CN111901410A - Server grouping method, device, equipment and computer readable storage medium - Google Patents

Abstract

The invention discloses a server grouping method. The method includes the following steps: analyzing a received server grouping request to obtain a target grouping task; obtaining parameters of each target server corresponding to the target grouping task in each server; using fuzzy C The mean clustering algorithm performs clustering and grouping of each server based on the parameters of each target server, and obtains each server group. By applying the technical solutions provided by the embodiments of the present invention, the influence of human subjective factors is greatly reduced, the grouping accuracy is improved, and the grouping efficiency is greatly improved. The invention also discloses a server grouping device, equipment and storage medium, which have corresponding technical effects.

Description

A server grouping method, apparatus, device and computer-readable storage medium

technical field

The present invention relates to the technical field of computer applications, and in particular, to a server grouping method, apparatus, device, and computer-readable storage medium.

Background technique

The scale of today's data centers is getting larger and larger, and there are thousands of machines in a computer room, which makes it extremely complicated for operation and maintenance personnel to manage. There are two scenarios in which a large number of servers need to be grouped for processing. The first scenario is to group all servers in the network according to configuration information to facilitate maintenance. For example, servers with the same configuration and the same computing power should belong as much as possible. For the same group, so that administrators can quickly find similar servers when performing system maintenance. Another situation is to group the running servers, and find out the machines with similar resources at the current moment to form a cluster. It is convenient for complex tasks to be distributed on the basis of this cluster. In general, the configuration and computing power of the servers are similar. Because the resources are balanced, the distributed tasks can be better executed. This distribution strategy is better than distributing tasks to machines with different configurations, because when dividing a temporary cluster, it is necessary to first group the entire server and filter out the resources. Relatively abundant, low-stress servers.

The above two scenarios both involve grouping each server in the server cluster. The existing server grouping method is to group each server in the server cluster by manual selection. However, in reality, the number of servers in a data center is often thousands. , The amount of server grouping tasks is huge, prone to errors, and is greatly affected by human subjective factors, the grouping accuracy is low, and the grouping efficiency is low.

To sum up, how to effectively solve the problems of low grouping accuracy and low grouping efficiency of the manually selected server grouping method is an urgent problem to be solved by those skilled in the art at present.

SUMMARY OF THE INVENTION

The purpose of the present invention is to provide a server grouping method, which greatly reduces the influence of human subjective factors, improves the accuracy of grouping, and greatly improves the efficiency of grouping; another purpose of the present invention is to provide a server grouping device , devices, and computer-readable storage media.

In order to solve the above-mentioned technical problems, the present invention provides the following technical solutions:

A server grouping method, comprising:

Analyze the received server grouping request to obtain the target grouping task;

Respectively obtain each target server parameter corresponding to the target grouping task in each server;

The servers are clustered and grouped based on the parameters of each target server by using a fuzzy C-means clustering algorithm to obtain each server group.

In a specific embodiment of the present invention, when the target grouping task is a server maintenance task, acquiring parameters of each target server corresponding to the target grouping task in each server, including:

Respectively obtain each original configuration parameter corresponding to the server maintenance task in each of the servers;

Use the configuration quantization function to perform quantization processing on each of the original configuration parameters to obtain each target configuration parameter;

Using the fuzzy C-means clustering algorithm to cluster and group each of the servers based on the parameters of each of the target servers, including:

The servers are clustered and grouped based on the target configuration parameters using the fuzzy C-means clustering algorithm.

In a specific embodiment of the present invention, using a configuration quantization function to perform quantization processing on each of the original configuration parameters, including:

Obtain the configuration quantization function preset through the function preset interface;

Each of the original configuration parameters is quantized by using the configuration quantization function.

In a specific embodiment of the present invention, after using the fuzzy C-means clustering algorithm to cluster and group each of the servers based on each of the target configuration parameters to obtain each of the server groups, the method further includes:

A parallel maintenance operation is performed on each of the servers in units of each of the server groups.

In a specific embodiment of the present invention, when the target grouping task is a service distribution task, acquiring parameters of each target server corresponding to the target grouping task in each server, including:

respectively acquiring each real-time performance parameter corresponding to the service distribution task in each of the servers;

Using the fuzzy C-means clustering algorithm to cluster and group each of the servers based on the parameters of each of the target servers, including:

The servers are clustered and grouped based on the real-time performance parameters using the fuzzy C-means clustering algorithm.

In a specific embodiment of the present invention, after using the fuzzy C-means clustering algorithm to cluster and group each of the servers based on each of the real-time performance parameters to obtain each of the server groups, the method further includes:

Selecting a target server group from each of the server groups;

Distributing the to-be-processed service to each target server in the target server group, so as to use each of the target servers to perform service processing.

In a specific embodiment of the present invention, clustering and grouping each of the servers based on the parameters of each of the target servers using a fuzzy C-means clustering algorithm, including:

Obtain the server subsets corresponding to each preset server class;

obtaining the centroids preset for each of the server subsets respectively;

Calculate each distance value between each of the servers and each of the centroids according to the parameters of each of the target servers;

Calculate the probability value that each of the servers belongs to each of the preset server classes according to the current distance values;

Recalculate each of the centroids according to each current probability value;

Calculate the distance values between each of the servers and the recalculated centroids according to the parameters of each of the target servers;

Calculate the value function by combining the current probability values and the current distance values;

Judging whether the value function satisfies a preset condition;

If not, repeat the step of calculating the probability value of each of the servers belonging to each of the preset server classes according to each of the distance values;

If so, each current centroid is obtained, and each of the servers is clustered and grouped according to each of the current centroids.

A server grouping device, comprising:

The task obtaining module is used to parse the received server grouping request to obtain the target grouping task;

a parameter obtaining module, used for obtaining each target server parameter corresponding to the target grouping task in each server;

The server group obtaining module is used for clustering and grouping each of the servers based on the parameters of each of the target servers by using a fuzzy C-means clustering algorithm to obtain each server group.

A server grouping device, comprising:

memory for storing computer programs;

The processor is configured to implement the steps of the aforementioned server grouping method when executing the computer program.

A computer-readable storage medium having a computer program stored on the computer-readable storage medium, when the computer program is executed by a processor, implements the steps of the foregoing server grouping method.

Apply the method provided by the embodiment of the present invention, analyze the received server grouping request, and obtain the target grouping task; obtain each target server parameter corresponding to the target grouping task in each server; use the fuzzy C-means clustering algorithm based on each The target server parameters are used to cluster and group each server to obtain each server group. By obtaining the target server parameters of each server according to the target grouping task, and using the fuzzy C-means clustering algorithm to automatically cluster and group each server based on the parameters of each target server, compared with the existing manual selection server grouping method, the present invention greatly reduces the The influence of human subjective factors is reduced, the grouping accuracy is improved, and the grouping efficiency is greatly improved.

Correspondingly, the embodiments of the present invention also provide a server grouping apparatus, device, and computer-readable storage medium corresponding to the above-mentioned server grouping method, which have the above-mentioned technical effects, and are not repeated here.

Description of drawings

In order to explain the embodiments of the present invention or the technical solutions in the prior art more clearly, the following briefly introduces the accompanying drawings that need to be used in the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description are only These are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can also be obtained according to these drawings without creative efforts.

FIG. 1 is an implementation flowchart of a server grouping method in an embodiment of the present invention;

Fig. 2 is another implementation flowchart of the server grouping method in the embodiment of the present invention;

Fig. 3 is a kind of server network topology diagram in the embodiment of the present invention;

4 is a schematic diagram of a server grouped according to an embodiment of the present invention;

Fig. 5 is another implementation flowchart of the server grouping method in the embodiment of the present invention;

6 is a structural block diagram of a server grouping apparatus in an embodiment of the present invention;

FIG. 7 is a structural block diagram of a server grouping device in an embodiment of the present invention.

Detailed ways

In order to make those skilled in the art better understand the solution of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. Obviously, the described embodiments are only some, but not all, embodiments of the present invention. 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.

Example 1:

Referring to FIG. 1, FIG. 1 is an implementation flowchart of a server grouping method in an embodiment of the present invention, and the method may include the following steps:

S101: Analyze the received server grouping request to obtain a target grouping task.

When the servers in the server cluster need to be clustered and grouped, a server grouping request is sent to the server grouping center, and the server grouping request includes the target grouping task. The server grouping center receives the server grouping request, parses the server grouping request, and obtains the target grouping task.

When it is necessary to maintain each server in the server cluster, a server grouping request can be generated, and the corresponding target grouping task is a server maintenance task; when a temporary business needs to be processed, a server grouping request can be generated, and the corresponding target grouping task can be generated. Distribute tasks for the business.

S102: Respectively obtain parameters of each target server corresponding to the target grouping task in each server.

After the target grouping task is obtained by parsing, parameters of each target server corresponding to the target grouping task in each server are obtained respectively. For example, when the target grouping task is a server maintenance task, the parameters of each target server may include CPU configuration information, memory configuration information, hard disk configuration information, network card configuration information, redundant disk array configuration information, operating system configuration information, etc.; when the target grouping task is When assigning tasks to services, the parameters of each target server can include the current CPU utilization, current memory utilization, current hard disk capacity utilization, current network bandwidth utilization, ratio of database connections, and ratio of hard disk IO bandwidth.

S103: Clustering and grouping each server based on the parameters of each target server by using a fuzzy C-means clustering algorithm to obtain each server group.

After acquiring the parameters of each target server corresponding to the target grouping task in each server, the fuzzy C-means clustering algorithm is used to cluster and group each server based on the parameters of each target server to obtain each server group. By obtaining the target server parameters of each server according to the target grouping task, and using the fuzzy C-means clustering algorithm to automatically cluster and group each server based on the parameters of each target server, compared with the existing manual selection server grouping method, the present invention greatly reduces the The influence of human subjective factors is reduced, the grouping accuracy is improved, and the grouping efficiency is greatly improved.

The fuzzy C-means clustering algorithm obtains the membership degree of each sample point to all class centers by optimizing the objective function, so as to determine the category of the sample point to achieve the purpose of automatically classifying the sample data.

Apply the method provided by the embodiment of the present invention, analyze the received server grouping request, and obtain the target grouping task; obtain each target server parameter corresponding to the target grouping task in each server; use the fuzzy C-means clustering algorithm based on each The target server parameters are used to cluster and group each server to obtain each server group. By obtaining the target server parameters of each server according to the target grouping task, and using the fuzzy C-means clustering algorithm to automatically cluster and group each server based on the parameters of each target server, compared with the existing manual selection server grouping method, the present invention greatly reduces the The influence of human subjective factors is reduced, the grouping accuracy is improved, and the grouping efficiency is greatly improved.

It should be noted that, based on the foregoing first embodiment, the embodiment of the present invention also provides a corresponding improvement solution. In subsequent embodiments, the same steps or corresponding steps in the above-mentioned first embodiment can be referred to each other, and corresponding beneficial effects can also be referred to each other, which will not be repeated in the following improved embodiments.

Embodiment 2:

Referring to FIG. 2, FIG. 2 is another implementation flowchart of a server grouping method in an embodiment of the present invention, and the method may include the following steps:

S201: Analyze the received server grouping request to obtain a target grouping task.

S202: When the target grouping task is a server maintenance task, obtain each original configuration parameter corresponding to the server maintenance task in each server.

When the target grouping task is the server maintenance task, each original configuration parameter corresponding to the server maintenance task in each server is obtained respectively. Each original configuration parameter may include CPU configuration information, memory configuration information, hard disk configuration information, network card configuration information, redundant disk array configuration information, operating system configuration information, and the like. Referring to FIG. 3, FIG. 3 is a topology diagram of a server network in an embodiment of the present invention. The management host needs to be in a network segment with the managed servers. The client can be connected to the management host through Ethernet. The client web page can display all servers in the current server cluster and their configuration parameters, mainly for interacting with users. Therefore, its main function is data display. The management host can establish a communication connection with the managed servers through the Gigabit switch management network.

S203: Obtain the configuration quantization function preset through the function preset interface.

Since the collected original configuration parameters are not necessarily suitable for clustering calculation, for example, the hard disk interface is an enumeration type and not a continuous number, so direct calculation is meaningless. Therefore, each original configuration parameter needs to be preprocessed. The function preset interface for setting the configuration quantization function required to quantize each original configuration parameter can be preset. Users can preset the configuration quantization function through the function preset interface according to actual needs, thereby improving the flexibility of parameter quantization and improving the user experience. experience.

After acquiring each original configuration parameter corresponding to the server maintenance task in each server, acquire the configuration quantization function preset through the function preset interface.

S204: Perform quantization processing on each original configuration parameter by using a configuration quantization function to obtain each target configuration parameter.

After the configuration quantization function preset through the function preset interface is obtained, each original configuration parameter is quantized by using the configuration quantization function to obtain each target configuration parameter.

The process of quantifying each original configuration parameter may include: firstly, quantizing each original configuration parameter into a digital class, and secondly, considering the difference between different enumerations and taking the difference into account in the quantization process. After quantification processing, each target configuration parameter of each server becomes a digital type that can be calculated, and each target configuration parameter constitutes multi-dimensional data. For example, D={d ₁ , d ₂ . . . d _j . . . d _n } can be used to represent the server cluster, and d _j represents the jth server of the server cluster. p _j is the quantized target configuration parameter set of the jth server, p _j ={cpu _1j ,mem _1j ,hd _1j ,net _1j ,raid _j ,os _j }, where cpu _1j is the CPU configuration of the jth server information, mem _1j is the memory configuration information of the jth server, hd _1j is the hard disk configuration information of the jth server, net _1j is the network card configuration information of the jth server, raid _j is the redundant disk array of the jth server The configuration information, os _j is the operating system configuration information of the jth server.

Taking the quantization process of the main frequency in the CPU as an example, the following quantization formula shows that when the main frequency is greater than 3Ghz, the quantization value of the CPU main frequency attribute is 100, when it is between 2.0 and 3.0Ghz, the quantization value is 60, and when it is less than 2.0Ghz, the quantization value is 100. The value is 30. This quantization formula provides the default quantization process, but provides a channel for users to copy, because various performance parameters in real scenarios will vary widely due to historical reasons, and a fixed quantization function cannot accurately describe the difference, and with the technology With the iterative development of , new performance indicators will be added, so only the framework layer solution is provided, and the specific measurement standards are handed over to the user, which can meet the actual situation to the greatest extent.

Among them, rate _1j is the CPU frequency, cache _1j is the CPU cache size, rate _2j is the memory frequency value, size _1j is the memory capacity value, size _2j is the hard disk capacity value, interface _j is the hard disk interface type, and speed _1j is the hard disk speed. , cache _2j is the hard disk data buffer capacity, speed _2j is the network card bandwidth, type _1j is the network card type, type _2j is the operating system type, and version _j is the operating system version.

S205: Acquire server subsets corresponding to each preset server class.

A total number of server categories is preset, and it is assumed that each server in the server cluster is divided into m categories, and the server _subsets corresponding to each preset server category are obtained.

S206: Acquire the centroids preset for each server subset respectively.

Set the centroid for each server subset in advance. After obtaining the server subset corresponding to each preset server class, obtain the centroid preset for each server subset, for example, use vi to represent the c _i - _th server subset Centroid.

S207: Calculate each distance value between each server and each centroid according to each target configuration parameter.

After acquiring the centroids preset for each server subset respectively, each distance value between each server and each centroid is calculated according to the parameters of each target server. For example, each distance value between each server and each centroid can be calculated by the following formula:

Among them, dist(d _j ,vi ) _represents the distance between the d _j server and the centroid vi of c _{i ,} q is the distance index, the default is 2, α, β, γ, λ, μ, η are the configuration parameters of each target, respectively The weight value, α+β+γ+λ+μ+η=1, generally set the weight value of each target configuration parameter as follows:

A standard API interface can be provided, allowing users to modify the weight parameters according to their own business.

S208: Calculate the probability value of each server belonging to each preset server class according to each current distance value.

After calculating the distance values between each server and each centroid, calculate the probability value of each server belonging to each preset server class according to the current distance values. For example, the following formula can be used to calculate that each server belongs to each preset server class The probability value of :

Among them, u _ij represents the probability that the d _j server belongs to the class c _i .

Get the membership matrix U:

The above matrix satisfies the following formula:

That is, the sum of the values of each row is equal to 1. Although the probability values of a certain server d _j belonging to the preset server classes c _i are different, the sum of their probabilities must be equal to 1.

S209: Recalculate each centroid according to each current probability value.

After calculating the probability value of each server belonging to each preset server class according to each current distance value, recalculate each centroid according to each current probability value. For example, each centroid can be recalculated according to each current probability value by the following formula:

S210: Calculate each distance value between each server and each recalculated centroid according to each target server parameter.

After each centroid is recalculated according to each current probability value, each distance value between each server and each recalculated centroid is calculated according to each target server parameter. For the calculation process, reference may be made to step S207, which is not repeated here.

S211: Calculate a value function in combination with each current probability value and each current distance value.

After calculating the distance values between each server and the recalculated centroids according to the parameters of each target server, and combining the current probability values and the current distance values to calculate the value function, the value function J can be calculated by the following formula:

S212: Determine whether the value function satisfies the preset condition, if yes, go to step S213, if not, go to step S208.

After calculating the value function in combination with the current probability values and the current distance values, it is judged whether the value function satisfies the preset conditions. Step S213, if not, it means that the currently recalculated centroid is not the best centroid position of the server subset represented by it, and the process returns to step S208.

When the value function is obtained from the first calculation, the preset condition can be that the value function is less than a preset value, such as ε ₁ , to determine whether the value function J is less than the preset value ε ₁ ; when the value function is not obtained from the first calculation, the preset condition can be In order to make the difference between the currently calculated value function and the last calculated value function J _pre smaller than a fixed value ε ₂ , judge whether the difference between the value function J and the last calculated value function J _pre is less than Fixed value ε ₂ .

S213: Acquire each current centroid, and perform clustering and grouping of each server according to each current centroid to obtain each server group.

After it is determined that the calculated value function satisfies the preset condition, each current centroid, that is, the newly calculated centroid, is obtained, and each server is clustered and grouped according to each current centroid to obtain each server group. The process of clustering and grouping each server may include: processing the membership degree matrix U, traversing the membership degree matrix U in a loop, each row represents the probability that the server d _j belongs to the server subset c _i , if u _ij >0.5, it is considered to belong to this A subset of servers, otherwise a subset of other servers. Referring to FIG. 4, FIG. 4 is a schematic diagram of a server grouped according to an embodiment of the present invention. It can be seen from Figure 4 that the servers are grouped into 4 server groups this time.

S214: Perform parallel maintenance operations on each server in units of each server group.

After clustering and grouping each server according to each current centroid to obtain each server group, parallel maintenance operations are performed on each server by taking each server group as a unit. By performing parallel maintenance operations on each server with each server group as a unit, the maintenance efficiency of each server in the server cluster is greatly improved.

It should be noted that the "current" mentioned in each step in the embodiment of the present invention refers to the corresponding parameter value newly obtained in the previous step.

Embodiment three:

Referring to FIG. 5, FIG. 5 is another implementation flowchart of the server grouping method in the embodiment of the present invention, and the method may include the following steps:

S501: Analyze the received server grouping request to obtain a target grouping task.

S502: When the target grouping task is a service distribution task, obtain each real-time performance parameter corresponding to the service distribution task in each server.

When the target grouping task is a service distribution task, each real-time performance parameter corresponding to the service distribution task in each server is obtained respectively. The real-time performance parameters may include current CPU utilization, current memory utilization, current hard disk capacity utilization, current network bandwidth utilization, database connection ratio, hard disk IO bandwidth ratio, and the like. For example, s _j can be used to represent the quantized real-time performance parameter set of the jth server, _sj ={cpu _2j ,mem _2j ,hd _2j ,net _2j ,db _j ,IO _j }, where cpu _2j is the jth server The current CPU utilization, mem _2j is the current memory utilization of the jth server, hd _2j is the current hard disk capacity utilization of the jth server, net _2j is the jth server The current network bandwidth utilization, db _j is the ratio of the number of database connections of the jth server, and IOj is the ratio of the hard disk IO bandwidth of the _jth server.

S503: Acquire server subsets corresponding to each preset server class.

S504: Acquire the centroids preset for each server subset respectively.

S505: Calculate each distance value between each server and each centroid according to each real-time performance parameter.

S506: Calculate the probability value of each server belonging to each preset server class according to each current distance value.

S507: Recalculate each centroid according to each current probability value.

S508: Calculate each distance value between each server and each recalculated centroid according to each target server parameter.

S509: Calculate a value function in combination with each current probability value and each current distance value.

S510: Determine whether the value function satisfies the preset condition, if yes, go to step S511, if not, go to step S506.

S511: Acquire each current centroid, and cluster and group each server according to each current centroid to obtain each server group.

The method of clustering and grouping each server in the server cluster according to each real-time performance parameter can be analogous to the method of clustering and grouping each server in the server cluster according to each target configuration parameter in the second embodiment, and only the target configuration parameter Set p _j is transformed into real-time performance parameter set s _j . The distance functions for solving dist(d _j ,vi ₎ in the two scenarios are different, and different distance functions need to be switched in different scenarios. The parameters of the distance function provide a standard restful interface to the outside world. When different temporary tasks come, users can Customize the parameters of the distance function to better meet the actual needs.

S512: Select a target server group from each server group.

After clustering and grouping each server according to each current centroid to obtain each server group, a target server group is selected from each server group.

The target server group is generally a relatively idle server with relatively abundant resources, low pressure, and low resource occupation.

S513: Distribute the to-be-processed service to each target server in the target server group, so as to use each target server to perform service processing.

After the target server group is selected, the to-be-processed business is distributed to each target server in the target server group, and each target server is used for business processing, thereby realizing efficient business processing.

It should be noted that the "current" mentioned in each step in the embodiment of the present invention refers to the corresponding parameter value newly obtained in the previous step.

Corresponding to the above method embodiments, an embodiment of the present invention further provides a server grouping apparatus, and the server grouping apparatus described below and the server grouping method described above may refer to each other correspondingly.

Referring to FIG. 6, FIG. 6 is a structural block diagram of a server grouping apparatus in an embodiment of the present invention, and the apparatus may include:

The task obtaining module 61 is used to parse the received server grouping request to obtain the target grouping task;

The parameter obtaining module 62 is used to obtain the parameters of each target server corresponding to the target grouping task in each server respectively;

The server group obtaining module 63 is used for clustering and grouping each server based on the parameters of each target server by using the fuzzy C-means clustering algorithm to obtain each server group.

The device provided by the embodiment of the present invention is applied to analyze the received server grouping request to obtain the target grouping task; the parameters of each target server corresponding to the target grouping task in each server are obtained respectively; The target server parameters are used to cluster and group each server to obtain each server group. By obtaining the target server parameters of each server according to the target grouping task, and using the fuzzy C-means clustering algorithm to automatically cluster and group each server based on the parameters of each target server, compared with the existing manual selection server grouping method, the present invention greatly reduces the The influence of human subjective factors is reduced, the grouping accuracy is improved, and the grouping efficiency is greatly improved.

In a specific embodiment of the present invention, the parameter acquisition module 62 includes:

The original configuration parameter obtaining sub-module is used to obtain the original configuration parameters corresponding to the server maintenance tasks in each server;

The parameter quantization sub-module is used to quantify each original configuration parameter by using the configuration quantization function to obtain each target configuration parameter;

The server group obtaining module is specifically a module for clustering and grouping each server based on each target configuration parameter by using the fuzzy C-means clustering algorithm.

In a specific embodiment of the present invention, the parameter quantization sub-module includes:

The quantization function acquisition unit is used to acquire the configuration quantization function preset through the function preset interface;

The parameter quantization unit is used for quantizing each original configuration parameter by using the configuration quantization function.

In a specific embodiment of the present invention, the device may further include:

The server maintenance module is used for clustering and grouping each server based on each target configuration parameter by using the fuzzy C-means clustering algorithm to obtain each server group, and performing parallel maintenance operations on each server with each server group as a unit.

In a specific embodiment of the present invention, the parameter acquisition module 62 includes:

Implementing a performance parameter acquisition sub-module for acquiring real-time performance parameters corresponding to service distribution tasks in each server;

The server group obtaining module is specifically a module for clustering and grouping each server based on each real-time performance parameter by using the fuzzy C-means clustering algorithm.

In a specific embodiment of the present invention, the device may further include:

The server group selection module is configured to use the fuzzy C-means clustering algorithm to cluster and group each server based on each real-time performance parameter, and after each server group is obtained, select a target server group from each server group;

The business processing module is used for distributing the business to be processed to each target server in the target server group, so as to use each target server for business processing.

In a specific embodiment of the present invention, the server group obtaining module 63 includes:

The server subset obtaining submodule is used to obtain the server subset corresponding to each preset server class;

The centroid obtaining submodule is used to obtain the centroids preset for each server subset;

The first distance calculation submodule is used to calculate each distance value between each server and each centroid according to each target server parameter;

The probability calculation submodule is used to calculate the probability value of each server belonging to each preset server class according to each current distance value;

The centroid calculation sub-module is used to recalculate each centroid according to the current probability values;

The second distance calculation submodule is used to calculate each distance value between each server and each recalculated centroid according to each target server parameter;

The value function calculation submodule is used to calculate the value function in combination with the current probability values and the current distance values;

The judgment sub-module is used to judge whether the value function satisfies the preset conditions;

The repeated execution submodule is used to repeatedly execute the step of calculating the probability value of each server belonging to each preset server class according to each distance value when it is determined that the value function does not meet the preset condition;

The server group obtaining sub-module is used to obtain each current centroid when it is determined that the value function satisfies the preset condition, and cluster and group each server according to each current centroid.

Corresponding to the above method embodiments, see FIG. 7 , which is a schematic diagram of a server grouping device provided by the present invention, and the device may include:

memory 71 for storing computer programs;

The processor 72 can implement the following steps when executing the computer program stored in the above-mentioned memory 71:

Analyze the received server grouping request to obtain the target grouping task; obtain the target server parameters corresponding to the target grouping task in each server respectively; use the fuzzy C-means clustering algorithm to cluster each server based on the target server parameters. to get each server group.

For the introduction of the device provided by the present invention, please refer to the above method embodiments, which will not be repeated in the present invention.

Corresponding to the above method embodiments, the present invention also provides a computer-readable storage medium, where a computer program is stored on the computer-readable storage medium, and when the computer program is executed by a processor, the following steps can be implemented:

Analyze the received server grouping request to obtain the target grouping task; obtain the target server parameters corresponding to the target grouping task in each server respectively; use the fuzzy C-means clustering algorithm to cluster each server based on the target server parameters. to get each server group.

The computer-readable storage medium may include: a USB flash drive, a removable hard disk, a read-only memory (Read-Only Memory, ROM), a random access memory (Random Access Memory, RAM), a magnetic disk or an optical disk, etc., which can store program codes. medium.

For the introduction of the computer-readable storage medium provided by the present invention, please refer to the foregoing method embodiments, which will not be repeated in the present invention.

The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments, and the same or similar parts of the various embodiments may be referred to each other. For the apparatuses, devices, and computer-readable storage media disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the descriptions are relatively simple, and reference may be made to the descriptions of the methods for related parts.

The principles and implementations of the present invention are described herein by using specific examples, and the descriptions of the above embodiments are only used to help understand the technical solutions and core ideas of the present invention. It should be pointed out that for those skilled in the art, without departing from the principle of the present invention, several improvements and modifications can also be made to the present invention, and these improvements and modifications also fall within the protection scope of the claims of the present invention.

Claims (10)

1. A server grouping method, comprising:

analyzing the received server grouping request to obtain a target grouping task;

respectively acquiring parameters of each target server corresponding to the target grouping task in each server;

and clustering and grouping the servers based on the parameters of the target servers by using a fuzzy C-means clustering algorithm to obtain each server group.

2. The server grouping method according to claim 1, wherein when the target grouping task is a server maintenance task, respectively obtaining each target server parameter corresponding to the target grouping task in each server comprises:

respectively acquiring each original configuration parameter corresponding to the server maintenance task in each server;

carrying out quantization processing on each original configuration parameter by using a configuration quantization function to obtain each target configuration parameter;

clustering and grouping the servers based on the parameters of the target servers by using a fuzzy C-means clustering algorithm, wherein the clustering and grouping method comprises the following steps:

and clustering and grouping the servers based on the target configuration parameters by using the fuzzy C-means clustering algorithm.

3. The server grouping method according to claim 2, wherein quantizing each of the original configuration parameters using a configuration quantization function includes:

acquiring a configuration quantization function preset through a function preset interface;

and quantizing each original configuration parameter by using the configuration quantization function.

4. The server grouping method according to claim 3, wherein after clustering and grouping the servers based on the target configuration parameters by using the fuzzy C-means clustering algorithm to obtain the server groups, further comprising:

and performing parallel maintenance operation on the servers by taking the server groups as units.

5. The server grouping method according to claim 1, wherein when the target grouping task is a service distribution task, respectively obtaining each target server parameter corresponding to the target grouping task in each server comprises:

respectively acquiring each real-time performance parameter corresponding to the service distribution task in each server;

clustering and grouping the servers based on the parameters of the target servers by using a fuzzy C-means clustering algorithm, wherein the clustering and grouping method comprises the following steps:

and clustering and grouping the servers based on the real-time performance parameters by using the fuzzy C-means clustering algorithm.

6. The method of claim 5, wherein after clustering and grouping the servers based on the real-time performance parameters by using the fuzzy C-means clustering algorithm to obtain the server groups, the method further comprises:

selecting a target server group from each server group;

and distributing the service to be processed to each target server in the target server group so as to utilize each target server to process the service.

7. The server grouping method according to any one of claims 1 to 6, wherein clustering groups the servers based on the target server parameters using a fuzzy C-means clustering algorithm, comprises:

acquiring a server subset corresponding to each preset server class;

obtaining centroids preset for each server subset respectively;

respectively calculating each distance value between each server and each centroid according to each target server parameter;

respectively calculating probability values of the servers belonging to the preset server classes according to the current distance values;

recalculating each centroid according to each current probability value;

calculating each distance value between each server and each recalculated centroid according to each target server parameter;

calculating a value function by combining the current probability values and the current distance values;

judging whether the value function meets a preset condition or not;

if not, the step of respectively calculating the probability value of each server belonging to each preset server class according to each distance value is repeatedly executed;

and if so, acquiring each current centroid, and clustering and grouping the servers according to each current centroid.

8. A server grouping apparatus, comprising:

the task obtaining module is used for analyzing the received server grouping request to obtain a target grouping task;

the parameter acquisition module is used for respectively acquiring parameters of each target server corresponding to the target grouping task in each server;

and the server group obtaining module is used for clustering and grouping the servers based on the target server parameters by using a fuzzy C-means clustering algorithm to obtain the server groups.

9. A server grouping apparatus, comprising:

a memory for storing a computer program;

a processor for implementing the steps of the server grouping method according to any one of claims 1 to 7 when executing the computer program.

10. A computer-readable storage medium, having stored thereon a computer program which, when being executed by a processor, carries out the steps of the server grouping method according to any one of claims 1 to 7.

Images