WO2019072171A1

WO2019072171A1 - Method and apparatus for network functions virtualization service chain mapping in annular network

Info

Publication number: WO2019072171A1
Application number: PCT/CN2018/109498
Authority: WO
Inventors: 卢华; 王延松; 吴少勇; 陈立全; 乔志
Original assignee: 中兴通讯股份有限公司
Priority date: 2017-10-09
Filing date: 2018-10-09
Publication date: 2019-04-18
Also published as: CN109639447A; CN109639447B

Abstract

Embodiments of this disclosure provide a method and apparatus for network functions virtualization service chain mapping in an annular network. The method comprises the following steps: performing server group region pre-partition on virtualized network functions of a service chain; allocating servers participating in resource allocation in network functions virtualization to a corresponding server group region; and during deployment of user traffic, selecting the corresponding virtualized network functions on the server group region corresponding to the virtualized network functions on the service chain required by the user traffic for deployment.

Description

Method and device for virtualizing service chain mapping of network function under ring network

Technical field

Embodiments of the present disclosure relate to, but are not limited to, the field of network communication technologies, and in particular, to a method and apparatus for virtualizing a service chain mapping of a network function under a ring network.

Background technique

The Network Functions Virtualization Infrastructure (NFV) technology is a technology for realizing network functions by virtualizing on the x86 server common platform in order to solve the problem of the versatility of the existing communication devices.

In some cases, the network function virtualization service chain mapping method may firstly obtain the percentage of the customer's demand for each virtualized network function (VNF) through historical data or experience, and randomly deploy from the largest to the smallest. On different servers; next time every user traffic comes, check if there is a virtual machine that can make the service chain not fold back; if it exists, select the path; if it does not exist, correct it by turning on the virtual machine Business chain path. The principle of virtual machine opening is based on the principle of maximization.

However, there may be the following problem: when the resources on a certain server are used, the newly opened virtual machine (VM) will have a larger distance to the existing VM running the VNF, which results in a business chain. Fold back.

Summary of the invention

According to an embodiment of the present disclosure, a method for virtualizing a service chain mapping of a network function under a ring network includes the following steps: pre-partitioning a server group area for a virtual network function of a service chain; and participating in virtualizing a network function The server of the resource allocation is allocated to the corresponding server group area; and when the user traffic is deployed, the virtual network corresponding to the server group area corresponding to each virtual network function in the service chain required by the user traffic is selected. Features are deployed.

According to an embodiment of the present disclosure, an apparatus for virtualizing a service chain mapping of a network function under a ring network includes: a dividing module configured to perform pre-partitioning of a server group area for a virtual network function of a service chain; and an allocation module, It is configured to allocate a server participating in resource allocation in the network function virtualization to the corresponding server group area; and a deployment module configured to select the service chain required by the user traffic when deploying the user traffic The corresponding virtual network function on the server group area corresponding to each virtual network function is deployed.

According to an embodiment of the present disclosure, an apparatus for virtualizing a service chain mapping of a network function under a ring network includes: a memory, a processor, and a computer program stored on the memory and operable on the processor, wherein The processor performs the following steps: performing server group area pre-division on the virtual network function of the service chain; assigning the server participating in the resource allocation in the network function virtualization to the corresponding server group area; When the user traffic is deployed, the corresponding network function on the server group area corresponding to each virtual network function in the service chain required by the user traffic is selected for deployment.

DRAWINGS

The drawings described herein are provided to provide a further understanding of the present disclosure and constitute a part of this application. The illustrative embodiments of the present disclosure and the description thereof are for explaining the present disclosure and do not constitute an undue limitation of the present disclosure. In the drawing:

FIG. 1 is a flowchart of a method for virtualizing a service chain mapping of a network function under a ring network according to an embodiment of the present disclosure;

2 is a schematic diagram of a ring networking mode in accordance with an embodiment of the present disclosure;

3 is a flowchart of a server group area pre-division phase for a virtual network function of a service chain according to an embodiment of the present disclosure;

4 is a diagram showing an example of VNF region pre-division according to an embodiment of the present disclosure;

5 is a flow chart of a distribution server phase in accordance with an embodiment of the present disclosure;

6 is a flowchart of server area repartitioning according to an embodiment of the present disclosure;

FIG. 7 is a schematic diagram of an apparatus for virtualizing a service chain mapping of a network function under a ring network according to an embodiment of the present disclosure.

Detailed ways

The embodiments of the present disclosure will be described in detail below with reference to the accompanying drawings. It should be noted that the embodiments in the present disclosure and the features in the embodiments may be arbitrarily combined with each other without conflict.

FIG. 1 is a flowchart of a method for virtualizing a service chain mapping of a network function under a ring network according to an embodiment of the present disclosure. As shown in FIG. 1, the method includes: step 11, step 12, and step 13.

In step 11, the server group area pre-division is performed on the VNF of the service chain.

In step 12, a server participating in resource allocation in the NFV is allocated to the corresponding server group area.

In step 13, when the user traffic is deployed, the corresponding network function on the server group area corresponding to each VNF in the service chain required by the user traffic is selected for deployment.

In an embodiment, the step of selecting a corresponding network function on the server group area corresponding to each VNF in the service chain required by the user traffic (step 13) includes the following steps: when determining the user traffic If the idle resources of the server group area corresponding to each VNF in the required service chain are insufficient, the server group area is re-divided, and the corresponding VNFs in the corresponding server group area after the re-division are selected for deployment.

In an embodiment, the step of re-division of the server group area comprises the steps of: determining a server group area with the most remaining current resources; and from the server group area with the most remaining current resources to the user traffic A server group area corresponding to each VNF in the required service chain, and sequentially assigns a specified resource from each server group area to the next server group area in the label order until the specified resource is allocated to the user The server group area corresponding to each VNF on the service chain required for traffic.

The method of the embodiment of the present disclosure can improve the efficiency of NFV resource mapping in the ring networking mode, reduce the return of network traffic, and comprehensively consider the CPU and memory resources to effectively improve the NFV bandwidth resource utilization.

Embodiments of the present disclosure will be described below with reference to the accompanying drawings.

2 is a schematic diagram of a ring networking mode in accordance with an embodiment of the present disclosure. As shown in Figure 2, during the NFV resource allocation process in the ring network, multiple x86-based servers exist to carry the VNF function. The data traffic originates from the starting point of the ring network, and the formed virtual network service chain traffic flows in the ring network. Each server has a VM running on it, and the VNF works on top of the VM. This is the application scenario of NFV resource allocation in ring networking mode.

The method for mapping the virtualized service chain of the network function in the ring network of the embodiment of the present disclosure includes: pre-sharing the service chain VNF area, allocating the server, deploying the traffic of the service chain, checking whether there are idle resources on the entire server group, and performing the server on the server. The area is subdivided into 4 parts.

The meanings of symbols such as num _xi , map _ij , x _k , n , VMNum _ij and the like used in the embodiments of the present disclosure are explained below.

Num _xi is the identifier of the i-th deployment mode of the service chain x. It is used to determine the mapping mode of each service chain to the area, and its initial value is zero. The longest service chain contains α VNFs, and the server group is also divided into α areas, and the number of the fields is recorded as 1, 2, ..., α. The numbers of the VNFs in the service chain are marked according to their service chain order. If a certain VNF is mapped to the area k, the VNF is said to exist on the area k.

The map _ij indicates the area in which the jth VNF of the i-th service chain is mapped. Ind _ij represents the number of VMs that are roughly required by the jth VNF on the i-th service chain. p _i represents the user's demand for the i-th business chain based on historical experience. Cap _j represents the number of customers that a single VM can accommodate the jth VNF.

x _k represents the number of VMs that need to be opened on the kth area.

The server number near the origin of the traffic is server 1, and the remaining numbers are 2, 3, 4...t.

n represents the total number of VMs that can be opened on the server group.

VMNum _ij represents the number of VMs required by the jth VNF in the i-th service chain.

The mapping mode of the service chain is a set of VNF mappings, so a certain mapping mode of a certain service chain is one of the permutation combinations of the VNF mapping. In addition, the idle network resource refers to the VNF running on the opened VM can continue to accommodate the resources of the customer traffic; and the idle resource refers to the resource that has the VM opened but not loaded with the VNF, that is, the CPU and memory that are not loaded with the VNF. Resources.

3 is a flow diagram of a server farm zone pre-partitioning phase of a virtual network function of a service chain in accordance with an embodiment of the present disclosure. As shown in FIG. 3, the server group pre-segmentation phase is performed on the VNF of the service chain, and each VNF is mapped to each area of the ring group network in order from long to short, so that the same VNF is obtained. Is mapped to the same area. The above process includes: Step 101 - Step 103.

At step 101, the maximum length α of the NFV service chain in the ring network is counted, and the server group is first divided into α areas, and the sequence mapping is started from the longest service chain.

At step 102, the remaining service chains are sequentially mapped according to the length of the service chain from long to short, and all the mapping modes are traversed without violating the order of the service chain, and the VNFs on the service chains and the server group are sequentially The group area is mapped.

At step 103, the sum of the number of VNF types in each server group area in all mapping modes is compared, and the mapping relationship between the VNF of each service chain and the server group area in the mapping mode with the smallest sum is selected.

The method of step 103 may be: when traversing to the i-th mapping mode, checking the mapping of all VNFs in the i-th mapping mode; when such a VNF already exists in the mapped region, adding the statistical value num _xi to 1 After all traversal is completed, select the maximum mapping mode of num _xi as the mapping method to be used; at the same time, remember that the number of VNF types on a certain area i is n _i , and the method of selecting the smallest sum of n _i as the mapping method to be used is exhausted. It is possible to have the same VNF mapped to the same area.

For example, suppose that the VNF types required by the user are the following two: one, DPI-FW-NAT, and second, FW-NAT. Fig. 4 shows an example of pre-division of the VNF region in this case.

In the embodiment of the present disclosure, there are three types of services that need to be mapped in the ring network: Deep Packet Inspection (DPI), Firewall (FW), and Network Address Translation (NAT). . As shown in FIG. 4, in the embodiment of the present disclosure, the server group is divided into three regions. For the mapping service chain DPI-FW-NAT, the service chain has one and only one mapping mode, that is, the DPI on the service chain will be mapped on the area 1, the FW will be mapped on the area 2, and the NAT will be mapped in the area. 3 on.

For the mapping service chain FW-NAT, under the premise that the service chain satisfies the mapping policy, there are three mapping modes:

1) FW mapping is on area 1, and NAT is mapped on area 2;

2) FW mapping is on area 1, NAT is mapped on area 3;

3) The FW map is on area 2 and the NAT is mapped on area 3.

Next, calculate the number and sum of VNFs in each of the three mapping modes:

1) There are two kinds of VNF in region 1, two kinds of VNF in region 2, and one kind of VNF in region 3, a total of five kinds;

2) There are two kinds of VNFs in region 1, one VNF in region 2, and one VNF in region 3, for a total of four species;

3) There is one VNF in the region 1, one VNF in the region 2, and one VNF in the region 3, and a total of three kinds.

By comparing the sum of the number of VNF types in each of the three types of mapping methods, the third embodiment of the present disclosure selects the third mapping method that minimizes the sum of the number of VNF types as the mapping method to be used.

FIG. 5 is a flow chart of a distribution server phase in accordance with an embodiment of the present disclosure. As shown in FIG. 5, in the allocation server stage, the service chain is first estimated according to the demand of the customer for each service chain and the size of the resources required by each VNF in the service chain when the user requests the service chain service. The size of the resources required by the VNF, and then according to the mapping relationship of the service chain to the area obtained by the above steps, further estimate the resource size required for each server group area. The estimating step includes: Step 201 - Step 204.

At step 201, the number of VMs ind _ij required for each VNF in each service chain is calculated. If the jth VNF does not exist on the i-th service chain, ind _ij =0; otherwise, according to the user p _{i i} demand chain lines of business, in conjunction with a single VM can accommodate VNF j th number of customers CAP _j, calculated _{_{_{ind ij = p i / cap j}}} , and rounding.

In the embodiment of the present disclosure, first, the DPI is labeled 1, the FW is 2, the NAT is 3, the service chain DPI-FW-NAT is the service chain 1, and the service chain FW-NAT is the service chain 2. . Then, ind _{11 is} first calculated, and ind ₁₂ , ind ₁₃ , ind ₂₁ , ind ₂₂ , ind _{23 are} calculated in turn, wherein the value of ind ₂₃ is 0, that is, the third VNF does not exist on the service chain 2.

At step 202, based on the ind _ij and the total number n of VMs that can be opened on the server group, the number of VMs VMNum _ij required for the jth VNF in the i-th service chain can be calculated:

And rounded up.

In the embodiment of the present disclosure, the number of VMs required for the first VNF in the first service chain is calculated.

among them,

Represents the sum of all ind _ij .

The above formula means that the number of VMs required for a single VNF is equal to the total number of VMs on the server equal to the sum of the values of the identifiers of the VNFs than the values of all the identifiers, that is, the VNFs required for each service chain. The resource size is proportional to the value of the identifier.

At step 203, the jth VNF defining the i-th service chain is mapped to the region value map _ij , and the kth is calculated according to the resource size VMNum _ij and the VNF mapping relationship map _ij required by the VNF. The number of VMs x _k to be opened on the area is first to make x _k =0, then traverse map _ij , if map _ij = k during traversal, then x _k = x _k + VMNum _ij .

In the embodiment of the present disclosure, the number of VMs to be turned on in the first region x _{1 is} calculated, first let x ₁ =0, then traverse map _ij , because only map ₁₁ =1, the final value of x ₁ is 2 .

At step 204, it is determined that the server group corresponding to the area k is placed in accordance with the order rule, and the area marked with a small number is preferentially placed on the server with a small label.

In the embodiment of the present disclosure, the server 1 is assigned to the area 1; the server 2 and the server 3 are assigned to the area 2; and the server 4 and the server 5 are assigned to the area 3.

For example, suppose that the number of users that can be carried by a VNF loaded on a single VM is the same, and each VM can carry 500 user traffic; it is assumed that the user has the same demand for the two service chains; assuming that there are a total of five servers in the scenario, Two VMs can be turned on per server. Based on experience, it is easy to calculate that each VNF on all service chains requires 2 VM-sized resources. According to the result of the pre-segmentation phase of the VNF region of the service chain, the number of VMs required for the region 1 is 2, the number of VMs required for the region 2 is 4, and the number of VMs required for the region 3 is 4.

After the server is allocated, whenever user traffic comes, the user traffic needs to be deployed. The deployment of user traffic is based on the principle that the VNF uses the resources of its mapped area. The deployment process can include the following four steps.

1) Check the area corresponding to each VNF in the service chain, assuming that the current service chain number is i, traverse the VNF with the label j∈[1...α], and α is the length of the service chain numbered i.

2) If map _ij is 0, the next traversal is performed; otherwise, it is checked whether the network resource of the jth VNF in the area map _ij is sufficient, and if not, the process jumps to step 3). If the resources are sufficient, proceed to the next traversal.

3) Check if there are still free resources in the map _ij area (that is, if there is still no VNF loaded on the VM). If there is no idle resource, the server area is repartitioned; if it exists, the corresponding VNF is loaded, and if j is smaller than α, j=j+1, and jumps to step 2).

4) Deploy user traffic. After the above traversal, there are sufficient network resources in the area corresponding to each VNF, and the user traffic will select the server mapped by each VNF in the required service chain as the route through which the traffic will pass.

6 is a flow diagram of server area repartitioning in accordance with an embodiment of the present disclosure. As shown in FIG. 6, if the idle resources of a certain area are insufficient, the server area is re-divided, and the steps include: Step 301-Step 303.

At step 301, a server group area with the most remaining resources is determined.

Determine the server group area with the most remaining resources, and record it as x _free ; judge the relative position of x _free and the server group area x _short lacking resources, and set the area close to the user traffic origin as the front, if x _free before x _short, step 302 is executed, otherwise, step 303.

From the server group area where the current resource remains the most to the server group area corresponding to each VNF on the service chain required by the user traffic, the specified resources are sequentially allocated from each server group area in the label order. A server group area, until the specified resource is allocated to the server group area corresponding to each VNF on the service chain required by the user traffic, as in step 302 and step 303.

At step 302, the server number closest to the origin of the traffic is server 1, and the remaining numbers are numbered 2, 3, 4...t. First, the server with the largest label in x _free is divided into resources that can open a VM to x _free +1. If there is no idle resource in the area with the largest server label, open a VNF running on the server with the highest label on the other server in the server group area, and the VNF on the server with the highest label and the traffic above it. Migrating in the past. Thus, in the label maximum server, there can be idle resources. Then, divide the resources that can open a VM to x _free +1.

The same operation is performed on the server group area such as x _free +1 and x _free +2 until the target area x _{short is} re-divided. At this point, it is equivalent to dividing x _free resources into x _short and guaranteeing regional continuity.

At step 303, the server number closest to the origin of the traffic is server 1, and the remaining numbers are numbered 2, 3, 4...t. First, divide the server with the smallest label in x _free into a resource that can open a VM to x _free -1. If there is no free resource in the area with the smallest server label, then open a VNF running on the server with the lowest label on the other server in the area, and migrate the VNF on the server with the lowest label and the traffic above it. past. Thus, in the minimum number of servers, there can be idle resources; the same operation is performed on the server group areas such as x _free -1 and x _free -2 until the target area x _{short is} re-divided. At this point, it is equivalent to dividing x _free resources into x _short and guaranteeing regional continuity.

For example, in the embodiment of the present disclosure, assuming that there are idle resources on the area 1, and there is no idle resource on the area 3, the step of re-allocating the idle resources on the area 1 to the area 3 may include the following steps a) and b).

In step a), the area 1 is denoted as x _free ; the relative position of the x _free and the resource lacking area x _short is judged, since x _free , that is, the area 1 is before the x _short area 3, step b) is performed.

In step b), there is no free resource on the server 2 with the highest number in the area 1, ie, the server 2. Therefore, the FW needs to be loaded on the server 1, and the user traffic carried by the FW on the server 2 is migrated to the FW of the server 1, and then the FW of the server 2 is closed. Next, the free VM on server 2 can be divided into zone 2. Then, since there is an idle resource on the server with the largest label in the area 2, the VM can be directly allocated to the area 3.

After the above steps are completed, the user traffic of the service chain needs to be re-deployed to implement the mapping of the specific VNF of the service chain. The steps of the foregoing operations include the following steps: traversing all VNFs that have network resources on the entire server group, arranging them together to generate a current traffic deployment plan, and selecting a deployment plan that satisfies the minimum total traffic as the deployment plan that we will use.

According to the present disclosure, a method for mapping a network function virtualization service chain under a ring network, by dividing a region and mapping the same VNF in the same region, and performing a regional re-division method, solves the service chain traffic in the ring network. The problem of reentry is generated, which improves the utilization and overall efficiency of bandwidth resources in the NFV resource allocation process.

FIG. 7 is a schematic diagram of an apparatus for virtualizing a service chain mapping of a network function under a ring network according to an embodiment of the present disclosure. As shown in FIG. 7 , the apparatus for mapping a network function virtualization service chain under the ring network provided by the embodiment of the present disclosure includes: a division module, an allocation module, and a deployment module.

The dividing module is configured to perform server group area pre-division on the VNF of the service chain.

The allocation module is configured to allocate a server participating in resource allocation in the NFV to the corresponding server group area.

The deployment module is configured to deploy the corresponding VNF on the server group area corresponding to each VNF in the service chain required by the user traffic when the user traffic is deployed.

In an embodiment, the deployment module is configured to select a network function corresponding to each VNF corresponding to each VNF in the service chain required by the user traffic to be deployed by: determining the user traffic. When the idle resources of the server group area corresponding to each VNF in the required service chain are insufficient, the server group area is re-divided, and the corresponding VNF on the corresponding divided server group area is selected for deployment.

In an embodiment, the deployment module is configured to re-divide the server group area by: determining a server group area with the most remaining current resources; from the server group area where the current resource remains the most a server group area corresponding to each VNF in the service chain required by the user traffic, and sequentially assigning specified resources from each server group area to the next server group area in the label order until the specified resources are divided. The server group area corresponding to each VNF on the service chain required for the user traffic is given.

In an embodiment, the deployment module is configured to sequentially allocate a specified resource from each server group area to a next server group area by: the server group area ratio of the current resource remaining most The server group area corresponding to each VNF in the service chain required by the user traffic is closer to the traffic origination, and the specified resource is sequentially allocated to the next server group from the server with the largest label in each server group area. If the server with the largest label does not have an idle resource, the VNF running in the server with the highest label is opened on the other servers in the server group area, and the VNF of the server with the largest label is set. And the traffic on the VMF is migrated to the open VNF, and the specified resource is allocated to the next server group area from the server with the largest label; if the server group area with the most remaining current resources is more than the user traffic The server group area corresponding to each VNF on the required service chain is far away from the traffic origination, and then from each The server with the lowest label in the server group area divides the specified resource into the next server group area. If the server with the lowest label does not have idle resources, then one of the other servers in the server group area is being opened. VNF running in the server with the smallest label, the VNF of the server with the lowest label and the traffic on the VNF are migrated to the open VNF, and the specified resource is divided from the server with the smallest label to the next Server group area.

In an embodiment, the dividing module comprises: a dividing unit, a mapping unit, and a selecting unit.

The dividing unit is configured to count the maximum length α of the NFV service chain in the ring network, and divide the server group into α server group areas.

The mapping unit is configured to traverse all the mapping manners according to the length of the service chain from long to short, and sequentially map the VNFs on the service chains and the server group regions.

The selecting unit is configured to compare the sum of the number of VNF types in each server group area in all mapping modes, and select a mapping relationship between the VNF of each service chain and the server group area in the mapping mode with the smallest sum.

In an embodiment, the allocation module comprises: an estimating unit and an allocating unit.

The estimating unit is configured to estimate resources required for each VNF of the service chain and resources required for each of the server group areas.

The allocating unit is configured to sequentially allocate the server to the corresponding server group area according to the estimated resource according to the label order of the server.

In an embodiment, the estimating unit is further configured to: estimate resources required by each virtual network function of the service chain and resources required for each of the server group areas by: following a customer-to-business The demand quantity of the chain, and the size of resources required by each VNF in the service chain when the user requests the service chain service, and estimate the resource size required by each VNF of the service chain; according to the resources required by each VNF of the service chain The size estimates the number of servers required for each of the server group areas.

An embodiment of the present disclosure further provides an apparatus for virtualizing a service chain mapping of a network function under a ring network, comprising: a memory, a processor, and a computer program stored on the memory and operable on the processor, wherein the processing When the program is executed, the following steps are implemented: pre-partitioning the server group area of the VNF of the service chain; assigning the server participating in the resource allocation in the NFV to the corresponding server group area; and deploying the user traffic, The network function corresponding to each VNF corresponding to each VNF in the service chain required by the user traffic is selected for deployment.

The embodiment of the present disclosure further provides a computer readable storage medium storing computer executable instructions, when the computer executable instructions are executed, implementing network function virtualization under the ring network described in the embodiments of the present disclosure. The method of mapping business chains.

One of ordinary skill in the art will appreciate that all or a portion of the steps described above can be accomplished by a program that instructs the associated hardware, such as a read-only memory, a magnetic or optical disk, and the like. All or part of the steps of the above embodiments may also be implemented using one or more integrated circuits. Correspondingly, each module/unit in the foregoing embodiment may be implemented in the form of hardware or in the form of a software function module. The present disclosure is not limited to any specific form of combination of hardware and software.

The above is merely an exemplary embodiment of the present disclosure. Of course, there are other various embodiments of the present disclosure. A person skilled in the art can make various corresponding changes and modifications in accordance with the present disclosure without departing from the spirit and scope of the disclosure, but the corresponding changes and modifications should be included in the claims appended to the present disclosure. The scope of protection.

Claims

A method for virtualizing a service chain mapping of a network function under a ring network includes the following steps:

Pre-partitioning the server group area for the virtual network function of the service chain;

Allocating a server participating in resource allocation in network function virtualization to the corresponding server group area;

When the user traffic is deployed, the corresponding virtual network function corresponding to each virtual network function in the service chain required by the user traffic is selected for deployment.
The method of claim 1, wherein the step of selecting the corresponding network function on the server group area corresponding to each virtual network function on the service chain required by the user traffic comprises the following steps:

When it is determined that the idle resource of the server group area corresponding to each virtual network function in the service chain required by the user traffic is insufficient, the server group area is re-divided, and the corresponding server group after the re-division is selected. The corresponding virtual network function in the area is deployed.
The method of claim 2 wherein said step of repartitioning said server group area comprises the steps of:

Determine the server group area where the current resources remain the most; and

From the server group area where the current resource remains the most to the server group area corresponding to each virtual network function on the service chain required by the user traffic, the designated resources are sequentially divided from each server group area in the label order. The next server group area is allocated until the specified resource is allocated to the server group area corresponding to each virtual network function on the service chain required by the user traffic.
The method of claim 3, wherein the step of sequentially dividing the designated resource from each server group area to the next server group area comprises the following steps:

If the server group area with the most remaining current resources is closer to the traffic originating area than the server group area corresponding to each virtual network function on the service chain required by the user traffic, then each server group area is sequentially obtained. The server with the largest label divides the specified resource into the next server group area. If the server with the largest label does not have idle resources, the server with the largest label is opened on the other servers in the server group area. The virtual network function running in the server migrates the virtual network function of the server with the largest label and the traffic on the virtual network function to the opened virtual network function, and then allocates the specified resource from the server with the largest label to the next one. Server group area;

If the server group area corresponding to the current resource remaining is greater than the server group area corresponding to each virtual network function in the service chain required by the user traffic, the server group area is away from the traffic origination, and then the server group area is sequentially marked. The server with the smallest number divides the specified resource into the next server group area. If the server with the lowest label does not have idle resources, open a server with the lowest label in other servers in the server group area. Running the virtual network function, migrating the virtual network function of the server with the lowest label and the traffic on the virtual network function to the opened virtual network function, and then dividing the specified resource from the server with the smallest label to the next A server group area.
The method of claim 1 wherein said step of pre-partitioning server farm regions for virtual network functions of the service chain comprises the steps of:

Counting the maximum length α of the network function virtualization service chain in the ring network, and dividing the server group into α server group areas;

Tracing all mapping modes according to the length of the service chain from long to short, and sequentially mapping the virtual network functions on each service chain with the server group area;

Comparing the sum of the number of virtual network function types in each server group area in all mapping modes, and selecting the mapping relationship between the virtual network function of each service chain and the server group area in the mapping mode with the smallest sum.
The method of claim 1, wherein the step of assigning a server participating in resource allocation in network function virtualization to the corresponding server group area comprises the steps of:

Estimating the resources required for each virtual network function of the service chain and the resources required for each of the server group areas;

And assigning the server to the corresponding server group area in sequence according to the estimated resources according to the labeling order of the server.
The method of claim 6 wherein said step of estimating resources required for each virtual network function of said service chain and resources required for each of said server group areas comprises the steps of:

Estimating the resource size required for each virtual network function of the service chain according to the demand of the customer for each service chain and the amount of resources required by each virtual network function in the service chain when the user requests the service chain service;

The number of servers required for each of the server group areas is estimated based on the resource size required for each virtual network function of the service chain.
A device for virtualizing a service chain mapping of a network function under a ring network includes:

a partitioning module, configured to pre-divide a server group area for a virtual network function of the service chain;

An allocation module configured to allocate a server participating in resource allocation in network function virtualization to the corresponding server group area;

The deployment module is configured to deploy the corresponding virtual network function on the server group area corresponding to each virtual network function in the service chain required by the user traffic when the user traffic is deployed.
The device of claim 8 wherein

The deployment module is configured to select a network function corresponding to each virtual network function corresponding to each virtual network function in the service chain required by the user traffic to be deployed: when determining the service required by the user traffic When the idle resource of the server group area corresponding to each virtual network function in the chain is insufficient, the server group area is re-divided, and the corresponding virtual network function on the re-divided corresponding server group area is selected for deployment.
The device according to claim 9, wherein

The deployment module is configured to re-divide the server group area by: determining a server group area with the most remaining current resources; and from the server group area with the most remaining current resources to the user traffic a server group area corresponding to each virtual network function in the required service chain, sequentially assigning specified resources from each server group area to the next server group area in the order of labels until the specified resources are allocated to the The server group area corresponding to each virtual network function on the service chain required for user traffic.
The device of claim 10, wherein

The deployment module is configured to sequentially allocate a specified resource from each server group area to a next server group area by: if the server group area with the most remaining current resources is required by the user traffic The server group area corresponding to each virtual network function in the service chain is closer to the traffic origination point, and then the specified resource is divided into the next server group area from the server with the largest label in each server group area. If the server with the largest label has no idle resources, the virtual network function running in the server with the largest label is opened on the other servers in the server group area, and the virtual network function of the server with the largest label is set. And the traffic on the network migrates to the opened virtual network function, and then allocates the specified resource from the server with the largest label to the next server group area;

If the server group area corresponding to the current resource remaining is greater than the server group area corresponding to each virtual network function in the service chain required by the user traffic, the server group area is away from the traffic origination, and then the server group area is sequentially marked. The server with the smallest number divides the specified resource into the next server group area. If the server with the lowest label does not have idle resources, open a server with the lowest label in other servers in the server group area. Running the virtual network function, migrating the virtual network function of the server with the lowest label and the traffic on the virtual network function to the opened virtual network function, and then dividing the specified resource from the server with the smallest label to the next A server group area.
The apparatus of claim 8 wherein said dividing module comprises:

The dividing unit is configured to count the maximum length α of the network function virtualization service chain in the ring network, and divide the server group into α server group areas;

The mapping unit is configured to traverse all the mapping manners according to the length of the service chain from long to short, and sequentially map the virtual network function on each service chain and the server group area;

The selecting unit is configured to compare the sum of the number of virtual network function categories in each server group area in all mapping modes, and select a virtual network function of each service chain and a mapping of the server group area in the mapping mode with the smallest sum. relationship.
The apparatus of claim 8 wherein said distribution module comprises:

An estimating unit, configured to estimate resources required for each virtual network function of the service chain and resources required for each of the server group areas;

And an allocating unit, configured to sequentially allocate the server to the corresponding server group area according to the estimated resource according to the label order of the server.
The device of claim 13 wherein

The estimating unit is further configured to: estimate resources required by each virtual network function of the service chain and resources required for each of the server group areas by: according to a previous customer demand for each service chain, and The size of the resources required by each virtual network function in the service chain when the user requests the service chain service, and the resource size required for each virtual network function of the service chain; and the virtual network function required according to the service chain. The resource size estimates the number of servers required for each of the server group areas.
An apparatus for virtualizing a service chain mapping of a network function under a ring network includes: a memory, a processor, and a computer program stored on the memory and operable on the processor, wherein when the processor executes the program Implement the following steps:

Pre-partitioning the server group area for the virtual network function of the service chain;

Allocating a server participating in resource allocation in network function virtualization to the corresponding server group area;

When the user traffic is deployed, the corresponding network function on the server group area corresponding to each virtual network function in the service chain required by the user traffic is selected for deployment.