I’ve stumbled upon so many terms and phrases about routing lately that I thought it would be interesting to list them up. It’s just a list of ‘routing stuff’, good things to know, but not an important topic for any certification (or in the field, for that matter). Well, here goes:

Selecting and using paths in a network to send data.

Inter-domain routing
Routing between different Autonomous Systems (AS). This usually involves BGP, or handles about inter-domain multicast.

Suboptimal routing
Having a working network that routes correctly, but not in the best way possible. Easiest example is RIP and different bandwidth: if a destination is two hops away on a 100Mbps link, and three hops on a 1Gbps link, RIP will choose the slowest link.

Multipath routing
Using multiple paths when routing, for load-balancing or redundancy. E.g. when two equal-cost (static) routes are present in the routing table, both paths will be used and receive about 50% of the data each.

Policy-based routing
Routing using policies. On Cisco routers, using ACLs to route based on port numbers and source addresses, not just destination addresses.

Static routing
The use of a fixed path when routing. A static route is usually entered manually into the router configuration, or hardcoded.

Not-so-static routing
A static route with a next hop set to an IP that’s not part of a connected subnet. On Cisco routers, this means a reverse lookup in the routing table will be done and the static route will follow the same route as a packet going to the next-hop IP would. If, for example, route has next-hop, and there’s an OSPF route to to interface FastEthernet0/1, then that interface will be used for too. Should the OSPF route change to another interface, the static route will follow.

Adaptive routing
Having the ability to adapt to changes in the network when routing. This can be through routing protocols which adapt to lost links, or tracked static routes too.

Dynamic routing
Synonym to adaptive routing, although it usually implies a routing protocol, not a tracked static route.

Black Hole routing
Dropping packets silently instead of forwarding them out of a physical interface. On a Cisco router, this is a route pointing to the Null interface. It can also be a route out of a physical interface to an empty subnet, where no host will ever respond. Usually accidental, but it can be useful to discard unwanted packets quickly. It’s called a black hole because it sucks up data, and you can’t see the actual black hole, you have to look for the missing data to find it.

Ad-hoc routing
Routing in a mobile (usually wireless), self configured network.

Hot-potato routing
Getting packets out of your AS as quickly as possible, sending them to the next AS with as few hops as possible. Since this is usually in a transit-AS, it’s typical terminology for an ISP. The name should be obvious.

Cold-potato routing
The opposite of hot-potato routing: keeping packets as long as possible in your AS, for example to give them better QoS.

Onion routing
A technique for anonymous communication, where each router adds or strips off a layer of encryption. Data is encrypted while traversing an onion router network so it’s unclear to any man-in-the-middle where data is coming from and going to. Like an onion has layers, so do packets have multiple layers of encryption.

Garlic routing
Like onion routing, but multiple messages are encrypted together so traffic analysis becomes difficult. Just like garlic, it’s a complex layered structure.

Ships-in-the-night routing
Having multiple routing protocols running on routers, but no redistribution between them takes place. Hence ships-in-the-night: everything is working, but the routing protocols can’t see each other. For example: domain A runs OSPF and needs to reach domain B, domain C runs EIGRP and needs to reach domain B also, but domain A and C may not see each other. You can run EIGRP and OSPF in domain B without redistribution to make this work.