Log alarm system based on streaming processing

Abstract

Provided is a log alarm system based on streaming processing which comprises a project matching module which configures a data source needing to be received, configures a system alarm receiver and configures log matching rules; a log matching module used for loading the received data source, formatting logs, carrying out log matching according to log matching rules and transmitting the matched logs to the log action module; a log alarm condition configuration module used for configuring alarm conditions of logs; and a log action module which triggers an alarm action corresponding to the log information to send an alarm to a system alarm receiver, and clears the message record of the log recorded before. Real-time matching alarm of mass logs is achieved, multi-source data are synchronouslymonitored, and the system has more efficient alarm receiver management, rich alarm implementation and a powerful self-repairing function.

Description

technical field [0001] The invention relates to the technical field of computers, in particular to a stream processing-based log alarm system. Background technique [0002] With the development of enterprise business, more and more applications are deployed, resulting in massive logs. As a tool to effectively record the running status of the application, the log can record external call encapsulation, status changes, system entry and exit, business exceptions, and unexpected execution situations. Real-time analysis of logs can improve the control over system operation and enhance the processing efficiency of abnormal system status and abnormal business. It greatly speeds up the crisis handling speed, improves the response speed of enterprise services, and ensures the stability of applications. [0003] The massive logs produced by running applications have been passed and cannot be monitored manually in real time. The normal use of logs is only after the exception occurs,...

AI Technical Summary

Problems solved by technology

[0003] The massive logs produced by running applications have been passed and cannot be monitored manually in real time

The routine use of the log is only after the exception occurs, and the technicians and operation and maintenance personnel will find the abnormal point through the log, and cannot quickly locate and deal with the problem at the first time

If you try to handle abnormal situations in a timely manner, you need to bury points in the application. Although it can respond quickly, this processing greatly increases the coupling degree of the application and increases the complexity of the code.

At present, many enterprises have used ELK for log management. Although kibana can analyze the logs and trigger corresponding alarms, the alarm method is relatively simple, and the alarm grouping function is simple, which cannot efficiently manage the alarm receivers.

Images