Network performance analysis and fault diagnosis method of distributed system

Abstract

The invention provides a network performance analysis and fault diagnosis method of a distributed system. The method includes the following steps that monitoring service is deployed to the monitored distributed system; management service runs according to characteristics of the distributed system; network topology discovery is carried out; a monitored node set of monitored nodes is determined; the management service collects and analyzes state information of the nodes; the network performance is detected; the network state is analyzed, and possible faults are determined. The states of all network equipment and links participating in a communication path and communication performance between the nodes are taken into consideration, the specific position of fault points can be analyzed and determined according to the network topology information, the fault detection accuracy is improved, and fault detection expenses are reduced. Meanwhile, according to analysis of the performance of the distributed system, the method can provide the actual performance between the nodes of the distributed system instead of the theoretical performance of the network system, and the performance estimation accuracy can be improved.

Description

technical field [0001] The invention relates to a diagnosis method, in particular to a distributed system network performance analysis and fault diagnosis method. Background technique [0002] A distributed system refers to a network system built on top of which different nodes cooperate to complete one or more services through message communication between nodes. Since distributed systems distribute services to different nodes, distributed systems have good scalability, fault isolation, and application transparency. It has been widely used in actual IT systems. Typical services include distributed file systems, distributed databases, and website services. [0003] Since the distributed system relies on network equipment to interconnect various service nodes, the performance and stability of network equipment play a decisive role in the performance and stability of the distributed system. With the expansion of the scale of the distributed system, the scale of the network, ...

AI Technical Summary

Problems solved by technology

[0008] ●The fault judgment source is simple: the hardware counter can only detect the fault source of the hardware itself, and cannot judge faults such as the state of the network link, software protocol layer errors, etc.; the software point-to-point test tool can only test the fault between two points Network performance, unable to quickly judge network faults through data

Images