An application team wants the same VLAN stretched between two sites. How do you respond?
Show the answer
Say this first
I first find the real requirement, usually VM mobility without re-addressing, a cluster that wants one subnet, or disaster recovery without changing IPs. Then I explain the cost: stretching layer 2 joins two failure domains, so a loop or storm at one site takes down both, which defeats the point of having two sites. If the need is real, I prefer solving it above the network (DNS, load balancers, application replication); if layer 2 is unavoidable, I use an overlay built for it, such as VXLAN with BGP EVPN, never a plain trunk with spanning tree across the link.
Go deeper
Why stretched layer 2 is risky
- Shared failure domain. Broadcast, unknown unicast and STP events cross the interconnect. One looped cable can take out both data centers.
- Traffic tromboning. A VM moved to site B still uses its gateway in site A unless the gateway is present in both sites.
- Split brain. If the interconnect fails, both halves of a cluster may think they are primary, and the same IP may be active in two places.
- Bandwidth. Flooded traffic and replication compete with real traffic on the interconnect.
Options, best first
- Remove the need: DNS or global load balancing to move users between sites, applications that replicate data themselves, automation that re-addresses during failover.
- Routed interconnect, separate subnets per site, and advertise service addresses where they live.
- If layer 2 is truly required, VXLAN with BGP EVPN between sites: distributed anycast gateway in both sites, ARP suppression, controlled flooding, no STP across the interconnect, only the VLANs that need it.
Questions I ask
- Which VLANs, how many hosts, and for how long?
- What should happen when the interconnect fails? Who is primary?
- Is this for live migration, or for disaster recovery where a short re-address would be fine?
They will ask next
What is a distributed anycast gateway, and why does it help here?
Every switch that hosts the VLAN answers as the gateway with the same IP and the same MAC. A host that moves between sites keeps its gateway configuration and is routed locally in its new site, so traffic does not cross the interconnect just to reach a gateway.
Traps
- Saying yes immediately because it is technically possible. The interviewer wants the risk and the alternatives.
- Saying no without understanding the business need. Ask why first.