Fortinet NSE4_FGT_AD-7.6 Practice Test 3
Question 20
Managed company browsers begin showing certificate warnings after FortiGate full SSL inspection is enabled for outbound HTTPS. Inspection otherwise works. Which remediation preserves the desired content control?
- Turn off HTTPS validation entirely on every employee browser.
- Import and trust the correct inspection CA certificate through managed endpoint trust configuration.
- Replace the remote server certificate with a new certificate from its public CA.
- Configure certificate-only inspection on every outbound policy so managed browsers avoid substitute certificates, abandoning the requirement to inspect decrypted content.
Explanation
Correct Answer
B:Full SSL inspection may present substituted server certificates signed by a FortiGate-controlled CA. Managed endpoints need the appropriate CA in their trusted stores; verify certificate-chain and hostname behavior after deployment. For certificate-pinned or exceptional destinations, assess narrow documented exemptions rather than disabling client validation broadly.
Incorrect Answers
A:Disabling browser certificate validation weakens authentication for all sites and allows impersonation risks far beyond the inspected traffic. It is not an acceptable way to handle a missing trusted inspection issuer.
C:A trusted upstream server certificate can be valid while the browser still distrusts the substitute certificate presented by a full-inspection firewall. Deploy and verify the specific FortiGate signing CA public certificate to managed endpoint trust stores; do not confuse that with the remote origin issuer.
D:Certificate-only inspection avoids some trust warnings by not decrypting payloads but removes content visibility the business explicitly wants to retain. Correct the managed client trust in the inspection CA and test representative HTTPS applications rather than globally reducing inspection depth.
Question 21
A cloud team has already steered client SYN packets through FortiGate-VM and confirms they match its inspection policy. The application server replies over a direct cloud-fabric route that bypasses the virtual firewall. The TLS session never completes even though inbound packet capture shows the requests. Which design change addresses the failure?
- Correct the application subnet and virtual-appliance route tables so both the request and return packets traverse the intended FortiGate-VM inspection path; validate session and NAT state.
- Review the application subnet security group ingress and return-port permissions, then allow the server response at that boundary.
- Enable session pickup and an additional IPS profile on the FortiGate-VM policy that already sees the client SYN.
- Increase the FortiGate TCP idle timer so the firewall treats missing return packets as though the three-way handshake had completed.
Explanation
Correct Answer
A:Stateful FortiGate inspection requires a valid return path for the session. The FortiGate already sees client ingress, so the missing step is to steer the server reply through the proper VM interface and ensure route tables, cloud forwarding permissions and translation state agree. Validate client SYN and return SYN/ACK with paired captures and session counters.
Incorrect Answers
B:Cloud security groups could block a TCP response, so inspecting them is a reasonable adjacent hypothesis. The scenario explicitly shows the server using a direct cloud-fabric return route, not an inspected return interface. Granting additional security-group permission would not steer the reply through the stateful FortiGate-VM path.
C:Session synchronization and IPS inspection have roles in HA and threat detection, but neither chooses the server subnet’s next hop. The return SYN/ACK still follows a direct cloud route that bypasses the VM. Resolve route symmetry and actual NAT/session path rather than adding unrelated inspection features.
D:A longer state timer can retain the original SYN entry, but does not deliver a SYN/ACK that cloud routing sends around the firewall. TCP session establishment and enforcement require the actual return traffic to traverse a valid path.
Question 22
A host-specific deny is installed below an existing broad HTTPS permit. Policy lookup for a new infected-host session still selects the broad permit. The operator plans to move the deny above the permit, but a previously established connection may remain open under cached session state. Which change and verification enforce the restriction without disrupting unrelated users?
- Add a second, narrower infected-host deny under the current broad HTTPS permit and clear the infected client’s existing session.
- Change the broad HTTPS permit to log all sessions and clear the infected host’s existing connection.
- Move the host-specific deny above the broad permit; test a NEW infected-host connection and, if the old connection persists, end only the matching established session so its next attempt is evaluated against the corrected deny.
- Assign a higher numerical policy ID to the host-specific deny and flush only that host’s cached session.
Explanation
Correct Answer
C:The specific deny is shadowed as long as the broad permit is checked first for new sessions. Move the deny above the general allow, then use a fresh connection to verify its policy hit and log action. Existing session-table entries may retain a previously accepted flow; identify and clear only that infected-host session if immediate revocation is necessary. Preserve unrelated existing business connections and confirm the expected deny ID for a new attempt.
Incorrect Answers
A:Clearing a cached connection forces a fresh policy lookup, but a narrower deny located below the matching broad permit is still shadowed for new traffic. The targeted restriction needs correct evaluation order before session cleanup; otherwise the next SYN continues to match the earlier permit.
B:New logging and a targeted session reset can help demonstrate the decision, but neither changes which ordered policy first matches the infected host. The broad permit remains above the host-specific deny, so the next connection is still accepted. Correct rule placement before using logs and session cleanup to prove enforcement.
D:It is tempting to treat a larger policy ID as stronger precedence, especially after noticing cached sessions. FortiGate normal firewall rule evaluation follows configured policy order, not numeric ID magnitude. Unless the deny actually moves above the broad permit, clearing the session simply reselects the earlier allow.
Question 23
An employee cannot open a legitimate vendor portal after a web-filter profile begins blocking its FortiGuard category. The policy permits HTTPS. What is the appropriate first investigation?
- Change the blocking FortiGuard category from Block to Monitor for the entire affected user group during vendor onboarding.
- Confirm the logged web-filter category, action and matching profile, then assess a narrowly scoped exception if the site is approved.
- Install a static URL-filter Allow entry for the portal host in the same web-filter profile and retain the category Block setting.
- Move the existing HTTPS firewall policy above other Internet rules so the portal traffic is matched earlier.
Explanation
Correct Answer
B:Confirm the exact FortiGuard rating, blocked profile, matched policy and user scope in logs before choosing any exception. For an approved individual portal, FortiOS web rating overrides can assign the site to an allowed local category without allowing unrelated destinations. A static URL-filter Allow alone does not defeat an otherwise blocking FortiGuard category; an Exempt action may also bypass other scanning. Retest the approved site and adjacent blocked categories.
Incorrect Answers
A:Monitor may allow users to reach sites while recording matches, but a broad category change also weakens enforcement for unrelated destinations. Because only one portal is reported, verify the actual category, matching profile and approved host before deciding whether any narrowly scoped override is justified.
C:A static URL Allow entry looks like a narrow exception, but FortiGuard category-based blocking can still apply; Allow does not necessarily bypass subsequent category enforcement. The investigator first needs the logged URL/category and actual web-filter action, then a deliberate local category override or other supported exception with known side effects.
D:Firewall policy ordering can change which security profile is applied, which makes this a plausible policy-level hypothesis. Here the matched HTTPS rule already permits traffic and the recorded block occurs in its web-filter category. Reordering policies without identifying the responsible profile would not prove or narrowly correct that category decision.
Question 24
Remote employees work on home networks. They must receive consistent corporate internet filtering without sending every device back through an on-premises FortiGate. Which service model most directly addresses that requirement?
- Deploy a branch FortiGate web-filter profile at headquarters, but configure no endpoint or site steering from remote home networks to that appliance.
- Distribute only an enterprise root CA to personal laptops.
- Install a split-DNS policy on the head-office LAN.
- FortiSASE policies with appropriate user or endpoint onboarding and traffic steering.
Explanation
Correct Answer
D:FortiSASE can provide centrally governed security controls for onboarded users away from the enterprise LAN. Configure supported endpoint or site connectivity, identity mappings and policy assignment, then test DNS, web traffic steering and logging for allowed and denied destinations across representative off-site networks. A deployed SASE client must actually forward eligible sessions through the intended inspection point.
Incorrect Answers
A:A central branch policy cannot inspect arbitrary home internet traffic that never traverses the deployed appliance. Remote users need an authenticated and steered path into an enforcement service, or a supported managed endpoint solution; merely configuring office firewall rules does not deliver that path.
B:Certificate trust is a dependency for certain inspection strategies, but installing a root CA alone does not establish a tunnel or secure web gateway to evaluate remote users’ sessions. FortiSASE requires client or site onboarding plus relevant policy assignment.
C:Split DNS can guide selected hostname resolution but does not by itself route and inspect every mobile internet session. Verify each user or device is onboarded, their web traffic goes through the intended security service and the appropriate identity/profile policy is enforced.
Question 25
An organization wants inbound public TCP 443 connections on 203.0.113.10 to reach a DMZ server at 10.4.0.20. The firewall must not expose all DMZ hosts. Which FortiGate configuration fits?
- Create a destination NAT virtual IP for the intended address and service, then reference it in a restrictive inbound policy.
- Enable source NAT on every internal LAN rule and assume inbound packets will find the server.
- Create a WAN-to-DMZ allow rule whose destination is the server private address, without a VIP or other valid public-to-private DNAT mapping.
- Create an unrestricted outbound policy with the DMZ network as source.
Explanation
Correct Answer
A:A FortiGate virtual IP can map the approved public address and port to a specific internal target. Pair the VIP with the matching WAN-to-DMZ security policy and limit allowed sources/services where appropriate. Verify routing to the internal server, the return path and which session fields show the pre- and post-DNAT destinations.
Incorrect Answers
B:Source NAT changes the source address of suitable outbound traffic; it does not publish a fixed internal server to the internet. The desired translation is on the destination of new inbound connections.
C:The external client connects to the specified public address, so an ordinary inbound rule targeting only the private server IP does not provide the needed destination translation. A virtual IP maps the public tuple to the intended DMZ service, and the matching firewall policy restricts access.
D:An outbound DMZ policy controls traffic initiated from the DMZ toward another zone. It neither publishes the public destination nor correctly limits new inbound TCP connections to one server. Publishing a server requires an explicit inbound destination and matching security rule.
Question 26
A new SD-WAN zone contains broadband and LTE with measured 23 ms versus 82 ms latency, respectively, and both are reachable. The business plans a voice service rule with a maximum 45 ms latency and bounded jitter while ordinary bulk traffic may use either link. Which policy design enforces the voice requirement?
- Select the lowest-cost link for voice irrespective of its measured 82 ms latency, without linking a performance SLA to the voice service rule.
- Use equal session-count load balancing for voice across both members.
- Define a voice-specific performance SLA with explicit latency/jitter thresholds and bind it to the matching SD-WAN service rule so only qualifying member paths carry new voice sessions.
- Permit voice in the firewall security policy and leave WAN member choice to an unrestricted default SD-WAN rule with no latency or jitter threshold.
Explanation
Correct Answer
C:Both WAN links can pass basic liveness checks, yet only broadband meets the current latency target. Configure a supported FortiOS health check with measured voice-relevant thresholds, attach it to the appropriate source/destination/service SD-WAN rule and preserve a separate strategy for bulk sessions. Validate rule order and actual member selection with fresh calls, including the intended fallback if neither link qualifies.
Incorrect Answers
A:Cost optimization may be appropriate for bulk traffic but cannot enforce the voice-specific maximum delay. The selected member must be tested against the configured latency/jitter SLA, not price alone.
B:Equal session splitting can place calls on the 82 ms LTE member despite the required 45 ms limit. Use an SLA-gated voice rule so only qualifying measured links are eligible.
D:The firewall accept policy authorizes the flow but does not itself select a performance-qualified WAN member. A dedicated SD-WAN service and health check must enforce voice quality.
Question 27
A FortiGate application-control profile blocks a broad collaboration category, including a sanctioned conferencing tool. Administrators must retain protection against other disallowed applications. What should they do?
- Change the entire collaboration application category action to Allow, removing blocks for every other application covered by that category.
- Replace the application sensor with a TCP/443-only allow rule.
- Identify the exact application signature and applicable rule, then create an approved narrow exception while monitoring other category traffic.
- Permit the conferencing service by one HTTPS FQDN object.
Explanation
Correct Answer
C:Inspect application-control logs to establish which signature and action classified the sanctioned service. Implement a suitable explicit exception at the correct profile or policy scope, then test both the allowed tool and known unwanted applications. Account for signatures, TLS/QUIC visibility and updates to service behavior.
Incorrect Answers
A:A broad category allow can restore the sanctioned tool but also admits unrelated prohibited collaboration software. Identify the actual blocked application signature and choose the narrowest sensor override or exemption while retaining other enforcement.
B:A transport-port allow cannot distinguish the approved conferencing tool from unrelated HTTPS or QUIC applications. It would undermine the existing application-specific controls instead of correcting the mistaken categorization. Distinct application signatures and policy actions are needed to preserve legitimate HTTPS uses.
D:One destination FQDN may fail to cover application-specific media endpoints and does not directly correct an application-control signature or action. Inspect the logged signature and appropriate policy rather than assuming transport destination identity is sufficient.
Question 28
Two FortiGate IPsec peers are configured for certificate-based IKE authentication instead of pre-shared keys. What must the administrator verify for trustworthy peer authentication?
- Confirm the peers accept the same IKE cipher proposal.
- That each peer presents a suitable certificate and the other validates its chain and expected identity.
- Install each peer certificate without its issuing CA chain, then rely on matching public IP addresses as the certificate trust decision.
- Disable IKE peer certificate validation while retaining encrypted ESP.
Explanation
Correct Answer
B:Each endpoint needs the appropriate trusted CA material, a usable private key and certificate, and identity checks consistent with the configured IKE peer. Confirm validity periods, certificate names and the negotiation log on both sides. Only after authentication succeeds should traffic selectors, protected routes and data policies be tested.
Incorrect Answers
A:A mutually supported cryptographic proposal lets IKE negotiate algorithms, but does not validate who owns the signing private key or whether the presented certificate chains to a trusted CA with the expected peer identity. Validate certificate trust and identity separately.
C:Certificate-based IKE needs a valid chain to the configured trusted issuer and the intended authenticated peer identity. Public address matching without issuing CA trust is not sufficient to validate a signed certificate; examine the actual IKE authentication logs.
D:ESP protects data after a negotiated security association, but encryption alone does not prove that an IKE peer presented a valid certificate for the expected identity. Disabling certificate validation undermines gateway authentication; restore CA and identifier checks.
Question 29
New workstations on the FortiGate port3 LAN receive no IPv4 lease, but a static-addressed test PC can reach port3. The business wants the firewall itself to allocate addresses, default gateway and DNS information to this subnet. Which configuration directly meets the requirement?
- Check whether a corporate DHCP server is reachable and configure port3 to relay workstation DHCPDISCOVER messages to that external server.
- Set port3 to obtain its own address by DHCP and leave workstation DHCP settings at automatic.
- Enable the FortiGate DHCP server on port3 with a nonoverlapping client pool, correct gateway/DNS options and suitable lease settings; then verify client leases.
- Review MAC-address reservations for existing employees and create static DHCP bindings for each known workstation on port3.
Explanation
Correct Answer
C:FortiGate can serve DHCP directly on a selected interface. Choose a pool inside the connected port3 subnet but outside reserved static addresses, assign the appropriate gateway and DNS options, and set a deliberate lease period. Renew a client lease and inspect the assigned address, gateway and server binding.
Incorrect Answers
A:DHCP relay is a supported way to serve clients from a remote address allocator; it does not make the FortiGate itself allocate leases, which the requirement explicitly demands. Configure FortiOS DHCP server mode on port3 with a correct pool, gateway and DNS options, then inspect the client lease.
B:Interface DHCP client mode helps the FortiGate receive its own leased address from another upstream server. It does not provide lease offers to attached workstations. For these users, the FortiGate must operate a DHCP server on the port3 LAN with an appropriate on-subnet allocation range and network options.
D:A reservation identifies addresses assigned by an operating DHCP server and can improve consistency for recognized devices. Reservations alone do not configure the missing local server service or provide leases for new, unlisted workstations. Enable the port3 server and range first, then optionally add reservations.
Question 30
On a FortiGate model with more than 2 GB RAM running FortiOS 7.6.0, SSL VPN uses an external RADIUS server for user authentication. Clients receive timeout errors, but the same users can authenticate directly to the identity service. The administrator must distinguish network silence from credential rejection. What should be checked first?
- Check FortiGate-to-RADIUS reachability and UDP service/return traffic, configured server address and shared secret, and server logs to establish whether an Access-Accept, Access-Reject or no response occurred.
- Test one known valid directory credential directly on the RADIUS server and reset failed user passwords.
- Extend the SSL VPN portal authentication window and increase idle timeout for the affected users.
- Temporarily change the SSL VPN portal to a local test-user group and rerun the login from the same client.
Explanation
Correct Answer
A:A timeout can arise when requests or replies never arrive, the server is misaddressed or the shared secret/protocol configuration prevents a usable response. Correlate a bounded FortiGate test and RADIUS debug with server-side accounting and firewall reachability, then distinguish an explicit reject from silence before changing user credentials or policy.
Incorrect Answers
B:Direct RADIUS authentication tests can reveal credential issues, but the report is a FortiGate-side timeout and users already authenticate directly to the identity service. Resetting accounts cannot distinguish absent UDP replies, misconfigured client secret or transport reachability. Compare FortiGate request/response behavior with server logs first.
C:A longer client login window can be relevant where interactive authentication times out after slow prompts, but it does not establish why the FortiGate RADIUS client receives no usable reply. The server address, shared secret, request and response packets need to be examined to distinguish silence from an Access-Reject.
D:A local-account test can isolate part of the VPN portal, but it deliberately bypasses the RADIUS path that is failing and cannot diagnose the server reply or shared-secret mismatch. The requested decision is to distinguish transport timeout from explicit RADIUS rejection using both ends’ logs and packet evidence.