Distributed monitoring method based on bidirectional information flow

Abstract

The invention discloses a distributed monitor and control method based on bi-directional information flow which comprises, receiving the monitoring information, receiving the monitoring information and proceeding analysis and decision-making, receiving the accepting file data and control command data, when the monitoring party finds the accident or potential fault on the monitored party, the corresponding processing scheme can be determined according to the different fault and clearing the fault.

Description

technical field [0001] The invention relates to the field of information technology, in particular to a distributed monitoring method based on bidirectional information flow. Background technique [0002] With the rapid development of network technology, various distributed application systems based on network interconnection are gradually increasing. Such as Web server clusters that provide WWW services, database server clusters that provide database services, computing server clusters that provide computing services, etc. In many applications, such as data centers, e-commerce, scientific computing, etc., the cluster contains a large number of server nodes and network devices, often ranging from dozens to hundreds, and the number of these server nodes and devices The distribution of physical locations may also be very wide. How to effectively monitor and control the above-mentioned distributed system, so as to detect and eliminate the existing and potential faults in the ...

AI Technical Summary

Problems solved by technology

This manual control method often leads to system failures that cannot be dealt with in time and affects the work of the entire system.

In cluster applications where there are a large number of nodes and the physical locations of nodes are widely distributed, this problem will be more prominent, resulting in the cost of manpower, material and financial resources

[0004] Analyzing the structure of the traditional distributed monitoring system, we found that an important reason for the above problems is that its information flow is mostly one-way, that is, the information mostly flows from the monitored party to the monitoring party, and lacks an effective monitoring system. control information flow from party to monitored party

Although the SNMP protocol also provides the information flow function from the monitoring party to the monitored party, its function is very weak. Generally, it can only transmit character or integer data through it, and it is difficult to meet the needs of some complex control.

Images