Prometheus exporter database monitoring system

Abstract

The invention provides a Prometheus exporter database monitoring system. The system comprises a factory class (generating different database index collectors), an interface realized by the factory class (obtaining indexes and exposing the indexes to a prometheus registry), different database index collector classes and different database index specific parameter value grabbing logic (similar to the existing open source exporter, capable of being modified as required). The factory class generates different index collectors, and the exporter end is separated from a plurality of target ends by connecting the public interface with the target ends to obtain parameter values. According to the invention, the index monitoring of the multi-target dynamic target can be realized by only deploying oneDB Exporter. The conditions of resource occupation and excessive processes of the target end are avoided. The problem of mirror image generation or docker frequent deployment does not need to be considered. The later operation and maintenance are simple.

Description

technical field [0001] The invention relates to a Prometheus exporter database monitoring system. Background technique [0002] Usually the Prometheus exporter is bundled with the monitoring target. Prometheus is based on the Pull working mode and regularly collects data from the Agent side (exporter). Prometheus Exporter exposes the endpoints of monitoring data collection to Prometheus Server in the form of HTTP services. Prometheus Server can obtain the monitoring data that needs to be collected by accessing the Endpoint endpoints provided by the Exporter. [0003] The general process of Exporter is as follows figure 1 As shown, the internal implementation of HTTPServer is as follows: when the Collector instance register() method is called, the instance will be saved to the CollectorRegistry, and the CollectorRegistry is responsible for maintaining all Collector instances in the current system. After the HTTPServer receives the HTTP request, it will get all the Collector...

AI Technical Summary

Problems solved by technology

Whether it is bundled with the target through the image or manually deployed to the target later, the exporter is relatively intrusive to the target, and the user's misoperation may affect the exporter process

At the same time, monitoring involves many aspects, such as node_exporter for monitoring linux machines, mysqld_exporter for monitoring mysql, network monitoring, etc. If various exporter processes on the official website and in the community run on the target side according to requirements, it will make the target side very bloated and consume resources , affecting target performance

Images