Solution overview
This lab guide, "Building Redundancy with vPC, MLAG, ESI, and creating your eBGP Underlay," is designed to provide you with hands-on experience in configuring and verifying resilient access layer connectivity in a multi-vendor VXLAN fabric. The lab environment will consist of three spine switches (Cisco, Arista, Juniper) and three pairs of leaf switches (Cisco Nexus, Arista, and Juniper vQFX), each connected to a dedicated Ubuntu server. This multi-vendor setup will highlight the nuances and best practices for interoperability. We will explore the practical implementation of various redundancy mechanisms, including:
Cisco vPC (Virtual Port-Channel): A proprietary multi-chassis LAG solution for Cisco Nexus switches.
Arista MLAG (Multi-Chassis Link Aggregation Group): Arista's implementation of multi-chassis LAG. *
Active/Standby Connectivity: A basic redundancy model for platforms that do not support advanced multi-chassis LAG.
In the first section, we will build a foundational eBGP underlay across all leaf and spine switches, which is essential for providing IP reachability for VXLAN tunnels and serves as the transport for the overlay control plane. This underlay will feature a multi-vendor spine layer and leaf pairs (Cisco, Arista, Juniper) to highlight real-world interoperability challenges and solutions. We will also configure common VXLAN fabric elements such as VRFs, VLANs, and SVIs to simulate a realistic data center environment.