FCP-FGT-AD domain - 20% of the exam

Routing

Routing is 20% of the FCP-FGT-AD exam. These are the objectives it covers, each with practice questions, with every answer explained.

The domain in numbers

70
Practice questions
2
Objectives
34
Easy
36
Hard

Where people go wrong

  • Reversing the order of distance, priority and metric

    The wrong answers reverse the tiebreakers. Some treat a higher administrative distance as the more trusted route, some check priority before distance, and some add distance and metric together. FortiGate compares administrative distance first and installs the route with the lower value, whatever its metric. Priority only separates static routes that already share a distance, and metric only ranks routes from the same source.

    See it in a sample question
  • Expecting ECMP when distance or priority differs

    The wrong answers install both routes, or share traffic across them, even though one value differs. Equal-cost multipath needs the same administrative distance and the same priority. If the distances differ, only the lower-distance route is installed. If the distances match but the priorities differ, the route with the lower priority value is installed alone. A matching prefix length is necessary but does not make two routes equal.

    See it in a sample question
  • Treating a missed SLA target as a dead link

    The wrong answers say a member that breaches a latency target is declared dead, loses its route or has its administrative distance rewritten. A breached target only leaves the member not meeting the SLA, so SLA-aware rules skip it while its route stays installed. A member is declared dead only after the configured number of consecutive probes fail, and it returns once enough recovery probes succeed.

    See it in a sample question
  • Swapping the SD-WAN load-balancing algorithms

    Sessions, Volume, spillover and Source-Destination IP are each credited with another algorithm's behaviour. Sessions places new sessions in proportion to member weights by count, and Volume does the same by bytes. Spillover fills the first member until a bandwidth threshold is exceeded. Source-Destination IP hashes both addresses, so one host's traffic to different servers can use different members, which a source-only hash does not allow.

    See it in a sample question

Objectives in this domain

What to study

The official documents this domain's practice questions cite most.

Sample questions from this domain

Free sampleRoutingeasy

An administrator has two static routes to the same destination subnet, each pointing at a different reachable gateway. The routes carry different administrative distance values but identical priority values. Which statement correctly describes how the FortiGate decides which route to use for forwarding?

  • AThe route with the lower administrative distance is selected as best and installed in the routing table, so it is the one used to forward traffic. Correct
  • BThe route with the higher administrative distance is selected as best, because a higher distance signals a more trusted manual configuration.
  • CBoth routes are installed and load balanced, because identical priority values make the two routes equal regardless of their distance.
  • DPriority decides the outcome first, so with equal priority the FortiGate falls back to picking the route with the higher metric value.
FortiGate selects the best route to a destination by choosing the lowest administrative distance before any other tie-breaker. Administrative distance ranks the trustworthiness of competing routes to the same prefix. FortiGate compares distance first; the lowest-distance route wins best-route selection, enters the RIB, and is programmed into the FIB used for actual forwarding. Priority only matters between routes that have already tied on distance.

Why A is correct: Administrative distance is the first criterion FortiGate uses to rate competing routes to the same destination; the lowest distance wins the best-route selection and that route is installed in the RIB and pushed to the FIB for forwarding.

Why B is wrong: This inverts the rule, which is tempting because higher numbers feel stronger; in reality a lower administrative distance means a more trusted route, so the higher-distance route loses.

Why C is wrong: Equal priority alone does not make routes equal; distance is compared before priority, so the differing distances break the tie and only one route is chosen rather than both being load balanced.

Why D is wrong: This wrongly puts priority ahead of distance and then misuses metric; distance is evaluated before priority, and a higher metric is worse, not better, so this reasoning fails on two counts.

Free sampleRoutingeasy

A FortiGate has two static routes to the same destination subnet, each pointing at a different reachable gateway. The administrator wants both routes installed at the same time so traffic load-shares across both links. Which condition must be true for the FortiGate to perform equal-cost multipath across both routes?

  • ABoth routes must have the same administrative distance but different priority values.
  • BBoth routes must have the same administrative distance and the same priority value. Correct
  • CBoth routes must have different administrative distances but the same priority value.
  • DBoth routes must have the same destination prefix length but may differ in distance.
Equal-cost multipath load-shares static routes to one destination only when they tie on both administrative distance and priority. The FortiGate compares administrative distance first, then priority. Only routes that are equal on both values are considered truly equal-cost and are installed together in the FIB, which is what enables ECMP load-sharing across the next hops.

Why A is wrong: This is tempting because both fields matter, but a lower priority value wins outright, so the route with the better priority is installed alone and no load-sharing occurs.

Why B is correct: Equal-cost multipath installs multiple routes to one destination only when they tie on both administrative distance and priority, so the FortiGate treats them as equal and load-shares.

Why C is wrong: Distance is compared first, so a lower distance is preferred and only that single route enters the forwarding table, which prevents multipath rather than enabling it.

Why D is wrong: Matching prefix length is necessary but not sufficient; differing distances cause one route to be preferred, so the FortiGate cannot load-share across them.

Free sampleRoutinghard

An SD-WAN zone has two members, port1 and port2, both bound to a performance SLA that probes a remote server. An SD-WAN rule uses the Lowest Cost (SLA) strategy and lists port1 as the preferred member. While port1 still has link and a valid default route, its measured latency rises above the SLA target and stays there. What causes new sessions matching the rule to move to port2?

  • AThe static default route on port1 is withdrawn from the routing table the moment its latency exceeds the SLA target, so only port2 remains routable for the rule.
  • BFortiGate raises the administrative distance of port1's route above port2's, so the routing table now prefers port2 for all destinations on that interface.
  • Cport1 fails the SLA target and is marked as not meeting SLA, so it becomes ineligible for the Lowest Cost (SLA) rule, which then selects the next eligible member, port2. Correct
  • Dport1 is declared dead by link health monitoring as soon as one probe exceeds the latency target, and a dead member is never offered to any SD-WAN rule.
A Lowest Cost (SLA) rule excludes members that are alive but failing the SLA target, steering new sessions to an eligible member. Lowest Cost (SLA) selection works in two stages: filter to members currently meeting the SLA target, then choose the cheapest. A latency breach changes the member's SLA status to not meeting target, which removes it from the eligible set without deleting its route or declaring it dead, so the rule fails over to the next compliant member.

Why A is wrong: Tempting because failover feels like a routing change, but a performance SLA target breach does not remove the static route; the route stays in the table and port1 can still carry non-SD-WAN traffic. The SLA only affects SD-WAN rule eligibility.

Why B is wrong: Tempting because admin distance does select between routes, but a performance SLA does not rewrite administrative distance when a target is missed. Member selection inside an SD-WAN rule is driven by SLA status, not by editing route attributes.

Why C is correct: Correct. Lowest Cost (SLA) first filters members to those currently meeting the SLA target, then picks the lowest-cost member among them. A member that is alive but missing the target is excluded from selection, so the rule moves new sessions to port2.

Why D is wrong: Tempting because dead members are indeed excluded, but exceeding a latency target does not make a member dead. Dead status comes from missing the configured number of consecutive probe responses; a latency breach makes the member 'not meeting SLA' while it stays alive.

Free sampleRoutinghard

An administrator changes an SD-WAN load-balancing rule from Source IP to Source-Destination IP because a single busy host was overloading one member. Both members are healthy. What behaviour does the Source-Destination IP algorithm produce that the Source IP algorithm did not?

  • AEach session is assigned strictly in turn between the two members, guaranteeing an exactly equal session count on each link.
  • BAll traffic shifts to the member with the lowest measured latency in the performance SLA for that destination.
  • CThe host's existing sessions are rehashed and immediately rebalanced onto the less-used member to even out the load.
  • DSessions from the same host to different destination servers can be distributed across both members, because the hash includes the destination address. Correct
The Source-Destination IP algorithm hashes both addresses, so one host's sessions to different servers can spread across SD-WAN members. Source-Destination IP load balancing keys the hash on the pair of source and destination addresses. A single source talking to many destinations therefore produces many distinct hash values, allowing its sessions to land on different members and relieving the single-source concentration that pure Source IP hashing causes.

Why A is wrong: Strict round-robin alternation is the Sessions algorithm; a source-destination hash distributes by pair value and does not guarantee an exactly equal count.

Why B is wrong: Selecting on measured latency is the Lowest Cost (SLA) or Best Quality strategy, not a hashing algorithm; Source-Destination IP never reads SLA metrics.

Why C is wrong: Established sessions stay on their current member; the algorithm only affects how new sessions are placed, so nothing rehashes traffic already flowing.

Why D is correct: Adding the destination to the hash key means each source-destination pair can map to a different member, so a single host's traffic to many servers spreads across the zone instead of pinning to one link.

Other domains in this exam

See also the FCP-FGT-AD cert hub, the study guide, and the cheat sheet.

Examworthy is not affiliated with or endorsed by Fortinet. Original, blueprint-aligned practice material only.