The scenario
Office sits behind the company router Customer (AS 65100), which has one uplink to ISP-A (AS 65001) and one to ISP-B (AS 65002). The web server Server is in a hosting network, 203.0.113.0/24, behind router Hosting (AS 65200), which is connected to both ISPs too. Every router runs eBGP with its neighbours and none has a static route or runs OSPF. So Customer learns 203.0.113.0/24 from both ISPs, Hosting learns the office network 172.16.1.0/24 from both ISPs, and each picks one. When the link to ISP-A fails, the routes through it are withdrawn and both routers are left with the path through ISP-B.
What you will work out
Learn how a multi-homed network learns the same destination from two ISPs, how BGP picks one, and what really happens when an uplink fails.
Why this matters
Businesses that cannot afford to be offline buy uplinks from two ISPs and run BGP with both. Whether that actually survives an outage depends on the details this lab makes visible: the destination must be reachable through both providers, both directions of a conversation must be able to move, and which link is primary should be a deliberate choice rather than an accident of router IDs.
Running it
Multi-Homed Customer: Two ISP Uplinks opens on a canvas, already wired and addressed. Trace a packet through it hop by hop, or run the same topology live on real Linux network namespaces. How the two modes differ.
Open Multi-Homed Customer: Two ISP Uplinks
Runs in your browser. No install, no signup, and the simulation starts instantly.
Open the lab →