on the performances of components, the reliability performance change of some of the components may play a more important role than other components, so that a classic Importance metric index, namely Birnbaum Importance metric Importance Measure, is introduced to characterize the Importance metric of the reliability of the components in the system; the formula for IM for component k is as follows:
Figure BDA0002053277310000057
IM represents the degree of importance of the reliability increment of the component k to the reliability of the system; inspired by IM, the invention defines a network function fijBackup Gains (BG):
Figure BDA0002053277310000061
wherein the content of the first and second substances,
Figure BDA0002053277310000062
representing network functions fijThe unit overhead of the backup, namely the ratio of the overhead of the backup resources of the network function to the increment of the network function after the backup:
Figure BDA0002053277310000063
for various resource occupancies in the backup overhead, different requirements can be met by adjusting the weight coefficient; the BG can reflect the importance of the virtual network function in the whole service function chain, and the larger the value of the BG is, the larger the influence on the reliability improvement of the service function chain is when the corresponding virtual network function is backed up; therefore, when backup is carried out, the virtual network function with a larger BG index is preferentially selected for backup;
firstly, initializing and placing a service function chain according to the service requirement of a network and a corresponding deployment method;
secondly, reliability evaluation of the service function chain is carried out, and whether the reliability of the service function chain meets the reliability requirement of the service is judged;
if not, selecting the virtual network function with the largest backup profit index and the second largest backup profit index by calculating the backup profit indexes of all the virtual network functions in the whole service function chain, and carrying out double sharing reliability backup on the nodes meeting the requirements by a local search method; updating the virtual network function after double sharing backup;
fourthly, reliability evaluation of the service function chain is carried out again, and then the third step is repeated until the reliability of the service function chain meets the requirements required by the service;
pseudo codes of the double sharing reliability deployment method based on the backup benefits are as follows:
Figure BDA0002053277310000064
Figure BDA0002053277310000071
the specific content of the step 3 is as follows:
the invention designs a relevant solution aiming at the condition that the reliability of the service function chain is changed after the node fails and is remapped or migrated to the backup,
when the virtual network function fails, firstly judging the backup type, and carrying out corresponding processing flow according to the judgment result; after the processing is finished, the deployment condition of the service function chain is input into the reliability deployment module, and multiple conditions may exist in the VNF fault:
firstly, migration to backup results in obvious reduction of the reliability of the original service function chain if migration to the backup node results in the need of re-checking whether the reliability of the service function chain meets the reliability requirement of the service; particularly, when one of the virtual network functions adopting the double shared backup fails, the failed virtual network function needs to be recovered on the backup node, and meanwhile, the reliability of the virtual network function sharing the backup resource with the backup node is reduced, and at the moment, the reliability of the virtual network function sharing the backup resource in the service function chain needs to be updated;
remapping the virtual network function, wherein when the remapping of the virtual network function is carried out by adopting a simulated annealing method, the reliability of the remapped virtual network function example may be changed, so that the overall reliability of a service function chain is changed;
any situation that results in a change in the reliability of the service function chain requires a new round of reliability assessment on the service function chain to ensure that the reliability of the service function chain can always meet the reliability requirements of the service.
Has the advantages that: the invention has the following advantages:
the method can effectively improve the reliability of the service function chain, simultaneously reduce the resource occupation of the reliability backup as much as possible, and maximize the backup benefit of the service function chain.
And the problem that the failed virtual network function has no backup can be solved. Reliability compensation is performed by remapping of virtual network functions.
The problem that the virtual network function with the fault has double sharing backup and the virtual network function sharing backup resources has the fault at the same time can be solved. The virtual network function recovery problem when the service function chain fails is solved by implementing a remapping method for the virtual network function in the service function chain.
And fourthly, carrying out a new round of reliability evaluation on the service function chain aiming at any condition which causes the reliability change of the service function chain in time so as to ensure that the reliability of the service function chain can always meet the reliability requirement of the business.
Drawings
Figure 1 is a reliability deployment module flow diagram,
figure 2 is a VNF failure handling flow diagram,
figure 3 is a basic model of a service function chain,
figure 4 is an equivalent model of the service function chain voting model r ═ n,
FIG. 5 is a r/n (G) model,
figure 6 is a diagram of dedicated backup protection,
figure 7 is a diagram of a double shared backup protection,
figure 8 is an equivalent model of the double sharing mechanism,
figure 9 is an equivalent service function chain after double sharing backup,
fig. 10 is a basic model of network function virtualization.
Detailed Description
In order to achieve the above object, the present invention adopts the following technical solutions:
firstly, a double sharing mechanism based on a backup profit index is adopted for reliability deployment, and the execution flow of the method is shown in fig. 1:
firstly, initializing and placing a service function chain according to the service requirement of a network and a corresponding deployment method;
secondly, reliability evaluation of the service function chain is carried out, and whether the reliability of the service function chain meets the reliability requirement of the service is judged;
if not, selecting the virtual network function with the largest backup profit index and the second largest backup profit index by calculating the backup profit indexes of all the virtual network functions in the whole service function chain, and carrying out double sharing reliability backup on the nodes meeting the requirements by a local search method; updating the virtual network function after double sharing backup;
and thirdly, reliability evaluation of the service function chain is carried out again, and then the step III is repeated until the reliability of the service function chain meets the requirement required by the service.
The dual-sharing reliability deployment method based on the backup benefits solves the problem of reliability of the service function chain in the initial deployment stage, but some problems still exist in the reliability of the service function chain and need to be solved; when a virtual network function fails in a service function chain, there may be a number of situations:
the virtual network function with fault has double sharing backup and the virtual network function sharing backup resource works normally.
The virtual network function with the fault has double backup, but the virtual network function sharing the backup resource has the fault at the same time.
And thirdly, the virtual network function with the fault is not backed up.
Performing corresponding processing aiming at various conditions by adopting a virtual network function reliability deployment method based on a backup profit index and remapping; and after the processing is finished, the problem that the reliability of the service function chain is reduced is effectively solved.
After the virtual network function is in fault, the virtual network function firstly enters a backup type judgment module for carrying out corresponding fault type judgment on the virtual network function in fault; when the type is judged, when the fault virtual network function has backup, the service flow is redirected to the node where the backup is located; if the network function for carrying out double sharing backup fails at the same time, the virtual network function which is firstly judged by the backup type judgment module is regarded as the existence of virtual network function backup, the virtual network function backup type sharing backup resources with the virtual network function is changed into no backup, and the later-entered virtual network function is directly regarded as the non-backup virtual network function for processing; and for the virtual network function which is not subjected to the reliability backup, remapping the virtual network function.
The remapping of the virtual network functions uses a simulated annealing method. And selecting a proper node to be allocated to the virtual network function for instantiation through a simulated annealing method. The solution space of the simulated annealing method is characterized in that all nodes are traversed, and the nodes meeting the resource requirements are added into the solution space
Figure BDA0002053277310000091
The initial solution n is in the solution space through a random function
Figure BDA0002053277310000092
Performing random selection; the evaluation function used in the present invention is C (n) ═ f (C)n,Mn,BnDelta tau), the service function chain end-to-end delay increment delta tau brought by the use condition of various resources of the computing node and the migration to the node n to be computed; receivingThe probability of the difference solution is
Figure BDA0002053277310000093
By implementing the remapping method for the virtual network function in the service function chain, the recovery problem of the virtual network function when the service function chain fails is solved. The invention designs a relevant solution aiming at the condition that the reliability of a service function chain is changed after a node fails and is remapped or migrated to a backup, and the flow chart is shown as figure 2:
when the virtual network function fails, firstly judging the backup type, and carrying out corresponding processing flow according to the judgment result; after the processing is finished, the deployment condition of the service function chain is input into the reliability deployment module, and multiple conditions may exist in the VNF fault:
firstly, migration to backup results in obvious reduction of the reliability of the original service function chain if migration to the backup node results in the need of re-checking whether the reliability of the service function chain meets the reliability requirement of the service; particularly, when one of the virtual network functions using the dual shared backup fails, the failed virtual network function needs to be recovered at the backup node, and at the same time, the reliability of the virtual network function sharing the backup resource with the failed virtual network function is reduced, and at this time, the reliability of the virtual network function sharing the backup resource in the service function chain needs to be updated.
And secondly, remapping the virtual network function, wherein when the virtual network function is remapped by adopting a simulated annealing method, the reliability of the remapped virtual network function instance may be changed, so that the overall reliability of the service function chain is changed.
Any situation that results in a change in the reliability of the service function chain requires a new round of reliability assessment on the service function chain to ensure that the reliability of the service function chain can always meet the reliability requirements of the service.
The invention is further described below with reference to the accompanying drawings.
The invention introduces a double-sharing backup mechanism, effectively improves the reliability of the service function chain, simultaneously reduces the occupation of the backup to the resource as much as possible, thereby improving the resource utilization rate of the system, then provides a reasonable and efficient virtual network function remapping method aiming at the virtual network function which is not subjected to the reliability backup in the service function chain, and solves the problems that the virtual network function in the service function chain is subjected to backup recovery after the fault occurs and the reliability of the service function chain is reduced due to the reliability change of the virtual network function after remapping.
The method for deploying the reliability of the virtual network function based on the backup income and the remapping mainly comprises the following steps:
1. reliability analysis of service function chains
1.1 virtual network functional reliability index
The index for measuring the reliability of the virtual network function is the probability of the normal service of the virtual network function, and the corresponding reliability of the virtual network function can be calculated through the normal working time length of the virtual network function and the abnormal working time length of the virtual network function. The reliability r of a virtual network function is equal to the ratio of the virtual network function normal operating time to the total time the virtual network function is running, which depends on the Mean Time Between Failure (MTBF) of the virtual network function and the mean time to failure recovery time (MTTR) of the virtual network function.
Figure BDA0002053277310000111
As shown in fig. 4, it is assumed that the service function chain SFC is composed of n virtual network functions, and the reliability of each virtual network function is riThe service function chain needs to be made to have overall reliability meeting the reliability requirements of the service. According to the voting model (r/n (G) model) in the reliability model, as shown in FIG. 5, the system will not fail if the number of normal units is not less than r (1 ≦ r ≦ n) among the n units constituting the system, and such a model is called a voting model. In the voting model, when r is equal to n, the model degenerates into a series model, i.e. the service function chain in the graph can be equivalent to a series of individual service function chainsLink of virtual network functions, hence reliability r of the entire service function chainSFCComprises the following steps:
Figure BDA0002053277310000112
1.2 redundant backup mechanism and virtual network functional reliability analysis
The method for improving the reliability is to perform redundancy backup, and when the virtual network function mapping is performed, corresponding redundancy resources can be allocated for corresponding virtual network functions to perform backup. As shown in fig. 6
The service function chain in fig. 6 is composed of three virtual network functions, namely a virtual network function vnfiAllocating corresponding backup resources to carry out dedicated backup vnfa，vnfaAnd vnfiRequires comparable computation, storage and network resources, and the resources can only be vnfaExclusive when vnfiDuring normal operation, assigned to vnfaAlso, the resources of (a) cannot be used by other virtual network functions. Assume pre-backup vnf before backupiHas a reliability of riThe reliability of the backup virtual network function is raThen vnf after backupaAnd vnfiThe overall reliability of (2) is:
ra+i＝1-(1-ri)(1-ra)＝ri+ra-rira (3)
by backup, the virtual network function gets a reliability increment of
Δri＝1-(1-ri)(1-ra)-ri＝ra(1-ri) (4)
The resource overhead of dedicated backup protection is too high, so that the overhead is reduced by adopting a double-sharing backup mode, and the structure schematic is shown in fig. 7. Different from the special backup protection, the double sharing backup adopts the sharing backup resources among a plurality of virtual network functions to reduce the occupation of the backup on the resources; however, these virtual network functions cannot fail simultaneously because the doubly shared backup cannot provide the same as the dedicated backupReliability; assuming virtual network functions vnfiHas a resource occupation of SiReliability is riVirtual network function vnfjHas a resource occupation of SjReliability is rjThe occupation of the backup node on the resource is Sa＝max(Si,Sj) Backup node for vnfiAnd vnfjRespectively in ra、rb: then the reliability of the service function chain after backup is:
r'＝rirj+(1-ri)rjra+rirb(1-rj) (5)
the reliability of the original service function chain is as follows:
rSFC＝rirj (6)
after the double sharing backup is performed, the reliability increment of the service function chain is as follows:
Figure BDA0002053277310000121
in order to reduce the occupation of resources by the service function chain backup, it is a common practice to perform a reliable backup on one or more virtual network functions with the lowest reliability among the virtual network functions constituting the service function chain, so that the reliability of the whole service function chain meets the service requirement. For the problem of selecting a virtual network function for backup, a conventional method provides that a virtual network function with the lowest reliability and a virtual network function with the next lowest reliability in a service function chain are selected for backup, and the reliability of the service function chain is improved most obviously.
Assume service function chain SFCjHas a reliability requirement of
Figure BDA0002053277310000122
How to make reasonable deployment so that groupsThe reliability of the virtual network function forming the service function chain meets the requirements of services, simultaneously reduces the occupation of system resources by redundant backup as much as possible, and ensures that the resources are effectively used.
1.3 double-shared equivalent model
After the double sharing mechanism backup is carried out, the virtual network function vnf of the backupaCan be regarded as two virtual network functions vnfb、vnfc(ii) a But because the two shares are assigned to the virtual network function vnfaSo that these two virtual network functions cannot function simultaneously. As shown in fig. 8
Investigating only virtual network functions vnfiReliability of (2):
ri'＝ri+(1-ri)(1-rj)rb (8)
the increment is as follows:
Δri'＝ri+(1-ri)(1-rj)rb-ri＝(1-ri)(1-rj)rb (9)
compared with the original exclusive backup protection, the reliability of the double shared backup is reduced as follows:
Δri-Δri'＝ra(1-ri)-(1-ri)(1-rj)rb (10)
to clarify the effect of backup, assume ra＝rb. The average reliability of the generic service is substituted here by 99.9%, and the reliability of the double shared backup compared to the dedicated backup is reduced by 0.0000999%, which is negligible even for carrier class services. But the backup overhead is reduced to almost half of the original backup overhead.
In order to reduce the complexity of the Model, the invention provides a Double-Shared Equivalent Model (DSEM) which enables the original backed-up virtual network function vnf to be realizediVnf with its doubly shared backupbViewed as a whole-a network function fiAs a new virtual network function vnfi', with a reliability of
ri'＝ri+(1-ri)(1-rj)rb (11)
Virtual network function vnfjVnf with its doubly shared backupcViewed as a whole-a network function fjAs a new virtual network function vnf'jWith a reliability of
rj'＝rj+(1-ri)(1-rj)rc (12)
At this time, vnf is set for the virtual network functioni' consider its occupation of resources. Backing up virtual network functions vnf hereinaThe resource occupation of (a) is taken as a special resource occupation and is not distributed to a virtual network function vnfi、vnfjThereby reducing subsequent network functions fjComplexity of resource allocation problem at backup time. The equivalent service function chain after double sharing backup is shown in fig. 9.
2. System model
2.1 problem statement
In the NFV network, the business requirement is realized by network service, and the entity for realizing the network service is a Service Function Chain (SFC); the service function chain is implemented by a group of virtual network functions VNF in a logical order. The network traffic flows sequentially through these network functions in a corresponding logical order. Each service function chain has its own performance requirements, such as end-to-end delay and reliability requirements of the service chain; the virtual network functions that make up the service function chain need to be deployed on nodes in the network. Nodes in the network have respective exclusive physical resources, including computing resources, storage resources and network resources; multiple VNFs may be placed on each node. When the VNF instantiates on a node, various required resources need to be distributed; meanwhile, when the virtual network function is deployed, the end-to-end performance requirement of the service function chain where the virtual network function is located needs to be considered. When deploying virtual network functions, how to select a proper virtual network function for backup needs to be considered to meet the reliability requirement of the service function chain where the virtual network function is located, and how to select a proper backup strategy can reduce the occupation of backup overhead on system resources as much as possible.
As shown in fig. 10, in the NFV network, when a request of a new service arrives, the NFV orchestration manager converts the service request into a service function chain request, which includes reliability requirement, delay requirement, and corresponding strength requirement of the service. The NFV orchestration manager converts these into the requirements of various resources required by the corresponding virtual network functions in the service function chain, and performs the next preparation for deployment. And performing initial deployment by a VNF virtual network function deployment method based on time delay, and outputting an initial deployment scheme. And calculating a service function chain backup scheme according to the deployment and the reliability of each virtual network function in the preliminary deployment scheme, and performing backup deployment on the service function chain until the reliability of the service function chain meets the service requirement. The redundant backup solves the reliability problem in the service function chain deployment in the survivable virtual network mapping; a method needs to be provided to solve the reliability problem in the running process of the virtual network function, that is, to solve the remapping problem of the virtual network function when the virtual network function that is not backed up fails. At present, no matter redundant backup failure recovery or virtual network function remapping failure recovery is carried out, the problem that the service reliability of a service function chain where a virtual network function is located is reduced after the virtual network function failure recovery is not considered.
2.2 network model
The NFV network G ═ N, E, where N denotes a set of nodes and E denotes a set of links between nodes. Multiple virtual network function instances may be placed on each node. Each node n has its own exclusive computing resource CnStorage resource MnAnd network resource Bn. Virtual network function vnf placed on a nodeiRequiring a node to allocate its own dedicated computing resource ciStorage resource miAnd network resource biAnd the sum of various resources occupied by the virtual network function deployed on the node must not exceed the resources owned by the node.
A service request in the NFV network is defined, the service request comprising a service type, a service reliability requirement and a service strength. Service request type in NFV networkI.e. service function chain request siThe set of service function chains is denoted S. Service function chains siIs a set NF of network functions ordered by a certain logiciThe components are mixed; virtual network function set NFiThe virtual network function in (1) is denoted as fijEach network function fijCorresponding to one or more virtual network functions vnfijWherein i denotes that the virtual network function belongs to a service function chain siJ indicates that the virtual network function is the jth virtual network function that constitutes a service function chain, with a reliability of rij. Defining a Boolean variable aijnWhen vnfijWhen placed on node n, aijn1 is ═ 1; when vnfijWhen not placed on node n, aijn0; the type of service request determines the service function chain siOf type (D), on which all network functions are denoted FiThe strength of the service request determines the size of the resources required by the virtual network functions that make up the service function chain; the reliability requirements of the service determine the service function chain siHas a reliability requirement of Ri。
After introducing the redundant backup of the virtual network function, the virtual network function vnfijThe network model of (2) needs to be changed to some extent; assuming deployment of a single virtual network function vnfijThe occupied computing, storage and network resources are respectively cij、mijAnd bijWith a reliability of rijFor virtual network functions vnfijDeploying redundant backups on node n
Figure BDA0002053277310000151
Need to be distributed
Figure BDA0002053277310000152
The required computing, storage and network resources are respectively ci、mi、biWhile for the network function fjIn other words, the reliability of the virtual network function vnf is improved to a certain extent, and in order to reduce the complexity of the model, the occupied resources are still regarded as the virtual network function vnfijOccupied resources, virtual network function backup
Figure BDA0002053277310000153
Network function f is not counted in occupied resourcesjOccupied resources, but including its service function chain siThe overhead of the backup occupied.
Redundant backups can increase the reliability of a virtual network function, thus defining a network function fijBack-up profit g (f)ij) Comprises the following steps:
Figure BDA0002053277310000154
the backup revenue for the service function chain is G(s)i) Comprises the following steps:
Figure BDA0002053277310000155
the virtual network function backup occupies a large amount of resources while increasing reliability, and when the virtual network function is not failed, the resources occupied by the backup cannot bring any benefit, so that the reliability of the service function chain is required to meet business requirements, the resource occupation of the backup is reduced as much as possible, and the backup benefit of the service function chain is maximized.
Optimizing the target:
Figure BDA0002053277310000161
constraint conditions are as follows:
Figure BDA0002053277310000162
Figure BDA0002053277310000163
Figure BDA0002053277310000164
Figure BDA0002053277310000165
Figure BDA0002053277310000166
the objective function (14) represents that maximizing backup of service function chains is an optimization objective for revenue text; the constraint (15) represents the reliability r of the service function chainSFCThe reliability requirement R of the service needs to be meti(ii) a The constraint conditions (16) to (18) represent that the sum of the occupation of all virtual network functions on the node computing, storing and network resources cannot exceed the resource allocated by the node; the constraint (19) indicates that the virtual network function must be deployed to one node and can only be deployed to one node.
3 virtual network function deployment method based on backup income index and remapping
In order to solve the reliability problem during service function chain deployment, the virtual network function failure Remapping and the service function chain reliability reduction, a comprehensive and reliable solution, namely a virtual network function deployment method based on Backup Revenue Index and Remapping (BRIR), is provided.
3.1 Dual sharing reliability deployment method based on backup benefits
The network service has a corresponding reliability requirement on the service function chain, and when the service function chain cannot meet the service reliability requirement, the virtual network functions forming the service function chain need to be redundantly backed up to improve the reliability of a single network function, so that the reliability of the service function chain meets the service requirement. How to backup the service function chain can effectively improve the reliability of the service function chain to improve the effectiveness of the service function chain as much as possible is a difficult problem to be solved.
In a reliability system, in which the overall performance depends on the performance of its components, the reliability performance variation of some of these components may play a more important role than other components, so a classic Importance metric index, i.e., Birnbaum Importance Metric (IM), is introduced to characterize the Importance metric of the reliability of the components in the system. The formula for IM for component k is as follows:
Figure BDA0002053277310000171
IM represents how important the incremental reliability of component k is to the reliability of the system. Inspired by IM, a network function f is defined hereinijBackup Gains (BG):
Figure BDA0002053277310000172
wherein the content of the first and second substances,
Figure BDA0002053277310000173
representing network functions fijThe unit overhead of the backup, namely the ratio of the overhead of the backup resources of the network function to the increment of the network function after the backup:
Figure BDA0002053277310000174
for various resource occupancies in the backup overhead, different requirements can be met by adjusting the weight coefficient. The BG can reflect the importance of the virtual network function in the whole service function chain, and the larger the value of the BG is, the larger the reliability improvement influence on the service function chain is when the corresponding virtual network function is backed up. Therefore, when backup is carried out, the virtual network function with the larger BG index is preferentially selected for backup.
The execution flow of the double sharing reliability deployment method based on the backup profit index is shown in fig. 1:
firstly, initializing and placing a service function chain according to the service requirement of a network and a corresponding deployment method;
secondly, reliability evaluation of the service function chain is carried out, and whether the reliability of the service function chain meets the reliability requirement of the service is judged;
if not, selecting the virtual network function with the largest backup profit index and the second largest backup profit index by calculating the backup profit indexes of all the virtual network functions in the whole service function chain, and carrying out double sharing reliability backup on the nodes meeting the requirements by a local search method; updating the virtual network function after double sharing backup;
fourthly, reliability evaluation of the service function chain is carried out again, and then the step III is repeated until the reliability of the service function chain meets the requirement required by the service.
Pseudo codes of the double sharing reliability deployment method based on the backup benefits are as follows:
Figure BDA0002053277310000175
Figure BDA0002053277310000181
3.2 virtual network function reliability deployment method based on backup revenue index and remapping
The dual-sharing reliability deployment method based on the backup benefits solves the problem of reliability of the service function chain in the initial deployment stage, but some problems still exist in the reliability of the service function chain and need to be solved; when a virtual network function fails in a service function chain, there may be a number of situations:
the virtual network function with fault has double sharing backup and the virtual network function sharing backup resource works normally.
The virtual network function with the fault has double backup, but the virtual network function sharing the backup resource has the fault at the same time.
And thirdly, the virtual network function with the fault is not backed up.
Performing corresponding processing aiming at various conditions by adopting a virtual network function reliability deployment method based on a backup profit index and remapping; and after the processing is finished, the problem that the reliability of the service function chain is reduced is effectively solved.
After the virtual network function is in fault, the virtual network function firstly enters a backup type judgment module for carrying out corresponding fault type judgment on the virtual network function in fault; when the type is judged, when the fault virtual network function has backup, the service flow is redirected to the node where the backup is located; if the network function for carrying out double sharing backup fails at the same time, the virtual network function which is firstly judged by the backup type judgment module is regarded as the existence of virtual network function backup, the virtual network function backup type sharing backup resources with the virtual network function is changed into no backup, and the later-entered virtual network function is directly regarded as the non-backup virtual network function for processing; and for the virtual network function which is not subjected to the reliability backup, remapping the virtual network function.
The remapping of the virtual network function uses a Simulated Annealing method (SAA), and a suitable node is selected by the Simulated Annealing method to be allocated to the virtual network function for instantiation. In the solution space of the simulated annealing method in this chapter, nodes satisfying the resource requirements are added to the solution space by traversing all the nodes
Figure BDA0002053277310000192
The initial solution n is in the solution space through a random function
Figure BDA0002053277310000193
Performing random selection; the evaluation function used in this chapter is C (n) ═ f (C)n,Mn,BnDelta tau), the service function chain end-to-end delay increment delta tau brought by the use condition of various resources of the computing node and the migration to the node n to be computed; the probability of accepting the difference solution is
Figure BDA0002053277310000194
The pseudo code of the simulated annealing method is as follows:
Figure BDA0002053277310000191
by implementing the remapping method for the virtual network function in the service function chain, the recovery problem of the virtual network function when the service function chain fails is solved. The invention designs a relevant solution aiming at the condition that the reliability of a service function chain is changed after a node fails and is remapped or migrated to a backup, and the flow chart is shown as figure 2:
when the virtual network function fails, firstly judging the backup type, and carrying out corresponding processing flow according to the judgment result; after the processing is finished, the deployment condition of the service function chain is input into the reliability deployment module, and multiple conditions may exist in the VNF fault:
firstly, migration to backup results in obvious reduction of the reliability of the original service function chain if migration to the backup node results in the need of re-checking whether the reliability of the service function chain meets the reliability requirement of the service; particularly, when one of the virtual network functions using the dual shared backup fails, the failed virtual network function needs to be recovered at the backup node, and at the same time, the reliability of the virtual network function sharing the backup resource with the failed virtual network function is reduced, and at this time, the reliability of the virtual network function sharing the backup resource in the service function chain needs to be updated.
And secondly, remapping the virtual network function, wherein when the virtual network function is remapped by adopting a simulated annealing method, the reliability of the remapped virtual network function instance may be changed, so that the overall reliability of the service function chain is changed.
Any situation that results in a change in the reliability of the service function chain requires a new round of reliability assessment on the service function chain to ensure that the reliability of the service function chain can always meet the reliability requirements of the service.

Claims (1)

1. A method for deploying virtual network function reliability based on backup income and remapping is characterized by comprising the following steps:
step 1, based on a double sharing backup mechanism of backup income, providing a backup income index virtual network function selection method with backup, and considering the reliability of a service function chain and the occupation effectiveness of system resources;
step 2, providing a solution based on a simulated annealing method for the mapping problem with the backup virtual network function;
step 3, a reliability compensation strategy is provided for the problem that the reliability of the service function chain is reduced after the backup recovery or remapping is carried out on the virtual network function with backup;
wherein the content of the first and second substances,
the concrete content of the step 1 is as follows:
double sharing equivalent model:
after the double sharing mechanism backup is carried out, the virtual network function vnf of the backupaCan be regarded as two virtual network functions vnfb、vnfc(ii) a Since two shares are assigned to the virtual network function vnfaSo that the two virtual network functions cannot function simultaneously,
only the virtual network function vnf is considered under the modeliReliability of (2):
ri′＝ri+(1-ri)(1-rj)rb (8)
the increment is as follows:
Δri′＝ri+(1-ri)(1-rj)rb-ri＝(1-ri)(1-rj)rb (9)
compared with the original exclusive backup protection, the reliability of the double shared backup is reduced as follows:
Δri-Δri′＝ra(1-ri)-(1-ri)(1-rj)rb (10)
wherein: r represents the reliability of the virtual network function, vnf before backupiHas a reliability of riVirtual network function vnfjHas a reliability of rj(ii) a Backup node for vnfiAnd vnfjRespectively has a reliability of ra、rb(ii) a Δ generally represents an increment, therefore, Δ riIs represented by riAn increment of (d);
to clarify the effect of backup, assume ra＝rb(ii) a The average reliability of the general service is substituted into 99.9%, the reliability of the double-sharing backup is reduced to 0.0000999% compared with the exclusive backup, even for the carrier-class service, the reduction of the reliability can be ignored, but the backup cost is reduced to almost half of the original backup cost;
in order to reduce the complexity of the model, a double-sharing equivalent model DSEM is provided for enabling the original backup virtual network function vnf to be realizediVnf with its doubly shared backupbViewed as a whole-a network function fiAs a new virtual network function vnfi', with a reliability of
ri′＝ri+(1-ri)(1-rj)rb (11)
Virtual network function vnfjVnf with its doubly shared backupcViewed as a whole-a network function fjAs a new virtual network function vnf'jWith a reliability of
rj′＝rj+(1-ri)(1-rj)rc (12)
Wherein the virtual network function vnfcHas a reliability of rc；
At this time, vnf is set for the virtual network functioni', the virtual network function vnf will be backed up by inspecting the occupation of the resourceaThe resource occupation of (a) is taken as a special resource occupation and is not distributed to a virtual network function vnfi、vnfjThereby reducing subsequent network functions fjComplexity of resource allocation problem during backup;
network model
NFV network G ═ N, E, G denotes an NFV network, where N denotes a set of nodes and E denotes a nodeA set of links between points; multiple virtual network function instances can be placed on each node, and each node n has own exclusive computing resource CnStorage resource MnAnd network resource BnVirtual network functions vnf placed on nodesiRequiring a node to allocate its own dedicated computing resource ciStorage resource miAnd network resource biAnd the sum of various resources occupied by the virtual network function deployed on the node must not exceed the resources owned by the node;
defining a service request in the NFV network, wherein the service request comprises a service type, a service reliability requirement and service strength; the service request type in the NFV network is the service function chain request siService function chain set denoted S, service function chain SiIs a set NF of network functions ordered by a certain logiciThe components are mixed; virtual network function set NFiThe virtual network function in (1) is denoted as fijEach network function fijCorresponding to one or more virtual network functions vnfijWherein i denotes that the virtual network function belongs to a service function chain siJ indicates that the virtual network function is the jth virtual network function that constitutes a service function chain, with a reliability of rijDefine a Boolean variable aijnWhen vnfijWhen placed on node n, aijn1 is ═ 1; when vnfijWhen not placed on node n, aijn0; the type of service request determines the service function chain siOf type (D), on which all network functions are denoted FiThe strength of the service request determines the size of the resources required by the virtual network functions that make up the service function chain; the reliability requirements of the service determine the service function chain siHas a reliability requirement of Ri；
After introducing the redundant backup of the virtual network function, the virtual network function vnfijThe network model of (2) needs to be changed to some extent; assuming deployment of a single virtual network function vnfijThe occupied computing, storage and network resources are respectively cij、mijAnd bijWith a reliability of rijNeedle, needleFor virtual network function vnfijDeploying redundant backups on node n
Figure FDA0003341665830000031
Need to be distributed
Figure FDA0003341665830000032
The required computing, storage and network resources are respectively ci、mi、biWhile for the network function fjIn other words, the reliability of the virtual network function vnf is improved to a certain extent, and in order to reduce the complexity of the model, the occupied resources are still regarded as the virtual network function vnfijOccupied resources, virtual network function backup
Figure FDA0003341665830000033
Network function f is not counted in occupied resourcesjOccupied resources, but including its service function chain siOccupied backup overhead;
redundant backups can increase the reliability of a virtual network function, thus defining a network function fijBack-up profit g (f)ij) Comprises the following steps:
Figure FDA0003341665830000034
wherein Δ r (f)ij) For increased network reliability after deployment of network functions,
the backup revenue for the service function chain is G(s)i) Comprises the following steps:
Figure FDA0003341665830000035
ΔR(si) Representing a chain of deployed service functions siIncreased network reliability
The virtual network function backup occupies a large amount of resources while increasing reliability, and the resources occupied by the backup cannot bring any benefit when the virtual network function is not failed, so that the resource occupation of the backup is reduced as much as possible while the reliability of the service function chain meets the service requirement, and the backup benefit of the service function chain is maximized;
optimizing the target:
Figure FDA0003341665830000036
constraint conditions are as follows:
Figure FDA0003341665830000041
Figure FDA0003341665830000042
Figure FDA0003341665830000043
Figure FDA0003341665830000044
Figure FDA0003341665830000045
the objective function (14) represents that maximizing backup of service function chains is an optimization objective for revenue text; the constraint (15) represents the reliability r of the service function chainSFCThe reliability requirements of the service need to be met
Figure FDA0003341665830000046
The constraint conditions (16) to (18) represent that the sum of the occupation of all virtual network functions on the node computing, storing and network resources cannot exceed the resource allocated by the node; the constraint (19) indicates that the virtual network function must be deployed to one node and can be deployed to only one node;
the specific content of the step 2 is as follows:
dual-sharing reliability deployment method based on backup benefits
The network service has corresponding reliability requirements on the service function chain, when the service function chain cannot meet the service reliability requirements, the reliability of a single network function is improved by performing redundant backup on the virtual network function forming the service function chain, so that the reliability of the service function chain meets the service requirements; how to backup can effectively improve the reliability of the service function chain so as to improve the effectiveness of the service function chain as much as possible, which is a difficult problem to be solved;
in a reliability system, in the system of which the overall performance depends on the performances of components, the reliability performance change of some of the components may play a more important role than other components, so that a classic Importance metric index, namely Birnbaum Importance metric Importance Measure, is introduced to characterize the Importance metric of the reliability of the components in the system; the formula for IM for component k is as follows:
Figure FDA0003341665830000047
rsystem(r) represents a reliability measure of the system r, rkThe incremental increase in the reliability of the component k,
IM represents the degree of importance of the reliability increment of the component k to the reliability of the system; inspired by IM, the method defines a network function fijBack-up yield BG:
Figure FDA0003341665830000051
wherein the content of the first and second substances,
Figure FDA0003341665830000052
representing a chain of network functions siThe incremental increase in reliability of the system (c),
Figure FDA0003341665830000053
representing network functions fijThe reliability of the operation of the system is improved,
Figure FDA0003341665830000054
representing network functions fijIncrement of reliability of fxy∈siRepresenting network functions fxyIn the network function chain si,
Wherein the content of the first and second substances,
Figure FDA0003341665830000055
representing network functions fijThe unit overhead of the backup, namely the ratio of the overhead of the backup resources of the network function to the increment of the network function after the backup:
Figure FDA0003341665830000056
Figure FDA0003341665830000057
and
Figure FDA0003341665830000058
respectively representing network functions fijThe occupation amount of various resources in the overhead of medium storage, calculation and bandwidth resource backup can adapt to different requirements by adjusting the weight coefficient; for various resource occupancies in the backup overhead, different requirements can be met by adjusting the weight coefficient; the BG can reflect the importance of the virtual network function in the whole service function chain, and the larger the value of the BG is, the larger the influence on the reliability improvement of the service function chain is when the corresponding virtual network function is backed up; therefore, when backup is carried out, a virtual network function with a large BG index is selected for backup;
firstly, initializing and placing a service function chain according to the service requirement of a network and a corresponding deployment method;
secondly, reliability evaluation of the service function chain is carried out, and whether the reliability of the service function chain meets the reliability requirement of the service is judged;
if not, selecting the virtual network function with the largest backup profit index and the second largest backup profit index by calculating the backup profit indexes of all the virtual network functions in the whole service function chain, and carrying out double sharing reliability backup on the nodes meeting the requirements by a local search method; updating the virtual network function after double sharing backup;
fourthly, reliability evaluation of the service function chain is carried out again, and then the third step is repeated until the reliability of the service function chain meets the requirements required by the service;
the specific content of the step 3 is as follows:
the method designs a relevant solution aiming at the condition that the reliability of the service function chain is changed after the node fails and is remapped or migrated to the backup,
when the virtual network function fails, firstly judging the backup type, and carrying out corresponding processing flow according to the judgment result; after the processing is finished, the deployment condition of the service function chain is input into the reliability deployment module, and multiple conditions may exist in the VNF fault:
firstly, migration to backup results in obvious reduction of the reliability of the original service function chain if migration to the backup node results in the need of re-checking whether the reliability of the service function chain meets the reliability requirement of the service; when one of the virtual network functions adopting the double shared backup fails, the failed virtual network function needs to be recovered on the backup node, and meanwhile, the reliability of the virtual network function sharing the backup resource with the backup node is reduced, and at the moment, the reliability of the virtual network function sharing the backup resource in the service function chain needs to be updated;
remapping the virtual network function, wherein when the remapping of the virtual network function is carried out by adopting a simulated annealing method, the reliability of the remapped virtual network function example may be changed, so that the overall reliability of a service function chain is changed;
any situation that results in a change in the reliability of the service function chain requires a new round of reliability assessment on the service function chain to ensure that the reliability of the service function chain can always meet the reliability requirements of the service.
CN201910380700.0A 2019-05-08 2019-05-08 Virtual network function reliability deployment method based on backup income and remapping Active CN110190987B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201910380700.0A CN110190987B (en) 2019-05-08 2019-05-08 Virtual network function reliability deployment method based on backup income and remapping

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201910380700.0A CN110190987B (en) 2019-05-08 2019-05-08 Virtual network function reliability deployment method based on backup income and remapping

Publications (2)

Publication Number Publication Date
CN110190987A CN110190987A (en) 2019-08-30
CN110190987B true CN110190987B (en) 2022-02-01

Family

ID=67715810

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201910380700.0A Active CN110190987B (en) 2019-05-08 2019-05-08 Virtual network function reliability deployment method based on backup income and remapping

Country Status (1)

Country Link
CN (1) CN110190987B (en)

Families Citing this family (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112749041B (en) * 2019-10-29 2023-12-26 中国移动通信集团浙江有限公司 Virtualized network function backup strategy self-decision method, device and computing equipment
CN111147307B (en) * 2019-12-30 2022-04-29 重庆邮电大学 Service function chain reliable deployment method based on deep reinforcement learning
CN111901170B (en) * 2020-07-29 2022-12-13 中国人民解放军空军工程大学 Reliability-aware service function chain backup protection method
CN114374608B (en) * 2020-10-15 2023-08-15 中国移动通信集团浙江有限公司 Slice instance backup task scheduling method and device and electronic equipment
CN112543119B (en) * 2020-11-27 2022-02-18 西安交通大学 Service function chain reliability deployment method based on deep reinforcement learning
CN113660131B (en) * 2021-10-18 2022-01-07 武汉绿色网络信息服务有限责任公司 Virtual network function unit switching method, device, controller and storage medium
CN113938930B (en) * 2021-12-16 2022-03-25 中国船舶重工集团公司第七二二研究所 Construction method of virtual network function forwarding graph adapting to 5G network multi-service scene
CN114448804A (en) * 2022-03-17 2022-05-06 长安大学 Service function chain mapping system and method based on information interaction connection duration optimization
CN115086175B (en) * 2022-04-26 2023-06-30 北京邮电大学 Resource optimization method and device
CN114944985B (en) * 2022-05-11 2023-11-24 重庆大学 Reliability-oriented service function chain backup method in resource heterogeneous environment
CN115378817B (en) * 2022-08-24 2023-09-22 广东电网有限责任公司 Resource backup method, device, equipment and storage medium of service function chain

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106603293A (en) * 2016-12-20 2017-04-26 南京邮电大学 Network fault diagnosis method based on deep learning in virtual network environment
EP3206335A1 (en) * 2014-11-03 2017-08-16 Huawei Technologies Co. Ltd. Virtual network function example migration method, device and system
CN107332913A (en) * 2017-07-04 2017-11-07 电子科技大学 A kind of Optimization deployment method of service function chain in 5G mobile networks
CN108965014A (en) * 2018-07-25 2018-12-07 北京智芯微电子科技有限公司 The service chaining backup method and system of QoS perception

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20070180509A1 (en) * 2005-12-07 2007-08-02 Swartz Alon R Practical platform for high risk applications

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP3206335A1 (en) * 2014-11-03 2017-08-16 Huawei Technologies Co. Ltd. Virtual network function example migration method, device and system
CN106603293A (en) * 2016-12-20 2017-04-26 南京邮电大学 Network fault diagnosis method based on deep learning in virtual network environment
CN107332913A (en) * 2017-07-04 2017-11-07 电子科技大学 A kind of Optimization deployment method of service function chain in 5G mobile networks
CN108965014A (en) * 2018-07-25 2018-12-07 北京智芯微电子科技有限公司 The service chaining backup method and system of QoS perception

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
朱晓荣.虚拟网络功能迁移研究综述.《 南京邮电大学学报(自然科学版)》.2018,(第3期), *

Also Published As

Publication number Publication date
CN110190987A (en) 2019-08-30

Similar Documents

Publication Publication Date Title
CN110190987B (en) Virtual network function reliability deployment method based on backup income and remapping
US7827435B2 (en) Method for using a priority queue to perform job scheduling on a cluster based on node rank and performance
US8112758B2 (en) Methods and apparatus for resource allocation in partial fault tolerant applications
CN102346460B (en) Transaction-based service control system and method
CN110535680B (en) Byzantine fault-tolerant method
Tsai et al. Two-tier multi-tenancy scaling and load balancing
CN110471820B (en) Cloud storage system disk fault prediction method based on cyclic neural network
US20130254590A1 (en) Real time database system
CN108965014A (en) The service chaining backup method and system of QoS perception
CN103139302A (en) Real-time copy scheduling method considering load balancing
CN108616394B (en) Virtual network function backup and deployment method
JP2007500895A (en) Dynamically configurable fault tolerance in autonomous computing with multiple service points
CN106339181A (en) Method and system for processing data in storage system
TW202133055A (en) Method for establishing system resource prediction and resource management model through multi-layer correlations
Zhai et al. Reliability-aware service function chain backup protection method
WO2012020698A1 (en) Primary-backup based fault tolerant method for multiprocessor systems
CN109165122B (en) Method for improving disaster recovery capability of application system same city multi-park deployment realized based on block chain technology
CN112749041A (en) Virtualized network function backup strategy self-decision method and device and computing equipment
CN106603645A (en) Large-scale cloud storage copy server consistency processing method and system
CN116302574B (en) Concurrent processing method based on MapReduce
CN108763312B (en) Slave data node screening method based on load
US7406486B1 (en) Transforming transactions to increase parallelism when replicating
CN106383995B (en) A kind of checkpoint laying method based on node failure relevance
CN113422664B (en) Multi-clock-source high-reliability time synchronization method
CN115048061A (en) Cold data storage method based on Raft

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