CN112994937A

CN112994937A - Deployment and migration system of virtual CDN in intelligent fusion identification network

Info

Publication number: CN112994937A
Application number: CN202110181613.XA
Authority: CN
Inventors: 陈佳; 张宏科; 高德云; 朱文兰; 邓悦达; 郭阔; 程鑫; 陈京
Original assignee: Beijing Jiaotong University
Current assignee: Beijing Jiaotong University
Priority date: 2021-02-10
Filing date: 2021-02-10
Publication date: 2021-06-18

Abstract

The invention provides a deployment and migration system of a virtual CDN in an intelligent identity network. The system comprises a CDN container virtualization module, a service management module, a node information collection module, a virtual CDN decision module, a deployment and migration issuing module and a route redirection module; the service management module is used for analyzing and obtaining a virtual CDN server which is matched with the request resource of the user and is closest to the user; the virtual CDN decision module makes an optimal decision according to the node information, the source IP address of the user and the virtual CDN server, and sends the virtual CDN arrangement decision information to the deployment migration issuing module; the deployment and migration issuing module is used for transmitting the newly deployed and migrated IP addresses of the virtual CDN to the route redirection module; the route redirection module realizes the function of route redirection. The invention provides a heuristic algorithm combining virtual CDN deployment and migration, which is used for deploying and migrating according to the whole network resources and user requests, thereby ensuring the quality of content delivery service and avoiding overload of a server.

Description

Deployment and migration system of virtual CDN in intelligent fusion identification network

Technical Field

The invention relates to the technical field of internet service quality, in particular to a deployment and migration system of a virtual CDN in an intelligent fusion identification network.

Background

With the increase of internet users and the expansion of application fields, the existing internet has more and more exposed defects, such as the problems that the service quality of the whole network is difficult to meet the requirements of users, the transmission efficiency is low, the resource utilization rate is low, and the like. Meanwhile, from the perspective of users, the service quality of network content is always the most concerned problem of internet providers, and due to the appearance of various novel applications and intelligent terminals, when various service forms access heterogeneous structures, the service quality requirements are different greatly, and the time delay of the traditional content transmission method is too high, so that the requirements of users are difficult to meet. Furthermore, due to the current trend of network traffic visualization, a new network architecture is needed to solve the application problem of massive data. Therefore, in the information field, there is an urgent need to design a new network architecture to solve the problems of the original design deficiency of the internet, which has also attracted attention from the academic and industrial circles, and the next generation of internet design and research has been developed in all countries. The Beijing university of transportation proposes an intelligent fusion identification network, which realizes the effective support of supply of personalized services according to the requirement and flexible networking through the intelligent fusion of the whole network with multi-space and multi-dimensional resources.

In the intelligent identification Network, a Virtualization technology is utilized to migrate the traditional Network function into a virtual service, and meanwhile, an NFV (Network function Virtualization) controller is utilized to perform virtual function arrangement operation, so that dynamic and flexible service supply is realized, and the user service quality and the server resource utilization rate are improved. Therefore, a CDN (Content Delivery Network) can be migrated into a Network virtual function through the NFV technology, the Network can solve the foregoing Content Delivery problem to a certain extent compared with a conventional CDN, the novel Network is a virtual Content Delivery Network (vdns), the virtual Content Delivery Network refers to a CDN Network operating in an NFV environment, the CDN components are virtualized by using the NFV technology to realize decoupling of CDN software and hardware, and the problem that the conventional CDN Network function is constrained on a physical server is overcome, so that a function of flexible deployment or migration of the virtualized CDN in a hardware server in a software form is realized, and the efficiency of deployment, migration, and upgrade update of the Content Delivery Network is improved. However, since there is no deployment/migration protocol introduced for the vdns in the research on the vdns today, there is very little research on solving the problem of Virtualized Network Function (VNF) placement and migration in the CDN using an accurate algorithm, and there is no consideration for optimization of actual system and Network parameters that meet delivery requirements. .

The CDN is a set in which elements in the internet cooperate with each other, where content is cached on multiple duplicate Web servers, and then a request of a user is redirected to an optimal duplicate Web server by route redirection, and the server responds to the request and provides the content. In accessing content, CDNs have overcome the inherent limitations of the traditional internet from the user's perspective, improving quality of service. Meanwhile, through the content distribution network, the service provider can improve the bandwidth to the maximum extent, and the correctness of transmission is ensured through the cache of the content so as to improve the network performance. Typical functions of a CDN include: the request routing redirection and content delivery function directs the request to the most appropriate CDN cache server by using a mechanism for avoiding congestion, thereby avoiding sudden access caused by congestion; a content outsourcing and distribution service to copy or cache content from an origin server into a distributed Web server; a content negotiation service to meet specific needs of each user or group of users; and the management service is used for managing all network components, monitoring and reporting the use condition of the content in the network and the like. The CDN is added into the intelligent identification network to reduce the content acquisition delay and improve the user service quality.

The CDN distributes the content to a group of CDN cache servers distributed on the whole world, so that a reliable and high-speed transmission mode can be realized, the end user can enjoy better service quality, and meanwhile, the content can be sent to cache nodes in various places in a pushing mode to be cached in advance. When a user requests, the user is redirected to the optimal edge CDN cache server through the request routing server. In this way, the user communicates with a nearby CDN server without awareness and obtains content from that server. However, due to the characteristics of the existing CDN, in the existing content delivery network, since the edge cache server is implemented by a special hardware device, the deployment, migration and upgrade costs of the edge cache server are high, and the entire network is inflexible. Secondly, acquiring the resource information of the edge cache server requires developing a program according to a corresponding interface of the hardware device, which is high in cost and lacks flexibility. Thirdly, in the arrangement scheme of the conventional content distribution network, in order to cope with the burst traffic, redundancy deployment is performed on each edge node server, and rapid capacity expansion cannot be performed.

In the white paper of the NFV standard system architecture published by ETSI in 2013, virtualization of the CDN is provided, and the problems of complex deployment, migration and upgrading operations of the traditional CDN are solved through a virtualization technology. However, in the current virtual CDN deployment scheme, a research team targets on a vndn deployment or migration problem, and a mainstream scheme is to optimize the deployment or migration, which inevitably brings about problems of server overload and service quality degradation. If only virtual CDN optimal deployment is performed for new traffic, server overload is likely to cause overload, and if only migration is considered, degradation of service quality may be caused to a certain extent in the migration process. How to guarantee the service quality of users while avoiding the overload of the server is a need to fully consider the content distribution network arrangement. Next, in the current research of the virtual CDN, an actual system is lacked to support migration and deployment of the virtualized CDN, and the virtual CDN server cannot be flexibly scheduled in the content delivery network. In the process of forwarding a data packet by a Network, a conventional content distribution Network is transmitted based on a conventional switch and a router, and an SDN (Software Defined Network) technology is not integrated in the Network to realize separation of control and forwarding of the Network, and to centrally control and monitor Network resources.

In the current research of vCDN deployment and migration, in the face of the theory of the algorithm complexity of the difficult combinatorial optimization problem, there are mainly two solutions of the virtual machine/virtual network function Placement Problem (PP), namely 1) a specific method (optimal algorithm) and 2) a heuristic method (for example, best fit decreasing, first fit decreasing, genetic and meta-heuristic). While some relevant solutions to solve VM placement using heuristic methods can be found, few studies have been made to solve the above problems using precise algorithms.

In the prior art, a method for controlling resources of different operators and managing memory space at a vndn node includes: the method comprises the steps of obtaining VCDN resource capacity information of at least one node, wherein the VCDN resource capacity information comprises VCDN space use conditions of different providers on the at least one node, and respectively controlling VCDN resources of the different providers on the at least one node according to the VCDN resource capacity information. The specific control operations include: and responding to the content distribution request, determining a distribution node according to the VCDN space use conditions of different providers on the at least one node, and distributing corresponding content to the distribution node. The problem that VCDN resources are not clearly managed and controlled in the related technology is solved, and the method has the beneficial effect that the VCDN resources can be finely controlled according to the VCDN space use conditions of different providers.

The method for controlling resources of different operators and managing memory space at the vndn node in the prior art has the following disadvantages: although the scheme provides a management method for the vCDN resources, the scheme is only limited to selection and use of virtual vCDN nodes, and a scheme capable of guaranteeing and improving the service quality of users is not provided, and the vCDN nodes cannot be flexibly migrated and deployed. And the management and control information on the virtual vCDN nodes is not collected in detail, and is only limited to resources such as storage space and bandwidth provided by an operator, and global scheduling and control cannot be realized due to lack of centralized management and control.

Another resource allocation method for a virtualized content distribution network in the prior art includes: the method is applied to an Operation Management System (OMS) of a virtualized content delivery network (vCDN), and comprises the following steps: sending a resource configuration adjustment request to a Virtual Network Function Manager (VNFM) of a telecom cloud TIC node; the resource allocation adjustment request is a capacity expansion request or a newly-built configuration request. Wherein, the method also comprises: sending a resource application query request to a VNFM of a telecom cloud TIC node; receiving resource application feedback information sent by the VNFM after resource query is carried out on a Network Function Virtualization Orchestrator (NFVO) according to the resource application query request; and sending the resource configuration adjustment request to a VNFM of a corresponding TIC node according to the resource application feedback information. Acquiring the capacity requirement of an edge service node SNS of all signed content provider CP customers of the CDN in a future specific time period; and sending the resource application query request to a VNFM of the TIC node according to the capacity requirement. And sending the resource configuration adjustment request to a VNFM of the corresponding TIC node according to the resource application feedback information.

Another prior art method for providing a resource allocation method for a virtualized content distribution network has the following disadvantages: although the deployment and management method provided by the method also combines NFV, deployment and migration are not combined, and the method used by statistical information is relatively resource-consuming.

Disclosure of Invention

The embodiment of the invention provides a deployment and migration system of a virtual CDN in an intelligent identification network, which aims to solve the problems that the traditional CDN needs redundant deployment, cannot quickly respond to a user request according to the whole network resource information and flexibly deploy and migrate a server.

In order to achieve the purpose, the invention adopts the following technical scheme.

A deployment and migration system of a virtual CDN in an intelligent identity network comprises the following steps: the system comprises a CDN container virtualization module, a service management module, a node information collection module, a virtual CDN decision module, a deployment and migration issuing module and a route redirection module;

the CDN container virtualization module is used for constructing executable files, configuration files, dependency libraries and file systems for containers according to the requirements of each CDN component, realizing Network Function Virtualization (NFV) of a CDN server by using a container virtualization technology, storing a standard mirror image of the virtual CDN server into a mirror image warehouse, and setting a container label for each virtual CDN server;

the service management module is used for receiving user request information transmitted by the edge computing node, analyzing the user request information to obtain a virtual CDN server which is matched with a request resource of a user and is closest to the user, transmitting a source IP address of the user and the virtual CDN server which are obtained through analysis to the virtual CDN decision module, and triggering the node information collection module;

the node information collection module is used for collecting node information under the triggering of the service management module and transmitting the collected node information to the virtual CDN decision module;

the virtual CDN decision module is used for making an optimal decision according to the node information transmitted by the node information collection module, the source IP address of the user and the virtual CDN server transmitted by the service management module, and issuing virtual CDN arrangement decision information to the deployment migration issuing module;

the deployment and migration issuing module is used for issuing deployment or migration commands to corresponding edge computing nodes according to the received virtual CDN deployment decision information, initializing hardware resources of the virtual CDN, starting corresponding virtual network functions, responding to content requests of users, and transmitting newly deployed and migrated IP addresses of the virtual CDN to the routing redirection module after the deployment and operation of the virtual CDN are completed;

and the route redirection module is used for interacting with the DNS server through the control node according to the received newly deployed and migrated IP addresses of the virtual CDN, and guiding the traffic accessing the virtual CDN to the newly deployed or migrated virtual CDN server.

Preferably, the system further comprises:

and the network integration module is used for controlling data forwarding of the virtual CDN server by utilizing the SDN controller, configuring a network plug-in of a newly deployed virtual CDN node through the CNI plug-in, distributing an IP address to the virtual CDN, designating the gateway as a Linux bridge and connecting the virtual CDN to the Linux bridge.

Preferably, the node information collected by the node information collection module includes allocable memory, allocable CPU, allocable storage space, allocable pos number, bandwidth size, network topology and traffic statistics information.

Preferably, the virtual CDN decision module is specifically configured to perform greedy deployment on an initial virtual CDN server through data traffic statistics of the virtual switch OVS, match an edge computing node closest to a user according to a received source IP address of the user, use the edge computing node as an optimal deployment node, determine that resources of the optimal deployment node can deploy a virtual CDN server transmitted by the service management module for the user requirement, and if yes, perform deployment operation at the optimal deployment node; otherwise, according to the network topology information, matching adjacent nodes through a greedy method, calculating the priority of the migration nodes according to the capacity limit, the migration cost and the content factors acquired nearby by the CDN, migrating the virtual CDN server from the overloaded optimal deployment node to the edge calculation node with the highest priority, and issuing the deployment and migration strategies of the virtual CDN to a deployment migration issuing module.

Preferably, the deployment and migration issuing module is specifically configured to invoke an NFV controller to perform deployment and migration operations on a designated edge computing node according to the decision-making layout information, initialize hardware resources of the virtual CDN, open a corresponding virtual network function, respond to a content request of a user, obtain IP addresses of all virtual CDN nodes by invoking the NFV controller after the deployment and migration operations, find an intersection between an IP address set before the deployment/migration operations and an IP address set after the deployment/migration operations, obtain a source IP address of the newly deployed or migrated virtual CDN server and a modified IP address, and transmit the source IP address of the virtual CDN server and the IP address after the deployment and migration as input data to the routing redirection module.

Preferably, the route redirection module is specifically configured to, after receiving a newly deployed IP address of the virtual CDN, process the file by a function rforrdpplying () according to the view file matched to the IP address, where the function is operated to automatically add an a record; after receiving the migrated IP address of the virtual CDN, calling an rRFormigrating () function, matching a corresponding DNS view file according to a source IP address of a user, traversing the field of the NS view file, finding an analysis A record of a virtual CDN server before migration, and modifying the IP address analyzed by the A record into the IP address of the virtual CDN after migration; after the deployed or migrated DNS configuration file modification function is executed, the DNS is reconfigured by the function reloadDNS ().

According to the technical scheme provided by the embodiment of the invention, the heuristic algorithm combining virtual CDN deployment and migration is provided, deployment and migration are carried out according to the whole network resources and the user request, the content delivery service quality can be ensured, and the overload of the server can be avoided.

Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings needed to be used in the description of the embodiments are briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts.

Fig. 1 is a structural diagram of a deployment and migration system of a virtual CDN in an intelligent identity network according to an embodiment of the present invention;

fig. 2 is a schematic diagram illustrating an implementation principle of a CDN container virtualization module according to an embodiment of the present invention;

fig. 3 is a schematic diagram illustrating an implementation principle of a service management module according to an embodiment of the present invention;

fig. 4 is a schematic diagram illustrating an implementation principle of a node information collection module according to an embodiment of the present invention;

fig. 5 is a schematic diagram illustrating an implementation principle of a virtual CDN decision module according to an embodiment of the present invention;

fig. 6 is a schematic diagram illustrating an implementation principle of deploying a migration issuing module according to an embodiment of the present invention;

fig. 7 is a schematic diagram illustrating an implementation principle of a network integration module according to an embodiment of the present invention;

fig. 8 is a schematic diagram illustrating an implementation principle of a route redirection module according to an embodiment of the present invention.

Detailed Description

Reference will now be made in detail to embodiments of the present invention, examples of which are illustrated in the accompanying drawings, wherein like reference numerals refer to the same or similar elements or elements having the same or similar function throughout. The embodiments described below with reference to the accompanying drawings are illustrative only for the purpose of explaining the present invention, and are not to be construed as limiting the present invention.

As used herein, the singular forms "a", "an", "the" and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will be further understood that the terms "comprises" and/or "comprising," when used in this specification, specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof. It will be understood that when an element is referred to as being "connected" or "coupled" to another element, it can be directly connected or coupled to the other element or intervening elements may also be present. Further, "connected" or "coupled" as used herein may include wirelessly connected or coupled. As used herein, the term "and/or" includes any and all combinations of one or more of the associated listed items.

It will be understood by those skilled in the art that, unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. It will be further understood that terms, such as those defined in commonly used dictionaries, should be interpreted as having a meaning that is consistent with their meaning in the context of the prior art and will not be interpreted in an idealized or overly formal sense unless expressly so defined herein.

For the convenience of understanding the embodiments of the present invention, the following description will be further explained by taking several specific embodiments as examples in conjunction with the drawings, and the embodiments are not to be construed as limiting the embodiments of the present invention.

The design of a virtual CDN deployment and migration mechanism system in an intelligent identification network must fully consider the defects of the existing content distribution network. The method can ensure the service quality of the user while avoiding the overload of the server, and the content distribution network arrangement needs to be fully considered. Secondly, the SDN technology is integrated in the network to realize the separation of the control and the forwarding of the network, and the centralized control and the monitoring of the network resources. Finally, considering the change of the IP address after the migration of the virtual CDN deployment node, how to obtain the IP address of the virtual CDN after the migration or deployment, and the automatic route redirection are also one of the requirements of system design.

In combination with the above analysis, deployment and migration of the virtual CDN in the intelligent identity network have the following functional requirements:

(1) CDN virtualization function: the system needs to realize an edge cache server through software, and utilizes a container virtualization technology to migrate a traditional CDN server into an NFV (network file virtualization) to realize decoupling of hardware and software.

(2) Service management function: the system needs to be able to parse the information requested by the user, match the virtual CDN server closest to the user, and send it as input to the decision making module.

(3) An information collection function: the system needs to be able to acquire real-time hardware and network information of the node through the NFV controller and the SDN controller. The hardware information comprises CPU, memory, storage space and Pods which can be allocated to each node. And the network information comprises topology information and bandwidth, and the effectiveness of the whole network resource information transmitted into the decision module is ensured.

(4) A decision making function: the system can make a decision of combination of deployment and migration according to user request information and whole network resource information, and compromise between server load and service quality is achieved.

(5) Deployment and migration functions: the system can call the NFV controller to perform virtual CDN deployment and migration operations according to a decision making result.

(6) Network integration function: the system can integrate SDN in a network part, monitor the network resources of the whole system through an SDN controller, and realize the separation of the control and forwarding of the system network, so that the system network is flexible and programmable.

(7) Routing redirection function: after the deployment and migration operations of the virtual CDN are completed, the system needs to acquire an IP address corresponding to a new virtual CDN, and automatically realize route redirection.

The structure diagram of the deployment and migration system of the virtual CDN in the intelligent convergence identity network provided by the embodiment of the present invention is shown in fig. 1, and in the virtual CDN environment of the intelligent convergence identity network, the system can make deployment and migration decisions of the virtual CDN cache server according to request information of a user and resources of the whole network, and redirect a route automatically. The whole system design mainly comprises seven functional modules, namely a CDN container virtualization module, a service management module, a node information collection module, a virtual CDN decision module, a deployment and migration issuing module, a network integration module and a route redirection module. The functions of the respective modules are described below.

1: CDN container virtualization module

In the CDN container virtualization module, all components of the CDN are migrated into network function virtualization, implemented by containers, and centrally stored in a mirror repository. This module is the basis for virtual CDN deployment and migration. The CDN components that need to perform container virtualization include an edge cache server, a route redirect server, and a content source station. The container virtualization module can operate the CDN component in a software mode, so that decoupling of hardware and software is realized; the assembly can be packaged into a standard mirror image and transmitted into a mirror image warehouse, executable files, environment dependence and configuration required by each assembly are stored, when the assembly needs to be deployed or migrated, the mirror image of the assembly is directly pulled from the mirror image warehouse, virtual service can be directly operated, the effect of 'one-time construction and everywhere operation' of the CDN assembly is achieved, deployment and migration are carried out based on second level based on container virtualization, user requests can be responded more quickly, and service quality is improved.

An implementation principle schematic diagram of a CDN container virtualization module provided in the embodiment of the present invention is shown in fig. 2, and first pull a standard centros operation mirror as an initial container; secondly, according to the requirements of each component, constructing an executable file, a configuration file, a dependency library and a file system for the container, and confirming that the container can realize the functions required by the component; then packaging the manufactured standard mirror image and transmitting the standard mirror image into a Docker Hub mirror image warehouse, and setting a container label for each component; and finally, deploying or migrating the virtual CDN components at each edge computing node by a Kubernetes container arrangement tool, directly pulling the mirror image of the CDN components from the mirror image warehouse according to the label and running corresponding virtual services.

2: service management module

The service management module has the main functions of analyzing user request information and triggering the node information collection module in real time. When a user requests content, each edge computing node collects information of the user request and transmits the information to the service management module, the service management module analyzes the information of the user request and inputs the result to the decision module, and meanwhile, the node information collection module is triggered to complete the collection of the real-time input information of the virtual CDN decision module. The service management module can analyze and obtain a source IP address of the user, and the source IP address is transmitted to the decision-making module to be used for matching an edge computing node closest to the user; the size of the request content of the user can be analyzed to obtain the resources needing to occupy the virtual CDN node, so that the decision module can judge whether the edge computing node has enough load to provide service.

Fig. 3 is a schematic diagram illustrating an implementation principle of a service management module according to an embodiment of the present invention, where when a user requests content, the service management module first collects user information from a trigger node information module, and then analyzes a source ip requested by the user, the requested content, and resources that need to be occupied, and transmits the source ip, the requested content, and the resources to a deployment and migration module of a control node as input. Triggering information collection after a user request can ensure the accuracy and real-time performance of the node physical resource information and the network information collected by a network manager, then transmitting the information into a virtual CDN decision module, and deploying and migrating the virtual CDN according to the user request and the whole network resource information.

3: node information collection module

The node information collection module is also a basic module of the whole system, the virtual CDN decision module needs the node information module to provide the whole network resource information, and the virtual CDN decision module cannot make a decision if the information of the module is not input. The node information module can provide hardware information (CPU, memory, storage space and assignable Pods) of each node, so as to judge the hardware load of the node; network resource information (network topology, bandwidth and traffic statistics) can be provided for the virtual decision module, wherein the network topology information can facilitate the decision module to match the edge computing node closest to the user, thereby performing deployment and selection of the migration node.

Therefore, the node information collection module needs to collect the following information in common:

the memory can be allocated: the virtual CDN is deployed in the edge computing node and needs to specify memory allocation, and if the memory of the node is not enough to meet the user requirement, the performance and the service quality of content distribution are affected. Therefore, before deployment and migration, the allocable memory of each edge computing node needs to be known, so as to avoid overload of the server memory.

The CPU can be allocated: the CPU is a processing core of each computing node, the number of CPUs needs to be assigned for distribution when the virtual CDN server is deployed and migrated in the system, the decision module needs to know the assignable CPU of each edge computing node before decision, and the processing speed of the virtual CDN is affected if the CPU is fully loaded.

The memory space can be allocated: the virtual CDN serves as a proxy cache server, needs a storage space to cache content requested by a user, and cannot implement a proxy cache function if the allocable storage space is smaller than the content requested by the user. It is therefore necessary to know the allocatable storage resources of the edge compute nodes to determine if a virtual CDN can be deployed and provide normal services.

The number of assignable Pods: during the data, the Kubernetes controller limits the container deployment quantity of each edge computing node, and the data is collected to judge whether each edge computing node can also deploy a new virtual CDN server.

The bandwidth size is as follows: the size of the bandwidth of each node is mainly used for judging the network flow condition, if the bandwidth of the node exceeds the upper limit or is nearly full, the node replies RST messages to other data packets, and a user cannot normally access the virtual CDN server.

Network topology: the network topology of the edge computing node can be used for matching the edge computing node closest to the user, the information is mainly provided for a virtual CDN decision module, and decision making of the node to be deployed and the node to be migrated is carried out through topology information.

And (3) flow statistics: when the virtual CDN is initially deployed, greedy deployment needs to be performed according to traffic statistics of each edge computing node, where the data is obtained by calling the OVS switch of each edge computing node through the SDN controller.

An implementation principle schematic diagram of a node information collection module provided in an embodiment of the present invention is shown in fig. 4, where a service management module triggers the node information collection module to collect resource information of a whole network, may invoke an NFV controller to obtain allocable memory, CPU, storage space, and Pods quantity information of each node, and invokes an SDN controller to obtain topology information of the whole network and bandwidth size and traffic statistical information of each node. The information of each node is generated into an object, the attribute corresponding to the object is assigned, and then the assigned value is transmitted into a virtual CDN decision module, so that the real-time performance and the accuracy of the input information are guaranteed.

4: virtual CDN decision module

The virtual CDN decision module makes an optimal decision according to the information of the node information collection module and the information input by the service management module, and issues a migration or deployment command. In the decision algorithm part, a heuristic algorithm combining deployment and migration is mainly considered. However, the migration of the virtual CDN causes extra forwarding delay and may cause interruption of the previous service or degradation of the service quality. The control plane needs to ensure that the virtual CDN servers are deployed at the optimal deployment node without overloading the servers. Wherein the optimal deployment node refers to the nearest edge computing node matched with the IP address of the user. If the optimal deployment node server appears or is about to be overloaded, the server of the node needs to be migrated. And it is very important to which server the CDN content is migrated to ensure the quality of service for CDN content delivery and the "nearby acquisition" principle and minimize migration cost. In addition, different control situations have different control targets, and during migration, various factors, including load of network functions, traffic size, migration delay of different traffic, requested content size and resources to be allocated, should be considered to determine the migration and deployment processes. Therefore, in order to make accurate decision making of deployment and migration combination, the following two points need to be fully considered: (1) the triggering condition (2) of the migration operation is effective migration/deployment flow selection.

Fig. 5 is a schematic diagram illustrating an implementation principle of a virtual CDN decision module according to an embodiment of the present invention. Firstly, before a user content request, a decision module performs initial deployment of a virtual CDN (content delivery network), performs greedy deployment of the virtual CDN through data flow statistics of an OVS (virtual switch), and then when a user requests information, the module matches an edge computing node closest to the user, namely an optimal deployment node, according to a user IP address input by a service management module. Secondly, according to the resource information collected by the node information module and calculating the resource limit condition of the optimal deployment node, if the optimal deployment node has no virtual CDN server and has enough hardware resources and network resources to deploy the virtual CDN server, the deployment operation is executed at the optimal deployment node, if the resources of the optimal deployment node are not enough to deploy the virtual CDN server required by the user, the adjacent nodes are matched by a greedy method according to the network topology information, the priority of the migration node is calculated according to three factors of capacity limit, migration cost and CDN nearby acquisition content, the virtual CDN server is migrated from the overloaded optimal deployment node to the edge computing node with the highest priority, finally the strategy of virtual CDN deployment and migration is issued to the deployment migration issuing module to perform the deployment or migration operation of the virtual CDN, and the work of routing redirection is completed, providing services to the user.

Pseudo code for a heuristic algorithm for policy making is as follows:

5: deployment migration issuing module

An implementation principle schematic diagram for deploying a migration issuing module provided in the embodiment of the present invention is shown in fig. 6, and its function is to, according to the virtual CDN orchestration decision information of the virtual CDN decision module, issue a deployment or migration command to a corresponding edge computing node in time, initialize hardware resources of the virtual CDN, simultaneously start a corresponding virtual network function, and respond to a content request of a user. Secondly, after the deployment and operation of the virtual CDN are completed, the deployment migration issuing module acquires an IP address newly deployed and migrated, transmits the IP address to the route redirection module, and performs a route redirection operation.

The working flow of the deployment, migration and issuing module is as follows: and the virtual CDN decision module issues the scheduling decision of the virtual CDN to the deployment migration issuing module, and the deployment migration issuing module calls the NFV controller to perform deployment and migration operations on the designated edge computing node according to the decision scheduling information. Secondly, after deployment and migration operations, acquiring IP addresses of all virtual CDN nodes by calling an NFV controller, solving an intersection of an IP address set before the deployment/migration operations and an IP address set after the deployment/migration operations, then acquiring a source IP address and a modified IP address of a newly deployed or migrated virtual CDN server, and finally transmitting the source IP address and the IP address after the deployment and migration as input data to a route redirection module for automatic route redirection.

6: network integration module

The network integration module has the main functions of integrating an SDN network into the virtual CDN system and controlling data forwarding of the virtual CDN server by utilizing an SDN controller. By calling the NFV controller, the deployment migration issuing module can perform deployment and migration of the virtual CDN according to the decision module. Then, newly deployed virtual CDN nodes need network plug-ins of all the nodes to be configured, an OVS switch is arranged in a bottom data plane of the system, the network plug-ins are responsible for connecting the virtual CDN to the OVS switch, and network resources are managed in a centralized mode through an SDN controller, so that the network of the whole system is more flexible, and forwarding control and separation are achieved.

Fig. 7 shows a schematic diagram of an implementation principle of a network integration module according to an embodiment of the present invention, where a processing procedure of the model includes: when a computing node is initialized, an OVS bridge is required to be connected with a physical Network port and a port of a Linux bridge, and when a deployment migration issuing module issues deployment or migration operation of a virtual CDN at a control node, a Kubelet of the computing node calls a CNI (Container Network Interface, API Interface of a Container Network) plug-in to allocate a Network of the virtual CDN, where the CNI plug-in needs to take charge of functions of:

1) an IP address is assigned to the virtual CDN.

2) The gateway is designated as a Linux bridge.

3) The virtual CDN is connected to a Linux bridge.

Through the network integration module, the virtual CDN server network can be automatically distributed, all cross-node data packets in the system pass through the OVS bridge, and therefore the cross-node data packets need to be forwarded according to flow table rules issued by the SDN controller. Therefore, centralized management of SDN controller on whole network resources and data forwarding and system integration of SDN and NFV are realized.

7: route redirection module

The route redirection module is mainly responsible for guiding all traffic accessing the virtual CDN to a newly deployed or migrated virtual CDN server after the deployment/migration operation is completed, thereby implementing a function of route redirection. The route redirection interacts with the DNS through the control node, automatically modifies the corresponding DNS analysis file through a control node calling program, and reloads the configuration to realize the real-time update of the DNS analysis under the condition of not restarting the DNS.

In the system implementation, a BIND (Berkeley Internet Name Domain service) server is adopted as a DNS server to undertake the routing redirection work. BIND is open source software realized by a DNS, the server can provide functions of domain name resolution, forwarding, sub-domain authorization, view and the like, the server is the most widely used DNS server software in the world at present, and more than half of DNS servers of the Internet are constructed and realized by the BIND. The view function of the server can provide different DNS resolution views for addresses of different source IPs, namely, the same domain name obtains different resolution results according to the source IPs, so that content requests accessing a source station can be redirected to edge virtual CDN nodes, an optimal server is selected for different users, and the delay of access service is reduced by reducing the geographical link distance.

Fig. 8 is a schematic diagram illustrating an implementation principle of a route redirection module according to an embodiment of the present invention, where a specific processing procedure includes: firstly, after the deployment and migration issuing module completes the deployment and migration operation, the deployed and migrated IP is transmitted to the route redirecting module. After the route redirection module receives the request of route redirection, it will make a judgment on the executed operation, if it is the deployment operation, it needs to match the view file corresponding to the IP address according to the IP address of the user, and the file is processed by the function rfordelay (), where the operation of the function is to automatically add an a record. And if the matching operation is migration, an rRForming () function is required to be called, the function is responsible for matching a corresponding DNS view file according to the source IP address of the user, then, fields of the file are traversed, an analysis A record of the virtual CDN server before migration is found, the IP address analyzed by the A record is modified, and the IP address is changed into the IP address of the virtual CDN after migration. After the deployed or migrated DNS configuration file modification function is executed, the DNS needs to be reconfigured through a function reloadDNS (), and a command used by the function is rndc reload, namely, the configuration can be imported into an analysis module of the DNS in real time, so that the routing redirection work after the virtual CDN is deployed is realized.

Through the above process, it can be obtained that after the virtual CDN node is deployed, the operation of the route redirection module is an operation of adding an a record to the view file corresponding to the virtual CDN, so addition is used instead of replacement, because after addition is completed, the DNS server performs polling analysis on a plurality of a records, thereby also achieving an effect of load balancing of the virtual CDN within one node, and avoiding a single point of failure. And when the migration operation is executed, the record A is directly replaced, so that the flow of the virtual CDN in the node can be redirected to the virtual CDN server in the migrated node, the overload condition of the source node is effectively avoided, and the operation of route redirection is also completed. The functions involved in the route redirection module are shown in tables 4-5.

Table 2-1 route redirection module main functions and description

The virtual CDN decision-making module is the core of the invention, if the virtual CDN decision-making module is deployed for the first time, the virtual CDN is deployed according to data statistical information of the OVS, if the virtual CDN decision-making module is not deployed for the first time, when the service management module receives a user request, the service management module and the node information collection module transmit collected information into the virtual CDN decision-making module, and the module makes a virtual CDN arrangement decision according to the whole network resource information and the user request. In the scheduling decision, a heuristic algorithm combining virtual CDN deployment and migration is mainly provided.

Firstly, the virtual CDN decision module receives the whole network resource information of the node information module, and needs to calculate whether node information of the virtual CDN close to the user is overloaded. If the node server is overloaded, the executed function is vcdnigitrestratgy (), the function is responsible for calculating the migration cost from the source node to the migration node based on the user request information, and then sequentially judging the load condition of the nodes to be selected according to the descending order. And if the migration node load with the highest priority can bear the migrated node, the node is the migrated node. Secondly, if the optimal node server is not overloaded, deployment of the virtual CDN instance is carried out, and the resource is appointed to be right, wherein the called function is vCDNDeployStrategy ().

In summary, the embodiment of the present invention provides a virtual CDN deployment and migration mechanism based on an intelligent identity network. Network resources are collected through an SDN controller, node hardware resources are collected through an NFV controller, a service management module collects user request information, so that decision and operation of virtual CDN deployment or migration are carried out, then automatic routing redirection is carried out, and a new user request is guided to a newly deployed or migrated virtual CDN node. Wherein, each module is realized by programming and the validity of each module is verified by a test environment. The virtual CDN deployment and migration are researched, a heuristic algorithm combining the virtual CDN deployment and migration is provided, and the deployment and migration are carried out according to the whole network resources and the user request, so that the server load overload can be avoided while the content delivery service quality can be ensured.

The system solves the problems that the traditional CDN needs redundant deployment, cannot quickly respond to user requests according to the whole network resource information and flexibly deploy and migrate the server. The system provided by the invention designs a virtual CDN system framework, combines a container virtualization technology and utilizes Docker as a carrier of a vCDN node component, realizes high virtualization of the vCDN node and flexible deployment of the vCDN node, and can support migration in an intelligent fusion identification network.

The embodiment of the invention designs a method for integrating a vCDN (virtual content distribution network) and an intelligent fusion identification network, namely integrating the vCDN as a network function into the intelligent fusion identification network. And the coupling of a network virtual function vCDN and an underlying network is realized through a network integration module. The method for realizing the combination of vCDN node migration and deployment by utilizing the NFV controller in the intelligent identity network is characterized in that the centralized control migration and deployment are realized by utilizing pod packed vCDN nodes.

The embodiment of the invention designs a heuristic algorithm for making a strategy in a vCDN system, which can be applied to an intelligent fusion identification network, comprehensively considers the problems of migration and deployment, proposes a migration and deployment strategy which is made by using a NFV controller for centrally controlling vCDN nodes to replace the migration and deployment by utilizing the deployment to replace the migration

The embodiment of the invention designs a mechanism for collecting and formulating vCDN deployment and migration strategy information, namely a mechanism for collecting network topology information by using an SDN controller and collecting node information including memory, computing resources and the like by using an NFV controller. And various resources can be considered more comprehensively when the strategy is formulated.

Those of ordinary skill in the art will understand that: the figures are merely schematic representations of one embodiment, and the blocks or flow diagrams in the figures are not necessarily required to practice the present invention.

From the above description of the embodiments, it is clear to those skilled in the art that the present invention can be implemented by software plus necessary general hardware platform. Based on such understanding, the technical solutions of the present invention may be embodied in the form of a software product, which may be stored in a storage medium, such as ROM/RAM, magnetic disk, optical disk, etc., and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the method according to the embodiments or some parts of the embodiments.

The embodiments in the present specification are described in a progressive manner, and the same and similar parts among the embodiments are referred to each other, and each embodiment focuses on the differences from the other embodiments. In particular, for apparatus or system embodiments, since they are substantially similar to method embodiments, they are described in relative terms, as long as they are described in partial descriptions of method embodiments. The above-described embodiments of the apparatus and system are merely illustrative, and the units described as separate parts may or may not be physically separate, and the parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment. One of ordinary skill in the art can understand and implement it without inventive effort.

The above description is only for the preferred embodiment of the present invention, but the scope of the present invention is not limited thereto, and any changes or substitutions that can be easily conceived by those skilled in the art within the technical scope of the present invention are included in the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.

Claims

1. A deployment and migration system of a virtual CDN in an intelligent identity network is characterized by comprising: the system comprises a CDN container virtualization module, a service management module, a node information collection module, a virtual CDN decision module, a deployment and migration issuing module and a route redirection module;

2. The system of claim 1, further comprising:

3. The system according to claim 1, wherein the node information collected by the node information collection module includes allocable memory, allocable CPU, allocable storage space, allocable quantities of Pods, bandwidth size, network topology and traffic statistics.

4. The system of claim 1, wherein:

the virtual CDN decision module is specifically used for greedy deploying an initial virtual CDN server through data flow statistics of a virtual switch OVS, matching an edge computing node closest to a user according to a received source IP address of the user, taking the edge computing node as an optimal deployment node, judging that resources of the optimal deployment node can deploy the virtual CDN server transmitted by the service management module of user requirements, and if so, executing deployment operation at the optimal deployment node; otherwise, according to the network topology information, matching adjacent nodes through a greedy method, calculating the priority of the migration nodes according to the capacity limit, the migration cost and the content factors acquired nearby by the CDN, migrating the virtual CDN server from the overloaded optimal deployment node to the edge calculation node with the highest priority, and issuing the deployment and migration strategies of the virtual CDN to a deployment migration issuing module.

5. The system of claim 1, wherein:

the deployment and migration issuing module is specifically configured to invoke an NFV controller to perform deployment and migration operations on designated edge computing nodes according to decision-making arrangement information, initialize hardware resources of the virtual CDN, start a corresponding virtual network function, respond to a content request of a user, obtain IP addresses of all virtual CDN nodes by invoking the NFV controller after the deployment and migration operations, find an intersection between an IP address set before the deployment/migration operations and an IP address set after the deployment/migration operations, obtain a source IP address of a newly deployed or migrated virtual CDN server and a modified IP address, and transmit the source IP address of the virtual CDN server and the IP address after the deployment and migration as input data to the route redirection module.

6. The system of claim 1, wherein:

the route redirection module is specifically configured to, after receiving a newly deployed IP address of the virtual CDN, process the file by a function rforrdploying () according to the view file matched to the IP address, where the function is operated by automatically adding an a record; after receiving the migrated IP address of the virtual CDN, calling an rRFormigrating () function, matching a corresponding DNS view file according to a source IP address of a user, traversing the field of the NS view file, finding an analysis A record of a virtual CDN server before migration, and modifying the IP address analyzed by the A record into the IP address of the virtual CDN after migration; after the deployed or migrated DNS configuration file modification function is executed, the DNS is reconfigured by the function reloadDNS ().