Question bank — Static routing vs dynamic routing

False. Every router forwards packets by consulting a Routing table; "static" only describes how the table got filled (by hand). The table is still fully present, just hand-written and unchanging.

False. Automation costs bandwidth, CPU, convergence delay, and opens an attack surface. For a tiny stub network with one exit, static is both faster and safer.

False. The reply packet travels the opposite direction and needs its own return route on the other router. Routing is per-direction; one-way config gives the classic "ping times out" bug.

True. Longest prefix match always prefers the more specific (longer) prefix, so 0.0.0.0/0 is only used when nothing more specific matches — no conflict, just a fallback.

True (by definition). There is nothing to recompute; the table is already set. But this "instant convergence" is only correct as long as the topology it assumes hasn't changed.

False. During the convergence window, routers temporarily disagree and can drop or loop packets. Dynamic routing recovers automatically, but recovery is not instantaneous.

False. RIP is distance-vector (Bellman-Ford algorithm) — it trusts a neighbour's hop count. OSPF is link-state (Dijkstra's algorithm) — every router builds the full map and computes shortest paths itself.

False. Simplicity moves the error to the human: a typo'd next-hop, a forgotten return route, or a stale route after a topology change all silently misroute traffic with no auto-correction.

The mask is wrong. A default route is 0.0.0.0/0 (prefix length 0, the shortest possible). /24 would only match a narrow block, defeating the catch-all purpose. See Default gateway.

The word trusted is the error. A malicious or compromised neighbour can inject false routes; distance-vector protocols especially will believe an advertised low cost. Trust is exactly the vulnerability, not a guarantee.

The error is comparing link cost alone. Bellman-Ford compares link + neighbour's advertised cost, i.e. $c(x,v)+D_v(d)$. A pricier first link can still yield a cheaper total path.

The error is ignoring manual labour and scale. Static routes grow with every destination × every router and can't self-heal; at that size the human cost and failure risk make dynamic the only practical choice.

Confuses prefix length with metric. The router first selects the row whose destination prefix matches the most bits of the packet's IP; metric only breaks ties among equally-specific routes. See Longest prefix match.

There is no central server. In link-state, every router floods its local links so all routers hold the same map, then each independently runs Dijkstra's algorithm locally.

Because a conversation is two-way: the request goes forward and the reply returns. Each direction needs a matching route, so a return route is required on the router the reply passes through.

Its prefix length is 0, so it matches last under longest-prefix match. Any more-specific route wins, meaning the default is used only when nothing else matches.

Many protocols (e.g. RIP) send periodic update messages to keep neighbours informed and to detect silent failures. These keep-alives run regardless of whether the topology moved.

No routes are learned from neighbours, so there is no channel through which a malicious party can inject fake routes — the attack surface of route exchange simply does not exist.

Because the destination might be reachable through several neighbours at different total costs; taking the minimum guarantees the router picks the genuinely cheapest first hop, per Bellman-Ford algorithm.

Convergence is the time routers need to agree after a change. Static tables never recompute, so there is nothing to converge — the trade-off is that they also never notice a change.

They are still forwarded toward the dead next hop and dropped (or blackholed). Static routing cannot detect the failure, so the wrong route stays "correct" until a human fixes it.

The /32 wins because it is the longer (more specific) prefix; longest-prefix match always favours the tighter match regardless of configuration order.

It may install both and load-balance across them (equal-cost multipath), or apply a tie-breaker. Equal cost is a valid, non-degenerate outcome, not an error.

Static (a single default route). There is nothing for a protocol to adapt to, so paying protocol overhead buys nothing — a textbook stub-network case.

No. With no matching route (and no default), the router simply drops the packet, typically returning a "destination unreachable" signal rather than looping.