METHOD FOR OPTIMAL CONTAINER PLACEMENT FOR WEB PORTALS BASED ON RESOURCES AND PERFORMANCE

Abstract

In the context of the rapid growth in the complexity of modern web portals and the increasing pressure on IT infrastructure, the issue of efficient resource allocation has become a central challenge in maintaining high system performance, reliability, and service availability. The dynamic and often unpredictable nature of user interactions, combined with fluctuating traffic patterns and the proliferation of microservices, places immense demands on container orchestration systems. Inefficient container placement strategies can result in suboptimal utilization of computational resources, disproportionate server load distribution, higher energy and operational costs, latency spikes, and ultimately, a degraded user experience.

This research focuses on enhancing the method of optimal container placement specifically tailored for web portal environments. The proposed approach builds upon traditional optimization techniques by integrating dynamic and context-aware metrics, including real-time load variability, application-specific performance indicators, environmental stability measures, system recovery time objectives (RTOs), hardware limitations of physical servers, and network topology constraints. By leveraging multi-criteria decision-making and optimization models, the method seeks to find a balance between performance efficiency and resource conservation.

The expected outcomes of the study are twofold: firstly, to significantly reduce the mean time to recovery (MTTR) in the event of system disruptions; and secondly, to ensure the consistent, high-quality operation of web portals even under volatile load conditions. The practical contribution of the proposed method lies in its applicability for automated infrastructure management, offering a scalable and adaptive solution for cloud and on-premise environments. This results in reduced IT operational costs, improved service delivery, and increased end-user satisfaction. Furthermore, the method provides a foundation for future advancements in intelligent orchestration and self-healing infrastructure systems.