Abstract

In 2017, the widespread adoption of container orchestration platforms such as Kubernetes introduced new complexities in managing network communication across microservices. Unlike traditional virtual machines, containers demanded lightweight, dynamic, and scalable networking solutions. This research explores the challenges of virtual networking in containerized environments, including service discovery, overlay network performance, east-west traffic bottlenecks, and multi-host communication. It evaluates leading network plugins (e.g., Flannel, Calico, Weave) and their trade-offs in terms of latency, security, and scalability. The study proposes an adaptive networking architecture that incorporates policy-based routing and service mesh integration to enhance microservice resilience and observability.