Computer IO HUNG event early warning method and device, equipment and medium

Abstract

The invention provides a computer IO HUNG event early warning method, device and equipment, and a medium. The method comprises the steps that collecting machines are deployed on all virtual machines;the acquisition machine acquires data on the virtual machine at regular time and writes the data into the monitoring message file; if the disk writing succeeds, a state information message is sent; ifnot, sending is not performed, wherein the state information message comprises acquisition time; the early warning aircraft regularly checks the message of the server, reads the acquisition time fromthe message closest to the current time, compares the acquisition time with the local standard time, and if the difference reaches a set deviation value, carries out early warning; the invention notonly is effective for known and unknown IO HUNG problems of various computers, but also is suitable for part of traditional faults, and is beneficial to discovering operating system level problems such as timing task failure, user password expiration and clock skew. Besides, through the message data collected in real time, a performance report of the client can be generated and used for regular analysis.

Description

technical field [0001] The present invention relates to the field of computer technology, in particular to an early warning method, device, equipment and medium for computer IO HUNG events. Background technique [0002] In today's era of turbulent IT technology, the general trend of cloud computing is unstoppable like the roaring Yangtze River. At present, various businesses of enterprises have increasingly strong demand for IT, and data centers are continuously integrated intensively. IT managers are increasingly aware of the operation and maintenance challenges brought about by new data centers, especially in the daily operation and maintenance of many failures. Most of the faults can be found by the conventional monitoring system, but some faults, such as IO HUNG causing the computer to "lose connection", are difficult to capture in time. IO HUNG is a kind of extremely weird fault, and it is a big problem in computer system monitoring, and traditional monitoring (such as...

AI Technical Summary

Problems solved by technology

When this kind of failure occurs, the operating system level of the machine has actually stopped working, and it is basically impossible to log in, but the machine is connected by pinging the machine at the network level, which makes the traditional monitoring system have the illusion that the machine is normal, so that The situation expanded and even caused production accidents

On March 3, 2019, an IO HUNG event occurred in the server of a North China cloud computing data center of a domestic IT giant. Due to the difficulty in locating the fault for a while, it affected the customer's various businesses for several hours, causing huge economic losses and reputational impact

[0005] Traditional BMC monitoring has some disadvantages: first, the software is huge and there are several CDs; second, the deployment is complicated, not only the server needs to install the console, configuration management, knowledge base module, forwarding service, etc., but also the client needs to install 680M software; The third is that the operation is complicated, there are many menus, and it is not easy to get started; the fourth is that the reliability is poor, and various client alarms often appear on the server, most of which are caused by the expiration of the password of the patrol user on the client. In this case, the patrol needs to be reconfigured on the server. The new password can continue to monitor, and there is a gap in monitoring during this outage period; fifth, the BMC client program consumes more system resources, and even affects the operation of other client applications; sixth, the copyright of BMC software is strict and expensive

Images