ANALYSIS OF THE EFFICIENCY OF CLOUD INFRASTRUCTURE MONITORING SYSTEMS BASED ON PROMETHEUS AND GRAFANA

Abstract

The article considers the effectiveness of cloud infrastructure monitoring systems based on Prometheus and Grafana in modern distributed environments. Modern cloud computing environments are widely used to deploy information systems and services, which leads to an increase in the complexity of their infrastructure and increases the requirements for monitoring its state. In this regard, ensuring effective monitoring of the performance, availability and reliability of system components becomes particularly relevant. The rapid development of cloud technologies and microservice architectures leads to an increase in the complexity of information systems. This requires reliable methods for monitoring the state of the infrastructure and early detection of potential problems. The main attention is paid to the analysis of approaches to collecting, processing, storing and visualizing service performance indicators in the cloud environment.
The architecture of the monitoring system implemented on the Amazon Web Services platform using the Terraform, Docker and CI/CD pipeline automation tools is analyzed. The mechanisms for collecting telemetry data using Node Exporter and CloudWatch Exporter are investigated, as well as the possibilities of aggregating indicators in Prometheus and their subsequent visualization using Grafana. Key performance indicators such as processor load, RAM usage, network bandwidth, query intensity, and database status are analyzed.
The results of the study showed that the integration of Prometheus and Grafana provides high transparency of the cloud infrastructure, allows for timely detection of anomalies, reduces response time to failures, and supports informed decisions on scaling services. It was found that the use of custom dashboards and automated notification systems increases the efficiency of infrastructure management and contributes to the rational use of computing resources. Particular attention was paid to assessing the impact of monitoring systems on the stability of cloud services and the ability to quickly identify infrastructure bottlenecks. The results confirm that the use of modern monitoring tools increases system fault tolerance, optimizes infrastructure costs, and improves the quality of digital services. The practical significance of this work lies in the possibility of applying the proposed approach to building complex monitoring systems in enterprise cloud environments, intensively used web applications, and distributed information systems.