Distributed cluster hardware fault early warning method and system

Abstract

The invention discloses a distributed cluster hardware fault early warning method and system. The method comprises: a performance monitoring module acquires hardware index data in real time and transmits the hardware index data to a data analysis module; the data analysis module performs neural network learning and analysis on all the acquired hardware index data, and predicts whether the hardware will have a fault; the data analysis module is constructed based on a neural network model; and if the hardware is predicted to have a fault, a hardware early warning management module controls an alarm module via a network server to start alarming. The data analysis module performs neural network analysis, so that the hardware fault can be detected out before the fault happens, then the fault solution efficiency and the availability of the cluster system are improved, and the accuracy of fault detection is improved.

Description

technical field [0001] The invention relates to the field of storage technology, in particular to a distributed cluster hardware failure early warning method and system. Background technique [0002] With the development of the Internet and the acceleration of the internal informationization process of enterprises, in order to meet the growing business needs, enterprises gradually use server clusters to process and store large amounts of data. However, the increasing number of server components increases the difficulty of cluster unified management, and at the same time increases the probability of cluster failure to a certain extent. Therefore, an effective cluster system hardware failure early warning scheme is of great significance to solve cluster reliability problems. [0003] A distributed cluster system generally consists of multiple servers, and these servers form a cluster to provide external services uniformly. At present, the traditional distributed cluster hardw...

AI Technical Summary

Problems solved by technology

At present, the traditional distributed cluster hardware failure early warning mechanism can only check the current status of the system, and cannot analyze the previous cluster monitoring data, so that it cannot complete reasonable system performance evaluation, optimization and system cluster hardware detection. The existing working method is relatively passive. The failure can only be resolved after the failure occurs, which not only prolongs the failure time, but also greatly reduces the availability of the cluster system

This method can only inspect the current status of the database running status, and does not have the early warning and analysis capabilities of the cluster system

Images