How VxLAN Powers Seamless Dual‑Active Data Center Failover

In dual‑active data‑center solutions, achieving high reliability across sites requires the business cluster, storage, and network to operate as a single logical entity, so that virtual machines can be migrated without users noticing any interruption.

Traditional Layer‑2 networks force a change of IP and MAC addresses when a server moves to another data‑center, breaking TCP connections and causing service outages. This makes seamless VM migration impossible.

To address this, a "big‑layer" (大二层) network architecture was created, specifically designed for the dynamic migration needs of virtualized environments. Major vendors provide implementations such as VxLAN, Huawei’s CSS/iStack/EVN/SVF, Cisco’s VSS/FEX, and H3C’s IRF.

VxLAN works by encapsulating Ethernet frames inside UDP packets, using a 24‑bit Virtual Network Identifier (VNI) that supports up to 16 million isolated networks. This overlay eliminates the 12‑bit VLAN ID limitation (max 4094 VLANs) and bypasses the constraints of the Spanning Tree Protocol, providing a loop‑free, scalable Layer‑2 extension across data‑center sites.

Traditional VLANs rely on a 12‑bit VLAN ID to segment broadcast domains, which quickly reaches its limit in large virtualized environments. Additionally, STP restricts the number of usable VLANs and can cause broadcast storms.

By adopting VxLAN, data‑center operators can achieve seamless cross‑site failover, maintain active‑active services, and ensure that virtual machines continue to run without IP/MAC changes or connection interruptions.