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

Firewall policies and authentication

Firewall policies and authentication 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

74
Practice questions
4
Objectives
19
Easy
38
Medium
17
Hard

Where people go wrong

  • Confusing source NAT with destination NAT

    The wrong answers swap the two translation directions. They credit an IP pool, or the NAT toggle on a policy, with publishing an inbound server, and they describe source NAT as if it could make a virtual IP work. A virtual IP performs destination NAT: once it is the destination address of a wan1-to-LAN policy, the FortiGate rewrites the public destination to the internal server. An IP pool and the NAT toggle only translate the source of outbound traffic.

    See it in a sample question
  • Mixing up IP pool types

    The wrong options attach the behaviour of one pool type to another. Overload shares addresses through port translation, so many hosts sit behind each address. One-to-one gives each internal host a whole public address with port translation disabled, so extra hosts get no mapping once the addresses are used. Fixed-port-range maps an internal range to an external range and gives each internal address a calculated port range. Port block allocation assigns port blocks and does not pin a host to one address.

    See it in a sample question
  • Expecting specificity or action to override list position

    The wrong answers treat the policy list like a routing table. They pick the more specific entry, give a deny action priority, or rely on policy ID or an identity-first pass. FortiGate reads the list from top to bottom and applies the first match, so a broad policy placed above a specific one shadows it. The Central SNAT table follows the same rule: a /16 rule above a /24 rule wins until the /24 rule is moved higher.

    See it in a sample question
  • Misplacing FSSO components and modes

    The wrong options move FSSO work to the wrong component or mode. DC agent mode needs the dcagent DLL on every monitored domain controller, and the DC agent forwards logons only to the collector agent, never to the FortiGate. The collector agent, running on a member server, aggregates events, resolves groups and is the only component that talks to the FortiGate. Agentless polling needs nothing on the domain controllers, and RSSO reads RADIUS accounting records instead of Windows logons.

    See it in a sample question
  • Treating passive FSSO as a login prompt

    The wrong answers blur passive and active authentication. FSSO learns identity from Active Directory logon events and never displays a prompt, so offering it to produce a login page fails. Active authentication needs a captive portal on the interface or policy, an accept policy with a user group, and a protocol the FortiGate can intercept, such as HTTP or HTTPS. Captive portal and two-factor both still prompt the user, so neither meets a requirement for no credential prompt.

    See it in a sample question

Objectives in this domain

Sample questions from this domain

Free sampleFirewall policies and authenticationmedium

An administrator publishes an internal web server (10.10.1.50:443) to the internet using a virtual IP that maps the WAN interface IP to the server on the same port. A firewall policy from wan1 to the internal LAN uses this VIP as the destination address with the action set to ACCEPT. Which translation does the FortiGate apply to the inbound traffic, and where in the policy is it configured?

  • ASource NAT, configured by enabling the NAT toggle and selecting an IP pool in the same firewall policy.
  • BDestination NAT, performed automatically by the virtual IP referenced as the destination address of the firewall policy. Correct
  • CDestination NAT, performed only after the NAT toggle is enabled and central NAT mode is turned on for the VDOM.
  • DSource NAT, performed by a static one-to-one IP pool that maps the server address to the WAN interface address.
Recognise that a virtual IP referenced as a firewall policy destination performs destination NAT automatically, independent of the NAT toggle. A virtual IP is FortiGate's DNAT object. When it is used as the destination address of an ACCEPT policy, FortiOS rewrites the packet's external destination IP and port to the mapped internal address before forwarding, so the inbound publishing happens through the VIP itself and does not depend on the source NAT toggle.

Why A is wrong: SNAT rewrites the source address of outbound sessions and is enabled with the NAT toggle plus an optional IP pool; this is tempting because both NAT types are policy-related, but inbound publishing of a server uses destination translation through the VIP, not source translation.

Why B is correct: A virtual IP is a DNAT object; once it is selected as the policy destination, the FortiGate rewrites the external destination address and port to the mapped internal server, so the translation is inherent to the VIP rather than the NAT toggle.

Why C is wrong: The NAT toggle governs source translation and central NAT changes where SNAT and DNAT rules live, not whether a VIP works; a VIP referenced as a destination performs DNAT regardless of the NAT toggle, making this a plausible but incorrect conflation.

Why D is wrong: A one-to-one IP pool does translate addresses, but it operates on the source of egress traffic; using it to describe inbound server publishing reverses the direction of translation, which is the common error this distractor captures.

Free sampleFirewall policies and authenticationmedium

An administrator configures an overload IP pool containing a small contiguous range of public addresses and applies it to an outbound firewall policy with NAT enabled. A security partner requires that each internal host always present the same single public address for the life of every session and across reconnects, while still allowing many hosts to share the pool. Which IP pool type meets this per-host consistency requirement?

  • AOverload, because it already maps each internal host to exactly one fixed public address chosen at session start.
  • BOne-to-one, because it maps each internal address to a dedicated public address with no port translation, keeping the mapping stable per host. Correct
  • CFixed port range, because it reserves a block of source ports per host so the public address never has to change.
  • DPort block allocation, because it hands each host a dedicated block of ports while rotating the host across pool addresses for capacity.
Select a one-to-one IP pool when each internal host must consistently present a dedicated public source address. A one-to-one IP pool maps internal addresses to public addresses on a fixed basis without port address translation, so a given host always egresses with the same dedicated public IP. Overload, fixed port range, and port block allocation share addresses across hosts or rotate them, breaking the per-host consistency the requirement demands.

Why A is wrong: Overload uses all pool addresses with port translation and does not pin a host to one address, so successive sessions from the same host may use different pool members; this is tempting because overload is the default many-to-one pool but it fails the consistency requirement.

Why B is correct: A one-to-one IP pool gives each internal host a dedicated public address and performs no port overloading, so the same host consistently presents the same public IP across sessions and reconnects, which is precisely what the partner requires.

Why C is wrong: A fixed port range pool constrains which source ports a host may use, which helps with predictable port mapping, but it does not guarantee a stable one-to-one mapping of host to public address, so it does not satisfy the stated requirement.

Why D is wrong: Port block allocation assigns port blocks to scale logging and capacity, but it can still place a host on different pool addresses, so the public address a host presents is not guaranteed to stay constant across sessions.

Free sampleFirewall policies and authenticationeasy

An administrator has several firewall policies between the same interface pair. A broad policy that matches a wide range of traffic sits above a more specific policy. Why is the more specific policy below it never used?

  • AFortiGate evaluates policies from bottom to top, so the lower specific policy is read last and discarded.
  • BFortiGate applies the policy with the highest numeric ID regardless of its position in the list.
  • CFortiGate evaluates policies in order from top to bottom and applies the first match, so the broad policy above consumes the traffic first. Correct
  • DFortiGate merges overlapping policies into one combined rule, so the specific policy is absorbed into the broad one.
FortiGate matches firewall policies top to bottom and applies the first match, so order determines which policy is used. Because policy lookup stops at the first match in the top-to-bottom list, a broad policy positioned above a specific one captures the traffic and the specific policy never applies.

Why A is wrong: The direction is reversed here; FortiGate reads policies top to bottom, not bottom to top, so this misstates the lookup order.

Why B is wrong: Policy ID is an identifier assigned at creation and does not determine evaluation order, so relying on it is a common misconception.

Why C is correct: Policy lookup proceeds top to bottom and stops at the first match, so a broad policy placed higher will match traffic before the specific policy below is ever reached.

Why D is wrong: FortiGate does not merge policies; each is evaluated independently in sequence, so this invented behaviour does not occur.

Free sampleFirewall policies and authenticationhard

A network administrator is deploying FSSO in DC agent mode for a Windows Active Directory domain with three domain controllers. Each user logon must be captured and forwarded to the collector agent so that the FortiGate can apply identity-based policies. Which component must be installed on each monitored domain controller for this mode to function?

  • AA DC agent (dcagent.dll) registered on every monitored domain controller, which intercepts logon events and forwards them to the collector agent. Correct
  • BA collector agent installed directly on every domain controller, so each controller independently sends user-to-IP mappings to the FortiGate.
  • CA TS agent on each domain controller to track per-session source ports for users sharing one host.
  • DThe FortiGate polling service enabled against each controller, removing the need for any agent on the controllers.
Identify that FSSO DC agent mode requires a DC agent installed on each monitored domain controller to capture logon events. In DC agent mode the dcagent.dll is loaded on each domain controller and intercepts user logon events in real time, forwarding them to the collector agent, which then sends consolidated user-to-IP-to-group mappings to the FortiGate. This per-controller agent is what separates DC agent mode from agentless polling.

Why A is correct: DC agent mode requires the FSSO DC agent to be installed on each domain controller; it hooks logon events at the source and pushes them to the collector agent, which is the defining characteristic of this mode.

Why B is wrong: It is tempting because the collector agent is central to FSSO, but in DC agent mode the collector is a single (or redundant) service that aggregates events from DC agents, not a per-controller component that talks to the FortiGate alone.

Why C is wrong: The TS agent is used for Citrix or Terminal Server environments to distinguish users on a shared host by port range, not for capturing standard domain controller logon events.

Why D is wrong: This describes agentless polling mode rather than DC agent mode; the question explicitly specifies DC agent mode, which depends on an installed DC agent rather than FortiGate-initiated polling.

Free sampleFirewall policies and authenticationmedium

An administrator builds a firewall policy that requires users to authenticate before their HTTP traffic is allowed out to the internet. Users report that they are never prompted for credentials and the traffic is simply blocked. The policy has a user group assigned and the action is set to accept. Which configuration detail is required so that unauthenticated HTTP sessions trigger an active authentication prompt?

  • AEnable a captive portal on the ingress interface or enable disclaimer-and-portal so the FortiGate can intercept the HTTP session and present a login page. Correct
  • BDisable the explicit web proxy on the ingress interface so that the FortiGate falls back to transparent captive portal prompting for the policy.
  • CChange the policy action to deny and add an authentication exemption so the FortiGate prompts before applying the deny verdict to each new flow.
  • DAdd an FSSO collector agent connector to the policy so that the agent injects the credentials and removes the need for any interactive prompt.
Understand that active firewall authentication requires a captive portal to intercept HTTP and present a login page. Firewall policies that reference a user group rely on active authentication, which is delivered through a captive portal; without the portal enabled on the interface or policy the FortiGate has no way to intercept the HTTP session and prompt for credentials, so matching traffic is dropped.

Why A is correct: Correct. Active authentication on a policy needs a captive portal mechanism, typically enabled on the ingress interface or via the policy, so the FortiGate can intercept the supported HTTP session and serve the login page.

Why B is wrong: Tempting because proxy modes affect prompting, but disabling a proxy that may not even be configured does not enable the policy-driven captive portal and does not address why no prompt appears.

Why C is wrong: Tempting because action and authentication feel linked, but a deny policy drops traffic and never authenticates, and an exemption removes authentication rather than enabling a prompt.

Why D is wrong: Tempting because FSSO authenticates users, but FSSO is passive single sign-on that maps existing logins and never displays a prompt, so it does not produce the requested active login page.

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.