CN114448810A - Bandwidth pulling method, device and electronic device - Google Patents

Abstract

Embodiments of the present invention relate to a bandwidth pulling method, device, and electronic device. The method includes: determining a server group to be migrated, a first amount of bandwidth used by each unit of bandwidth using bandwidth in the server group to be migrated, and a to-be-migrated server group. The second bandwidth amount to be migrated for the server group of and determine the redundant bandwidth amount of each candidate server group; according to the candidate server group, the amount of redundant bandwidth, each unit individual in the unit individual combination, and the first unit corresponding to each unit individual in the unit individual combination With a bandwidth, an integer programming model is constructed; the integer programming model is solved to generate a migration scheme for the unit individual combination. Compared with the prior art, the method of the present invention has a faster processing speed.

Description

Bandwidth pulling method, device and electronic device

technical field

Embodiments of the present invention relate to the field of computer technologies, and in particular, to a bandwidth pulling method, device, and electronic device.

Background technique

Content Delivery Network (CDN for short) is a kind of network between service providers and consumers, through the establishment of nodes to allow users to obtain the desired content nearby, so as to relieve network congestion and improve the response speed of users accessing websites. .

Generally speaking, CDN services are provided by a group of servers. Since the server itself has an upper limit of capacity, in the actual online operation process, if the server group exceeding the upper limit of capacity continues to provide services, the machine load will often be high, resulting in packet loss. or other circumstances that affect the quality of service. At this time, the customers of the service on this group need to be migrated to other groups to ensure the normal operation of the service by reducing the amount. When such problems are encountered in online operations, operation and maintenance personnel or programs generally conduct exhaustive searches to reduce the amount. At present, these two types of methods have the following problems:

1. Since the volume reduction migration involves a large amount of data calculation, manual adjustment can often only handle some simple scenarios (such as the case where the number of customers served on the server group is small, the number of groups that can be undertaken is large, and the redundancy is sufficient), and it is impossible to Manual bandwidth migration for complex scenarios;

2. A computer program is used to calculate the migration amount, and the existing methods generally use an exhaustive search method to solve the problem. Since the number of customers served on the server group may be large, the combination method involved is huge when performing downscaling, and multi-level downscaling may be required, so it is often impossible to solve within a limited time.

SUMMARY OF THE INVENTION

The present application provides a bandwidth pulling method, device and electronic device to solve the above problems in the prior art.

In a first aspect, the present application provides a bandwidth pulling method, which includes:

Determine the server group to be migrated, the first bandwidth amount used by each unit of bandwidth used in the server group to be migrated, and the second bandwidth amount to be migrated by the server group to be migrated;

According to the first bandwidth amount and the second bandwidth amount, at least one unit individual is selected from the unit individuals using the bandwidth in the server group to be migrated to form the unit individual combination to be migrated;

From the server groups other than the server group to be migrated, screen candidate server groups corresponding to each unit individual in the unit-individual combination, and determine the redundant bandwidth amount of each candidate server group;

Build an integer programming model according to the candidate server group, the redundant bandwidth amount of the candidate server group, each unit individual in the unit individual combination, and the first bandwidth amount corresponding to each unit individual in the unit individual combination;

Solve the integer programming model to generate the migration scheme of the unit individual combination.

In a second aspect, the present application provides a bandwidth pulling device, the device comprising:

a determining module, configured to determine the server group to be migrated, the first bandwidth amount used by each unit of bandwidth used in the server group to be migrated, and the second bandwidth amount to be migrated by the server group to be migrated;

A selection module, configured to select at least one unit individual from the unit individuals using bandwidth in the server group to be migrated according to the first bandwidth amount and the second bandwidth amount to form a unit individual combination to be migrated;

The screening module is used for screening the candidate server groups corresponding to each unit individual in the unit individual combination from the server groups other than the server group to be migrated, and determining the redundant bandwidth amount of each candidate server group;

A building module for constructing an integer programming model according to the candidate server group, the redundant bandwidth amount of the candidate server group, each unit individual in the unit individual combination, and the first bandwidth amount corresponding to each unit individual in the unit individual combination;

The generation module is used to solve the integer programming model and generate the migration scheme of the unit individual combination.

In a third aspect, an electronic device is provided, including a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus;

memory for storing computer programs;

The processor is configured to implement the steps of the bandwidth pulling method of any embodiment of the first aspect when executing the program stored in the memory.

In a fourth aspect, a computer-readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, implements the steps of the bandwidth pulling method according to any one of the embodiments of the first aspect.

Compared with the prior art, the above-mentioned technical solutions provided in the embodiments of the present application have the following advantages:

The method provided by the embodiment of the present application determines the server group to be migrated, the first bandwidth amount used by each unit of bandwidth used in the server group to be migrated, and the second bandwidth amount to be migrated by the server group to be migrated. Then, according to the first bandwidth amount and the second bandwidth amount, at least one unit individual is extracted from the unit individuals using the bandwidth in the server group to be migrated to form a combination of unit individuals to be migrated, from the server group to be migrated except the server group to be migrated. In the server group, the candidate server group corresponding to each unit individual in the unit individual combination is screened, and then according to the candidate server group, the redundant bandwidth of the candidate server group, the unit individual combination, and each unit individual in the unit individual combination corresponds to The first bandwidth amount of , constructs an integer programming model, solves the integer programming model, and finally generates a migration scheme for the unit individual combination. In this method, the bandwidth migration problem is transformed into an integer model solution problem, and a migration scheme is obtained through a mathematical algorithm. Compared with the manual adjustment of the bandwidth migration scheme, the method provided by the embodiment of the present application can more conveniently and quickly deal with a relatively complex scenario. Moreover, the method of the present invention obtains a better bandwidth migration scheme through a certain algorithm solution method, and compared with the exhaustive method in the prior art, the processing speed is faster, and the solution can be performed in a limited time.

Description of drawings

1 is a schematic flowchart of a bandwidth pulling method according to an embodiment of the present invention;

2 is a schematic diagram of a multi-bandwidth pulling structure of a specific application example of a bandwidth pulling method provided by the present invention;

3 is a schematic structural diagram of a bandwidth pulling device according to an embodiment of the present invention;

FIG. 4 is a schematic structural diagram of an electronic device according to an embodiment of the present invention.

Detailed ways

In order to make the purposes, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments These are some embodiments of the present invention, but not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

In order to facilitate the understanding of the embodiments of the present invention, further explanation will be given below with specific embodiments in conjunction with the accompanying drawings, and the embodiments do not constitute limitations to the embodiments of the present invention.

In response to the technical problems mentioned in the background art, an embodiment of the present application provides a bandwidth pulling method, as shown in FIG. 1 . FIG. 1 is a schematic flowchart of a bandwidth pulling method provided by an embodiment of the present invention. The method steps include:

Step 110: Determine the server group to be migrated, the first bandwidth amount used by each unit of bandwidth used in the server group to be migrated, and the second bandwidth amount to be migrated by the server group to be migrated.

Specifically, an individual unit can be understood as a combination of the client using the bandwidth in the server group + the identification information of the region granularity, which is used to refer to the bandwidth under a certain operator used by a certain user in a certain region. The first bandwidth amount, that is, the currently counted amount of bandwidth used by the user.

When a user accesses a domain name, a server in the server group will record the access request in the log system, where the request includes the user IP, the location of the machine that executes the access request, and the operator corresponding to the bandwidth used. and so on. Among them, the customer refers to the user IP, and the region includes the operator and the specific location of the machine (such as the province, region, etc.). The method in this embodiment of the present application may be executed by a terminal device, and the terminal device may collect the real-time bandwidth of each server group in real time, and the collection channel includes obtaining log information, or the terminal device directly establishes a connection with a server in the server group , so as to collect the real-time bandwidth on the server in real time.

After collecting the real-time bandwidth of each server, the terminal device can know the current total bandwidth of each server group. When the total bandwidth exceeds the preset threshold, for example, after exceeding a certain percentage of the rated bandwidth of the server group, in order to prevent the server group from continuing to provide services resulting in excessive machine load, packet loss or even other situations that affect the quality of service, it is necessary to set the The server group is added to the queue to be migrated, which is also the reduction of the server group (some units are migrated to other server groups). The server group whose current bandwidth usage exceeds a certain percentage of the rated bandwidth is the server group to be migrated, and the first bandwidth amount of each unit in each server group and the second bandwidth amount to be migrated of the server group to be migrated are obtained.

Further optional, considering that the terminal equipment obtains the user IP, the area where the machine is located, and the operator, and these information cannot be directly used by the scheduling system in the terminal equipment, it is also necessary to information is mapped. Including mapping the customer domain name to the dispatching system through CNAME, and converting it into a dispatching domain name. If the scheduling domain name is defined as a resolution group, the information obtained after the client + area granularity mapping is the resolution group + area granularity.

Step 120 , according to the first bandwidth amount and the second bandwidth amount, select at least one unit individual from the unit individuals using the bandwidth in the server group to be migrated to form a unit individual combination to be migrated.

Specifically, in a practical application environment, there are usually many units using bandwidth in any server group. Then it is necessary to make certain choices as to which individual units will perform the migration work.

During specific implementation, at least one unit individual may be selected from a plurality of unit individuals according to the first bandwidth amount and the second bandwidth amount to form a unit individual combination.

For example, the second bandwidth amount is 5G, the unit individuals in the server group include A, the bandwidth amount is 1G, B is 2G, C is 3G, D is 1G, and E is 2G. Then, the combination of unit individuals can include A+B+E, or A+C+D, or B+D+E and so on.

Specifically, as long as the sum of the first bandwidth amount corresponding to at least one of the unit units using bandwidth in the server group to be migrated is greater than or equal to the second bandwidth amount, it can be determined that at least one unit unit constitutes the unit unit combination.

During specific execution, one of the combinations will be randomly selected, and the following operations will be performed.

Step 130: From the server groups other than the server group to be migrated, screen candidate server groups corresponding to each unit individual in the unit individual combination, and determine the redundant bandwidth amount of each candidate server group.

Optionally, in order to find a candidate server group corresponding to each unit individual in the unit-individual combination from other server groups, the following methods can be used to implement, for details, see the following:

According to the redundant carrying capacity and/or pre-configured constraints of each server group in the server group except the server group to be migrated, and the first bandwidth amount corresponding to each unit individual in the unit individual combination, The server groups other than the server group to be migrated are screened to obtain a candidate server group corresponding to each unit individual in the unit-individual combination.

That is, firstly, it is judged whether the redundancy capability of the server groups in the remaining server groups except the server group to be migrated is able to undertake a certain unit individual. If it can be accepted, it will be regarded as the candidate server group corresponding to the individual unit. If it cannot be accepted, it is excluded.

And/or, when the server group also includes certain constraints, some resources can only be used in some regions and cannot be used as a candidate server group for an individual unit, so they need to be excluded.

In the above manner, candidate server groups corresponding to each unit individual in the unit individual combination are screened out.

In a possible implementation, although each individual unit can only select one candidate server group, in some special cases, it is possible for the same candidate server group to satisfy the conditions of being a candidate server for multiple individual units at the same time .

Step 140: Construct an integer programming model according to the candidate server group, the redundant bandwidth of the candidate server group, each individual unit in the unit individual combination, and the first bandwidth amount corresponding to each unit individual in the unit individual combination.

Optionally, before performing the method steps of the above step 140, the method may further include: randomly migrating at least one unit individual in the unit individual combination to the corresponding candidate server group in the at least one candidate server group, and refreshing the candidate server. Redundancy information for the group.

Specifically, since there may be a large number of parsing groups + regions carried on the server group, in order to reduce the computational load for solving the integer programming problem, and the bandwidth migration problem itself may have multiple feasible solutions, before performing step 140, you can first With the Las Vegas algorithm (the Las Vegas algorithm is a randomized algorithm that always gives the correct solution, the Las Vegas algorithm does not get an incorrect solution. Once found with the Las Vegas algorithm A solution, this solution must be the correct solution) Randomly select part of the analysis group + area and randomly assign it to the candidate server group, and refresh the redundancy information of the candidate server group.

Of course, as described above, this operation is not an operation that must be performed. This operation is to reduce the amount of subsequent computations. The amount of calculation will be increased.

After the above operations are completed, use the candidate server group, the redundant bandwidth of the candidate server group, each unit individual in the unit individual combination, and the first bandwidth amount corresponding to each unit individual in the unit individual combination to construct an integer plan. Model.

Specifically, the idea of this step is that migrating the parsing group + area on the server group to the candidate server group is essentially a resource allocation problem, that is, in the case where the number of each available resource is limited, the resources are carefully distribution to maximize profits. In addition, since the transfer of the analytical group + region does not satisfy the separability, the problem can be characterized as a binary integer programming problem, and the model is established as follows:

1. By introducing binary variables (or 0-1 variables), the objective function of the problem is designed as:

max Z=A ₁ x ₁₁ +A ₂ x ₁₂ +…+A _n x _nm

Z表示能够迁移的解析组+区域的带宽量总和，max Z表示通过相关算法能够让Z最大化。Among them, A _i is the bandwidth of the ith parsing group + area, x _ij is whether the ith parsing group + area is migrated to the j th candidate server group,

Z represents the sum of the bandwidth of the analytic group + region that can be migrated, and max Z represents that Z can be maximized by the correlation algorithm.

Further optionally, after constructing the data model, the method may further include: according to the candidate server group, the redundant bandwidth amount of the candidate server group, each unit individual in the unit individual combination, and each unit in the unit individual combination. The first bandwidth amount corresponding to the individual generates constraints corresponding to the integer programming model.

此外一个候选服务器组仅能够承载不超过自身冗余能力的解析组+区域带宽量，即：Specifically, for the same parsing group + area, it is necessary to satisfy and can only be migrated to a certain candidate server group, that is, to satisfy

In addition, a candidate server group can only carry the amount of resolution group + area bandwidth that does not exceed its own redundancy capacity, namely:

Therefore, when building an integer programming model, the above execution conditions can also be configured.

In step 150, the integer programming model is solved to generate a migration scheme of the unit individual combination.

Specifically, when the solution process is specifically executed, a branch-and-bound algorithm may be used to solve the bandwidth migration problem.

Specifically include the following steps:

1) Initialization: set Z ^* =-∞, set the optimal solution of the objective function to infinitesimal, and then use the delimiting and pruning method to solve the original problem.

2) Bounding: For each subproblem, first use the simplex method to calculate the boundary value of the linear relaxation problem of the subproblem, and then round down its Z value.

3) Pruning: For each sub-problem, the following three methods are used for pruning;

1. If the boundary value of the sub-problem ≤ Z ^* , the sub-problem can be pruned;

2. If the linear relaxation problem of the sub-problem does not contain a feasible solution, the sub-problem can be pruned;

3. If the optimal solution of the linear relaxation problem of the sub-problem is an integer and is better than the original best feasible solution, the solution of the sub-problem is the current best feasible solution.

The specific solution process is realized by the existing technology, so it will not be repeated here.

When all analytic groups + regions are assigned, it means that the current solution is regarded as a feasible solution to the original problem. The migration of the unit individuals in the unit individual combination can be completed directly according to the current migration method.

Optionally, considering as described above, it is possible that the candidate server groups corresponding to different units are the same candidate server group, but the current redundancy of the same candidate server group cannot carry multiple server groups at the same time. The unit individual, or the redundancy in the candidate server group changes suddenly, so that after the binary integer programming model is solved by the above method, there is still a situation that the analytical group + area is not allocated. At this time, the method may also include:

Perform multi-order reduction operations on individual units that fail to migrate;

The process ends when it is determined that all unit individuals in the unit individual combination have successfully migrated, or the number of iterations for performing multi-order reductions exceeds a preset number of times threshold.

Specifically, performing the multi-order volume reduction operation may include the following steps:

1. Initialize the hierarchy depth, that is, the recursive search depth.

2. Calculate the second-order server group that needs to be reduced (assuming that the above operation is the first-order migration of the server group to be migrated, this operation is the second-order (second-order) migration), for the unassigned analysis group + area, extract Its candidate server group is pruned (that is, the server group that has appeared on the recursive link in all stages before the current stage is removed).

3. Randomly select a candidate server group and perform all the above operation steps except the multi-order drop operation.

It ends when it is determined that all unit individuals in the unit individual combination have successfully migrated, or the number of iterations for performing multi-order reduction (performing multi-order migration) exceeds a preset number of times threshold.

When the number of iterations for performing multi-order reduction reaches a certain threshold, it means that there is no suitable solution for the migration scheme of the unit-individual combination. One or more unit combination schemes can be changed to repeat some or all of the above operations as needed.

If no suitable solution is found after all the unit combinations have performed the above operations, it means that the original problem may not have a feasible solution in a certain probability, that is, the bandwidth traction of the server group may not be able to find a solution that can be undertaken. The list of candidate server groups needs to be resolved through resource re-planning. On the contrary, if a suitable output scheme can be found, it means that there is an optimal solution.

In a specific application example, assuming that there is a certain server group g0 that needs to be reduced by 10MB, a set of dv (parse group + area) lists to be migrated are generated through calculation. Among them, the situation of the dv to be migrated in g0 is shown in Table 1, and the redundancy capability of the candidate server group g1-g3 is shown in Table 2.

Table 1

dv1 dv2 dv3 dv4 g0 2MB 2MB 3MB 3MB

Table 2

server group Redundancy capability g1 1MB g2 3MB g3 3MB

During specific execution, the partial parsing group + area is randomly assigned: for example, dv4 is assigned to g3 in a random manner.

The integer programming method is used to solve the sub-problem: because the redundancy of the candidate server groups g1 and g2 cannot fully undertake the amount of dv1, dv2 and dv3, it is necessary to recursively reduce the amount of g1.

Suppose that g1 is processed through recursive reduction, so that the redundancy of g1 becomes 4MB. At this time, the redundancy of the server group changes, as shown in Table 3:

table 3

server group Redundancy capability g1 4MB g2 3MB g3 0MB

At this point, use integer programming to solve:

max Z = 2x ₁₁ +2x ₁₂ +2x ₂₁ +2x ₂₂ +3x ₃₁ +3x ₃₂

stx ₁₁ +x ₁₂ ≤1

x ₂₁ +x ₂₂ ≤1

x ₃₁ +x ₃₂ ≤1

2x ₁₁ +2x ₂₁ +2x ₃₁ ≤4

2x ₁₂ +2x ₂₂ +2x ₃₂ ≤3

x ₁₁ , x ₁₂ , x ₁₃ , x ₂₁ , x ₂₂ , x ₂₃ ∈ {0, 1}

As shown in Table 4, the final feasible solution of the original problem is:

Table 4

dv1 dv2 dv3 dv4 g1 1 1 0 0 g2 0 0 1 0 g3 0 0 0 1

Among them, dv1 is assigned to g1, dv2 is assigned to g1, dv3 is assigned to g2, dv4 is assigned to g3, and server group g0 achieves the goal of pulling 10MB of bandwidth.

Referring specifically to FIG. 2 , FIG. 2 shows a flowchart of the migration of dvs in the g0 group except dv4 after multi-order reduction processing is performed, and the amount of redundant bandwidth in each candidate server in the candidate server group. where s represents the remaining dv combinations, including dv1, dv2, and dv3. The bandwidth of dv1 is 2M, the bandwidth of dv2 is 2M, and the bandwidth of dv3 is 3m. dv1 migrates to g1, dv2 migrates to g1, dv3 migrates to g2, and finally outputs the result t.

The bandwidth pulling method provided by the embodiment of the present invention determines the server group to be migrated, the first bandwidth amount used by each unit of bandwidth used in the server group to be migrated, and the second bandwidth amount to be migrated by the server group to be migrated . Then, according to the first bandwidth amount and the second bandwidth amount, at least one unit individual is extracted from the unit individuals using the bandwidth in the server group to be migrated to form a combination of unit individuals to be migrated, from the server group to be migrated except the server group to be migrated. In the server group, the candidate server group corresponding to each unit individual in the unit individual combination is screened, and then according to the candidate server group, the redundant bandwidth of the candidate server group, the unit individual combination, and each unit individual in the unit individual combination corresponds to The first bandwidth amount of , constructs an integer programming model, solves the integer programming model, and finally generates a migration scheme for the unit individual combination. In this method, the bandwidth migration problem is transformed into an integer model solution problem, and a migration scheme is obtained through a mathematical algorithm. Compared with the manual adjustment of the bandwidth migration scheme, the method provided by the embodiment of the present application can more conveniently and quickly deal with a relatively complex scenario. Moreover, the method of the present invention obtains a better bandwidth migration scheme through a certain algorithm solution method, and compared with the exhaustive method in the prior art, the processing speed is faster, and the solution can be performed in a limited time.

In the above, several method embodiments of the bandwidth pulling provided by the present application are described, and other embodiments of the bandwidth pulling provided by the present application will be introduced and described in the following. For details, refer to the following.

FIG. 3 is a bandwidth pulling device provided by an embodiment of the present invention. The device includes: a determination module 301 , a selection module 302 , a screening module 303 , a construction module 304 , and a generation module 305 .

A determination module 301, configured to determine the server group to be migrated, the first bandwidth amount used by each unit of bandwidth used in the server group to be migrated, and the second bandwidth amount to be migrated by the server group to be migrated;

The selection module 302 is configured to select at least one individual unit from the unit individuals using the bandwidth in the server group to be migrated according to the first bandwidth amount and the second bandwidth amount to form a unit individual combination to be migrated;

The screening module 303 is used to screen the candidate server groups corresponding to each unit individual in the unit individual combination from the server groups other than the server group to be migrated, and determine the redundant bandwidth amount of each candidate server group;

The building module 304 is configured to construct an integer programming model according to the candidate server group, the redundant bandwidth amount of the candidate server group, each unit individual in the unit individual combination, and the first bandwidth amount corresponding to each unit individual in the unit individual combination ;

The generating module 305 is configured to solve the integer programming model and generate the migration scheme of the unit individual combination.

Optionally, the selection module 302 is specifically configured to determine that at least one unit individual constitutes a unit when the sum of the first bandwidth amount corresponding to at least one unit individual in the unit individuals using bandwidth in the server group to be migrated is greater than or equal to the second bandwidth amount. individual combination.

Optionally, the screening module 303 is specifically configured to be based on the redundant carrying capacity and/or preconfigured constraints of each server group in the server group except the server group to be migrated, and each of the unit individual combinations. For the first bandwidth amount corresponding to the unit individual, the server groups other than the server group to be migrated are screened, and the candidate server group corresponding to each unit individual in the unit individual combination is obtained.

Optionally, the device further includes: a processing module configured to randomly migrate at least one unit individual in the unit individual combination to a corresponding candidate server group in the at least one candidate server group, and refresh the redundancy information of the candidate server group. .

Optionally, the processing module is further configured to, when the result of solving the integer programming model is that at least one unit individual in the unit individual combination fails to migrate, perform a multi-order reduction operation on the unit individual that fails to migrate;

The process ends when it is determined that all unit individuals in the unit individual combination have successfully migrated, or the number of iterations for performing multi-order reductions exceeds a preset number of times threshold.

Optionally, the building module 304 is further configured to, according to the candidate server group, the redundant bandwidth amount of the candidate server group, each unit individual in the unit individual combination, and the first bandwidth amount corresponding to each unit individual in the unit individual combination. , which generates constraints corresponding to the integer programming model.

Optionally, the number of unit-individual combinations to be migrated includes multiple; the screening module 303 is also used for, when selecting from server groups other than the server group to be migrated, each unit-individual in the current unit-individual combination. When the corresponding candidate server groups fail, re-select a new unit individual combination, and perform screening of the candidate server groups corresponding to each unit individual in the new individual combination from the server groups other than the server group to be migrated. .

The functions performed by each component in the bandwidth pulling device provided in the embodiment of the present invention have been described in detail in any of the foregoing method embodiments, and thus are not repeated here.

A bandwidth pulling device provided by an embodiment of the present invention determines a server group to be migrated, a first amount of bandwidth used by each unit of bandwidth used in the server group to be migrated, and a second bandwidth of the server group to be migrated to be migrated. amount of bandwidth. Then, according to the first bandwidth amount and the second bandwidth amount, extract at least one unit individual from the unit individuals using the bandwidth in the server group to be migrated to form a combination of unit individuals to be migrated, from the server group to be migrated except the server group to be migrated. In the server group, the candidate server group corresponding to each unit individual in the unit individual combination is screened, and then according to the candidate server group, the redundant bandwidth of the candidate server group, the unit individual combination, and each unit individual in the unit individual combination corresponds to The first bandwidth amount of , constructs an integer programming model, solves the integer programming model, and finally generates a migration scheme for the unit individual combination. In this method, the bandwidth migration problem is transformed into an integer model solution problem, and a migration scheme is obtained through a mathematical algorithm. Compared with the manual adjustment of the bandwidth migration scheme, the method provided by the embodiment of the present application can more conveniently and quickly deal with a relatively complex scenario. Moreover, the method of the present invention obtains a better bandwidth migration scheme through a certain algorithm solution method, and compared with the exhaustive method in the prior art, the processing speed is faster, and the solution can be performed in a limited time.

As shown in FIG. 4 , an embodiment of the present application provides an electronic device, including a processor 111 , a communication interface 112 , a memory 113 and a communication bus 114 , wherein the processor 111 , the communication interface 112 , and the memory 113 are completed through the communication bus 114 communication with each other.

a memory 113 for storing computer programs;

In an embodiment of the present application, the processor 111 is configured to implement the bandwidth pulling method provided by any one of the foregoing method embodiments when executing the program stored in the memory 113, including:

Determine the server group to be migrated, the first bandwidth amount used by each unit of bandwidth used in the server group to be migrated, and the second bandwidth amount to be migrated by the server group to be migrated;

According to the first bandwidth amount and the second bandwidth amount, at least one unit individual is selected from the unit individuals using the bandwidth in the server group to be migrated to form the unit individual combination to be migrated;

From the server groups other than the server group to be migrated, screen candidate server groups corresponding to each unit individual in the unit-individual combination, and determine the redundant bandwidth amount of each candidate server group;

Build an integer programming model according to the candidate server group, the redundant bandwidth amount of the candidate server group, each unit individual in the unit individual combination, and the first bandwidth amount corresponding to each unit individual in the unit individual combination;

Solve the integer programming model to generate the migration scheme of the unit individual combination.

Optionally, when the sum of the first bandwidth amount corresponding to at least one unit individual in the unit individuals using bandwidth in the server group to be migrated is greater than or equal to the second bandwidth amount, it is determined that at least one unit individual constitutes a unit individual combination.

Optionally, according to the redundant carrying capacity and/or preconfigured constraints of each server group in the server group except the server group to be migrated, and the first bandwidth corresponding to each unit individual in the unit individual combination The number of server groups other than the server group to be migrated is screened, and a candidate server group corresponding to each unit individual in the unit-individual combination is obtained.

Optionally, before generating the migration scheme for the unit-individual combination according to the candidate server group, the unit-individual combination, and the first bandwidth amount corresponding to each unit-individual in the unit-individual combination, the method further includes:

Randomly migrate at least one unit individual in the unit individual combination to the corresponding candidate server group in the at least one candidate server group, and refresh the redundancy information of the candidate server group.

Optionally, when the result of solving the integer programming model is that at least one unit individual in the unit individual combination fails to migrate, the method further includes:

Perform multi-order reduction operations on individual units that fail to migrate;

The process ends when it is determined that all unit individuals in the unit individual combination have successfully migrated, or the number of iterations for performing multi-order reductions exceeds a preset number of times threshold.

Optionally, according to the candidate server group, the redundant bandwidth amount of the candidate server group, each unit individual in the unit individual combination, and the first bandwidth amount corresponding to each unit individual in the unit individual combination, after constructing the integer programming model, Methods also include:

Constraints corresponding to the integer programming model are generated according to the candidate server group, the redundant bandwidth amount of the candidate server group, each unit individual in the unit individual combination, and the first bandwidth amount corresponding to each unit individual in the unit individual combination.

Optionally, the number of unit individual combinations to be migrated includes multiple; when selecting the candidate server groups corresponding to each unit individual in the current unit individual combination from the server groups other than the server group to be migrated, it fails. When , a new unit individual combination is reselected, and a candidate server group corresponding to each unit individual in the new individual combination is selected from the server groups other than the server group to be migrated.

Embodiments of the present application further provide a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, implements the steps of the bandwidth pulling method provided by any one of the foregoing method embodiments.

It should be noted that, in this document, relational terms such as "first" and "second" etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply these There is no such actual relationship or sequence between entities or operations. Moreover, the terms "comprising", "comprising" or any other variation thereof are intended to encompass a non-exclusive inclusion such that a process, method, article or device that includes a list of elements includes not only those elements, but also includes not explicitly listed or other elements inherent to such a process, method, article or apparatus. Without further limitation, an element qualified by the phrase "comprising a..." does not preclude the presence of additional identical elements in the process, method, article, or device that includes the element.

The above are only specific embodiments of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features claimed herein.

Claims (10)

1. a bandwidth pulling method, is characterized in that, described method comprises: determining the server group to be migrated, the first bandwidth amount used by each unit of bandwidth used in the server group to be migrated, and the second bandwidth amount to be migrated by the server group to be migrated; According to the first bandwidth amount and the second bandwidth amount, at least one unit individual is selected from the unit individuals using the bandwidth in the to-be-migrated server group to form a unit-individual combination to be migrated; From the server groups other than the server group to be migrated, screen candidate server groups corresponding to each unit individual in the unit individual combination, and determine the redundant bandwidth amount of each candidate server group; According to the candidate server group, the redundant bandwidth amount of the candidate server group, each unit individual in the unit individual combination, and the first bandwidth amount corresponding to each unit individual in the unit individual combination, Build an integer programming model; The integer programming model is solved to generate a migration scheme of the unit individual combination. 2 . The method according to claim 1 , wherein, according to the first bandwidth amount and the second bandwidth amount, at least one unit individual is selected from the unit individuals using bandwidth in the server group to be migrated. 3 . , which constitutes the unit individual combination to be migrated, including: When the sum of the first bandwidth amount corresponding to at least one of the unit individuals using bandwidth in the server group to be migrated is greater than or equal to the second bandwidth amount, it is determined that at least one of the unit individuals constitutes the unit individual combination. 3 . The method according to claim 1 , wherein the candidate server corresponding to each unit individual in the unit individual combination is selected from the server groups other than the server group to be migrated. 3 . group, and determine the amount of redundant bandwidth for each candidate server group, including: According to the redundant carrying capacity and/or pre-configured constraints of each server group in the server group other than the server group to be migrated, and the For a bandwidth amount, the server groups other than the server group to be migrated are screened, and a candidate server group corresponding to each unit individual in the unit-individual combination is obtained. The method according to claim 1, wherein, according to the candidate server group, the unit individual combination, and the first bandwidth amount corresponding to each unit individual in the unit individual combination, Before generating the migration scheme of the unit individual combination, the method further includes: Randomly migrate at least one of the unit individuals in the unit individual combination to a corresponding candidate server group in at least one of the candidate server groups, and refresh the redundancy information of the candidate server group. 5. The method according to claim 4, wherein when the result of solving the integer programming model is that at least one unit individual in the unit individual combination fails to migrate, the method further comprises: performing a multi-order reduction operation on the unit individual that fails to migrate; The process ends when it is determined that all the unit individuals in the unit individual combination are successfully migrated, or the number of iterations for performing multi-order reduction exceeds a preset number of times threshold. 6. The method according to any one of claims 1-5, wherein the method is based on each unit in the candidate server group, the redundant bandwidth amount of the candidate server group, and the unit individual combination. The individual and the first bandwidth amount corresponding to each unit individual in the unit individual combination, after the integer programming model is constructed, the method further includes: According to the candidate server group, the redundant bandwidth amount of the candidate server group, each unit individual in the unit individual combination, and the first bandwidth amount corresponding to each unit individual in the unit individual combination, Constraints corresponding to the integer programming model are generated. 7. The method according to any one of claims 1-5, wherein the number of unit individual combinations to be migrated includes a plurality; , when the selection of the candidate server group corresponding to each unit individual in the current unit individual combination fails, re-select a new unit individual combination, and execute from the server group other than the server group to be migrated, A candidate server group corresponding to each unit individual in the new individual combination is screened. 8. A bandwidth pulling device, wherein the device comprises: a determining module, configured to determine a server group to be migrated, a first bandwidth amount used by each unit of bandwidth used in the server group to be migrated, and a second bandwidth amount to be migrated by the server group to be migrated; A selection module, configured to select at least one unit individual from the unit individuals using bandwidth in the to-be-migrated server group according to the first bandwidth amount and the second bandwidth amount to form a unit-individual combination to be migrated; A screening module, configured to screen candidate server groups corresponding to each unit individual in the unit individual combination from server groups other than the server group to be migrated, and determine the redundancy of each candidate server group amount of bandwidth; The building module is configured to, according to the candidate server group, the redundant bandwidth amount of the candidate server group, each unit individual in the unit individual combination, and the corresponding unit individual in the unit individual combination The first bandwidth is used to construct an integer programming model; The generating module is configured to solve the integer programming model and generate the migration scheme of the unit individual combination. 9. An electronic device, characterized in that it comprises a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface, and the memory complete mutual communication through the communication bus; memory for storing computer programs; The processor is configured to implement the steps of the bandwidth pulling method according to any one of claims 1-7 when executing the program stored in the memory. 10. A computer-readable storage medium on which a computer program is stored, characterized in that, when the computer program is executed by a processor, the steps of the bandwidth pulling method according to any one of claims 1-7 are implemented.

Images