Transforming Enterprise Office Networks with EVPN-VXLAN: A BGP-Based Approach to Layer 2 Elimination

Abstract

As new demands for scalability, flexibility, cloud integration, and operational efficiency arise in enterprise office networks, there is a need for significant transformation. While using traditional Layer 2 spanning tree protocols may give some enterprises lower traffic utilization, slower convergence, higher broadcast overheads, and limited scalability, it can also add complexity when it comes to managing them.Traditional Layer 2 spanning tree protocols can also limit an enterprise's traffic utilization, convergence rate, broadcast overheads, and scalability in a large-scale network, while also increasing management complexity. In this study, the authors present an enterprise network modernization approach that removes legacy L2 dependencies by implementing Ethernet VPN (EVPN) and Virtual Extensible LAN (VXLAN) technologies in combination with a Border Gateway Protocol (BGP)-based control plane.
The proposed architecture will bring EVPN-VXLAN over-lay technology to the enterprise office infrastructures to meet the needs of scalable segmentation, optimized traffic forwarding, enhanced redundancy, and rapid converging. The model enhances network resiliency, operational management, and seamlessly integrates into SDN and cloud networking environments, all while replacing legacy Layer 2 mechanisms with a BGP-based approach. The study also considers the architectural benefits of EVPN-VXLAN in the areas of throughput, fault tolerance, scalability, and broadcast traffic reduction when compared to traditional enterprise networking solutions.
Furthermore, the study sheds light on the importance of programmable networking, automation, and intelligent monitoring for enhancing enterprise network performance and adaptability. The results prove that EVPN-VXLAN is a powerful, future-proof architecture that can enable enterprise digital transformation projects while solving the problems of traditional Layer 2 architectures.