A static route is typed in; a dynamic route is learned from a routing protocol. Static routing is simple, predictable and never adapts. Dynamic routing adapts and costs configuration, CPU and a protocol to understand. Small networks want static routes, with tracking so they can actually fail over.
- A static route is installed while its interface is up. That is the only check
- Static routes do not fail over on their own
- A floating static route plus tracking is the small network answer
- Administrative distance breaks ties, and does not beat prefix length
- A static route outranks every routing protocol, at distance 1
On this page
- What static and dynamic routes actually are
- Administrative distance, and the thing it does not do
- The two families of dynamic routing protocol
- The trap: a static route that points at a corpse
- Floating static routes, which is how failover actually works
- When a small network genuinely needs routing protocols
- Where people go wrong
- Comparison
- FAQ
The differenceWhat static and dynamic routes actually are
The whole of static vs dynamic routing is where the routing table entry came from, and everything else follows from that.
A static route is an instruction. Somebody decided that traffic for the 10.20.0.0/16 network goes to 192.168.1.2, and typed it in. The router installs it in the routing table and keeps it there. It does not check whether 192.168.1.2 is a real router, whether it knows a way to that network, or whether it is switched on.
A dynamic route is a conclusion. The router formed a relationship with a routing neighbor, received information about which networks that neighbor can reach, ran an algorithm over everything it heard, and installed the result. When the neighbor stops talking, or reports that something changed, dynamic routing recomputes the conclusion.
That difference produces every advantage and disadvantage in the comparison table, and one thing that is not usually in it: static routing has no way to know that anything on the network is wrong.
Administrative distanceAdministrative distance, and the thing it does not do
Administrative distance is how a router chooses when two different routing sources offer a route to the same network. It is a trust ranking, and lower wins.
| Source | Distance |
|---|---|
| Directly connected | 0 |
| Static route | 1 |
| External BGP | 20 |
| EIGRP | 90 |
| OSPF | 110 |
| IS-IS | 115 |
| RIP | 120 |
| Internal BGP | 200 |
Two things about those distances are worth stating clearly, because both are commonly misread.
A static route beats every dynamic routing protocol by default. Distance 1 outranks the 110 of OSPF and the 20 of BGP. If somebody types a static route for a network a routing protocol is already advertising, the static route wins, silently, and the protocol keeps advertising a path nobody is using.
Administrative distance only breaks ties for the same prefix. It is not a general priority system. A more specific network always wins regardless of where the route came from, because longest prefix match happens first.
A static route for 10.0.0.0/8 does not override an OSPF route for 10.20.5.0/24, no matter what the distances are. Getting this backward is the source of a whole class of confusing routing behavior.
The two familiesThe two families of dynamic routing protocol
Dynamic routing protocols split into two designs, and knowing which family a protocol belongs to explains most of how it behaves.
Distance vector protocols share their routing table. Each router tells its neighbors which networks it can reach and how far away they are, and each neighbor adds its own distance and passes it on.
No router ever sees the whole topology; it sees a list of destinations and distances reported by the routers next to it. RIP is the classic example, and EIGRP is a more capable descendant of the idea.
Link state protocols share the topology. Each router describes its own links to every other router in the area, so every router builds an identical map of the network and then computes its own shortest paths across it. OSPF and IS-IS work this way. It converges faster and costs more memory and CPU, because every router holds the whole map.
BGP is neither, or a distance vector protocol with a very long memory, depending on who you ask. It carries the list of networks each path traverses rather than a distance, which is what lets it make policy decisions and detect loops between organizations.
For a network small enough that this page is relevant, the family matters less than the fact that all of them are more machinery than a handful of static routes needs.
The trapThe trap: a static route that points at a corpse
This is the part of the comparison that matters in practice, and it is usually left out.
A static route is installed in the routing table as long as its outgoing interface is up. That is the only condition.
If the next hop router loses power, or its far side fails, or it is perfectly alive and has lost its own path onward, the link on your router stays up and the static route stays in the routing table. Packets are forwarded to a device that cannot deliver them, and they disappear.
Everything looks correct while this happens. The route is in the routing table. The link is up. The configuration is exactly as designed. And nothing on the network works.
Two situations produce it regularly.
A backup internet connection. The primary link stays physically up while the provider network beyond it fails. The static default route keeps pointing at it, and no packets reach the internet, because the failover was never going to happen.
A next hop router at another site. The tunnel is up, so the interface is up, so the route stays, while the router on the far end is off.
The fix is not a dynamic routing protocol. It is to make the static route conditional on something other than the state of the link.
FailoverFloating static routes, which is how failover actually works
A floating static route is a static route with a deliberately raised administrative distance, so it sits out of the routing table until the preferred route disappears.
ip route 0.0.0.0 0.0.0.0 198.51.100.1 primary, distance 1
ip route 0.0.0.0 0.0.0.0 203.0.113.1 10 backup, distance 10The second route is present in the configuration and absent from the routing table, because the first one is better. Remove the first and the second installs itself immediately. That is the whole failover mechanism, and it costs no bandwidth and no protocol.
It only helps if the first route can actually leave. Combining it with tracking is what makes it work:
Track something on the far side, not the local link. Most platforms can ping a reliable address through a specific route on a schedule and withdraw the route when the pings stop. On Cisco this is an IP SLA with object tracking; other vendors have equivalents under names like link monitors or health checks.
Track something past the provider's first hop. Pinging the provider's own gateway proves the link is up, which was never in doubt. Pinging a public address proves the path works.
Give it sensible timers. Aggressive tracking flaps the route on a brief packet loss. Three failed probes at five second intervals is a reasonable starting point.
A static route, a floating static route and tracking give a small network automatic failover without any dynamic routing protocol at all, which is the answer for most businesses with two internet connections.
When to switchWhen a small network genuinely needs routing protocols
There are cases where static routing stops being enough and a dynamic routing protocol earns its place. They are recognizable.
| Your network | Static routes | A routing protocol |
|---|---|---|
| One router, one internet link | The only sane choice | Pointless |
| Two internet links, one router | Floating static plus tracking | Only for BGP multi homing |
| Three or four routers, one path each | Still simpler | Not yet worth it |
| Ten routers, or more than one path | A spreadsheet nobody maintains | The right answer |
| Frequent additions and renumbering | Manual edits on every router | The right answer |
| Your own address space, two providers | No substitute exists | BGP, necessarily |
More than a handful of routers. The configuration burden of static routing grows with the number of routers times the number of networks. At three routers it is trivial. At ten routers it is a spreadsheet nobody maintains.
Any topology with more than one path. If there are genuinely two network paths between two points and you want both used, or automatic selection between them, that is what dynamic routing protocols are for.
A network that changes. Sites opening, networks being added, addressing being restructured. Every change is a manual routing edit on every router when routing is static.
Anything multi homed to the internet with your own address space. Two providers and your own addresses means BGP, and no arrangement of static routes substitutes for it.
Below those thresholds, a dynamic routing protocol adds a thing to understand, a thing to secure and a thing to debug, in exchange for adaptability the network does not need. The usual mistake in a small network is not choosing static routing. It is choosing static routing and not adding the tracking.
PitfallsWhere people go wrong
Expecting failover from plain static routes. They do not fail over. They have no mechanism to notice anything, and they will forward traffic into a hole indefinitely.
Tracking the wrong thing. Pinging the interface, or the provider's gateway, proves the cable works. The failure is usually further away than that.
Typing a static route over a dynamic route. Distance 1 wins against everything, so the static route silently overrides what the routing protocol computed, and the protocol keeps advertising a path that is not in use.
Assuming administrative distance overrides prefix length. It does not. More specific routes always win first, and distance only decides between equally specific routes from different routing sources.
Using an exit interface instead of a next hop on Ethernet. The router then treats every destination on the network as directly attached and sends an ARP request for each one. Static routes should name the next hop address.
Adding a dynamic routing protocol to a three router network. It works and it adds a subsystem to understand, secure and troubleshoot for adaptability nobody needed.
Leaving old static routes in the routing table. They never expire, they never complain, and a stale route for a network that was renumbered two years ago sits there quietly waiting to be the reason something fails.
ComparisonTwo rows decide it, and the rest is detail
| Criterion | Static routing | Dynamic routing |
|---|---|---|
| Configured by | A person, once | A protocol, continuously |
| Adapts to a failure | No, unless tracked | Yes, automatically |
| Bandwidth used | None | Small but constant |
| CPU and memory | Negligible | Real, and grows with the network |
| Effort at 3 routers | Trivial | Overkill |
| Effort at 30 routers | Unmanageable | Trivial |
| Predictable | Completely | Mostly |
| Something to secure | No | Yes, neighbors can be spoofed |
Rows two and five are the whole decision. Small and stable takes static routes with tracking bolted on. Large or changing takes dynamic routing protocols, and the crossover happens somewhere around the point where you stop being able to hold the network topology in your head.
FAQFrequently asked questions
What is the difference between static and dynamic routing?
A static route is configured by hand and never changes on its own. A dynamic route is learned from a routing protocol and updated automatically when the network changes.
Which is faster, static or dynamic routing?
Neither, once a packet is moving. The routing table lookup is the same either way. Dynamic routing is faster to react to a change and static routing is faster to configure on a small network.
Which is better for a small network?
Static routing, in almost every case, with a floating static route and tracking to handle failover. A dynamic routing protocol adds a subsystem that a small stable network does not need.
Does a static route fail over?
Not by itself. It stays in the routing table as long as its interface is up, regardless of whether the next hop can still deliver anything.
What is a floating static route?
A static route with a raised administrative distance, so it stays out of the routing table until the preferred route disappears. It is the standard way to configure a backup path.
What is administrative distance?
How much a router trusts each source of routing information. Lower wins. A static route defaults to 1, which beats every routing protocol.
Does administrative distance override a more specific route?
No. Longest prefix match happens first, so a more specific route always wins regardless of distance. Distance only chooses between routes to the same prefix from different sources.
How do I make a static route fail over properly?
Combine a floating static route with tracking that probes something beyond the provider's first hop, so the primary route is withdrawn when the path stops working rather than when the cable is unplugged.
What are the common dynamic routing protocols?
OSPF and EIGRP inside an organization, BGP between organizations and on the internet, IS-IS in large carrier networks, and RIP mostly in history.
When should a small business use a dynamic routing protocol?
Once there are more routers than you can hold in your head, more than one network path between points, frequent change, or multi homing to two providers with your own address space.
Is static routing more secure?
Slightly, in that there are no routing neighbors to spoof and nothing is advertised across the network. It is a small advantage and rarely the deciding one.
What is the most common static route?
The default route, 0.0.0.0/0, pointing at the internet connection. On most small networks it is the only entry in the routing table that somebody typed.
Why does my static route not work even though it is in the table?
Because being in the table only means the interface is up. If the next hop cannot deliver, the traffic is forwarded and lost, and nothing reports it.
Can I use static and dynamic routing together?
Yes, and most real networks do. A dynamic routing protocol handles general reachability and static routes pin the specific paths that should never be recomputed, with the administrative distance set deliberately rather than by accident.
Keep readingRelated concepts
Read next · Routing What Is BGP? Multi homing to two providers with your own address space is the one case with no static alternative. Open this next12 min- Routing · 12 min OSPF Explained The link state protocol most enterprises reach for when static routes stop being enough.
- Addressing · 10 min Default Gateway A default gateway is a static default route seen from the host side, which is why every device has one.
- Routing · 10 min The RIP Protocol, and Why You Still Meet It The decision that comes before choosing a protocol.
- Routing · 11 min EIGRP Configuration, and the Feasible Successor Logic That Makes It Work Where EIGRP sits in the choice between fixed routes and a routing protocol.