An AI-based server comprehensive management cloud platform system and management method

Abstract

The present invention discloses a server comprehensive management cloud platform system and management method based on AI, and the system includes: a collection module collects the normal operation parameters, authority parameters and data parameters of each management server and uploads them to the cloud platform; a classification module classifies all management servers according to the business information and function information of each management server, and sets labels for the classification results; a training module constructs an AI deep learning model and trains the AI ​​deep learning model using the set labels of each management server and the normal operation parameters, authority parameters and data parameters of the management server; a management module receives the business request instruction issued by the target account and inputs it into the trained AI deep learning model for parsing, and clusters each management server according to the parsing result. It can realize cross-company or cross-domain server synchronous management, improve management efficiency and management adaptability between servers.

Description

Technical Field

[0001] The present invention relates to the field of server integrated technology, and in particular to an AI-based server integrated management cloud platform system and management method. Background Art

[0002] At present, with the rapid growth of data processing capacity, the scale of servers is also growing exponentially, so the requirements for server management systems (i.e., automatic operation and maintenance systems) are getting higher and higher. However, the overall framework of the existing server management system is very restrictive, which seriously affects the improvement of system performance. Server centers usually deploy a large number of servers to provide various data services. In the case of a large number of servers deployed in the server center, it is usually inconvenient for administrators to quickly find a specific server. The existing technology is that each server manufacturer uses its own server cabinet management module to connect the servers in each server cabinet to the network, so that the server cabinet management module can read the information of the server and then capture the server on the switch. If the server cabinet management module is equipped with servers from other manufacturers, it is impossible to read the server information underneath, which makes the management adaptability between different servers very poor, greatly reducing the management efficiency. Summary of the invention

[0003] In response to the above-mentioned problems, the present invention provides an AI-based server comprehensive management cloud platform system and management method to solve the problem mentioned in the background technology that if the server cabinet management module is equipped with servers from different manufacturers, the server information underneath cannot be read, resulting in poor management compatibility between different servers and greatly reducing management efficiency.

[0004] An AI-based server integrated management cloud platform system, the system comprising:

[0005] A collection module, used to collect the normal operation parameters, permission parameters and data parameters of each management server and upload them to the cloud platform;

[0006] A classification module is used to classify all management servers according to the business information and function information of each management server and set labels for the classification results;

[0007] A training module, used to build an AI deep learning model and train the AI ​​deep learning model using the set labels of each management server and the normal operating parameters and permission parameters and data parameters of the management server;

[0008] The management module is used to receive business request instructions from the target account and input them into the trained AI deep learning model for analysis. According to the analysis results, cluster access, cluster scheduling and cluster operation and maintenance management are performed on each management server.

[0009] Preferably, the acquisition module includes:

[0010] An acquisition submodule, used for acquiring the operation code parameters of each management server, and decoding the operation code parameters to acquire the normal operation parameters of the management server;

[0011] A first retrieving submodule, used to retrieve the operation parameters corresponding to each level of authority from the authority program of each management server as the authority parameters of the management server;

[0012] The second calling submodule is used to connect to the cloud database of each management server to call the cache data, user data and server application data of the management server and use them as data parameters of the management server;

[0013] The upload submodule is used to connect to the cloud platform through a preset communication method and upload the normal operation parameters, permission parameters and data parameters of each management server.

[0014] Preferably, the classification module comprises:

[0015] An extraction submodule, used to extract multiple business attributes of each management server and associated functions of each business attribute;

[0016] A determination submodule, used to determine the business information and function information of each management server according to multiple business attributes of each management server and the associated function of each business attribute;

[0017] A classification submodule is used to classify all management servers based on the business information and function information of each management server based on the same type of standard classification mode, and obtain classification results;

[0018] The setting submodule is used to set a reference label for each category according to the core functional attributes of the category in the classification results.

[0019] Preferably, the training module includes:

[0020] An acquisition submodule is used to acquire a standard model framework and acquire a model engine architecture according to the model framework;

[0021] A construction submodule, used to write a preset AI deep learning algorithm into the model engine architecture to construct the AI ​​deep learning model;

[0022] A training submodule, used to write the setting label of each management server and the normal operation parameters, permission parameters and data parameters of the management server into the model node of the AI ​​deep learning model for deep learning, and train the AI ​​deep learning model using the first number of business requirements and their corresponding expected matching management servers as a training set;

[0023] The verification submodule is used to obtain a second number of business requirements as a verification set and input them into the trained AI deep learning model for verification. According to the verification results, it is determined whether the trained AI deep learning model is qualified. If not, the AI ​​deep learning model is retrained until it is qualified.

[0024] Preferably, the management module includes:

[0025] The parsing submodule is used to receive the business request instruction sent by the target account and input it into the trained AI deep learning model to determine the instruction type corresponding to the business request instruction and the current matching management server;

[0026] An activation submodule, used to activate a corresponding operation program based on the instruction type;

[0027] A deployment submodule, used to deploy cluster operation parameters using the operation program;

[0028] The management submodule is used to perform cluster access management and / or cluster scheduling management and / or cluster operation and maintenance management on the current matching management server based on the cluster operation parameters.

[0029] An AI-based server integrated management method comprises the following steps:

[0030] Collect the normal operation parameters, permission parameters and data parameters of each management server and upload them to the cloud platform;

[0031] Classify all management servers according to the business information and function information of each management server, and set labels for the classification results;

[0032] Building an AI deep learning model and training the AI ​​deep learning model using the set labels of each management server and the normal operating parameters and permission parameters and data parameters of the management server;

[0033] Receive the business request instructions from the target account and input them into the trained AI deep learning model for analysis. Based on the analysis results, perform cluster access, cluster scheduling, and cluster operation and maintenance management on each management server.

[0034] Preferably, the collecting of normal operation parameters, authority parameters and data parameters of each management server and uploading them to the cloud platform includes:

[0035] Obtaining the operation code parameters of each management server, and decoding the operation code parameters to obtain the normal operation parameters of the management server;

[0036] Retrieving the operation parameters corresponding to each level of authority from the authority program of each management server as the authority parameters of the management server;

[0037] Connecting to the cloud database of each management server, retrieving the cache data, user data and server application data of the management server and using them as data parameters of the management server;

[0038] The cloud platform is connected through a preset communication method and the normal operating parameters, authority parameters and data parameters of each management server are uploaded.

[0039] Preferably, the step of classifying all management servers according to the business information and function information of each management server and setting labels for the classification results includes:

[0040] Extracting multiple business attributes of each management server and associated functions of each business attribute;

[0041] Determine the business information and function information of the management server according to the multiple business attributes of each management server and the associated function of each business attribute;

[0042] Classify all management servers based on the business information and function information of each management server based on the same type of standard classification mode, and obtain classification results;

[0043] A reference label is set for each category based on the core functional attributes of the category in the classification results.

[0044] Preferably, the step of constructing an AI deep learning model and training the AI ​​deep learning model using the setting labels of each management server and the normal operating parameters, authority parameters, and data parameters of the management server includes:

[0045] Obtain a standard model framework, and obtain a model engine architecture according to the model framework;

[0046] Writing a preset AI deep learning algorithm into the model engine architecture to construct the AI ​​deep learning model;

[0047] Writing the setting label of each management server and the normal operation parameters, permission parameters and data parameters of the management server into the model node of the AI ​​deep learning model for deep learning, and using the first number of business requirements and their corresponding expected matching management servers as a training set to train the AI ​​deep learning model;

[0048] A second number of business requirements are obtained as a verification set and input into the trained AI deep learning model for verification. Whether the trained AI deep learning model is qualified is determined based on the verification results. If not, the AI ​​deep learning model is retrained until it is qualified.

[0049] Preferably, the receiving of the service request instruction sent by the target account and inputting it into the trained AI deep learning model for parsing, and performing cluster access, cluster scheduling and cluster operation and maintenance management on each management server according to the parsing result, including:

[0050] Receive the business request instruction sent by the target account and input it into the trained AI deep learning model to determine the instruction type corresponding to the business request instruction and the current matching management server;

[0051] activating a corresponding operating procedure based on the instruction type;

[0052] Deploying cluster operating parameters using the operating procedure;

[0053] Cluster access management and / or cluster scheduling management and / or cluster operation and maintenance management are performed on the current matching management server based on the cluster operation parameters.

[0054] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description and the accompanying drawings.

[0055] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] The accompanying drawings are used to provide further understanding of the present invention and constitute a part of the specification. They are used to explain the present invention together with the embodiments of the present invention and do not constitute a limitation of the present invention.

[0057] Figure 1 This is a structural diagram of an AI-based server comprehensive management cloud platform system provided by the present invention;

[0058] Figure 2 This is a structural diagram of a collection module in an AI-based server comprehensive management cloud platform system provided by the present invention;

[0059] Figure 3 This is a workflow diagram of the AI-based server comprehensive management method provided by the present invention;

[0060] Figure 4Another workflow diagram of the AI-based server comprehensive management method provided by the present invention. DETAILED DESCRIPTION

[0061] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present disclosure. Instead, they are merely examples of devices and methods consistent with some aspects of the present disclosure as detailed in the appended claims.

[0062] At present, with the rapid growth of data processing capacity, the scale of servers is also growing exponentially, so the requirements for server management systems (i.e., automatic operation and maintenance systems) are getting higher and higher. However, the overall framework of the existing server management system is very restrictive, which seriously affects the improvement of system performance. Server centers usually deploy a large number of servers to provide various data services. In the case where a large number of servers are deployed in the server center, it is usually inconvenient for administrators to quickly find a specific server. The prior art is that each server manufacturer uses its own server cabinet management module to connect the servers in each server cabinet to the network, so that the server cabinet management module can read the information of the server and then capture the server in the switch. If the server cabinet management module is equipped with servers from other manufacturers, it is impossible to read the server information underneath, which makes the management adaptability between different servers very poor, greatly reducing the management efficiency. In order to solve the above problems, this embodiment discloses an AI-based server comprehensive management cloud platform system.

[0063] An AI-based server integrated management cloud platform system, such as Figure 1 As shown, the system includes:

[0064] The collection module 101 is used to collect the normal operation parameters, authority parameters and data parameters of each management server and upload them to the cloud platform;

[0065] The classification module 102 is used to classify all management servers according to the business information and function information of each management server, and set labels for the classification results;

[0066] A training module 103, used to build an AI deep learning model and train the AI ​​deep learning model using the set tags of each management server and the normal operating parameters, permission parameters and data parameters of the management server;

[0067] The management module 104 is used to receive the business request instructions issued by the target account and input them into the trained AI deep learning model for analysis, and perform cluster access, cluster scheduling and cluster operation and maintenance management on each management server according to the analysis results.

[0068] In this embodiment, the normal operating parameters are represented by the performance index parameters and physical parameters of each management server during normal operation;

[0069] In this embodiment, the authority parameters represent the function parameters of each management server under each operation authority;

[0070] In this embodiment, the data parameters represent cache data, user data, and server application data for each management server;

[0071] In this embodiment, the business information represents the specific information of the relevant design business of each management server;

[0072] In this embodiment, the function information represents the online functions that can be implemented by each management server, such as online data query, etc.;

[0073] In this embodiment, setting a label means identifying a business function label for management servers of the same category.

[0074] The working principle of the above technical solution is as follows: first, the normal operating parameters, permission parameters and data parameters of each management server are collected through the collection module and uploaded to the cloud platform, wherein the normal operating parameters can be server status data, server network connection data, etc., the permission parameters are the operation permissions of each level of users in different servers, and the data parameters can be user usage records and server cache data, etc. Then, the classification module is used to classify all management servers according to the business information and functional information of each management server, and labels are set for the classification results; then, the AI ​​deep learning model is constructed through the training module and the AI ​​deep learning model is trained using the set labels of each management server and the normal operating parameters, permission parameters and data parameters of the management server; finally, the management module is used to receive the business request instructions issued by the target account and input them into the trained AI deep learning model for analysis, and cluster access, cluster scheduling and cluster operation and maintenance management are performed on each management server according to the analysis results.

[0075] The beneficial effects of the above technical solution are as follows: by uploading various parameters of different servers to the cloud platform for unified management, synchronous management of servers across companies or fields can be achieved, thereby improving management efficiency and management adaptability between servers. Furthermore, by setting up an AI deep learning model, a reasonable server can be quickly selected for use according to user needs, thereby improving user experience. Furthermore, by performing cluster unified management of management servers, synchronous access and operation and maintenance between servers can be achieved, thereby improving stability between servers. At the same time, it can also ensure that users can maximize their enjoyment of the data and services they need, thereby improving practicality. It solves the problem in the prior art that if there are servers from different manufacturers under the server cabinet management module, the server information underneath cannot be read, resulting in poor management adaptability between different servers and greatly reducing management efficiency.

[0076] In this embodiment, before performing cluster access, cluster scheduling, and cluster operation and maintenance management on each management server according to the analysis result, the following steps are also included:

[0077] Establish a communication connection with each management server, and obtain the host data corresponding to each management server after the connection is successful;

[0078] generating a server management task for each management server according to the host data;

[0079] Reading pre-stored server information of each management server based on the server management task of the management server;

[0080] Acquire the server centralized management item of each management server according to the pre-stored server information of the management server;

[0081] Retrieving an operation signal corresponding to each server centralized management item of each management server, and generating a dynamic management instruction related to the operation signal;

[0082] Executing multiple dynamic management instructions of each management server to collect multiple dynamic information fed back by the management server;

[0083] Parsing the plurality of dynamic information to determine a change in the server operation state of each management server under each dynamic management instruction;

[0084] Determine the dynamic operation state change parameter and static operation state parameter of the management server under each dynamic management instruction according to the change of the server operation state of each management server under each dynamic management instruction;

[0085] Generating a static monitoring instruction for each management server based on the static operation status parameters of each management server under each dynamic management instruction;

[0086] Integrate the dynamic management instructions and static monitoring instructions of each management server to obtain the remote management authority instructions of the management server;

[0087] The remote management authority instructions of all management servers are stored on the cloud platform to enable cluster access, cluster scheduling and cluster operation and maintenance management of each management server.

[0088] The beneficial effect of the above technical solution is: by generating dynamic management instructions and static monitoring instructions for each management server, dynamic monitoring parameters and static monitoring parameters can be accurately screened out for the management items of each management server, so as to uniformly manage the dynamic parameters in a targeted manner, thereby improving management efficiency while reducing management load, and improving practicality and reliability.

[0089] In one embodiment, Figure 2 As shown, the acquisition module 101 includes:

[0090] The acquisition submodule 1011 is used to acquire the operation code parameters of each management server, and decode the operation code parameters to acquire the normal operation parameters of the management server;

[0091] The first retrieving submodule 1012 is used to retrieve the operation parameters corresponding to each level of authority from the authority program of each management server as the authority parameters of the management server;

[0092] The second retrieval submodule 1013 is used to connect to the cloud database of each management server to retrieve the cache data, user data and server application data of the management server and use them as data parameters of the management server;

[0093] The upload submodule 1014 is used to connect to the cloud platform through a preset communication method and upload the normal operation parameters, authority parameters and data parameters of each management server.

[0094] The beneficial effect of the above technical solution is that by acquiring server parameters of different dimensions step by step, the stability and data reliability in the data acquisition process can be guaranteed, the occurrence of data disorder caused by mixing of various parameters can be avoided, and the stability and practicality can be further improved.

[0095] In one embodiment, the classification module comprises:

[0096] An extraction submodule, used to extract multiple business attributes of each management server and associated functions of each business attribute;

[0097] A determination submodule, used to determine the business information and function information of each management server according to multiple business attributes of each management server and the associated function of each business attribute;

[0098] A classification submodule is used to classify all management servers based on the business information and function information of each management server based on the same type of standard classification mode, and obtain classification results;

[0099] The setting submodule is used to set a reference label for each category according to the core functional attributes of the category in the classification results.

[0100] The beneficial effects of the above technical solution are: the main functional information and business information of each management server can be obtained without comprehensive testing, thereby improving work efficiency. Furthermore, by setting labels for the management servers in each classification category, management servers with different functions can be quickly distinguished, laying the foundation for subsequent model training and further improving practicality.

[0101] In one embodiment, the training module includes:

[0102] An acquisition submodule is used to acquire a standard model framework and acquire a model engine architecture according to the model framework;

[0103] A construction submodule, used to write a preset AI deep learning algorithm into the model engine architecture to construct the AI ​​deep learning model;

[0104] A training submodule, used to write the setting label of each management server and the normal operation parameters, permission parameters and data parameters of the management server into the model node of the AI ​​deep learning model for deep learning, and train the AI ​​deep learning model using the first number of business requirements and their corresponding expected matching management servers as a training set;

[0105] The verification submodule is used to obtain a second number of business requirements as a verification set and input them into the trained AI deep learning model for verification. According to the verification results, it is determined whether the trained AI deep learning model is qualified. If not, the AI ​​deep learning model is retrained until it is qualified.

[0106] The beneficial effects of the above technical solution are: by writing the preset AI deep learning algorithm into the model engine architecture to construct the AI ​​deep learning model, the constructed model can have a stable deep learning function, ensuring the training accuracy and reliability of the model. Furthermore, by verifying the model, the convergence of the model can be determined in a timely manner to ensure that the accuracy of its output results is maintained at the highest level, further improving the practicality.

[0107] In one embodiment, the management module includes:

[0108] The parsing submodule is used to receive the business request instruction sent by the target account and input it into the trained AI deep learning model to determine the instruction type corresponding to the business request instruction and the current matching management server;

[0109] An activation submodule, used to activate a corresponding operation program based on the instruction type;

[0110] A deployment submodule, used to deploy cluster operation parameters using the operation program;

[0111] The management submodule is used to perform cluster access management and / or cluster scheduling management and / or cluster operation and maintenance management on the current matching management server based on the cluster operation parameters.

[0112] The beneficial effects of the above technical solution are: by automatically generating and deploying cluster operation parameters, intelligent multi-server cluster scheduling and operation and maintenance can be achieved, which improves stability and intelligence, reduces the difficulty of use, and further improves the user experience.

[0113] In this embodiment, when cluster access management and / or cluster scheduling management and / or cluster operation and maintenance management are performed on the current matching management server, the operation load of the current matching management server is calculated and the operation load is adjusted in real time according to the calculation result, specifically:

[0114] Detect multiple calling nodes and the operating frequency of each calling node when performing cluster access management and / or cluster scheduling management and / or cluster operation and maintenance management for the current matching management server;

[0115] Obtaining job sequence parameters of operation tasks for cluster access management and / or cluster scheduling management and / or cluster operation and maintenance management for the current matching management server;

[0116] Acquire the system call resource amount of each job subsequence according to the job sequence parameters;

[0117] The predicted load when performing cluster access management and / or cluster scheduling management and / or cluster operation and maintenance management on the current matching management server is calculated based on the multiple calling nodes when performing cluster access management and / or cluster scheduling management and / or cluster operation and maintenance management on the current matching management server, the running frequency of each calling node, and the system call resource amount of each job subsequence:

[0118]

[0119] Wherein, δ represents the predicted load when the current matching management server performs cluster access management and / or cluster scheduling management and / or cluster operation and maintenance management, μ1 represents the first weight value, which is 0.4, M represents the number of calling nodes when the current matching management server performs cluster access management and / or cluster scheduling management and / or cluster operation and maintenance management, i represents the i-th calling node, and D i It is represented as the running frequency of the i-th call node, D ′ represents the operating frequency threshold of the preset maximum load, A1 represents the performance index of the current matching management server, μ2 represents the second weight, which is 0.6, N represents the number of job subsequences in the job sequence of the operation task when the current matching management server performs cluster access management and / or cluster scheduling management and / or cluster operation and maintenance management, j represents the jth job subsequence, F j It is represented by the system call resource amount of the jth job subsequence, F ′ It represents the threshold value set for the amount of virtual resources in the cloud platform, q represents the resource response coefficient of the current matching management server, and A2 represents the average resource utilization rate of the current matching management server;

[0120] The predicted load is confirmed as the operating load of the current matching management server and compared with the preset load. When the predicted load is greater than or equal to the preset load, the operating load of the current matching management server is reduced. When the predicted load is less than the preset load, the operating load of the current matching management server is increased.

[0121] The beneficial effect of the above technical solution is: by calculating the predicted load when performing cluster access management and / or cluster scheduling management and / or cluster operation and maintenance management on the current matching management server, the load parameters of the current matching manager can be timely evaluated and then adaptively adjusted to avoid damage to the system due to excessive load while ensuring stable management operations for each server, further improving stability and practicality.

[0122] This embodiment also discloses an AI-based server comprehensive management method, such as Figure 3 As shown, the following steps are included:

[0123] Step S301, collecting normal operation parameters, authority parameters and data parameters of each management server and uploading them to the cloud platform;

[0124] Step S302: classify all management servers according to the business information and function information of each management server, and set labels for the classification results;

[0125] Step S303: construct an AI deep learning model and train the AI ​​deep learning model using the setting tags of each management server and the normal operating parameters, authority parameters and data parameters of the management server;

[0126] Step S304: Receive the business request instruction sent by the target account and input it into the trained AI deep learning model for analysis, and perform cluster access, cluster scheduling, and cluster operation and maintenance management on each management server according to the analysis results.

[0127] The working principle of the above technical solution is: collect the normal operating parameters, authority parameters and data parameters of each management server and upload them to the cloud platform; classify all management servers according to the business information and functional information of each management server, and set labels for the classification results; build an AI deep learning model and use the set labels of each management server and the normal operating parameters, authority parameters and data parameters of the management server to train the AI ​​deep learning model; receive the business request instructions issued by the target account and input them into the trained AI deep learning model for analysis, and perform cluster access, cluster scheduling and cluster operation and maintenance management for each management server according to the analysis results.

[0128] The beneficial effects of the above technical solution are as follows: by uploading various parameters of different servers to the cloud platform for unified management, synchronous management of servers across companies or fields can be achieved, thereby improving management efficiency and management adaptability between servers. Furthermore, by setting up an AI deep learning model, a reasonable server can be quickly selected for use according to user needs, thereby improving user experience. Furthermore, by performing cluster unified management of management servers, synchronous access and operation and maintenance between servers can be achieved, thereby improving stability between servers. At the same time, it can also ensure that users can maximize their enjoyment of the data and services they need, thereby improving practicality. It solves the problem in the prior art that if there are servers from different manufacturers under the server cabinet management module, the server information underneath cannot be read, resulting in poor management adaptability between different servers and greatly reducing management efficiency.

[0129] In one embodiment, collecting the normal operating parameters, permission parameters and data parameters of each management server and uploading them to the cloud platform includes:

[0130] Obtaining the operation code parameters of each management server, and decoding the operation code parameters to obtain the normal operation parameters of the management server;

[0131] Retrieving the operation parameters corresponding to each level of authority from the authority program of each management server as the authority parameters of the management server;

[0132] Connecting to the cloud database of each management server, retrieving the cache data, user data and server application data of the management server and using them as data parameters of the management server;

[0133] The cloud platform is connected through a preset communication method and the normal operating parameters, authority parameters and data parameters of each management server are uploaded.

[0134] The beneficial effect of the above technical solution is that by acquiring server parameters of different dimensions step by step, the stability and data reliability in the data acquisition process can be guaranteed, the occurrence of data disorder caused by mixing of various parameters can be avoided, and the stability and practicality can be further improved.

[0135] In one embodiment, classifying all management servers according to the business information and function information of each management server and setting labels for the classification results includes:

[0136] Extracting multiple business attributes of each management server and associated functions of each business attribute;

[0137] Determine the business information and function information of the management server according to the multiple business attributes of each management server and the associated function of each business attribute;

[0138] Classify all management servers based on the business information and function information of each management server based on the same type of standard classification mode, and obtain classification results;

[0139] A reference label is set for each category based on the core functional attributes of the category in the classification results.

[0140] The beneficial effects of the above technical solution are: the main functional information and business information of each management server can be obtained without comprehensive testing, thereby improving work efficiency. Furthermore, by setting labels for the management servers in each classification category, management servers with different functions can be quickly distinguished, laying the foundation for subsequent model training and further improving practicality.

[0141] In one embodiment, the step of building an AI deep learning model and training the AI ​​deep learning model using the setting tags of each management server and the normal operating parameters, permission parameters, and data parameters of the management server includes:

[0142] Obtain a standard model framework, and obtain a model engine architecture according to the model framework;

[0143] Writing a preset AI deep learning algorithm into the model engine architecture to construct the AI ​​deep learning model;

[0144] Writing the setting label of each management server and the normal operation parameters, permission parameters and data parameters of the management server into the model node of the AI ​​deep learning model for deep learning, and using the first number of business requirements and their corresponding expected matching management servers as a training set to train the AI ​​deep learning model;

[0145] A second number of business requirements are obtained as a verification set and input into the trained AI deep learning model for verification. Whether the trained AI deep learning model is qualified is determined based on the verification results. If not, the AI ​​deep learning model is retrained until it is qualified.

[0146] The beneficial effects of the above technical solution are: by writing the preset AI deep learning algorithm into the model engine architecture to construct the AI ​​deep learning model, the constructed model can have a stable deep learning function, ensuring the training accuracy and reliability of the model. Furthermore, by verifying the model, the convergence of the model can be determined in a timely manner to ensure that the accuracy of its output results is maintained at the highest level, further improving the practicality.

[0147] In one embodiment, Figure 4 As shown, the business request instruction sent by the target account is received and input into the trained AI deep learning model for parsing, and cluster access, cluster scheduling and cluster operation and maintenance management are performed on each management server according to the parsing result, including:

[0148] Step S401: Receive a service request instruction from a target account and input it into a trained AI deep learning model to determine the instruction type corresponding to the service request instruction and the current matching management server;

[0149] Step S402: activating a corresponding operation program based on the instruction type;

[0150] Step S403: Deploy cluster operation parameters using the operation program;

[0151] Step S404: performing cluster access management and / or cluster scheduling management and / or cluster operation and maintenance management on the current matching management server based on the cluster operation parameters.

[0152] The beneficial effects of the above technical solution are: by automatically generating and deploying cluster operation parameters, intelligent multi-server cluster scheduling and operation and maintenance can be achieved, which improves stability and intelligence, reduces the difficulty of use, and further improves the user experience.

[0153] Other embodiments of the present disclosure will be readily apparent to those skilled in the art after considering the specification and practicing the disclosure disclosed herein. This application is intended to cover any variations, uses, or adaptations of the present disclosure that follow the general principles of the present disclosure and include common knowledge or customary techniques in the art that are not disclosed in the present disclosure. The specification and examples are to be considered exemplary only, and the true scope and spirit of the present disclosure are indicated by the following claims.

[0154] It should be understood that the present disclosure is not limited to the exact structures that have been described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present disclosure is limited only by the appended claims.

Claims

1. An AI-based server integrated management cloud platform system, characterized in that: The system includes: A collection module, used to collect the normal operation parameters, permission parameters and data parameters of each management server and upload them to the cloud platform; A classification module is used to classify all management servers according to the business information and function information of each management server and set labels for the classification results; A training module, used to build an AI deep learning model and train the AI ​​deep learning model using the set labels of each management server and the normal operating parameters and permission parameters and data parameters of the management server; The management module is used to receive the business request instructions from the target account and input them into the trained AI deep learning model for analysis. According to the analysis results, cluster access, cluster scheduling and cluster operation and maintenance management are performed on each management server. Before performing cluster access, cluster scheduling, and cluster operation and maintenance management on each management server based on the analysis results, it also includes: Establish a communication connection with each management server, and obtain the host data corresponding to each management server after the connection is successful; generating a server management task for each management server according to the host data; Reading pre-stored server information of each management server based on the server management task of the management server; Acquire the server centralized management item of each management server according to the pre-stored server information of the management server; Retrieving an operation signal corresponding to each server centralized management item of each management server, and generating a dynamic management instruction related to the operation signal; Executing multiple dynamic management instructions of each management server to collect multiple dynamic information fed back by the management server; Parsing the plurality of dynamic information to determine a change in the server operation state of each management server under each dynamic management instruction; Determine the dynamic operation state change parameter and static operation state parameter of the management server under each dynamic management instruction according to the change of the server operation state of each management server under each dynamic management instruction; Generating a static monitoring instruction for each management server based on the static operation status parameters of each management server under each dynamic management instruction; Integrate the dynamic management instructions and static monitoring instructions of each management server to obtain the remote management authority instructions of the management server; The remote management authority instructions of all management servers are stored on the cloud platform to enable cluster access, cluster scheduling and cluster operation and maintenance management of each management server.

2. The AI-based server integrated management cloud platform system according to claim 1 is characterized in that: The acquisition module comprises: An acquisition submodule, used for acquiring the operation code parameters of each management server, and decoding the operation code parameters to acquire the normal operation parameters of the management server; A first retrieving submodule, used to retrieve the operation parameters corresponding to each level of authority from the authority program of each management server as the authority parameters of the management server; The second calling submodule is used to connect to the cloud database of each management server to call the cache data, user data and server application data of the management server and use them as data parameters of the management server; The upload submodule is used to connect to the cloud platform through a preset communication method and upload the normal operation parameters, permission parameters and data parameters of each management server.

3. The AI-based server integrated management cloud platform system according to claim 1 is characterized in that: The classification module comprises: An extraction submodule, used to extract multiple business attributes of each management server and associated functions of each business attribute; A determination submodule, used to determine the business information and function information of each management server according to multiple business attributes of each management server and the associated function of each business attribute; A classification submodule is used to classify all management servers based on the business information and function information of each management server based on the same type of standard classification mode, and obtain classification results; The setting submodule is used to set a reference label for each category according to the core functional attributes of the category in the classification results.

4. The AI-based server integrated management cloud platform system according to claim 1 is characterized in that: The training module comprises: An acquisition submodule is used to acquire a standard model framework and acquire a model engine architecture according to the model framework; A construction submodule, used to write a preset AI deep learning algorithm into the model engine architecture to construct the AI ​​deep learning model; A training submodule, used to write the setting label of each management server and the normal operation parameters, permission parameters and data parameters of the management server into the model node of the AI ​​deep learning model for deep learning, and train the AI ​​deep learning model using the first number of business requirements and their corresponding expected matching management servers as a training set; The verification submodule is used to obtain a second number of business requirements as a verification set and input them into the trained AI deep learning model for verification. According to the verification results, it is determined whether the trained AI deep learning model is qualified. If not, the AI ​​deep learning model is retrained until it is qualified.

5. The AI-based server integrated management cloud platform system according to claim 1 is characterized in that: The management module comprises: The parsing submodule is used to receive the business request instruction sent by the target account and input it into the trained AI deep learning model to determine the instruction type corresponding to the business request instruction and the current matching management server; An activation submodule, used to activate a corresponding operation program based on the instruction type; A deployment submodule, used to deploy cluster operation parameters using the operation program; The management submodule is used to perform cluster access management and / or cluster scheduling management and / or cluster operation and maintenance management on the current matching management server based on the cluster operation parameters.

6. An AI-based integrated server management method, characterized in that: The following steps are involved: Collect the normal operation parameters, permission parameters and data parameters of each management server and upload them to the cloud platform; Classify all management servers according to the business information and function information of each management server, and set labels for the classification results; Building an AI deep learning model and training the AI ​​deep learning model using the set labels of each management server and the normal operating parameters and permission parameters and data parameters of the management server; Receive the business request instructions from the target account and input them into the trained AI deep learning model for analysis. Based on the analysis results, perform cluster access, cluster scheduling, and cluster operation and maintenance management for each management server. Before performing cluster access, cluster scheduling, and cluster operation and maintenance management on each management server based on the analysis results, it also includes: Establish a communication connection with each management server, and obtain the host data corresponding to each management server after the connection is successful; generating a server management task for each management server according to the host data; Reading pre-stored server information of each management server based on the server management task of the management server; Acquire the server centralized management item of each management server according to the pre-stored server information of the management server; Retrieving an operation signal corresponding to each server centralized management item of each management server, and generating a dynamic management instruction related to the operation signal; Executing multiple dynamic management instructions of each management server to collect multiple dynamic information fed back by the management server; Parsing the plurality of dynamic information to determine a change in the server operation state of each management server under each dynamic management instruction; Determine the dynamic operation state change parameter and static operation state parameter of the management server under each dynamic management instruction according to the change of the server operation state of each management server under each dynamic management instruction; Generating a static monitoring instruction for each management server based on the static operation status parameters of each management server under each dynamic management instruction; Integrate the dynamic management instructions and static monitoring instructions of each management server to obtain the remote management authority instructions of the management server; The remote management authority instructions of all management servers are stored on the cloud platform to enable cluster access, cluster scheduling and cluster operation and maintenance management of each management server.

7. The AI-based server integrated management method according to claim 6, characterized in that: The collecting of normal operation parameters, authority parameters and data parameters of each management server and uploading them to the cloud platform includes: Obtaining the operation code parameters of each management server, and decoding the operation code parameters to obtain the normal operation parameters of the management server; Retrieving the operation parameters corresponding to each level of authority from the authority program of each management server as the authority parameters of the management server; Connecting to the cloud database of each management server, retrieving the cache data, user data and server application data of the management server and using them as data parameters of the management server; The cloud platform is connected through a preset communication method and the normal operating parameters, authority parameters and data parameters of each management server are uploaded.

8. The AI-based server integrated management method according to claim 6, characterized in that: The step of classifying all management servers according to the business information and function information of each management server and setting labels for the classification results includes: Extracting multiple business attributes of each management server and associated functions of each business attribute; Determine the business information and function information of the management server according to the multiple business attributes of each management server and the associated function of each business attribute; Classify all management servers based on the business information and function information of each management server based on the same type of standard classification mode, and obtain classification results; A reference label is set for each category based on the core functional attributes of the category in the classification results.

9. The AI-based server integrated management method according to claim 6, characterized in that: The step of building an AI deep learning model and training the AI ​​deep learning model using the setting labels of each management server and the normal operating parameters, permission parameters, and data parameters of the management server includes: Obtain a standard model framework, and obtain a model engine architecture according to the model framework; Writing a preset AI deep learning algorithm into the model engine architecture to construct the AI ​​deep learning model; Writing the setting label of each management server and the normal operation parameters, permission parameters and data parameters of the management server into the model node of the AI ​​deep learning model for deep learning, and using the first number of business requirements and their corresponding expected matching management servers as a training set to train the AI ​​deep learning model; A second number of business requirements are obtained as a verification set and input into the trained AI deep learning model for verification. Whether the trained AI deep learning model is qualified is determined based on the verification results. If not, the AI ​​deep learning model is retrained until it is qualified.

10. The AI-based server integrated management method according to claim 6, characterized in that: The receiving of the service request instruction sent by the target account and inputting it into the trained AI deep learning model for parsing, and performing cluster access, cluster scheduling and cluster operation and maintenance management on each management server according to the parsing result, including: Receive the business request instruction sent by the target account and input it into the trained AI deep learning model to determine the instruction type corresponding to the business request instruction and the current matching management server; activating a corresponding operating procedure based on the instruction type; Deploying cluster operating parameters using the operating procedure; Cluster access management and / or cluster scheduling management and / or cluster operation and maintenance management are performed on the current matching management server based on the cluster operation parameters.

Images