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

Deployment and system configuration

Deployment and system configuration 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

71
Practice questions
4
Objectives
18
Easy
36
Medium
17
Hard

Where people go wrong

  • Crediting the Security Fabric with FortiManager's work

    The wrong answers hand the Security Fabric jobs that belong to other products. Pulling a full policy package, pushing firmware and editing policy on every member appear as fabric features, and joined members are assumed to share admin passwords, local users and routing tables. The Fabric supplies visibility, topology, Security Rating and synchronisation of selected objects such as an address group. FortiManager manages configuration and firmware as a fleet, and FortiAnalyzer stores historical logs.

    See it in a sample question
  • Treating the fabric root as elected or promotable

    Promoting the downstream unit to root, electing the root by model throughput or priority, and making the root a forced data path all appear as wrong options. A fabric has exactly one root, set by configuration at the top of the tree. Downstream units join by pointing at the upstream device, and a downstream unit can have members of its own beneath it. The root aggregates telemetry and is typically where the fabric connects to FortiAnalyzer.

    See it in a sample question
  • Giving session pickup other HA jobs

    Session pickup is offered as the answer to link-failure detection, per-unit management access, configuration synchronisation and primary election. It does none of these. It synchronises the session table so established connections survive a failover. Monitoring the WAN interface triggers failover on link loss, a reserved HA management interface gives each member its own management address, matching configuration checksums govern sync, and device priority with override decides which unit becomes primary.

    See it in a sample question
  • Choosing a diagnostic command that answers a different question

    The wrong answers pick a real command that reports something adjacent. Session statistics show counts but never name the process using CPU, and a CPU and memory summary does not show how full the session table is. The sniffer shows packets on the wire, not whether a session was created. diagnose sys top ranks processes, diagnose sys session list shows the policy id and interfaces a session is bound to, and diagnose sys session stat gives the count against the table limit.

    See it in a sample question
  • Swapping one administrator access control for another

    An administrator profile, two-factor authentication or an interface binding is offered to restrict the login source, and a trusted host is offered to add a second factor. A pre-login banner or password expiry is offered for idle logout, and interface role or captive portal for GUI access. Trusted hosts limit source networks, two-factor authentication adds the one-time code, the idle timeout closes inactive sessions, and the allowaccess list must include HTTPS on the interface.

    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 sampleDeployment and system configurationmedium

A root FortiGate and its downstream FortiGate are members of the same Security Fabric. The administrator needs an object created on the root, such as an address group, to be usable in policies on the downstream device without recreating it manually. Which fabric capability satisfies this requirement?

  • APolicy package installation, where the downstream FortiGate pulls the root's entire firewall policy table on a schedule.
  • BLog forwarding, which copies the object definitions to FortiAnalyzer so the downstream can import them from logs.
  • CSD-WAN rule replication, which distributes the root's path-selection rules and any referenced objects to members.
  • DFabric synchronisation of selected objects, so the root can share configuration objects like addresses with downstream members. Correct
Recognise that the Security Fabric can synchronise selected configuration objects from the root to downstream members. Centralised control in the fabric includes the ability to synchronise certain configuration objects, such as addresses and address groups, from the root FortiGate to downstream members. This avoids duplicating objects manually and keeps shared definitions consistent across the fabric.

Why A is wrong: Pulling a full policy package describes FortiManager-style management, not native fabric object sharing; the requirement is to reuse a single object, not to overwrite the downstream policy table.

Why B is wrong: Log forwarding moves event data to FortiAnalyzer for analytics; logs do not carry reusable configuration objects, so the downstream cannot build an address group from forwarded logs.

Why C is wrong: SD-WAN rules govern path selection on a single device and are not the mechanism for sharing a generic address object across the fabric, so this misapplies an unrelated feature.

Why D is correct: The Security Fabric can synchronise certain configuration objects from the root to downstream members, letting an address group defined once on the root be referenced in downstream policies without manual recreation.

Free sampleDeployment and system configurationmedium

Two FortiGates are connected back to back. The administrator wants the lower device to become a downstream member of the fabric anchored by the upper device. On the downstream FortiGate, which setting establishes its place in the fabric tree?

  • ASet the downstream FortiGate to join the fabric and specify the upstream FortiGate's interface address as its upstream connection. Correct
  • BSet the downstream FortiGate's fabric role to root and point all member FortiGates at it as the new top of the tree.
  • CSet the downstream FortiGate to use the same fabric name but leave the upstream field empty so it discovers the root by broadcast only.
  • DSet the downstream FortiGate into HA standby for the upstream device so it inherits the fabric configuration on failover.
Configure a downstream FortiGate by pointing its upstream connection at the root's fabric interface to join the fabric tree. The fabric tree is built top down. A downstream FortiGate becomes a member by being told which upstream device to connect to, identified by the upstream FortiGate's fabric interface address. The root then authorises it and shows it beneath itself in the topology.

Why A is correct: A downstream member joins by being configured to connect to the upstream device's fabric interface, which places it below the root and lets the root authorise and display it in the topology.

Why B is wrong: Promoting the lower device to root inverts the intended hierarchy; only one root anchors a fabric, and here the upper device is meant to be that anchor, so the lower unit must not be root.

Why C is wrong: Matching the fabric name alone without defining the upstream connection leaves the device with no anchor; the join relies on pointing at the upstream interface, not on name-only broadcast discovery.

Why D is wrong: HA pairs two units into one logical device for redundancy; it does not make one a downstream fabric member of the other, so this confuses high availability with fabric membership.

Free sampleDeployment and system configurationmedium

An administrator wants the FGCP cluster to fail over automatically if the WAN-facing interface on the primary unit loses its link, but is unsure how to make the cluster react to that specific port. Which configuration achieves this?

  • AEnable session pickup so the cluster tracks the WAN interface state and migrates sessions when the link drops.
  • BRaise the device priority on the WAN interface so the cluster favours whichever unit has that link up.
  • CConfigure a second heartbeat interface on the WAN port so heartbeat loss there forces the cluster to switch units.
  • DAdd the WAN interface to the HA monitored interfaces list so a link failure on it lowers the unit's score and triggers failover. Correct
Know that adding an interface to HA monitored interfaces makes a link failure on it lower the unit score and trigger an FGCP failover. HA interface monitoring lets the administrator nominate critical interfaces. FGCP factors a monitored interface going down into each unit's failover score, so if the monitored WAN link drops on the primary, its score falls below the secondary and the cluster fails over to keep that connection available.

Why A is wrong: Tempting because session pickup is an HA feature, but it only synchronises session tables for smoother failover; it does not monitor interface link state or trigger a failover when a port goes down.

Why B is wrong: Plausible because priority influences primary selection, but device priority is a static administrative preference and is not evaluated per interface link state, so it does not cause an automatic failover on link loss.

Why C is wrong: This confuses heartbeat with port monitoring; heartbeat interfaces detect peer reachability between cluster members, not the up or down state of a data interface like the WAN port.

Why D is correct: Interface monitoring (port monitoring) adds the chosen interface to the HA health calculation, so a link-down event on the monitored WAN port reduces that unit's cluster score and causes the cluster to fail over to the healthy unit.

Free sampleDeployment and system configurationhard

Users complain of intermittent slowness, and the FortiGate dashboard shows CPU spiking. An administrator needs to identify which running process is consuming CPU in real time and watch it update, similar to a Linux top view, before deciding whether IPS or the management daemon is responsible. Which command provides this live per-process CPU and memory breakdown?

  • Adiagnose sys top Correct
  • Bget system performance status
  • Cdiagnose sys session stat
  • Dget system status
Select the FortiGate command that gives a live per-process CPU and memory ranking to attribute high load to a specific daemon. diagnose sys top behaves like the Linux top utility, continuously listing processes sorted by resource consumption so the administrator can attribute CPU spikes to a specific process such as the IPS engine, which summary or session-count commands cannot do.

Why A is correct: This command lists running processes ranked by CPU and memory usage and refreshes continuously like top, letting the administrator see in real time whether the IPS engine or another daemon is consuming the CPU.

Why B is wrong: This gives a useful one-time summary of CPU, memory, and session averages over the last minute, but it does not break the load down per process and does not refresh live, so it cannot pinpoint the offending daemon.

Why C is wrong: Session statistics report counts such as active and ephemeral sessions and setup rate, which can correlate with load but never attribute CPU consumption to a named process, so it cannot identify the culprit daemon.

Why D is wrong: This returns firmware, serial number, and licence information about the unit, with no live process or resource view at all, making it irrelevant for diagnosing which process is burning CPU.

Free sampleDeployment and system configurationeasy

A new administrator account is created on a FortiGate for a contractor who will only ever connect from a single management workstation on the 10.20.30.0/24 subnet. Which configuration restricts that account so it can authenticate only from that subnet?

  • ASet a trusted host of 10.20.30.0/255.255.255.0 on the administrator account. Correct
  • BAssign the account the prof_admin profile so its privileges are limited to a predefined set of system features.
  • CEnable two-factor authentication on the account so a token is required from the contractor.
  • DBind the account to the management interface so it can only connect through that physical port.
Use trusted hosts to restrict an administrator account to logging in only from specified source networks. Trusted hosts are evaluated against the source IP of an administrative login attempt; if the source falls outside every configured trusted host the FortiGate silently drops the connection, which is why a 10.20.30.0/24 trusted host confines the account to that subnet.

Why A is correct: Trusted hosts define the source IP networks from which the account may authenticate, so configuring 10.20.30.0/24 as the trusted host limits logins to that subnet exactly as required.

Why B is wrong: An administrator profile controls which features the account may read or write, but it places no restriction on the source address an admin may log in from, so it does not meet the requirement.

Why C is wrong: Two-factor authentication strengthens proof of identity, but it permits the admin to log in from any reachable network, so it does not confine the login source to one subnet.

Why D is wrong: There is no per-account binding to a physical interface; administrative access is governed by interface allowaccess and per-account trusted hosts, so this option describes a control that does not exist in FortiOS.

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.