Remote control method and system for server

Abstract

The invention discloses a remote control method and system for a server, and relates to the technical field of server remote control, and the method comprises the steps: receiving first voice data inputted by a user, analyzing the first voice data, generating a preliminary instruction chain, building a simulation environment through the current state of the server, obtaining a first performance change result, and transmitting the first performance change result to the server; and generating an operation event stream, converting the update event stream into an update instruction chain, inputting the update instruction chain into the time prediction model, and obtaining a second performance change result. Through the voice recognition and natural language processing technology, the user instruction is analyzed in real time, and the influence of the user instruction on the server performance is predicted; the refresh rate is dynamically adjusted according to the importance of the operation, server resources are optimized, a user can trace back to change instructions, the influence after change is evaluated through a prediction model, and the safety and accuracy of the operation are ensured. Therefore, the user experience is improved, the risk of misoperation is reduced, and the intelligence and high efficiency of remote management of the server are realized.

Description

Technical Field

[0001] The present invention relates to the technical field of server remote control, and in particular to a remote control method and system for a server. Background Art

[0002] With the rapid development of cloud computing and big data technology, remote server management has become an indispensable part of enterprise operations. Traditional server management methods often rely on complex graphical user interfaces (GUIs) or command line interfaces (CLIs), which are not only cumbersome to operate, but also difficult to monitor and predict the impact of operations on server performance in real time.

[0003] At present, the Chinese invention patent with application number CN202410047368.7 discloses a remote control system based on a cloud server, which includes: a client and a cloud processor, the client is connected to the cloud processor, the client is used to send remote control requests to the cloud processor, and receive the remote control results returned by the cloud processor; the cloud processor is used to receive the remote control request sent by the client, and perform corresponding operations according to the request, and then return the operation results to the client; the client includes a login module, a display module, a central processing unit, a hardware control module, a command control module, a file transfer module, a file storage module, an execution signal transmitting module and a feedback signal receiving module, the login module is used to log in to the user account and obtain user information, and the hardware control module is used for the customer to input control information.

[0004] The above technology makes it difficult to make timely adjustments to changes in users' remote operation instructions and simulate the degree of risk, which affects the effect of remote control. Summary of the invention

[0005] The technical problem solved by the present invention is that it is difficult in the prior art to timely adjust changes in user's remote operation instructions and simulate the risk level, which affects the effect of remote control.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0007] A remote control method for a server comprises the following steps:

[0008] Step S1, receiving first voice data input by a user, parsing the first voice data and generating a preliminary instruction chain;

[0009] Step S2, using the current server status to create a simulation environment, inputting the preliminary instruction chain into the simulation environment, recording performance change information, training a time prediction model based on historical performance indicator data and historical instruction parameters, inputting real-time performance data and the preliminary instruction chain into the time prediction model, and obtaining a first performance change result;

[0010] Step S3, converting each step instruction of the preliminary instruction chain and the corresponding execution state into an event according to the first performance change result, wherein the event includes a timestamp, and sorting the events in chronological order to generate an operation event stream, and dividing the events in the operation time stream into important operations and unimportant operations according to a preset event importance table, and assigning refresh frequencies to the important operations and unimportant operations respectively;

[0011] Step S4, receiving the second voice data input by the user, parsing the second voice data, generating an update instruction with a traceable time point, inputting the time point into the operation time stream to find a timestamp corresponding to the time point, replacing the corresponding event with the event corresponding to the update instruction, deleting all events after the corresponding timestamp, updating the event stream, and converting the updated event stream into an update instruction chain;

[0012] Step S5, inputting the update instruction chain into the time prediction model, obtaining the second performance change result, and outputting a prompt signal according to the second performance result.

[0013] Preferably, the step S1 includes the following sub-steps:

[0014] Step S101, receiving first voice data input by a user, recognizing the first voice data, converting the first voice data into a first instruction text, using a natural language processing model to recognize the core intent of the first instruction text, and outputting the first core intent;

[0015] Step S102, input the first core intention into a preset intention instruction comparison table, search for corresponding instructions, integrate the instructions, and generate a preliminary instruction chain.

[0016] Preferably, step S2 includes the following sub-steps:

[0017] Step S201, copying the current server state to an isolated simulation environment, wherein the current server state includes the current CPU, current memory and current network traffic;

[0018] Step S202, inputting the preliminary instruction chain into the simulation environment, executing the preliminary instruction chain in the simulation environment, and recording performance change information;

[0019] Step S203, training a time prediction model according to the performance change information, historical performance index data and historical instruction parameters, obtaining real-time performance data, inputting the real-time performance data and the preliminary instruction chain into the time prediction model, and obtaining a first performance change result.

[0020] Preferably, the mathematical expression of the performance change of the time prediction model is:

[0021]

[0022] in, is the performance data at the kth time point in the future, x t ,x t-1 ,,,x t-n is the performance data of the current time point and historical event points, f is the time prediction model, and k is the order of time points;

[0023] The mathematical expression of the performance change result is:

[0024] ΔP=P after -P before ;

[0025] Among them, ΔP is the change in performance data, P before is the current server status, P after is the server status after simulation;

[0026] The performance data change amount is output as a first performance change result.

[0027] Preferably, step S3 includes the following sub-steps:

[0028] Step S301, converting each step instruction of the preliminary instruction chain and the corresponding execution state into an event according to the first performance change result;

[0029] If the performance data change amount is within the executable performance data change threshold, each step instruction of the preliminary instruction chain and the corresponding execution state are converted into an event;

[0030] The event includes a timestamp, an event type, an operation target, and an execution result;

[0031] Step S302, sorting the events in chronological order, generating and storing an operation event stream;

[0032] Step S303, classifying events in the operation time stream into important operations and unimportant operations according to a preset event importance table, wherein the event importance table includes events and corresponding importance, wherein the importance includes important operations and unimportant operations having different weight values ​​and impact scores, and refresh frequencies are assigned to important operations and unimportant operations respectively.

[0033] Preferably, the refresh frequencies assigned to the important operations and the unimportant operations in step S303 are:

[0034] The refresh frequency of important operations is stored until the end of the instruction chain;

[0035] The refresh frequency of the unimportant operation is the first refresh frequency, and the mathematical expression of the first refresh frequency is:

[0036]

[0037] Among them, R is the first refresh frequency, W is the event weight value, I is the event impact score, and T is the preset adjustment coefficient.

[0038] Preferably, step S4 includes the following sub-steps:

[0039] Step S401, receiving second voice data input by a user, recognizing the second voice data, converting the second voice data into a second instruction text, using a natural language processing model to recognize the core intent of the second instruction text, and outputting it as a second core intent;

[0040] Step S402, inputting the second core intent into a preset intent instruction comparison table, searching for corresponding instructions, integrating the instructions, and generating an update instruction with a traceable time point;

[0041] Step S403, input the time point into the operation time stream to find the timestamp corresponding to the time point, and replace the corresponding event with the event corresponding to the update instruction;

[0042] Step S404, delete all events after the corresponding timestamp, update the event stream, and convert the updated event stream into an update instruction chain.

[0043] Input the update instruction chain into the time prediction model, obtain the second performance change result, and output a prompt signal according to the second performance result

[0044] Preferably, step S5 includes the following sub-steps:

[0045] Step S501, inputting the update instruction chain into the time prediction model to obtain a second performance change result, where the second performance change result is a risk score;

[0046] Step S502, obtaining a performance data baseline value, and outputting a prompt signal according to a second performance result;

[0047] If the risk score is within the preset high risk score threshold, a warning signal is output and remote control is terminated;

[0048] If the risk score is within the preset low risk score threshold, remote control continues.

[0049] Preferably, the mathematical expression of the second performance result is:

[0050]

[0051] Among them, R is the risk score, W is the event weight value, P i is the current performance data, P baseline,iis the performance data baseline value, σ i is the standard deviation of the current performance, n is a natural number greater than 1, and i is the value of n in each calculation.

[0052] A remote control system for a server, which is applied to a remote control method for a server, comprises a first speech analysis module, a first performance change module, an event stream establishment storage module, an instruction integration and update module and a prompt signal generation module;

[0053] The first voice analysis module is used to receive first voice data input by a user, analyze the first voice data and generate a preliminary instruction chain;

[0054] The first performance change module is used to create a simulation environment using the current server state, input the preliminary instruction chain into the simulation environment, record performance change information, train a time prediction model based on historical performance indicator data and historical instruction parameters, input the real-time performance data and the preliminary instruction chain into the time prediction model, and obtain a first performance change result;

[0055] The event stream establishment and storage module is used to convert each step instruction of the preliminary instruction chain and the corresponding execution state into an event according to the first performance change result, wherein the event includes a timestamp, and the events are sorted in chronological order to generate an operation event stream, and the events in the operation time stream are divided into important operations and unimportant operations according to a preset event importance table, and refresh frequencies are respectively assigned to important operations and unimportant operations;

[0056] The instruction integration and update module is used to receive the second voice data input by the user, parse the second voice data, generate an update instruction with a traceable time point, input the time point into the operation time stream to find a timestamp corresponding to the time point, replace the corresponding event with the event corresponding to the update instruction, delete all events after the corresponding timestamp, update the event stream, and convert the update event stream into an update instruction chain;

[0057] The prompt signal generating module is used to input the update instruction chain into the time prediction model, obtain the second performance change result, and output the prompt signal according to the second performance result.

[0058] Beneficial effects of the present invention: The present invention uses speech recognition and natural language processing technology to analyze user instructions in real time and predict their impact on server performance, dynamically adjust the refresh rate according to the importance of the operation, optimize server resources, and users can backtrack to change instructions and evaluate the impact of the changes through the prediction model to ensure the safety and accuracy of the operation. This not only improves the user experience, but also reduces the risk of operational errors, and realizes the intelligence and efficiency of remote server management. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 A flowchart of a remote control method for a server provided by an embodiment of the present invention;

[0060] Figure 2 A basic flow chart of a remote control system for a server provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0061] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings. It is obvious that the described embodiments are only part of the embodiments of the present invention, but not all of the embodiments.

[0062] Example 1, reference Figure 1 , provides a remote control method for a server, comprising the following steps:

[0063] Step S1, receiving first voice data input by a user, parsing the first voice data and generating a preliminary instruction chain.

[0064] Step S2, create a simulation environment using the current server status, input the preliminary instruction chain into the simulation environment, record performance change information, train the time prediction model based on historical performance indicator data and historical instruction parameters, input the real-time performance data and the preliminary instruction chain into the time prediction model, and obtain the first performance change result.

[0065] Step S3, according to the first performance change result, each step instruction of the preliminary instruction chain and the corresponding execution state are converted into events, the events include a timestamp, the events are sorted in chronological order, an operation event stream is generated, and the events in the operation time stream are divided into important operations and unimportant operations according to a preset event importance table, and refresh frequencies are assigned to important operations and unimportant operations respectively.

[0066] Step S4, receiving the second voice data input by the user, parsing the second voice data, generating an update instruction with a traceback time point, inputting the time point into the operation time stream to find the timestamp corresponding to the time point, replacing the corresponding event with the event corresponding to the update instruction, deleting all events after the corresponding timestamp, updating the event stream, and converting the updated event stream into an update instruction chain.

[0067] Step S5, inputting the update instruction chain into the time prediction model, obtaining the second performance change result, and outputting a prompt signal according to the second performance result.

[0068] Step S1 includes the following sub-steps:

[0069] Step S101, receiving first voice data input by a user, recognizing the first voice data, converting the first voice data into a first instruction text, using a natural language processing model to recognize the core intent of the first instruction text, and outputting it as the first core intent.

[0070] Step S101 accurately receives and recognizes the user's voice input, converts it into a first instruction text in text form, uses a natural language processing model to deeply analyze the first instruction text, accurately identifies the user's operation intention, that is, the first core intention, and provides key information for the generation of subsequent instructions.

[0071] Step S102, input the first core intention into a preset intention instruction comparison table, search for corresponding instructions, integrate the instructions, and generate a preliminary instruction chain.

[0072] Step S102 matches the identified first core intent with the preset intent instruction comparison table, quickly finds the corresponding server operation instructions, integrates and optimizes the found instructions, and forms a preliminary instruction chain with clear logic and strong executableness, providing input for subsequent performance prediction and simulation execution.

[0073] The main effect of step S1 is to convert the user's voice input into a preliminary instruction chain executable by the server. Through advanced speech recognition and natural language processing technology, the user's operation intention is accurately captured and converted into specific server operation instructions, laying a solid foundation for subsequent execution and performance prediction.

[0074] Step S2 includes the following sub-steps:

[0075] Step S201, copying the current server state to an isolated simulation environment, wherein the current server state includes the current CPU, current memory and current network traffic.

[0076] Step S201 copies the key status information of the current server to an independent simulation environment to ensure that the simulation process does not affect the normal operation of the actual server, ensure the consistency of the simulation environment and the actual server in the initial state, and provide an accurate basis for subsequent instruction execution and performance prediction.

[0077] Step S202: input the preliminary instruction chain into the simulation environment, execute the preliminary instruction chain in the simulation environment, and record performance change information.

[0078] Step S202 safely executes the preliminary instruction chain in the simulation environment, simulates the execution process of the actual operation, monitors and records the performance change information after the preliminary instruction chain is executed in the simulation environment in real time, and provides data support for subsequent analysis and prediction.

[0079] Step S203, training a time prediction model according to the performance change information, historical performance index data and historical instruction parameters, obtaining real-time performance data, inputting the real-time performance data and the preliminary instruction chain into the time prediction model, and obtaining a first performance change result.

[0080] The mathematical expression of the performance change of the time prediction model is:

[0081]

[0082] in, is the performance data at the kth time point in the future, x t ,x t-1 ,,,x t-n is the performance data of the current time point and historical event points, f is the time prediction model, and k is the time point sequence.

[0083] The mathematical expression of the performance change result is:

[0084] ΔP=P after -P before ;

[0085] Among them, ΔP is the change in performance data, P before is the current server status, P after This is the server status after simulation.

[0086] The performance data change amount is output as a first performance change result.

[0087] Step S203 uses historical performance indicator data, historical instruction parameters and performance change information of the current simulation environment to train a time prediction model so that it can accurately predict the impact of instruction execution on server performance. The real-time performance data and the preliminary instruction chain are input into the trained time prediction model, and the performance data change is output as the first performance change result, providing key information for subsequent operational decisions and event sorting.

[0088] The main effect of step S2 is to predict the impact of the preliminary instruction chain on server performance through a simulated environment without interfering with the actual server operation. By copying the current server status to an isolated simulated environment and executing the preliminary instruction chain therein, the performance changes that may be brought about by the operation can be safely and accurately evaluated, providing a basis for subsequent decision-making.

[0089] Step S3 includes the following sub-steps:

[0090] Step S301: convert each step instruction of the preliminary instruction chain and the corresponding execution state into an event according to the first performance change result.

[0091] If the performance data change amount is within the executable performance data change threshold, each step instruction of the preliminary instruction chain and the corresponding execution state are converted into an event.

[0092] An event contains a timestamp, event type, operation target, and execution result.

[0093] Step S301 converts each step of the preliminary instruction chain and its execution status into a specific event according to the first performance change result. This conversion process ensures that each operation and its result can be accurately recorded and tracked. By judging whether the performance data change is within the executable performance data change threshold, it is determined which instructions' execution status needs to be converted into events, which helps to filter out operations with less impact on performance and reduce unnecessary event records. Each event contains key information such as timestamp, event type, operation target and execution result, providing sufficient data support for subsequent event sorting and importance assessment.

[0094] Step S302, sorting the events in chronological order, generating and storing an operation event stream.

[0095] Step S302 sorts the converted events in chronological order, generates an operation event stream, and stores it, thereby ensuring the orderliness and traceability of the events and providing a basis for subsequent analysis and decision-making.

[0096] Step S303, classifying events in the operation time stream into important operations and unimportant operations according to a preset event importance table, wherein the event importance table includes events and corresponding importance, wherein the importance includes important operations and unimportant operations having different weight values ​​and impact scores, and refresh frequencies are assigned to important operations and unimportant operations respectively.

[0097] In step S303, the refresh frequencies assigned to important operations and unimportant operations are:

[0098] The refresh rate of important operations is stored until the end of the instruction chain.

[0099] The refresh frequency of the unimportant operation is the first refresh frequency, and the mathematical expression of the first refresh frequency is:

[0100]

[0101] Among them, P is the first refresh frequency, W is the event weight value, I is the event impact score, and T is the preset adjustment coefficient.

[0102] Step S303 divides the events in the operation event stream into important operations and unimportant operations according to the preset event importance table, which helps to distinguish between critical operations and unimportant operations, provides a basis for subsequent resource allocation, and assigns different refresh frequencies to important operations and unimportant operations. Important operations are assigned a higher refresh frequency to ensure that they receive continuous attention; unimportant operations are assigned a lower first refresh frequency to reduce resource usage, improve overall operational efficiency, and ensure timely response to critical operations.

[0103] Step S3 converts the execution process of the preliminary instruction chain into a series of ordered events, and assigns different refresh frequencies according to the importance of the events, so as to efficiently manage and monitor these events, ensuring that key operations receive continuous attention, while reducing the resource usage of unimportant operations and improving overall operational efficiency.

[0104] Step S4 includes the following sub-steps:

[0105] Step S401, receiving second voice data input by the user, recognizing the second voice data, converting the second voice data into a second instruction text, using a natural language processing model to recognize the core intent of the second instruction text, and outputting it as a second core intent.

[0106] Step S401 accurately receives and recognizes the user's second voice data, converts it into a second instruction text in text form, uses a natural language processing model to deeply analyze the second instruction text, accurately recognizes the user's operation intention, that is, the second core intention, and provides key information for the generation of subsequent instructions.

[0107] Step S402, input the second core intent into the preset intent instruction comparison table, search for the corresponding instruction, integrate the instructions, and generate an update instruction with a traceable time point.

[0108] Step S402 matches the identified second core intent with the preset intent instruction comparison table, quickly finds the corresponding server operation instructions, integrates the found instructions, and adds tracing time point information to generate update instructions with tracing time points, thereby ensuring the accuracy and traceability of the update instructions.

[0109] Step S403: input the time point into the operation time stream to search for a timestamp corresponding to the time point, and replace the corresponding event with the event corresponding to the update instruction.

[0110] Step S403 inputs the time point into the operation time stream, searches for the timestamp corresponding to the time point, and replaces the corresponding event with the event corresponding to the update instruction, thereby ensuring the real-time and accuracy of the operation event stream.

[0111] Step S404, delete all events after the corresponding timestamp, update the event stream, and convert the updated event stream into an update instruction chain.

[0112] Step S404 deletes all events after the corresponding timestamp to reflect the user's latest operation intention. Then, the event stream is updated and converted into an updated instruction chain, ensuring the consistency of the instruction chain and the operation event stream, providing an accurate basis for subsequent execution.

[0113] The main effect of step S4 is to allow the user to modify or update the operation event flow through voice input, ensuring the flexibility and real-time performance of the operation instructions. By recognizing the user's voice instructions, converting them into specific operation intentions, and updating the operation event flow accordingly, a new instruction chain is generated to adapt to changing operation requirements.

[0114] Step S5 includes the following sub-steps:

[0115] Step S501: input the update instruction chain into the time prediction model to obtain a second performance change result, where the second performance change result is a risk score.

[0116] Step S501 inputs the update instruction chain into the time prediction model, uses the model to predict the performance changes that may be brought about by executing the instruction chain, and outputs it as a risk score, providing key performance risk information for subsequent decision-making.

[0117] Step S502, obtaining a performance data baseline value, and outputting a prompt signal according to the second performance result.

[0118] If the risk score is within the preset high risk score threshold, a warning signal is output and remote control is terminated.

[0119] If the risk score is within the preset low risk score threshold, remote control continues.

[0120] The mathematical expression of the second performance result is:

[0121]

[0122] Among them, R is the risk score, W is the event weight value, P i is the current performance data, P baseline,i is the performance data baseline value, σ i is the standard deviation of the current performance, n is a natural number greater than 1, and i is the value of n in each calculation.

[0123] Step S502 first obtains the performance data baseline value of the server as a benchmark for evaluating the current operation risk. Then, the second performance change result is compared with the preset risk score threshold. If the risk score is within the preset high risk score threshold, it indicates that executing the update instruction chain may bring a greater performance risk. At this time, a warning signal is output and remote control is terminated to avoid potential performance problems. If the risk score is within the preset low risk score threshold, it indicates that executing the update instruction chain has little impact on the server performance. At this time, the remote control operation can continue to ensure the continuity and efficiency of the operation.

[0124] The main effect of step S5 is to perform a performance risk assessment on the updated instruction chain and decide whether to continue the remote control operation based on the assessment results. The risk score is obtained by inputting the updated instruction chain into the time prediction model and compared with the preset risk score threshold, thereby ensuring the safety and stability of the remote operation.

[0125] The present invention generates a preliminary instruction chain by real-time analysis of the voice data input by the user, and predicts its performance impact in a simulated environment. This enables the user to understand the performance changes that may be caused by the operation in a timely manner before performing the actual operation, so as to make a more informed decision. If the user needs to change the instruction midway, it can be convenient to go back to a specific time point, replace or delete the corresponding event, and ensure the accuracy and flexibility of the operation instruction. The present invention uses a time prediction model to predict the performance change results of the preliminary instruction chain based on historical performance indicator data and the current server status. According to the prediction results, the operation events are divided into important operations and unimportant operations, and different refresh frequencies are assigned to each of them. For unimportant operations that have little effect on the instruction results, a lower refresh frequency is used to reduce system overhead; for important operations that have a significant impact on the instruction results, a higher refresh frequency is maintained until the operation is completed to ensure the real-time and accuracy of key operations.

[0126] The present invention allows the user to trace back to any time point during the remote control process and change the previous operation instructions. This greatly improves the user experience and enables users to more conveniently correct errors or adjust strategies without having to re-execute the entire operation sequence from the beginning. At the same time, by predicting the impact of the changed instructions on server performance, users can make decisions more confidently and avoid potential risks. On the basis of allowing users to trace back to change instructions, the present invention also predicts the performance impact of the changed instructions through a time prediction model. This enables users to intuitively understand the performance changes that may be brought about by the changed instructions, so as to make more reasonable decisions. This not only improves the accuracy of the operation, but also reduces the risk of server performance degradation or failure due to operational errors.

[0127] Example 2, reference Figure 2, provides a remote control system for a server, including a first voice analysis module, a first performance change module, an event stream establishment storage module, an instruction integration and update module and a prompt signal generation module.

[0128] The first voice analysis module is used to receive first voice data input by a user, analyze the first voice data and generate a preliminary instruction chain.

[0129] The first performance change module is used to create a simulation environment using the current server status, input the preliminary instruction chain into the simulation environment, record performance change information, train the time prediction model based on historical performance indicator data and historical instruction parameters, input the real-time performance data and the preliminary instruction chain into the time prediction model, and obtain the first performance change result.

[0130] The event stream establishment storage module is used to convert each step instruction of the preliminary instruction chain and the corresponding execution status into an event according to the first performance change result. The event includes a timestamp, and the events are sorted in chronological order to generate an operation event stream. The events in the operation time stream are divided into important operations and unimportant operations according to a preset event importance table, and refresh frequencies are assigned to important operations and unimportant operations respectively.

[0131] The instruction integration and update module is used to receive the second voice data input by the user, parse the second voice data, generate an update instruction with a traceable time point, input the time point into the operation time stream to find the timestamp corresponding to the time point, replace the corresponding event with the event corresponding to the update instruction, delete all events after the corresponding timestamp, update the event stream, and convert the update event stream into an update instruction chain.

[0132] The prompt signal generation module is used to input the update instruction chain into the time prediction model, obtain the second performance change result, and output the prompt signal according to the second performance result.

[0133] Through the first voice analysis module, users can directly input commands through voice without manual input, which greatly improves the operation efficiency. The voice analysis module can accurately identify and analyze user commands, generate preliminary command chains, and reduce the possibility of human operation errors. The first performance change module uses the simulation environment and time prediction model to accurately predict the impact of the preliminary command chain on server performance and provide data support for decision-making. Through the combination of real-time performance data and historical data, the accuracy of the time prediction model is improved, further enhancing the risk control ability. The event flow establishment storage module converts the instruction execution process into an orderly event flow for easy tracking and analysis. It classifies events according to the event importance table and assigns different refresh frequencies to ensure that key operations are processed in a timely manner and reduce the resource occupation of unimportant operations. The instruction integration and update module allows users to update instructions through voice input to achieve flexible adjustment of operations. The updated instructions can accurately replace the original events and delete subsequent unnecessary events to ensure the consistency and accuracy of the operation process. The prompt signal generation module can evaluate the impact of the updated instruction chain on server performance in real time and output corresponding prompt signals according to the risk score. The output of the warning signal can terminate potential high-risk operations in time, protect the server from damage, and ensure the security of remote control.

[0134] It should be understood by those skilled in the art that the embodiments of the present invention can be provided as methods, systems or computer program products. Therefore, the present invention can take the form of a complete hardware embodiment, a complete software embodiment or an embodiment combining software and hardware. Moreover, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media containing computer-usable program codes. Among them, the storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (Static Random Access Memory, referred to as SRAM), electrically erasable programmable read-only memory (Electrically Erasable Programmable Read-Only Memory, referred to as EEPROM), erasable programmable read-only memory (Erasable Programmable Read Only Memory, referred to as EPROM), programmable read-only memory (Programmable Read-Only Memory, referred to as PROM), read-only memory (Read-Only Memory, referred to as ROM), magnetic memory, flash memory, disk or optical disk. These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0135] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A remote control method for a server, characterized in that: The steps include: Step S1, receiving first voice data input by a user, parsing the first voice data and generating a preliminary instruction chain; Step S2, using the current server status to create a simulation environment, inputting the preliminary instruction chain into the simulation environment, recording performance change information, training a time prediction model based on historical performance indicator data and historical instruction parameters, inputting real-time performance data and the preliminary instruction chain into the time prediction model, and obtaining a first performance change result; Step S3, converting each step instruction of the preliminary instruction chain and the corresponding execution state into an event according to the first performance change result, wherein the event includes a timestamp, and sorting the events in chronological order to generate an operation event stream, and dividing the events in the operation time stream into important operations and unimportant operations according to a preset event importance table, and assigning refresh frequencies to the important operations and unimportant operations respectively; Step S4, receiving the second voice data input by the user, parsing the second voice data, generating an update instruction with a traceable time point, inputting the time point into the operation time stream to find a timestamp corresponding to the time point, replacing the corresponding event with the event corresponding to the update instruction, deleting all events after the corresponding timestamp, updating the event stream, and converting the updated event stream into an update instruction chain; Step S5, inputting the update instruction chain into the time prediction model, obtaining the second performance change result, and outputting a prompt signal according to the second performance result.

2. A remote control method for a server as claimed in claim 1, characterized in that: The step S1 includes the following sub-steps: Step S101, receiving first voice data input by a user, recognizing the first voice data, converting the first voice data into a first instruction text, using a natural language processing model to recognize the core intent of the first instruction text, and outputting the first core intent; Step S102, input the first core intention into a preset intention instruction comparison table, search for corresponding instructions, integrate the instructions, and generate a preliminary instruction chain.

3. A remote control method for a server as claimed in claim 2, characterized in that: The step S2 includes the following sub-steps: Step S201, copying the current server state to an isolated simulation environment, wherein the current server state includes the current CPU, current memory and current network traffic; Step S202, inputting the preliminary instruction chain into the simulation environment, executing the preliminary instruction chain in the simulation environment, and recording performance change information; Step S203, training a time prediction model according to the performance change information, historical performance index data and historical instruction parameters, obtaining real-time performance data, inputting the real-time performance data and the preliminary instruction chain into the time prediction model, and obtaining a first performance change result.

4. A remote control method for a server as claimed in claim 3, characterized in that: The mathematical expression of the performance change of the time prediction model is: in, is the performance data at the kth time point in the future, x t ,x t-1 ,,,x t-n is the performance data of the current time point and historical event points, f is the time prediction model, and k is the order of time points; The mathematical expression of the performance change result is: ΔP=P after -P before ; Among them, ΔP is the change in performance data, P before is the current server status, P after is the server status after simulation; The performance data change amount is output as a first performance change result.

5. A remote control method for a server as claimed in claim 4, characterized in that: The step S3 includes the following sub-steps: Step S301, converting each step instruction of the preliminary instruction chain and the corresponding execution state into an event according to the first performance change result; If the performance data change amount is within the executable performance data change threshold, each step instruction of the preliminary instruction chain and the corresponding execution state are converted into an event; The event includes a timestamp, an event type, an operation target, and an execution result; Step S302, sorting the events in chronological order, generating and storing an operation event stream; Step S303, classifying events in the operation time stream into important operations and unimportant operations according to a preset event importance table, wherein the event importance table includes events and corresponding importance, wherein the importance includes important operations and unimportant operations having different weight values ​​and impact scores, and refresh frequencies are assigned to important operations and unimportant operations respectively.

6. A remote control method for a server as claimed in claim 5, characterized in that: The refresh frequencies assigned to important operations and unimportant operations in step S303 are: The refresh frequency of important operations is stored until the end of the instruction chain; The refresh frequency of the unimportant operation is the first refresh frequency, and the mathematical expression of the first refresh frequency is: Among them, R is the first refresh frequency, W is the event weight value, I is the event impact score, and T is the preset adjustment coefficient.

7. A remote control method for a server as claimed in claim 6, characterized in that: The step S4 includes the following sub-steps: Step S401, receiving second voice data input by a user, recognizing the second voice data, converting the second voice data into a second instruction text, using a natural language processing model to recognize the core intent of the second instruction text, and outputting it as a second core intent; Step S402, inputting the second core intent into a preset intent instruction comparison table, searching for corresponding instructions, integrating the instructions, and generating an update instruction with a traceable time point; Step S403, input the time point into the operation time stream to find the timestamp corresponding to the time point, and replace the corresponding event with the event corresponding to the update instruction; Step S404, delete all events after the corresponding timestamp, update the event stream, and convert the updated event stream into an update instruction chain. The update instruction chain is input into the time prediction model to obtain a second performance change result, and a prompt signal is output according to the second performance result.

8. A remote control method for a server as claimed in claim 7, characterized in that: The step S5 comprises the following sub-steps: Step S501, inputting the update instruction chain into the time prediction model to obtain a second performance change result, where the second performance change result is a risk score; Step S502, obtaining a performance data baseline value, and outputting a prompt signal according to a second performance result; If the risk score is within the preset high risk score threshold, a warning signal is output and remote control is terminated; If the risk score is within the preset low risk score threshold, remote control continues.

9. A remote control method for a server as claimed in claim 8, characterized in that: The mathematical expression of the second performance result is: Among them, R is the risk score, W is the event weight value, P i is the current performance data, P baseline,i is the performance data baseline value, σ i is the standard deviation of the current performance, n is a natural number greater than 1, and i is the value of n in each calculation.

10. A remote control system for a server, applied to a remote control method for a server as claimed in any one of claims 1 to 8, characterized in that: It includes a first speech analysis module, a first performance change module, an event stream establishment and storage module, an instruction integration and update module, and a prompt signal generation module; The first voice analysis module is used to receive first voice data input by a user, analyze the first voice data and generate a preliminary instruction chain; The first performance change module is used to create a simulation environment using the current server state, input the preliminary instruction chain into the simulation environment, record performance change information, train a time prediction model based on historical performance indicator data and historical instruction parameters, input the real-time performance data and the preliminary instruction chain into the time prediction model, and obtain a first performance change result; The event stream establishment and storage module is used to convert each step instruction of the preliminary instruction chain and the corresponding execution state into an event according to the first performance change result, wherein the event includes a timestamp, and the events are sorted in chronological order to generate an operation event stream, and the events in the operation time stream are divided into important operations and unimportant operations according to a preset event importance table, and refresh frequencies are respectively assigned to important operations and unimportant operations; The instruction integration and update module is used to receive the second voice data input by the user, parse the second voice data, generate an update instruction with a traceable time point, input the time point into the operation time stream to find a timestamp corresponding to the time point, replace the corresponding event with the event corresponding to the update instruction, delete all events after the corresponding timestamp, update the event stream, and convert the update event stream into an update instruction chain; The prompt signal generating module is used to input the update instruction chain into the time prediction model, obtain the second performance change result, and output the prompt signal according to the second performance result.

Images