Static routing vs BGP
Static routing vs BGP: static is hand-configured routes, BGP exchanges routes dynamically. We cover when static suffices and when you need BGP.
Last updated:
Static routing means routes you configure on a router by hand, which never change until an administrator changes them. BGP (Border Gateway Protocol, RFC 4271) is a dynamic protocol in which networks automatically exchange route information and react to changes in topology. In practice static routing is enough for simple, single-homed connections, and BGP only becomes essential once you need redundancy, multihoming and your own routing policy at the edge of the internet.
| Aspect | Static routing | BGP |
|---|---|---|
| Configuration | Manually entered routes | Dynamic route exchange with a neighbour |
| Response to failure | None, needs intervention or BFD | Automatic switch to another path |
| Scalability | A handful of routes, one location | Hundreds of thousands of routes, many locations |
| Multihoming | Not workable in practice | The protocol's core use case |
| What you need | Nothing extra | Your own ASN and address space |
| Load on hardware | Minimal | Memory and CPU for the routing table |
| When it is enough | One link and a fixed topology | Two links or your own traffic policy |
When static routing is enough
Static routing works well wherever the topology is simple and stable. The typical case is a single link to one provider: you configure a default route (0.0.0.0/0 or ::/0) pointing at the provider's router and you have full internet reachability without any dynamic protocol. Static routing uses no memory or CPU to maintain a session. It is deterministic and easy to troubleshoot, because the route is exactly what you typed in.
- A single link to one provider (single-homed) with a default route to the upstream.
- Small, stable networks where the topology barely changes.
- Links where you want full, deterministic control over every route.
- Scenarios where you have neither your own ASN nor independent address space.
What static routing cannot do
The main weakness of static routing is that it does not react to failures. A configured route stays active even if the network on the other side has stopped working, as long as the local interface is up, so traffic is black-holed. You can partly work around this with path-failure detection such as BFD (RFC 5880) or IP SLA with reachability-tracked routes, but those are workarounds, not real dynamic routing. Static routing also does not scale to large numbers of prefixes and gives you no way to advertise your own networks to the internet.
When you need BGP
BGP is necessary the moment you want to be multihomed, that is, connected to at least two independent providers at once. Only BGP lets you advertise your prefixes to the world with your own ASN and automatically reroute traffic when one upstream fails. BGP also gives you policy control: with attributes such as local-preference, AS-path prepending, MED and BGP communities you shape where traffic enters and leaves. It is the foundation of an operator's resilience and independence.
- Multihoming: advertising your own prefixes to two or more providers at once.
- Automatic failover when one upstream or link stops working.
- Steering inbound and outbound traffic through BGP policy.
- Peering at internet exchange points (IXPs), where BGP is the only option.
- Owning your own public ASN and independent address space (PI).
Routing with AS202520 SkyPass
AS202520 SkyPass supports both approaches. If you have a simple, single-homed connection, we configure a static default route into our network and you are up immediately. When you need redundancy, full multihoming or your own policy, we establish a BGP session with you for full IP transit or peering, with PoPs in Warsaw and Wrocław and peering at Polish IXPs (THINX, TPIX, WRIX, POZIX, 1-IX). We also support migration from static routing to BGP as your network grows and needs its own ASN.
Frequently asked questions
Do I need BGP for a single link?
No. For one link to one provider, a static route with a default route to the upstream is enough. BGP only starts to make sense with multihoming or your own routing policy.
Can I multihome with static routes?
Not in a full sense. Static routing does not advertise your prefixes to the internet and does not react automatically to route failures. True multihoming at the internet edge requires BGP, your own ASN and independent address space.
How does static routing handle a link failure?
Poorly by default. The route stays active as long as the interface is up, so traffic is black-holed. You can mitigate this with mechanisms like BFD (RFC 5880) or IP SLA, but those are workarounds, not full dynamic routing.
Can I start with static and move to BGP later?
Yes. It is a common path: you start with a static route, and once you obtain your own ASN and address space plus a second provider, you migrate to a BGP session without disrupting service.