Financial industry network change script checking method based on machine learning

Abstract

The invention discloses a financial industry network change script checking method based on machine learning. An offline training module, an online detection module and a model updating module are disclosed. Wherein the offline training module uses a machine learning algorithm and a statistical method to train models for network equipment (such as a Huawean firewall) of different manufacturers and equipment types respectively by using corresponding historical successfully changed script data, that is, the algorithms used during training of different manufacturers and equipment types are the same but the data are different, and the network equipment can be trained by using the machine learning algorithm and the statistical method. And different models are obtained. According to the technical scheme, the maintenance cost is low, the fault recall rate is high, the universality is high, the manual recheck efficiency is improved, and the success rate of network change is greatly improved.

Description

technical field [0001] The invention relates to the field of checking methods for network change scripts, in particular to a method for checking network change scripts in financial industries based on machine learning. Background technique [0002] In network changes in the financial industry, many change operations require system administrators to manually write change scripts, and then send the change scripts to different network devices through an automated system to make network changes to the corresponding devices. In the actual usage scenario, due to the different manufacturers supported by different network devices, the syntax commands corresponding to the scripts are different, and the administrators are not familiar with the system, resulting in spelling errors, command format errors, IP and other variables spelling errors in the actual change script. , the command sequence of the change script is wrong, etc., which causes the change to fail. [0003] The detection...

AI Technical Summary

Problems solved by technology

However, in actual usage scenarios, due to the fact that different network devices support different manufacturers, scripts correspond to different grammatical commands, and administrators are not familiar with the system, the actual change script has command spelling errors, command format errors, and IP and other variables spelling errors. , the command sequence of the change script is wrong, etc., which leads to the change failure

[0003] The detection of the existing technology is based on the manufacturer's classification of existing commands and logical rule checking. Since different manufacturers of different devices have different script commands, these rules are difficult to standardize uniformly.

The differences between manufacturers and commands exist objectively, resulting in multiple versions of these rules from different manufacturers, and there is a large time cost and learning cost for the management and maintenance of operators

The rules themselves rely on the continuous development and maintenance of the manufacturer's manual. The method is relatively traditional and rigid, and it is difficult to meet the timeliness of customer business changes.

Images