A Fault Detection Method for Big Data Computer System Based on Deep Recurrent Network

Abstract

The invention relates to a computer system fault detection method, in particular to a large data computer system fault detection method based on a deep recursive network. It introduces the deep recursive network into the fault detection link, uses historical data, conducts training and learning, replaces the part of manual feature extraction, uses the network to automatically extract fault features, and can update weights in real time, eliminating the difficulty of manually selecting fault features. At the same time, through real-time data learning and feature extraction, faults can be found quickly and accurately, and the types of faults can be predicted to improve system performance.

Description

technical field [0001] The invention relates to a computer system fault detection method, in particular to a large data computer system fault detection method based on a deep recursive network. Background technique [0002] At present, with the increasing scale and structural complexity of big data computing systems, their overall system performance has been affected in many ways. If a state node fails, other nodes connected to it will also be affected and cannot operate normally. At the same time, if the fault cannot be detected in time, it will continue to spread and even lead to the paralysis of the entire system. [0003] In the existing big data storage system, the use of fault detection technology can detect the possibility of system failure in advance according to the detection results, and make corresponding rescue preparations. The system state data is a typical time series data, which is the key to establish the system fault model. Establishing a system fault mo...

AI Technical Summary

Problems solved by technology

[0005] However, the existing fault detection and processing methods mainly rely on the experience knowledge of domain experts, which puts forward higher requirements on the experience level of experts. Negligence can also lead to failure

At the same time, in the fault prediction process, it is necessary to manually extract and select fault features. Using the method of artificial neural network and relying on a large amount of historical data, the rapidity and accuracy of prediction also need to be improved.

With the development of big data computing systems, their structures are becoming more and more complex, and people cannot quickly detect the cause of failures, let alone predict failures

Images