Routing is the process of deciding how traffic gets from one network to another. Inside a single subnet, devices can talk directly. The moment traffic needs to leave that subnet, something has to decide the next hop — that's a router's job, and the logic it uses to make that decision is where routing protocols come in.

Static routing: simple, but doesn't scale

A static route is a manually configured instruction: "to reach this network, send traffic to this next-hop address." It's simple, predictable, and perfectly adequate for small networks with one or two paths in and out. The problem is that it doesn't adapt — if a link goes down, a static route has no way to know and reroute around it. Someone has to notice and fix it manually.

Most small business networks with a single internet connection and no complex internal routing never need anything beyond static routes and a default gateway. This is worth knowing, because a lot of routing protocol complexity is solving a problem that a given network doesn't actually have.

Dynamic routing: letting routers figure it out

Dynamic routing protocols let routers automatically discover the network topology and adjust when something changes. They generally fall into two families: distance-vector protocols, which share "how far away is this network" information with neighbors, and link-state protocols, which build a more complete map of the entire network topology.

OSPF: the internal workhorse

OSPF (Open Shortest Path First) is a link-state protocol used to route traffic within a single organization's network — think a company with multiple offices and internal data centers that all need to reach each other efficiently. It organizes a network into "areas" to keep things manageable at scale, and calculates the best path using a cost metric, typically based on link bandwidth.

If you work in networking long enough to touch enterprise-scale internal routing, OSPF is the protocol you're most likely to encounter and configure directly.

BGP: how the internet actually holds together

BGP (Border Gateway Protocol) is the protocol that routes traffic between separate networks — specifically, between the large independent networks (called autonomous systems) run by ISPs, cloud providers, and large enterprises. It's a path-vector protocol, meaning it doesn't just consider distance, it considers the actual path a route takes and can apply policy decisions about which paths are preferred.

BGP is genuinely how the internet holds together — every major provider announces which address ranges it can reach, and BGP is the mechanism by which that information propagates globally. It's also infamous for how badly things go when a misconfiguration leaks — a bad BGP announcement has taken down major services and even entire countries' internet access in the past.

Which one will you actually touch?

  • Static routes: almost every IT professional configures these at some point, even in small environments
  • OSPF: relevant once you're managing internal routing for a mid-size or larger organization with multiple sites or complex internal topology
  • BGP: mostly the domain of ISPs, large enterprises with multiple internet connections, and cloud infrastructure teams — most IT careers touch this rarely, if ever, but understanding what it does is genuinely useful context for anyone working with cloud networking or multi-site connectivity