Системи моніторингу та прогнозування навантаження в інфокомунікаціях

Abstract

This work explores the importance of load forecasting systems in infocommunications and implementation of monitoring systems according to modern system requirements to ensure constant availability and prevent incidents. Tools for ensuring constant availability and monitoring were proposed and explained their functionality, this includes ELK stack for log monitoring, Prometheus for system physical and program specific metrics, auto-scaling tools such as AWS Auto Scaling which can be used with Prometheus to create specific rules of auto scaling. Load Balancing tools were proposed to ensure availability even on incident appearance by redirecting traffic from broken instances to working ones without downtime.