A method and system for intelligent operation and maintenance of application system

Abstract

The present invention provides an application system intelligent operation and maintenance method and system in the field of intelligent operation and maintenance technology, the method comprising the following steps: step S1, real-time collection of application system including performance indicators and operation log operation data, encrypting the operation data into encrypted data and uploading it to a server; step S2, the server decrypts the received encrypted data to obtain operation data, analyzes the operation data through a machine learning algorithm, and obtains an analysis report; step S3, generates a corresponding fault handling process based on the analysis report, and automatically handles the fault of the application system based on the fault handling process; step S4, real-time recording of the fault handling log of the application system, and encrypting and storing the fault handling log. The advantages of the present invention are: greatly improving the accuracy and timeliness of application system fault handling.

Description

Technical Field

[0001] The present invention relates to the field of intelligent operation and maintenance technology, and in particular to an application system intelligent operation and maintenance method and system. Background Art

[0002] An application system is a computer system used to implement specific applications / functions. Some failures are inevitable during the operation of the application system, so the operation and maintenance of the application system involves the monitoring and handling of failures.

[0003] Intelligent operation and maintenance is an operation and maintenance method based on application system early warning and artificial intelligence technology. Its advantages are that it can effectively handle complex situations of multiple application systems and multiple early warnings, improve the efficiency and accuracy of fault handling, and promptly notify relevant personnel, effectively reduce the time and cost of human intervention, and improve the stability of the application system and customer experience. However, with the increase in the number of application system functions and the increase in complexity, the early warning information has become larger and more complex, resulting in a gradual decrease in the accuracy of fault judgment and the accuracy of automated processing, which in turn affects the timeliness of fault handling. If the fault is not handled in a timely manner, it may affect the normal operation of the application system, and thus affect the customer experience.

[0004] Therefore, how to provide an intelligent operation and maintenance method and system for application systems to improve the accuracy and timeliness of application system fault handling has become a technical problem that needs to be solved urgently. Summary of the invention

[0005] The technical problem to be solved by the present invention is to provide an intelligent operation and maintenance method and system for an application system, so as to improve the accuracy and timeliness of application system fault handling.

[0006] In a first aspect, the present invention provides an application system intelligent operation and maintenance method, comprising the following steps:

[0007] Step S1, collecting the operation data of the application system including performance indicators and operation logs in real time, encrypting the operation data into encrypted data and uploading it to the server;

[0008] Step S2: The server decrypts the received encrypted data to obtain operation data, and analyzes the operation data using a machine learning algorithm to obtain an analysis report;

[0009] Step S3: generating a corresponding fault handling process based on the analysis report, and automatically handling the fault of the application system based on the fault handling process;

[0010] Step S4: Record the fault processing log of the application system in real time, and encrypt and store the fault processing log.

[0011] Furthermore, the step S1 is specifically as follows:

[0012] Real-time collection of operating data of the application system including performance indicators and operation logs; the performance indicators at least include response time, throughput, number of concurrent users, resource utilization, system availability, system stability, error rate, time processing time, page loading time, data throughput, latency, peak performance, fault tolerance, recovery time and energy consumption; the operation log at least includes log level, message, user information, operation details, error code, exception information, system status, resource usage, transaction ID, network information, security events and log ID;

[0013] The operation data is hashed by the SHA-3 algorithm to obtain a hash value, a current timestamp is obtained, the operation data, the hash value and the timestamp are encrypted by the AES-256 algorithm to obtain a first encrypted string, each character in the first encrypted string is shifted rightward based on a preset first displacement to obtain a second encrypted string, the second encrypted string is encrypted by the 3DES algorithm to obtain a third encrypted string, each character in the third encrypted string is shifted rightward based on a preset second displacement to obtain a fourth encrypted string, the fourth encrypted string is encrypted by the RC6 algorithm to obtain a fifth encrypted string, each character in the fifth encrypted string is shifted rightward based on a preset third displacement to obtain a sixth encrypted string, and the sixth encrypted string is encrypted by the ECDSA algorithm to obtain encrypted data.

[0014] Furthermore, the step S2 is specifically as follows:

[0015] The server creates a fault recognition model based on a recurrent neural network for identifying faults that have occurred and predicting faults that are about to occur, sets the loss function of the fault recognition model to be a cross-entropy loss function, obtains a large amount of historical operation data for preprocessing and fault annotation to construct a data set, and divides the data set into a training set, a validation set, and a test set based on a preset ratio; trains the fault recognition model with the training set until the loss value of the cross-entropy loss function is less than a preset loss threshold, and then uses the validation set to calculate the recognition accuracy to verify the trained fault recognition model, and uses the test set to calculate the confidence to test the verified fault recognition model;

[0016] The server receives the encrypted data in real time, decrypts the encrypted data through the ECDSA algorithm to obtain a sixth encrypted string, shifts each character in the sixth encrypted string to the left based on a preset third displacement to obtain a fifth encrypted string, decrypts the fifth encrypted string through the RC6 algorithm to obtain a fourth encrypted string, shifts each character in the fourth encrypted string to the left based on a preset second displacement to obtain a third encrypted string, decrypts the third encrypted string through the 3DES algorithm to obtain a second encrypted string, shifts each character in the second encrypted string to the left based on a preset first displacement to obtain a first encrypted string, decrypts the first encrypted string through the AES-256 algorithm to obtain running data, a hash value, and a timestamp, performs a time validity check through the timestamp, and then performs an integrity check on the running data through the hash value;

[0017] The server inputs the verified operating data into a fault identification model for analysis to obtain an analysis report including at least abnormal data, fault type, number of faults that have occurred, number of predicted faults, and analysis time.

[0018] Furthermore, the step S3 is specifically as follows:

[0019] Obtain a large number of historical fault processing logs, learn each of the historical fault processing logs through machine learning technology, obtain the fault processing process for correctly handling faults of different fault types, and bind and store the fault types and fault processing processes in a pre-created fault management table;

[0020] Based on the fault type carried in the analysis report, a corresponding fault handling process is matched from the fault management table, and the fault handling process is sent to the application system for automatic execution, so as to automatically handle the fault of the application system.

[0021] Furthermore, the step S4 is specifically as follows:

[0022] Record the fault handling log of the application system in real time, encrypt the fault handling log by SM3 algorithm to obtain level one ciphertext data, swap the number 1 with the letter Q in the level one ciphertext data to obtain level two ciphertext data, encrypt the level two ciphertext data by SM4 algorithm to obtain level three ciphertext data, swap the number 2 with the letter W in the level three ciphertext data to obtain level four ciphertext data, encrypt the level four ciphertext data by SM7 algorithm to obtain level five ciphertext data, swap the number 3 with the letter E in the level five ciphertext data to obtain level six ciphertext data, encrypt the level six ciphertext data by 3DES algorithm to obtain an encrypted log, and store and distribute the encrypted log.

[0023] In a second aspect, the present invention provides an application system intelligent operation and maintenance system, including the following modules:

[0024] An operation data encryption upload module is used to collect the operation data of the application system including performance indicators and operation logs in real time, and encrypt the operation data into encrypted data and upload it to the server;

[0025] An operation data analysis module, used for the server to decrypt the received encrypted data to obtain operation data, and analyze the operation data through a machine learning algorithm to obtain an analysis report;

[0026] A fault handling module, used to generate a corresponding fault handling process based on the analysis report, and automatically handle the fault of the application system based on the fault handling process;

[0027] The fault processing log management module is used to record the fault processing log of the application system in real time and encrypt and store the fault processing log.

[0028] Furthermore, the operation data encryption upload module is specifically used for:

[0029] Real-time collection of operating data of the application system including performance indicators and operation logs; the performance indicators at least include response time, throughput, number of concurrent users, resource utilization, system availability, system stability, error rate, time processing time, page loading time, data throughput, latency, peak performance, fault tolerance, recovery time and energy consumption; the operation log at least includes log level, message, user information, operation details, error code, exception information, system status, resource usage, transaction ID, network information, security events and log ID;

[0030] The operation data is hashed by the SHA-3 algorithm to obtain a hash value, a current timestamp is obtained, the operation data, the hash value and the timestamp are encrypted by the AES-256 algorithm to obtain a first encrypted string, each character in the first encrypted string is shifted rightward based on a preset first displacement to obtain a second encrypted string, the second encrypted string is encrypted by the 3DES algorithm to obtain a third encrypted string, each character in the third encrypted string is shifted rightward based on a preset second displacement to obtain a fourth encrypted string, the fourth encrypted string is encrypted by the RC6 algorithm to obtain a fifth encrypted string, each character in the fifth encrypted string is shifted rightward based on a preset third displacement to obtain a sixth encrypted string, and the sixth encrypted string is encrypted by the ECDSA algorithm to obtain encrypted data.

[0031] Furthermore, the operation data analysis module is specifically used for:

[0032] The server creates a fault recognition model based on a recurrent neural network for identifying faults that have occurred and predicting faults that are about to occur, sets the loss function of the fault recognition model to be a cross-entropy loss function, obtains a large amount of historical operation data for preprocessing and fault annotation to construct a data set, and divides the data set into a training set, a validation set, and a test set based on a preset ratio; trains the fault recognition model with the training set until the loss value of the cross-entropy loss function is less than a preset loss threshold, and then uses the validation set to calculate the recognition accuracy to verify the trained fault recognition model, and uses the test set to calculate the confidence to test the verified fault recognition model;

[0033] The server receives the encrypted data in real time, decrypts the encrypted data through the ECDSA algorithm to obtain a sixth encrypted string, shifts each character in the sixth encrypted string to the left based on a preset third displacement to obtain a fifth encrypted string, decrypts the fifth encrypted string through the RC6 algorithm to obtain a fourth encrypted string, shifts each character in the fourth encrypted string to the left based on a preset second displacement to obtain a third encrypted string, decrypts the third encrypted string through the 3DES algorithm to obtain a second encrypted string, shifts each character in the second encrypted string to the left based on a preset first displacement to obtain a first encrypted string, decrypts the first encrypted string through the AES-256 algorithm to obtain running data, a hash value, and a timestamp, performs a time validity check through the timestamp, and then performs an integrity check on the running data through the hash value;

[0034] The server inputs the verified operating data into a fault identification model for analysis to obtain an analysis report including at least abnormal data, fault type, number of faults that have occurred, number of predicted faults, and analysis time.

[0035] Furthermore, the fault handling module is specifically used for:

[0036] Obtain a large number of historical fault processing logs, learn each of the historical fault processing logs through machine learning technology, obtain the fault processing process for correctly handling faults of different fault types, and bind and store the fault types and fault processing processes in a pre-created fault management table;

[0037] Based on the fault type carried in the analysis report, a corresponding fault handling process is matched from the fault management table, and the fault handling process is sent to the application system for automatic execution, so as to automatically handle the fault of the application system.

[0038] Furthermore, the fault handling log management module is specifically used for:

[0039] Record the fault handling log of the application system in real time, encrypt the fault handling log by SM3 algorithm to obtain level one ciphertext data, swap the number 1 with the letter Q in the level one ciphertext data to obtain level two ciphertext data, encrypt the level two ciphertext data by SM4 algorithm to obtain level three ciphertext data, swap the number 2 with the letter W in the level three ciphertext data to obtain level four ciphertext data, encrypt the level four ciphertext data by SM7 algorithm to obtain level five ciphertext data, swap the number 3 with the letter E in the level five ciphertext data to obtain level six ciphertext data, encrypt the level six ciphertext data by 3DES algorithm to obtain an encrypted log, and store and distribute the encrypted log.

[0040] The advantages of the present invention are:

[0041] 1. Collect the operation data of the application system including performance indicators and operation logs in real time, encrypt the operation data into encrypted data and upload it to the server. The server decrypts the received encrypted data to obtain the operation data, analyzes the operation data through a machine learning algorithm to obtain an analysis report, generates a corresponding fault handling process based on the analysis report, automatically handles the failure of the application system based on the fault handling process, and records the fault handling log of the application system in real time, and encrypts and stores the fault handling log; that is, collect the operation log in real time and upload it to the server for fault analysis. The machine learning algorithm is combined in the fault analysis process, and the machine learning technology is combined in the fault handling process matching process to accurately locate the fault and trigger the correct fault handling process, so as to avoid delays in fault handling due to incorrect fault location or fault handling process, thereby greatly improving the accuracy and timeliness of application system fault handling.

[0042] 2. Real-time collection of operating data of the application system including performance indicators and operation logs; performance indicators at least include response time, throughput, number of concurrent users, resource utilization, system availability, system stability, error rate, time processing time, page loading time, data throughput, latency, peak performance, fault tolerance, recovery time and energy consumption; operation logs at least include log level, message, user information, operation details, error code, exception information, system status, resource usage, transaction ID, network information, security events and log ID; that is, the collection of multi-dimensional operating data is convenient for fault location and proofreading from a multi-dimensional perspective, thereby greatly improving the accuracy of fault location.

[0043] 3. Perform hash calculation on the running data using the SHA-3 algorithm to obtain a hash value, obtain the current timestamp, encrypt the running data, the hash value and the timestamp using the AES-256 algorithm to obtain a first encrypted string, shift each character in the first encrypted string to the right based on the first displacement to obtain a second encrypted string, encrypt the second encrypted string using the 3DES algorithm to obtain a third encrypted string, shift each character in the third encrypted string to the right based on the second displacement to obtain a fourth encrypted string, encrypt the fourth encrypted string using the RC6 algorithm to obtain a fifth encrypted string, shift each character in the fifth encrypted string to the right based on the third displacement to obtain a sixth encrypted string, and The DSA algorithm encrypts the sixth encrypted string to obtain encrypted data; subsequently, the integrity can be checked through the hash value, and the timeliness can be checked through the timestamp. It also combines multiple encryption and data transformation operations. If the corresponding encryption algorithm or data transformation rule is not known, the encrypted data cannot be cracked. At least 12 security measures have been taken (SHA-3 algorithm, timestamp, AES-256 algorithm, first displacement, second displacement, third displacement, first displacement direction, second displacement direction, third displacement direction, 3DES algorithm, RC6 algorithm, ECDSA algorithm), which greatly improves the security of uploading running data, avoids fault analysis based on tampered running data, and further improves the accuracy of fault location.

[0044] 4. Create a fault recognition model for identifying faults that have occurred and predicting faults that are about to occur through a recurrent neural network, set the loss function to the cross-entropy loss function, obtain a large amount of historical operation data to build a data set, and divide the data set into a training set, a validation set, and a test set based on a preset ratio; train the fault recognition model through the training set until the loss value of the cross-entropy loss function is less than the preset loss threshold, and then use the validation set to calculate the recognition accuracy to verify the trained fault recognition model, and use the test set to calculate the confidence to test the verified fault recognition model; that is, train the fault recognition model through a large amount of historical data, and continuously verify and test during the training process until the requirements are met, so that the trained fault recognition model can effectively identify faults that have occurred and predict faults that are about to occur, thereby greatly improving the accuracy of fault handling in the application system.

[0045] 5. Obtain a large number of historical fault handling logs, learn from each historical fault handling log through machine learning technology, obtain the fault handling process for correctly handling faults of different fault types, bind the fault type and fault handling process and store them in a pre-created fault management table. That is, obtain the fault handling rules from a large number of historical fault handling logs through machine learning technology. When encountering the same fault, the corresponding fault handling process can be quickly matched from the fault management table to automatically handle the fault, thereby greatly improving the timeliness of fault handling in the application system.

[0046] 6. Encrypt the fault handling log by the SM3 algorithm to obtain the first-level ciphertext data, swap the number 1 and the letter Q in the first-level ciphertext data to obtain the second-level ciphertext data, encrypt the second-level ciphertext data by the SM4 algorithm to obtain the third-level ciphertext data, swap the number 2 and the letter W in the third-level ciphertext data to obtain the fourth-level ciphertext data, encrypt the fourth-level ciphertext data by the SM7 algorithm to obtain the fifth-level ciphertext data, swap the number 3 and the letter E in the fifth-level ciphertext data to obtain the sixth-level ciphertext data, encrypt the sixth-level ciphertext data by the 3DES algorithm to obtain an encrypted log, and store and distribute the encrypted log; that is, the encryption of the fault handling log combines multiple encryption algorithms and data transformation rules. If the corresponding encryption algorithm or data transformation rule is not known, the encrypted log cannot be cracked, which prevents the fault handling log from being stolen and tampered with in plain text, greatly improves the security of the fault handling log storage, and facilitates later tracing. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] The present invention will be further described below in conjunction with embodiments with reference to the accompanying drawings.

[0048] Figure 1 It is a flow chart of an application system intelligent operation and maintenance method of the present invention.

[0049] Figure 2 It is a structural schematic diagram of an application system intelligent operation and maintenance system of the present invention. DETAILED DESCRIPTION

[0050] The technical solution in the embodiments of the present application has the following overall idea: real-time collection of operation logs and uploading them to the server for fault analysis. Machine learning algorithms are combined in the fault analysis process, and machine learning technology is combined in the fault handling process matching process to accurately locate the fault and trigger the correct fault handling process, so as to avoid delays in fault handling due to incorrect fault location or fault handling process, so as to improve the accuracy and timeliness of application system fault handling.

[0051] Please refer to Figure 1 to Figure 2 As shown, a preferred embodiment of an application system intelligent operation and maintenance method of the present invention includes the following steps:

[0052] Step S1, collecting the operation data of the application system including performance indicators and operation logs in real time, encrypting the operation data into encrypted data and uploading it to the server;

[0053] Step S2: The server decrypts the received encrypted data to obtain operation data, analyzes the operation data using a machine learning algorithm to obtain an analysis report, and pushes the analysis report to a pre-associated management terminal in real time;

[0054] Step S3: The server generates a corresponding fault handling process based on the analysis report, and automatically handles the fault of the application system based on the fault handling process, such as automatically restarting, switching to a standby node, etc.;

[0055] Step S4: Record the fault processing log of the application system in real time, and encrypt and store the fault processing log.

[0056] The step S1 is specifically as follows:

[0057] Real-time collection of operating data of the application system including performance indicators and operation logs; the performance indicators at least include response time, throughput, number of concurrent users, resource utilization, system availability, system stability, error rate, time processing time, page loading time, data throughput, latency, peak performance, fault tolerance, recovery time and energy consumption; the operation log at least includes log level, message, user information, operation details, error code, exception information, system status, resource usage, transaction ID, network information, security events and log ID;

[0058] By collecting the operating data of the application system including performance indicators and operation logs in real time; the performance indicators at least include response time, throughput, number of concurrent users, resource utilization, system availability, system stability, error rate, time processing time, page loading time, data throughput, latency, peak performance, fault tolerance, recovery time and energy consumption; the operation log at least includes log level, message, user information, operation details, error code, exception information, system status, resource usage, transaction ID, network information, security events and log ID; that is, collecting multi-dimensional operating data, it is convenient to locate and proofread faults from a multi-dimensional perspective, thereby greatly improving the accuracy of fault location.

[0059] The operation data is hashed by the SHA-3 algorithm to obtain a hash value, a current timestamp is obtained, the operation data, the hash value and the timestamp are encrypted by the AES-256 algorithm to obtain a first encrypted string, each character in the first encrypted string is shifted rightward based on a preset first displacement to obtain a second encrypted string, the second encrypted string is encrypted by the 3DES algorithm to obtain a third encrypted string, each character in the third encrypted string is shifted rightward based on a preset second displacement to obtain a fourth encrypted string, the fourth encrypted string is encrypted by the RC6 algorithm to obtain a fifth encrypted string, each character in the fifth encrypted string is shifted rightward based on a preset third displacement to obtain a sixth encrypted string, and the sixth encrypted string is encrypted by the ECDSA algorithm to obtain encrypted data.

[0060] The operation data is hashed by the SHA-3 algorithm to obtain a hash value, the current timestamp is obtained, the operation data, the hash value and the timestamp are encrypted by the AES-256 algorithm to obtain a first encrypted string, each character in the first encrypted string is shifted rightward based on the first displacement to obtain a second encrypted string, the second encrypted string is encrypted by the 3DES algorithm to obtain a third encrypted string, each character in the third encrypted string is shifted rightward based on the second displacement to obtain a fourth encrypted string, the fourth encrypted string is encrypted by the RC6 algorithm to obtain a fifth encrypted string, each character in the fifth encrypted string is shifted rightward based on the third displacement to obtain a sixth encrypted string, and the ECD The SA algorithm encrypts the sixth encrypted string to obtain encrypted data; subsequently, the integrity can be checked through the hash value, and the timeliness can be checked through the timestamp. It also combines multiple encryption and data transformation operations. If the corresponding encryption algorithm or data transformation rule is not known, the encrypted data cannot be cracked. At least 12 security measures have been taken (SHA-3 algorithm, timestamp, AES-256 algorithm, first displacement, second displacement, third displacement, first displacement direction, second displacement direction, third displacement direction, 3DES algorithm, RC6 algorithm, ECDSA algorithm), which greatly improves the security of uploading running data, avoids fault analysis based on tampered running data, and further improves the accuracy of fault location.

[0061] The step S2 is specifically as follows:

[0062] The server creates a fault recognition model based on a recurrent neural network for identifying faults that have occurred and predicting faults that are about to occur, sets the loss function of the fault recognition model to be a cross-entropy loss function, obtains a large amount of historical operation data for preprocessing and fault annotation to construct a data set, and divides the data set into a training set, a validation set, and a test set based on a preset ratio; trains the fault recognition model with the training set until the loss value of the cross-entropy loss function is less than a preset loss threshold, and then uses the validation set to calculate the recognition accuracy to verify the trained fault recognition model, and uses the test set to calculate the confidence to test the verified fault recognition model;

[0063] A fault recognition model for identifying faults that have occurred and predicting faults that are about to occur is created through a recurrent neural network, the loss function is set to a cross-entropy loss function, a large amount of historical operation data is obtained to build a data set, and the data set is divided into a training set, a validation set and a test set based on a preset ratio; the fault recognition model is trained through the training set until the loss value of the cross-entropy loss function is less than a preset loss threshold, and then the recognition accuracy is calculated using the validation set to verify the trained fault recognition model, and the confidence is calculated using the test set to test the verified fault recognition model; that is, the fault recognition model is trained through a large amount of historical data, and verification and testing are continuously performed during the training process until the requirements are met, so that the trained fault recognition model can effectively identify faults that have occurred and predict faults that are about to occur, thereby greatly improving the accuracy of fault handling in the application system.

[0064] The server receives the encrypted data in real time, decrypts the encrypted data through the ECDSA algorithm to obtain a sixth encrypted string, shifts each character in the sixth encrypted string to the left based on a preset third displacement to obtain a fifth encrypted string, decrypts the fifth encrypted string through the RC6 algorithm to obtain a fourth encrypted string, shifts each character in the fourth encrypted string to the left based on a preset second displacement to obtain a third encrypted string, decrypts the third encrypted string through the 3DES algorithm to obtain a second encrypted string, shifts each character in the second encrypted string to the left based on a preset first displacement to obtain a first encrypted string, decrypts the first encrypted string through the AES-256 algorithm to obtain running data, a hash value, and a timestamp, performs a time validity check through the timestamp, and then performs an integrity check on the running data through the hash value;

[0065] The server inputs the verified operating data into a fault identification model for analysis to obtain an analysis report including at least abnormal data, fault type, number of faults that have occurred, number of predicted faults, and analysis time.

[0066] The step S3 is specifically as follows:

[0067] Obtain a large number of historical fault processing logs, learn each of the historical fault processing logs through machine learning technology, obtain the fault processing process for correctly handling faults of different fault types, and bind and store the fault types and fault processing processes in a pre-created fault management table;

[0068] Based on the fault type carried in the analysis report, a corresponding fault handling process is matched from the fault management table, and the fault handling process is sent to the application system for automatic execution, so as to automatically handle the fault of the application system.

[0069] Obtain a large number of historical fault handling logs, learn from each historical fault handling log through machine learning technology, obtain the fault handling process for correctly handling faults of different fault types, bind the fault type and fault handling process and store them in a pre-created fault management table. That is, obtain the fault handling rules from a large number of historical fault handling logs through machine learning technology. When encountering the same fault, the corresponding fault handling process can be quickly matched from the fault management table to automatically handle the fault, thereby greatly improving the timeliness of fault handling in the application system.

[0070] The step S4 is specifically as follows:

[0071] Record the fault handling log of the application system in real time, encrypt the fault handling log by SM3 algorithm to obtain level one ciphertext data, swap the number 1 with the letter Q in the level one ciphertext data to obtain level two ciphertext data, encrypt the level two ciphertext data by SM4 algorithm to obtain level three ciphertext data, swap the number 2 with the letter W in the level three ciphertext data to obtain level four ciphertext data, encrypt the level four ciphertext data by SM7 algorithm to obtain level five ciphertext data, swap the number 3 with the letter E in the level five ciphertext data to obtain level six ciphertext data, encrypt the level six ciphertext data by 3DES algorithm to obtain an encrypted log, and store and distribute the encrypted log.

[0072] The fault handling log is encrypted by the SM3 algorithm to obtain the first-level ciphertext data, the number 1 in the first-level ciphertext data is swapped with the letter Q to obtain the second-level ciphertext data, the second-level ciphertext data is encrypted by the SM4 algorithm to obtain the third-level ciphertext data, the number 2 in the third-level ciphertext data is swapped with the letter W to obtain the fourth-level ciphertext data, the fourth-level ciphertext data is encrypted by the SM7 algorithm to obtain the fifth-level ciphertext data, the number 3 in the fifth-level ciphertext data is swapped with the letter E to obtain the sixth-level ciphertext data, the sixth-level ciphertext data is encrypted by the 3DES algorithm to obtain an encrypted log, and the encrypted log is stored and distributedly backed up; that is, the encryption of the fault handling log combines multiple encryption algorithms and data transformation rules. If the corresponding encryption algorithm or data transformation rule is not known, the encrypted log cannot be cracked, which prevents the fault handling log from being stolen and tampered with in plain text, greatly improves the security of the fault handling log storage, and facilitates later tracing.

[0073] A preferred embodiment of an application system intelligent operation and maintenance system of the present invention includes the following modules:

[0074] An operation data encryption upload module is used to collect the operation data of the application system including performance indicators and operation logs in real time, and encrypt the operation data into encrypted data and upload it to the server;

[0075] An operation data analysis module is used for the server to decrypt the encrypted data received to obtain operation data, analyze the operation data through a machine learning algorithm to obtain an analysis report; and push the analysis report to a pre-associated management terminal in real time;

[0076] A fault handling module, which is used for the server to generate a corresponding fault handling process based on the analysis report, and automatically handle the fault of the application system based on the fault handling process, such as automatically restarting, switching to a standby node, etc.;

[0077] The fault processing log management module is used to record the fault processing log of the application system in real time and encrypt and store the fault processing log.

[0078] The operation data encryption upload module is specifically used for:

[0079] Real-time collection of operating data of the application system including performance indicators and operation logs; the performance indicators at least include response time, throughput, number of concurrent users, resource utilization, system availability, system stability, error rate, time processing time, page loading time, data throughput, latency, peak performance, fault tolerance, recovery time and energy consumption; the operation log at least includes log level, message, user information, operation details, error code, exception information, system status, resource usage, transaction ID, network information, security events and log ID;

[0080] By collecting the operating data of the application system including performance indicators and operation logs in real time; the performance indicators at least include response time, throughput, number of concurrent users, resource utilization, system availability, system stability, error rate, time processing time, page loading time, data throughput, latency, peak performance, fault tolerance, recovery time and energy consumption; the operation log at least includes log level, message, user information, operation details, error code, exception information, system status, resource usage, transaction ID, network information, security events and log ID; that is, collecting multi-dimensional operating data, it is convenient to locate and proofread faults from a multi-dimensional perspective, thereby greatly improving the accuracy of fault location.

[0081] The operation data is hashed by the SHA-3 algorithm to obtain a hash value, a current timestamp is obtained, the operation data, the hash value and the timestamp are encrypted by the AES-256 algorithm to obtain a first encrypted string, each character in the first encrypted string is shifted rightward based on a preset first displacement to obtain a second encrypted string, the second encrypted string is encrypted by the 3DES algorithm to obtain a third encrypted string, each character in the third encrypted string is shifted rightward based on a preset second displacement to obtain a fourth encrypted string, the fourth encrypted string is encrypted by the RC6 algorithm to obtain a fifth encrypted string, each character in the fifth encrypted string is shifted rightward based on a preset third displacement to obtain a sixth encrypted string, and the sixth encrypted string is encrypted by the ECDSA algorithm to obtain encrypted data.

[0082] The operation data is hashed by the SHA-3 algorithm to obtain a hash value, the current timestamp is obtained, the operation data, the hash value and the timestamp are encrypted by the AES-256 algorithm to obtain a first encrypted string, each character in the first encrypted string is shifted rightward based on the first displacement to obtain a second encrypted string, the second encrypted string is encrypted by the 3DES algorithm to obtain a third encrypted string, each character in the third encrypted string is shifted rightward based on the second displacement to obtain a fourth encrypted string, the fourth encrypted string is encrypted by the RC6 algorithm to obtain a fifth encrypted string, each character in the fifth encrypted string is shifted rightward based on the third displacement to obtain a sixth encrypted string, and the ECD The SA algorithm encrypts the sixth encrypted string to obtain encrypted data; subsequently, the integrity can be checked through the hash value, and the timeliness can be checked through the timestamp. It also combines multiple encryption and data transformation operations. If the corresponding encryption algorithm or data transformation rule is not known, the encrypted data cannot be cracked. At least 12 security measures have been taken (SHA-3 algorithm, timestamp, AES-256 algorithm, first displacement, second displacement, third displacement, first displacement direction, second displacement direction, third displacement direction, 3DES algorithm, RC6 algorithm, ECDSA algorithm), which greatly improves the security of uploading running data, avoids fault analysis based on tampered running data, and further improves the accuracy of fault location.

[0083] The operation data analysis module is specifically used for:

[0084] The server creates a fault recognition model based on a recurrent neural network for identifying faults that have occurred and predicting faults that are about to occur, sets the loss function of the fault recognition model to be a cross-entropy loss function, obtains a large amount of historical operation data for preprocessing and fault annotation to construct a data set, and divides the data set into a training set, a validation set, and a test set based on a preset ratio; trains the fault recognition model with the training set until the loss value of the cross-entropy loss function is less than a preset loss threshold, and then uses the validation set to calculate the recognition accuracy to verify the trained fault recognition model, and uses the test set to calculate the confidence to test the verified fault recognition model;

[0085] A fault recognition model for identifying faults that have occurred and predicting faults that are about to occur is created through a recurrent neural network, the loss function is set to a cross-entropy loss function, a large amount of historical operation data is obtained to build a data set, and the data set is divided into a training set, a validation set and a test set based on a preset ratio; the fault recognition model is trained through the training set until the loss value of the cross-entropy loss function is less than a preset loss threshold, and then the recognition accuracy is calculated using the validation set to verify the trained fault recognition model, and the confidence is calculated using the test set to test the verified fault recognition model; that is, the fault recognition model is trained through a large amount of historical data, and verification and testing are continuously performed during the training process until the requirements are met, so that the trained fault recognition model can effectively identify faults that have occurred and predict faults that are about to occur, thereby greatly improving the accuracy of fault handling in the application system.

[0086] The server receives the encrypted data in real time, decrypts the encrypted data through the ECDSA algorithm to obtain a sixth encrypted string, shifts each character in the sixth encrypted string to the left based on a preset third displacement to obtain a fifth encrypted string, decrypts the fifth encrypted string through the RC6 algorithm to obtain a fourth encrypted string, shifts each character in the fourth encrypted string to the left based on a preset second displacement to obtain a third encrypted string, decrypts the third encrypted string through the 3DES algorithm to obtain a second encrypted string, shifts each character in the second encrypted string to the left based on a preset first displacement to obtain a first encrypted string, decrypts the first encrypted string through the AES-256 algorithm to obtain running data, a hash value, and a timestamp, performs a time validity check through the timestamp, and then performs an integrity check on the running data through the hash value;

[0087] The server inputs the verified operating data into a fault identification model for analysis to obtain an analysis report including at least abnormal data, fault type, number of faults that have occurred, number of predicted faults, and analysis time.

[0088] The fault processing module is specifically used for:

[0089] Obtain a large number of historical fault processing logs, learn each of the historical fault processing logs through machine learning technology, obtain the fault processing process for correctly handling faults of different fault types, and bind and store the fault types and fault processing processes in a pre-created fault management table;

[0090] Based on the fault type carried in the analysis report, a corresponding fault handling process is matched from the fault management table, and the fault handling process is sent to the application system for automatic execution, so as to automatically handle the fault of the application system.

[0091] Obtain a large number of historical fault handling logs, learn from each historical fault handling log through machine learning technology, obtain the fault handling process for correctly handling faults of different fault types, bind the fault type and fault handling process and store them in a pre-created fault management table. That is, obtain the fault handling rules from a large number of historical fault handling logs through machine learning technology. When encountering the same fault, the corresponding fault handling process can be quickly matched from the fault management table to automatically handle the fault, thereby greatly improving the timeliness of fault handling in the application system.

[0092] The fault handling log management module is specifically used for:

[0093] Record the fault handling log of the application system in real time, encrypt the fault handling log by SM3 algorithm to obtain level one ciphertext data, swap the number 1 with the letter Q in the level one ciphertext data to obtain level two ciphertext data, encrypt the level two ciphertext data by SM4 algorithm to obtain level three ciphertext data, swap the number 2 with the letter W in the level three ciphertext data to obtain level four ciphertext data, encrypt the level four ciphertext data by SM7 algorithm to obtain level five ciphertext data, swap the number 3 with the letter E in the level five ciphertext data to obtain level six ciphertext data, encrypt the level six ciphertext data by 3DES algorithm to obtain an encrypted log, and store and distribute the encrypted log.

[0094] The fault handling log is encrypted by the SM3 algorithm to obtain the first-level ciphertext data, the number 1 in the first-level ciphertext data is swapped with the letter Q to obtain the second-level ciphertext data, the second-level ciphertext data is encrypted by the SM4 algorithm to obtain the third-level ciphertext data, the number 2 in the third-level ciphertext data is swapped with the letter W to obtain the fourth-level ciphertext data, the fourth-level ciphertext data is encrypted by the SM7 algorithm to obtain the fifth-level ciphertext data, the number 3 in the fifth-level ciphertext data is swapped with the letter E to obtain the sixth-level ciphertext data, the sixth-level ciphertext data is encrypted by the 3DES algorithm to obtain an encrypted log, and the encrypted log is stored and distributedly backed up; that is, the encryption of the fault handling log combines multiple encryption algorithms and data transformation rules. If the corresponding encryption algorithm or data transformation rule is not known, the encrypted log cannot be cracked, which prevents the fault handling log from being stolen and tampered with in plain text, greatly improves the security of the fault handling log storage, and facilitates later tracing.

[0095] In summary, the advantages of the present invention are:

[0096] 1. Collect the operation data of the application system including performance indicators and operation logs in real time, encrypt the operation data into encrypted data and upload it to the server. The server decrypts the received encrypted data to obtain the operation data, analyzes the operation data through a machine learning algorithm to obtain an analysis report, generates a corresponding fault handling process based on the analysis report, automatically handles the failure of the application system based on the fault handling process, and records the fault handling log of the application system in real time, and encrypts and stores the fault handling log; that is, collect the operation log in real time and upload it to the server for fault analysis. The machine learning algorithm is combined in the fault analysis process, and the machine learning technology is combined in the fault handling process matching process to accurately locate the fault and trigger the correct fault handling process, so as to avoid delays in fault handling due to incorrect fault location or fault handling process, thereby greatly improving the accuracy and timeliness of application system fault handling.

[0097] 2. Real-time collection of operating data of the application system including performance indicators and operation logs; performance indicators at least include response time, throughput, number of concurrent users, resource utilization, system availability, system stability, error rate, time processing time, page loading time, data throughput, latency, peak performance, fault tolerance, recovery time and energy consumption; operation logs at least include log level, message, user information, operation details, error code, exception information, system status, resource usage, transaction ID, network information, security events and log ID; that is, the collection of multi-dimensional operating data is convenient for fault location and proofreading from a multi-dimensional perspective, thereby greatly improving the accuracy of fault location.

[0098] 3. Perform hash calculation on the running data using the SHA-3 algorithm to obtain a hash value, obtain the current timestamp, encrypt the running data, the hash value and the timestamp using the AES-256 algorithm to obtain a first encrypted string, shift each character in the first encrypted string to the right based on the first displacement to obtain a second encrypted string, encrypt the second encrypted string using the 3DES algorithm to obtain a third encrypted string, shift each character in the third encrypted string to the right based on the second displacement to obtain a fourth encrypted string, encrypt the fourth encrypted string using the RC6 algorithm to obtain a fifth encrypted string, shift each character in the fifth encrypted string to the right based on the third displacement to obtain a sixth encrypted string, and The DSA algorithm encrypts the sixth encrypted string to obtain encrypted data; subsequently, the integrity can be checked through the hash value, and the timeliness can be checked through the timestamp. It also combines multiple encryption and data transformation operations. If the corresponding encryption algorithm or data transformation rule is not known, the encrypted data cannot be cracked. At least 12 security measures have been taken (SHA-3 algorithm, timestamp, AES-256 algorithm, first displacement, second displacement, third displacement, first displacement direction, second displacement direction, third displacement direction, 3DES algorithm, RC6 algorithm, ECDSA algorithm), which greatly improves the security of uploading running data, avoids fault analysis based on tampered running data, and further improves the accuracy of fault location.

[0099] 4. Create a fault recognition model for identifying faults that have occurred and predicting faults that are about to occur through a recurrent neural network, set the loss function to the cross-entropy loss function, obtain a large amount of historical operation data to build a data set, and divide the data set into a training set, a validation set, and a test set based on a preset ratio; train the fault recognition model through the training set until the loss value of the cross-entropy loss function is less than the preset loss threshold, and then use the validation set to calculate the recognition accuracy to verify the trained fault recognition model, and use the test set to calculate the confidence to test the verified fault recognition model; that is, train the fault recognition model through a large amount of historical data, and continuously verify and test during the training process until the requirements are met, so that the trained fault recognition model can effectively identify faults that have occurred and predict faults that are about to occur, thereby greatly improving the accuracy of fault handling in the application system.

[0100] 5. Obtain a large number of historical fault handling logs, learn from each historical fault handling log through machine learning technology, obtain the fault handling process for correctly handling faults of different fault types, bind the fault type and fault handling process and store them in a pre-created fault management table. That is, obtain the fault handling rules from a large number of historical fault handling logs through machine learning technology. When encountering the same fault, the corresponding fault handling process can be quickly matched from the fault management table to automatically handle the fault, thereby greatly improving the timeliness of fault handling in the application system.

[0101] 6. Encrypt the fault handling log by the SM3 algorithm to obtain the first-level ciphertext data, swap the number 1 and the letter Q in the first-level ciphertext data to obtain the second-level ciphertext data, encrypt the second-level ciphertext data by the SM4 algorithm to obtain the third-level ciphertext data, swap the number 2 and the letter W in the third-level ciphertext data to obtain the fourth-level ciphertext data, encrypt the fourth-level ciphertext data by the SM7 algorithm to obtain the fifth-level ciphertext data, swap the number 3 and the letter E in the fifth-level ciphertext data to obtain the sixth-level ciphertext data, encrypt the sixth-level ciphertext data by the 3DES algorithm to obtain an encrypted log, and store and distribute the encrypted log; that is, the encryption of the fault handling log combines multiple encryption algorithms and data transformation rules. If the corresponding encryption algorithm or data transformation rule is not known, the encrypted log cannot be cracked, which prevents the fault handling log from being stolen and tampered with in plain text, greatly improves the security of the fault handling log storage, and facilitates later tracing.

[0102] Although the specific implementation modes of the present invention are described above, those skilled in the art should understand that the specific implementation modes described are only illustrative and are not intended to limit the scope of the present invention. Equivalent modifications and changes made by those skilled in the art in accordance with the spirit of the present invention should be included in the scope of protection of the claims of the present invention.

Claims

1. An intelligent operation and maintenance method for an application system, characterized in that: The steps include: Step S1, collecting the operation data of the application system including performance indicators and operation logs in real time, encrypting the operation data into encrypted data and uploading it to the server; Step S2: The server decrypts the received encrypted data to obtain operation data, and analyzes the operation data using a machine learning algorithm to obtain an analysis report; Step S3: generating a corresponding fault handling process based on the analysis report, and automatically handling the fault of the application system based on the fault handling process; Step S4: Record the fault processing log of the application system in real time, and encrypt and store the fault processing log.

2. The intelligent operation and maintenance method of an application system according to claim 1, characterized in that: The step S1 is specifically as follows: Real-time collection of operating data of the application system including performance indicators and operation logs; the performance indicators at least include response time, throughput, number of concurrent users, resource utilization, system availability, system stability, error rate, time processing time, page loading time, data throughput, latency, peak performance, fault tolerance, recovery time and energy consumption; the operation log at least includes log level, message, user information, operation details, error code, exception information, system status, resource usage, transaction ID, network information, security events and log ID; The operation data is hashed by the SHA-3 algorithm to obtain a hash value, a current timestamp is obtained, the operation data, the hash value and the timestamp are encrypted by the AES-256 algorithm to obtain a first encrypted string, each character in the first encrypted string is shifted rightward based on a preset first displacement to obtain a second encrypted string, the second encrypted string is encrypted by the 3DES algorithm to obtain a third encrypted string, each character in the third encrypted string is shifted rightward based on a preset second displacement to obtain a fourth encrypted string, the fourth encrypted string is encrypted by the RC6 algorithm to obtain a fifth encrypted string, each character in the fifth encrypted string is shifted rightward based on a preset third displacement to obtain a sixth encrypted string, and the sixth encrypted string is encrypted by the ECDSA algorithm to obtain encrypted data.

3. The intelligent operation and maintenance method of an application system according to claim 1, characterized in that: The step S2 is specifically as follows: The server creates a fault recognition model based on a recurrent neural network for identifying faults that have occurred and predicting faults that are about to occur, sets the loss function of the fault recognition model to be a cross-entropy loss function, obtains a large amount of historical operation data for preprocessing and fault annotation to construct a data set, and divides the data set into a training set, a validation set, and a test set based on a preset ratio; trains the fault recognition model with the training set until the loss value of the cross-entropy loss function is less than a preset loss threshold, and then uses the validation set to calculate the recognition accuracy to verify the trained fault recognition model, and uses the test set to calculate the confidence to test the verified fault recognition model; The server receives the encrypted data in real time, decrypts the encrypted data through the ECDSA algorithm to obtain a sixth encrypted string, shifts each character in the sixth encrypted string to the left based on a preset third displacement to obtain a fifth encrypted string, decrypts the fifth encrypted string through the RC6 algorithm to obtain a fourth encrypted string, shifts each character in the fourth encrypted string to the left based on a preset second displacement to obtain a third encrypted string, decrypts the third encrypted string through the 3DES algorithm to obtain a second encrypted string, shifts each character in the second encrypted string to the left based on a preset first displacement to obtain a first encrypted string, decrypts the first encrypted string through the AES-256 algorithm to obtain running data, a hash value, and a timestamp, performs a time validity check through the timestamp, and then performs an integrity check on the running data through the hash value; The server inputs the verified operating data into a fault identification model for analysis to obtain an analysis report including at least abnormal data, fault type, number of faults that have occurred, number of predicted faults, and analysis time.

4. The intelligent operation and maintenance method of an application system according to claim 1, characterized in that: The step S3 is specifically as follows: Obtain a large number of historical fault processing logs, learn each of the historical fault processing logs through machine learning technology, obtain the fault processing process for correctly handling faults of different fault types, and bind and store the fault types and fault processing processes in a pre-created fault management table; Based on the fault type carried in the analysis report, a corresponding fault handling process is matched from the fault management table, and the fault handling process is sent to the application system for automatic execution, so as to automatically handle the fault of the application system.

5. The intelligent operation and maintenance method of an application system according to claim 1, characterized in that: The step S4 is specifically as follows: Record the fault handling log of the application system in real time, encrypt the fault handling log by SM3 algorithm to obtain level one ciphertext data, swap the number 1 with the letter Q in the level one ciphertext data to obtain level two ciphertext data, encrypt the level two ciphertext data by SM4 algorithm to obtain level three ciphertext data, swap the number 2 with the letter W in the level three ciphertext data to obtain level four ciphertext data, encrypt the level four ciphertext data by SM7 algorithm to obtain level five ciphertext data, swap the number 3 with the letter E in the level five ciphertext data to obtain level six ciphertext data, encrypt the level six ciphertext data by 3DES algorithm to obtain an encrypted log, and store and distribute the encrypted log.

6. An intelligent operation and maintenance system for an application system, characterized in that: Includes the following modules: An operation data encryption upload module is used to collect the operation data of the application system including performance indicators and operation logs in real time, and encrypt the operation data into encrypted data and upload it to the server; An operation data analysis module, used for the server to decrypt the encrypted data received to obtain operation data, and analyze the operation data through a machine learning algorithm to obtain an analysis report; A fault handling module, used to generate a corresponding fault handling process based on the analysis report, and automatically handle the fault of the application system based on the fault handling process; The fault processing log management module is used to record the fault processing log of the application system in real time and encrypt and store the fault processing log.

7. An application system intelligent operation and maintenance system as claimed in claim 6, characterized in that: The operation data encryption upload module is specifically used for: Real-time collection of operating data of the application system including performance indicators and operation logs; the performance indicators at least include response time, throughput, number of concurrent users, resource utilization, system availability, system stability, error rate, time processing time, page loading time, data throughput, latency, peak performance, fault tolerance, recovery time and energy consumption; the operation log at least includes log level, message, user information, operation details, error code, exception information, system status, resource usage, transaction ID, network information, security events and log ID; The operation data is hashed by the SHA-3 algorithm to obtain a hash value, a current timestamp is obtained, the operation data, the hash value and the timestamp are encrypted by the AES-256 algorithm to obtain a first encrypted string, each character in the first encrypted string is shifted rightward based on a preset first displacement to obtain a second encrypted string, the second encrypted string is encrypted by the 3DES algorithm to obtain a third encrypted string, each character in the third encrypted string is shifted rightward based on a preset second displacement to obtain a fourth encrypted string, the fourth encrypted string is encrypted by the RC6 algorithm to obtain a fifth encrypted string, each character in the fifth encrypted string is shifted rightward based on a preset third displacement to obtain a sixth encrypted string, and the sixth encrypted string is encrypted by the ECDSA algorithm to obtain encrypted data.

8. An application system intelligent operation and maintenance system as claimed in claim 6, characterized in that: The operation data analysis module is specifically used for: The server creates a fault recognition model based on a recurrent neural network for identifying faults that have occurred and predicting faults that are about to occur, sets the loss function of the fault recognition model to be a cross-entropy loss function, obtains a large amount of historical operation data for preprocessing and fault annotation to construct a data set, and divides the data set into a training set, a validation set, and a test set based on a preset ratio; trains the fault recognition model with the training set until the loss value of the cross-entropy loss function is less than a preset loss threshold, and then uses the validation set to calculate the recognition accuracy to verify the trained fault recognition model, and uses the test set to calculate the confidence to test the verified fault recognition model; The server receives the encrypted data in real time, decrypts the encrypted data through the ECDSA algorithm to obtain a sixth encrypted string, shifts each character in the sixth encrypted string to the left based on a preset third displacement to obtain a fifth encrypted string, decrypts the fifth encrypted string through the RC6 algorithm to obtain a fourth encrypted string, shifts each character in the fourth encrypted string to the left based on a preset second displacement to obtain a third encrypted string, decrypts the third encrypted string through the 3DES algorithm to obtain a second encrypted string, shifts each character in the second encrypted string to the left based on a preset first displacement to obtain a first encrypted string, decrypts the first encrypted string through the AES-256 algorithm to obtain running data, a hash value, and a timestamp, performs a time validity check through the timestamp, and then performs an integrity check on the running data through the hash value; The server inputs the verified operating data into a fault identification model for analysis to obtain an analysis report including at least abnormal data, fault type, number of faults that have occurred, number of predicted faults, and analysis time.

9. An application system intelligent operation and maintenance system as claimed in claim 6, characterized in that: The fault processing module is specifically used for: Obtain a large number of historical fault processing logs, learn each of the historical fault processing logs through machine learning technology, obtain the fault processing process for correctly handling faults of different fault types, and bind and store the fault types and fault processing processes in a pre-created fault management table; Based on the fault type carried in the analysis report, a corresponding fault handling process is matched from the fault management table, and the fault handling process is sent to the application system for automatic execution, so as to automatically handle the fault of the application system.

10. An application system intelligent operation and maintenance system according to claim 6, characterized in that: The fault handling log management module is specifically used for: Record the fault handling log of the application system in real time, encrypt the fault handling log by SM3 algorithm to obtain level one ciphertext data, swap the number 1 with the letter Q in the level one ciphertext data to obtain level two ciphertext data, encrypt the level two ciphertext data by SM4 algorithm to obtain level three ciphertext data, swap the number 2 with the letter W in the level three ciphertext data to obtain level four ciphertext data, encrypt the level four ciphertext data by SM7 algorithm to obtain level five ciphertext data, swap the number 3 with the letter E in the level five ciphertext data to obtain level six ciphertext data, encrypt the level six ciphertext data by 3DES algorithm to obtain an encrypted log, and store and distribute the encrypted log.

Images