Automation Methods of Data Collection and Analysis for Managing Complex IT Infrastructures

The article discusses an approach to the organization of monitoring and management of server equipment, aimed at ensuring the stable operation of the IT infrastructure. Modern methods based on threshold analysis cannot cope with the volume and variety of data, which requires automation of the collection, processing and analysis of metrics. A computing cluster of servers using the Proxmox VE virtualization platform is presented as an object of research. The features of time series storage and data processing using the InfluxDB database, which has a convenient query language called Flux, are described. Implemented software architecture for integrating monitoring and data analysis tools, including Prometheus PVE Exporter, Telegraf and Flux. The advantages of using Python for metric analysis and query development using the Builder pattern are shown, which simplifies working with the system. Methods of data preprocessing and visualization for anomaly analysis and forecasting with the aim of maintaining the stable operation of server equipment are discussed. The proposed approach ensures the efficiency and ease of monitoring, while also enhancing the accuracy of decision-making given the increasing complexity of IT infrastructure.