Fault early warning method and system for virtual environment

Abstract

The invention discloses a fault early warning method and system for a virtual environment, and belongs to the technical field of virtual environment fault detection. The fault early warning system for the virtual environment comprises an entity fault early warning server located outside the virtual environment, a virtual system log server located inside the virtual environment and an independent high-speed communication link connecting the entity fault early warning server and the virtual system log server. The virtual system log server is used for providing real-time log data of all virtual units in the virtual environment; the fault early warning server is used for receiving the virtual log data, processing the virtual log data by utilizing a neural network; after the neural network classifies and identifies the virtual log data, output results corresponding to a fault state of the virtual system are obtained, and early warning can be performed in advance; the communication link utilizes an exclusive-bandwidth more-than-gigabit Ethernet / optical-fiber link; due to the fault early warning method and system, which are provided by the invention, reliability of fault early warning is higher, and early warning results are more reliable.

Description

technical field [0001] The invention belongs to the technical field of fault detection in a virtualized environment, and in particular relates to a fault early warning method and system for a virtualized environment. Background technique [0002] With the innovation and development of information technology, higher requirements are put forward for the computing and service capabilities of enterprise data centers (computing centers). virtualization technology. [0003] The use of virtualization technology has improved the utilization rate of data center equipment, but it also brings some potential problems. If the resource requirements of tasks in the system do not match the resource allocation, and some resources are overloaded for a long time, resulting in resource failure, that is, the server or virtual machine cannot accept the task or the task is terminated unexpectedly when it is not completed. The failure of a virtual resource will inevitably affect other resource re...

AI Technical Summary

Problems solved by technology

However, these fault detection schemes all work inside the virtualization environment or are too closely related to the virtualization server. If a serious failure or even crash occurs in the virtualization environment, the fault detection cannot work normally; more importantly, these schemes cannot provide fault warning and cannot Enable managers to intervene and respond in advance

[0005] In addition to the fault recording function of the virtualization system itself, in the prior art, there are also other devices associated with the virtualization environment for fault detection. These fault detection solutions are also lack of early warning just like the above-mentioned fault detection solutions of the virtualization system itself function, and lacks independence from the virtualization environment; there are individual solutions that focus on methods and processes, and lack specific system structure and topology design

Images