Learn
A2.3.4 | STATIC AND DYNAMIC ROUTING (HL ONLY)
01 | ROUTING CHOOSES THE NEXT NETWORK STEP
A student laptop needs a file on another subnet. Its local switch can deliver Ethernet frames within its VLAN, but reaching the remote subnet requires a routing function. That function must know where to send the packet next.
Static routing uses administrator-defined routes. Dynamic routing uses configured routing protocols to learn and update route information. Both ultimately provide information used to forward IP packets toward destination networks.
This HL-only lesson explains their processes and trade-offs. We will compare configuration, maintenance, complexity, resource usage, convergence, scalability and network size. You do not need the operation of a named routing protocol for this objective.
The essential distinction is how routes are obtained and maintained. Static and dynamic routes are not different kinds of application data, and dynamic routing is not the same thing as dynamically assigning an IP address.
02 | ROUTING BETWEEN LOCAL NETWORK SECTIONS
Suppose student devices use 10.10.10.0/24 and a service uses 10.20.30.0/24. A student host sees that the service is outside its local subnet and sends toward its configured gateway under its routing rules.
The host’s Ethernet frame targets the gateway’s local interface, while the enclosed IP packet targets the service. The router selects an outgoing interface or next hop and creates suitable link framing for that part of the journey.
A Layer 3 switch can perform this routing between VLANs. An ordinary Layer 2 switch does not learn IP routes merely by learning MAC addresses. The same physical product may support both functions, but they remain different operations.
Hosts in the same ordinary subnet can usually communicate through local switching without crossing a router. The routing topic concerns movement between the relevant IP networks, including connected LANs and VLAN subnets.
03 | READ A ROUTING ENTRY
A route identifies a destination prefix and how to reach it, for example through a next-hop router or a directly connected interface. The prefix describes a network address block, not a list of every possible host.
The next hop is the next forwarding destination. It is not necessarily the final server. Each router normally makes its own forwarding decision from its installed information rather than asking the original sender to specify the entire path.
A route can also record its source and preference information. Dynamic protocols may compare candidates using a metric, such as a protocol-defined cost. Different protocols use different rules, so do not assume all metrics mean physical distance or current congestion.
A router can hold several candidate routes, then install selected forwarding information. Introductory descriptions often call the whole thing a routing table. In real implementations, the routing information base and forwarding table may be separate structures.
| Illustrative destination | Forwarding instruction | Meaning |
|---|---|---|
| 10.10.10.0/24 | Directly connected: student LAN | Deliver on that local interface |
| 10.20.30.0/24 | Next hop R2 on the router-to-router link | Send toward R2 for this remote subnet |
| 0.0.0.0/0 | Next hop R5 | Default for IPv4 destinations without a more specific match |
04 | MATCH THE DESTINATION PREFIX FIRST
Ordinary IP forwarding uses the longest matching prefix: the most specific installed route matching the destination address. A /24 matching route is more specific than a matching /16, which is more specific than a /0 default.
For example, a packet for 10.20.30.8 matches both 10.20.0.0/16 and 10.20.30.0/24. The /24 determines the forwarding choice in this example. A route does not win merely because its printed metric is a smaller number than a route to a different, less-specific prefix.
Preference between route sources and protocol metrics helps select candidates for a destination prefix. Keep that separate from choosing the longest matching prefix when forwarding a packet. You do not need vendor-specific preference values here.
The activity uses a fixed illustrative installed table. “R2” and “R3” are symbolic next-hop labels, not addresses to type into a real router.
| Installed prefix | Instruction |
|---|---|
| 10.10.10.0/24 | Connected student LAN |
| 10.20.0.0/16 | Via R2 |
| 10.20.30.0/24 | Via R3 |
| 0.0.0.0/0 | Via R5 |
05 | STATIC ROUTING: CONFIGURE THE INTENDED NEXT HOP
With static routing, an administrator defines the route, typically including a destination prefix and an appropriate next hop or outgoing interface. The router can use that instruction while its installation and reachability conditions are satisfied.
For a simple branch with one exit, a static default route may be enough for destinations beyond the branch. A central router also needs suitable routes back toward the branch. Forward connectivity alone does not guarantee that replies can return.
A static route does not discover the whole remote topology. The administrator must choose a valid next step, avoid inconsistent paths and maintain the configuration when the network changes. A wrong next hop can send traffic into a dead end or a forwarding loop.
You do not need command syntax for this lesson. Be able to describe the instruction: “For destination subnet X, forward through reachable neighbour Y.”
06 | STATIC ROUTING: SIMPLE DOES NOT MEAN NO WORK
A small, stable network can be easy to understand with a few static routes. Administrators choose predictable next hops, and there is no dynamic routing-protocol exchange needed just to learn those routes. This can reduce control-plane bandwidth and processing overhead.
Static routing still uses memory, route processing and packet-forwarding resources. “Static uses no CPU” is incorrect. Its resource advantage concerns the absence of ordinary dynamic protocol learning and update work, not the absence of all router activity.
As networks grow, manual configuration and checking can become difficult. New subnets, changed links and return paths may require edits on several devices. Repeated manual work increases the opportunity for mistakes and stale routes.
A useful comparison is effort over time. Static configuration may be straightforward initially for two routers, while maintaining hundreds of related instructions across a changing campus can be costly and risky.
07 | WHAT HAPPENS WHEN A STATIC PATH FAILS?
An ordinary configured static route does not automatically learn a new remote path because another route exists somewhere on a diagram. If its selected path fails beyond the locally reachable next hop, packets may continue toward that neighbour and then be dropped.
Static does not mean a route always remains active regardless of every condition. A directly failed interface or unresolved next hop can affect route installation. Administrators can also configure backup static routes, reachability tracking or other mechanisms for specific failover designs.
However, these are extra planned mechanisms, not proof that basic static routing discovers arbitrary topology changes. In our activity, the primary path fails remotely and no backup or tracking has been configured, so an administrator must change the route.
State the assumptions in an exam explanation. “The single static path has no configured failover, therefore this remote failure requires intervention” is more accurate than “static can never support any automatic failover”.
08 | DYNAMIC ROUTING: CONFIGURE HOW ROUTERS LEARN
With dynamic routing, administrators configure a routing protocol and the networks, interfaces, neighbours and policies that should participate. Compatible routers exchange the relevant reachability or topology information under the protocol’s rules.
Routers use that information and the protocol’s selection rules to calculate or select suitable routes. They install usable forwarding choices and update them as information changes. The application data packets still use ordinary IP forwarding; they do not need to contain the routing protocol’s entire calculation.
This reduces the need to type every remote route manually. It does not mean zero configuration: participation, addressing, policy, protection and capacity still need careful planning. Incorrect advertisements can create widespread problems.
Specific protocols vary in what they exchange, how they select paths and when they send updates. Some use periodic messages, others triggered changes or a mixture. Avoid a universal claim that every router broadcasts its entire table all the time.
09 | CONVERGENCE: ADAPTATION TAKES TIME
Convergence is the process of reaching a sufficiently consistent, stable routing state after information or topology changes. It includes detecting the change, communicating relevant information, recalculating or selecting routes and updating forwarding.
During the transition, different routers may temporarily hold different information. Some packets can be dropped, delayed or loop until the relevant state settles. A working alternative physical path does not by itself guarantee an uninterrupted service.
The time taken depends on the protocol, detection mechanisms, network design, policies, load and equipment. It is not one fixed number for all dynamic networks. Faster detection and calculation can also involve resource and design trade-offs.
Dynamic routing can adapt only if a valid permitted alternative exists and the devices can learn and use it. It cannot create a missing link, restore a failed destination service or ignore an intentional policy restriction.
10 | TEST A REMOTE LINK FAILURE
Our campus model has two routes from R1 toward the service LAN behind R4. Initially the upper route R1 → R2 → R4 is selected. The lower route R1 → R3 → R4 is a usable alternative under the activity’s policy. The costs are illustrative design values, not measurements of a real routing protocol.
The R2–R4 link then fails while R1–R2 remains up. In static mode, no tracked backup is configured: R1 continues to have its old instruction until the administrator changes it. In dynamic mode, advance through detection, update, calculation and installation to observe convergence.
The intermediate screens are simplified teaching phases. They do not prescribe the exact message order of a named protocol or model every router’s forwarding state. The result shown is R1’s selected direction and whether the illustrated complete path is usable. Return routes are assumed correctly arranged.
11 | ROUTE LEARNING AND PACKET FORWARDING ARE DIFFERENT
A dynamic routing protocol exchanges control information so routers can maintain their route knowledge. Ordinary packet forwarding uses installed information to move user traffic. Routers do not run a complete routing-protocol exchange from scratch for every web packet.
Likewise, a static route can be used many times without an administrator selecting each packet’s path manually. The human configures the instruction; the device then forwards matching traffic using that instruction.
This distinction explains resource usage. Dynamic control processes consume processing time, memory and some link capacity for neighbour state, updates and calculations. Both approaches still need resources to store forwarding information and move traffic.
Do not equate route updates with the application’s payload, or with TCP acknowledgements. Those messages have different purposes even when they share the same physical network.
12 | METRICS DO NOT ALWAYS MEAN “CURRENTLY FASTEST”
A metric is a protocol-defined value or collection of values used to compare candidate paths under that protocol’s rules. It might reflect hop count, configured cost or other supported characteristics. Policies and route-source preference can also affect what is selected.
A route with fewer hops is not necessarily lower latency or higher capacity. A path with a high-capacity link may still be busy. Many routing designs deliberately prefer stable paths and do not continuously chase instantaneous congestion.
Therefore “dynamic routing always chooses the fastest route” is not a safe explanation. Say it selects routes according to the configured protocol and policy using the information available.
Similarly, the numbers in our diagram represent an illustrative initial preference. The required objective is the general learning and adaptation process, not a named protocol or a universal path-cost formula.
13 | COMPARE ALL SEVEN REQUIRED FACTORS
| Factor | Static routing | Dynamic routing |
|---|---|---|
| Configuration | Explicit destination/next-hop instructions; manageable for simple designs | Configure protocol participation, neighbours/networks and policy |
| Maintenance | Manual route changes often needed as remote paths and networks change | Learns supported changes automatically; configuration and monitoring still needed |
| Complexity | Few routes can be easy to reason about; large manual sets become complex | More control-plane behaviour and troubleshooting knowledge required |
| Resource usage | No dynamic protocol learning/update overhead for these routes; forwarding still costs resources | Additional state, update traffic and calculations; amount varies by design |
| Convergence / response to change | Basic routes do not learn new paths; configured tracking/backup can change behaviour | Detection, updates and selection/install work take time; transient loss possible |
| Scalability | Manual effort rises with relevant route/device changes; aggregation/defaults can help | Often better suited to many changing routes, with hierarchy/policy and capacity planning |
| Network size | Often suitable for small, stable or single-exit sections | Often useful for larger or changing multi-path networks; not mandatory everywhere |
Both approaches can exist in one network. A campus may learn internal routes dynamically while using a configured static default or selected special-purpose routes. The choice is about the appropriate mechanism for each part of the design.
14 | CHOOSE FOR A SMALL OFFICE OR A LARGE CAMPUS
A small office has one router and one predictable exit. A static default route can be an economical and understandable choice because there are few alternative paths to discover. It still needs correct provider information, return connectivity where relevant and dependable equipment.
A campus with many routed VLANs and alternative building links changes more often. Dynamic learning can reduce repetitive manual route updates and adapt to supported link changes, making it attractive as the topology and route count grow.
Size alone is not a rule. A large network can contain stable stub sections with static routes; a small network with several important alternative paths may benefit from dynamic routing. Consider change frequency, availability requirements and support expertise.
Do not install complexity without purpose. A design the team cannot securely configure, monitor and recover can be less dependable despite having more automatic features.
15 | MAINTENANCE AND SECURITY STILL MATTER
Routing controls reachability, but it is not the same as application permission or encryption. A valid route toward a service does not authorise every user to access every record. Firewalls, authentication and application checks remain separate requirements.
Protect router administration and route configuration. For dynamic routing, restrict appropriate participation and advertisements and use supported protection mechanisms where applicable. A misleading advertisement or faulty policy can send traffic toward an unintended path.
Static routing avoids a dynamic protocol exchange for its routes, which can reduce that particular exposure, but a wrong or compromised static configuration can still redirect or blackhole traffic. Neither method is inherently a complete security solution.
Maintain documented prefixes, next hops, expected alternatives and recovery procedures. Test both permitted forwarding and failure behaviour after changes, including the return direction and shared dependencies.
16 | APPLY THE PROCESS TO A SITUATION
Identify the mechanism or concept most directly described. Some situations deliberately separate automatic address allocation, route learning and packet forwarding so their meanings stay clear.
17 | CORRECT THE COMMON CONFUSIONS
| Claim | More accurate explanation |
|---|---|
| Static routing requires a human for each packet | The configured route is automatically used for matching packets |
| Dynamic routing means DHCP | DHCP configures addresses; dynamic routing learns routes |
| Dynamic always picks the fastest path | Selection follows protocol and policy, not a universal speed detector |
| Any alternative link makes recovery instantaneous | Detection and convergence take time, and the path must be usable |
| Static routes can never have failover | Configured tracking or backup routes can provide specified failover behaviour |
| A route guarantees a complete successful exchange | Return paths, policy, services and transport/application handling also matter |
Use the distinction between route management and forwarding to explain why both methods can move the same kinds of user packets while responding differently to a topology change.
18 | EXPLAIN THE MECHANISM, THEN THE CONSEQUENCE
For explain, state how the method works and link it to its effect with because, therefore, so that or resulting in. Name the required factor rather than giving a vague judgement such as “dynamic is better”.
For static routing: “An administrator configures a destination prefix and next hop. Matching packets follow the installed instruction, so a stable single-exit network can be managed with little routing-protocol overhead. If a remote path changes and there is no tracked backup, the instruction does not discover a new path, therefore manual intervention may be required.”
For dynamic routing: “Configured routers exchange relevant information and use protocol rules to select routes. When a failure is detected, information and forwarding choices can be updated, so a valid alternative can be used after convergence. This reduces repetitive manual changes but adds control traffic, state and processing work.”
Then apply network size and change frequency. Explain why the operational benefit is worthwhile in the stated scenario, and include limits such as transient loss, configuration complexity or missing alternatives.