Fortinet NSE4_FGT_AD-7.6 Practice Test 31
Question 300
During a burst of archived downloads, FortiGate enters conserve mode and AV workers consume unusually high memory. The operator proposes disabling all malware scanning. What should be examined?
- Correlate AV worker memory and archive limits with conserve thresholds, then apply a narrowly tested inspection or capacity adjustment.
- Check current proxy antivirus archive recursion depth and raise the decompression limit to the model’s supported maximum.
- Raise the proxy antivirus archive-scanning size and buffer limits during the memory-conserve incident.
- Inspect the FortiGate conserve-mode alarm threshold and increase its notification severity during large archived downloads.
Explanation
Correct Answer
A:Large archive inspection can increase proxy buffering and antivirus process memory, potentially entering FortiOS conserve mode. Before disabling security, inspect process and memory diagnostics, active AV profiles, protocol oversize handling and download size patterns. Test a safely constrained size/capacity adjustment or appropriate policy scope, confirm memory returns below conserve thresholds and preserve detection on unrelated traffic.
Incorrect Answers
B:Archive-depth configuration can affect inspection completeness, but raising it under measured AV worker memory pressure can increase buffering and worsen conserve mode. Compare actual worker and session memory, supported limits and oversize dispositions before choosing narrowly bounded tuning or extra capacity; retain essential malware detection.
C:Larger buffering or scanning limits can increase memory demand during large or nested archive inspection. Before raising them, profile actual AV worker consumption and conserve thresholds, assess oversize/decompression actions and test a scoped change that preserves malware controls while remaining within model capacity.
D:Alarm severity influences notification, not process allocation or the memory thresholds that protect packet processing. The incident already shows unusually high AV worker memory under archive load. Correlate the worker, file handling limits and conserve state, then select a tested workload or profile correction without suppressing visibility.
Question 301
High-volume IPS logs overwhelm local storage during a scanning burst. The SOC must preserve actionable critical alerts without losing all visibility. Which approach is safest?
- Disable the currently noisy IPS sensor on its matched rule during the scan burst, leaving logging enabled only for forwarding traffic and administrator events.
- Record only administrator login events and stop all IPS telemetry because it has too many duplicate signature matches.
- Raise every IPS log’s severity to Critical in FortiAnalyzer so the SOC can retain fewer categories without inspecting the underlying duplicate signature or session evidence.
- Preserve critical IPS events centrally and tune only proven-noisy signatures after correlating repeated hits without disabling prevention.
Explanation
Correct Answer
D:During an IPS alert flood, preserve signature ID, severity, enforcement action, source/destination and policy ID in centrally retained security events. Use FortiAnalyzer or another supported remote collector, correlate repeated scans and investigate an actual noisy sensor before narrowing log filters. Retest that critical block events remain searchable and detection remains enabled.
Incorrect Answers
A:Sensor removal can reduce log volume but sacrifices prevention precisely during suspicious scanning. Preserve critical security evidence centrally and tune only corroborated high-noise detections after investigating actual signatures.
B:Login events do not describe exploit attempts. Use filtered aggregation, rate thresholds and suitable centralized retention instead of eliminating the required threat evidence. Authentication events cannot reconstruct exploit signatures, source/destination flows or IPS actions; keep relevant security telemetry.
C:Relabeling severity does not create storage capacity or distinguish repeated events from important prevention evidence. Filter/aggregate appropriately while preserving source, signature, action and correlation fields.
Question 302
An AWS FortiGate-VM deployed under BYOL is being replaced after an instance failure. A copied disk image boots on the new instance, but FortiGuard services are inactive and the replacement instance type has not been checked for support. What must be established before this replacement can safely resume production inspection?
- Import the original instance’s operating-system image and recover the previous security policy backup without activating BYOL on the replacement.
- Inspect replacement EC2 CPU and RAM headroom and increase the instance size before restoring the original FortiGuard inspection profiles.
- Review whether the replacement belongs to the original AWS account and Fortinet support organization, then use that account association as license proof.
- Validate the replacement FortiGate-VM BYOL activation, licensed security entitlements and supported AWS instance compatibility before restoring production traffic.
Explanation
Correct Answer
D:Fortinet requires a suitable BYOL license to activate a FortiGate-VM deployment, and valid FortiGuard service entitlements are separate from seeing a VM boot. A failed original instance does not establish that the replacement has an active eligible license, correct subscription services or supported instance sizing. Confirm license activation and the relevant contract, supported platform capacity and update reachability for the replacement before restoring the inspected production path. A migration between on-demand and BYOL licensing types is not a same-instance toggle; Fortinet documents a separate deployment path for that change.
Incorrect Answers
A:Bootstrapping a previously configured image or restoring policy can save time, but neither proves that the new FortiGate-VM instance has the valid activated license required for security services. BYOL activation, entitlement and supported instance properties must be checked on the replacement itself before it is treated as production-ready.
B:Adequate virtual resources are necessary for capacity, but the scenario explicitly reports inactive FortiGuard services under BYOL and unverified instance compatibility. Increasing size alone neither transfers the entitlement nor activates the new virtual serial. Confirm approved platform, license activation, contracts and policy behavior.
C:Account membership can be an administrative prerequisite for obtaining a license, but it is not proof the replacement FortiGate-VM has a valid activated BYOL license or security-service entitlement. Check the new instance and virtual serial against Fortinet support records and test actual FortiGuard status.
Question 303
A central SNAT policy was intended for finance clients on port4, but traffic arrives through finance-vlan and a general SNAT rule matches. Where should the administrator inspect?
- Change the Central SNAT finance rule ingress selector from port4 to finance-vlan while preserving its approved public pool.
- Move the Finance Accept policy above the catch-all and retain Central SNAT’s incoming selector port4 while Finance sessions now ingress finance-vlan.
- Add a new VIP for finance-vlan so destination translation rewrites the outgoing source interface condition.
- Expand the finance public SNAT pool with additional addresses without correcting the rule’s obsolete source ingress-interface match.
Explanation
Correct Answer
A:The finance traffic enters on finance-vlan, but the desired central SNAT mapping is restricted to port4, so it does not match and a more general entry translates the flow. Compare ingress/egress selectors, source/destination objects and rule precedence, then correct the specific source interface in the relevant map and verify fresh sessions use the intended public IP.
Incorrect Answers
B:An earlier forwarding rule does not alter which Central SNAT map matches. This translation rule still expects port4, so the new VLAN source cannot hit it; repair the SNAT ingress selector.
C:VIPs perform DNAT and cannot replace the central SNAT source-interface match. Check the actual session ingress, outgoing interface and map order. A destination VIP does not change which ingress interface the central source-NAT rule sees.
D:Increasing addresses can help capacity only after a source session reaches the correct map. The packet now arrives through finance-vlan, so the wrong port4 selector must be fixed first.
Question 304
An operator disables one SD-WAN WAN member during a failover rehearsal. Some existing sessions drop while new flows select backup. What should the test conclude?
- Enable session pickup in FGCP and expect an existing public-NATed application flow to continue on the backup ISP.
- Disable all health checks after the rehearsal because they were responsible for switching traffic away from the down member.
- Check that the standby WAN member reports link-up and count the rehearsal successful once new connection tests pass.
- Test backup egress for new sessions separately from established sessions whose NAT source and return path may change; capture translated source and server return path before claiming session continuity through failover.
Explanation
Correct Answer
D:SD-WAN health and route selection can redirect new flows to a healthy backup while existing connections fail if their egress IP, NAT mapping or stateful path changes. The observed drops do not by themselves prove the failover route is wrong. Inspect member health, selected rule, session/NAT state and application reconnect behavior under a controlled outage, documenting any required session-recovery expectations.
Incorrect Answers
A:FGCP session pickup is an HA feature between synchronized cluster members, not an automatic continuity guarantee when a standalone SD-WAN circuit changes its public source address. Existing return packets may target the old NAT tuple even while new backup sessions succeed. Test these behaviors separately.
B:Health checks provide controlled failure detection. Removing them can leave real outages invisible; fix any false targets and validate expected steering. Health checks are needed to detect member failure; disabling them after a failover test sacrifices resilience.
C:A healthy backup interface and working new flows prove only partial failover success. Previously established sessions can fail because the public source, return route and remote peer state changed. Capture the existing-session tuple and test application continuity separately before claiming no interruption.
Question 305
A site-to-site IPsec tunnel rekeys a child SA. Shortly afterward, a few inbound ESP packets are logged with an unknown Security Parameter Index, even as fresh protected sessions mostly work. Which evidence should the engineer compare before rotating credentials or disabling replay checks?
- Inspect the IKE authentication records and rotate the remote peer PSK after the child-SA rekey that generated the Invalid SPI events.
- Correlate inbound Invalid SPI events with child-SA rekey times and old/new SPI packet captures before declaring desynchronization.
- Disable anti-replay checks globally to suppress invalid-SPI messages.
- Review the established IKE phase-one association and suppress Invalid SPI logging for an otherwise reachable peer.
Explanation
Correct Answer
B:Correlate the invalid SPI events with IKE rekey timestamps, inbound/outbound child SAs and delayed packets from the prior SA. Intermittent post-rekey messages can be stale traffic; changing encryption or clearing the whole VPN without evidence could worsen availability.
Incorrect Answers
A:An Invalid SPI after a rekey can result from late ESP packets referencing a retired child-SA SPI, even when IKE authentication remains valid and new sessions work. Changing PSKs does not explain old-versus-new SPI timing. Correlate the inbound sequence with the installed child-SA lifetimes and counters.
C:Anti-replay and SA integrity controls protect against different attacks. Broadly disabling them is not justified by a single lifecycle log and does not prove the peers agree on current keys.
D:Successful phase-one authentication does not prove that every received ESP SPI maps to an installed child SA. An unknown SPI must be interpreted against rekey timing, packet capture and current SA tables before deciding whether it is benign late traffic or a real synchronization fault. Do not hide evidence prematurely.
Question 306
An internal application requires mutual TLS, with the client presenting a certificate to the server. Deep inspection breaks the handshake. Which feasibility check should guide the design?
- Examine the mTLS client-certificate exchange between endpoint and origin through the inspection path; preserve origin authentication and use only a documented, approved destination-scoped decryption exception if interception is incompatible.
- Review the TLS inspection profile’s server-certificate handling while preserving transparent decryption and testing only its downstream trust chain.
- Disable the server's mTLS requirement because the FortiGate already authenticates its own administrator.
- Deploy the FortiGate inspection CA to the application clients and renew their server trust before testing mutual TLS again.
Explanation
Correct Answer
A:Mutual TLS includes client-certificate authentication; full SSL interception may break the handshake unless supported exceptions or an approved alternative preserve mutual trust. Mutual TLS authenticates a client certificate at the origin, so middlebox decryption may interrupt this end-to-end handshake. Compare supported FortiOS 7.6 handling, the actual TLS failure and an approved per-app inspection exception rather than assuming the client identity can be transparently reissued.
Incorrect Answers
B:Certificate validation is important, but mutual TLS also requires the origin server to authenticate a client certificate. A proxy that intercepts the handshake may be unable to transparently forward the client credential as the server expects. Analyze the client-authentication handshake and supported exception design before declaring ordinary CA trust sufficient.
C:Administrative authentication of the firewall is not equivalent to application-level client authentication. Removing mTLS changes the application's trust model and does not solve inspection compatibility safely. FortiGate administrator authentication does not substitute for the application server verifying the client TLS certificate.
D:A trusted interception CA can solve conventional server-certificate warnings for managed endpoints, but the stated failure is the client-certificate authentication exchange through an intermediary. Installing a root CA does not necessarily preserve the original client key proof at the origin. Verify the mTLS handshake and apply a documented narrow exception if required.
Question 307
An engineer's policy change removed remote management access. The team has a recent configuration backup and console access. Which recovery plan is defensible?
- Enable unrestricted administrative HTTPS on each WAN interface in the backup configuration to recover faster after the accidental management-rule change.
- Recover through authenticated console access, identify the exact management-policy regression and restore a compatible verified backup or bounded change; retest trusted management hosts and preserve rollback.
- Power-cycle the appliance repeatedly until the management policy reverts to an earlier running configuration instead of examining the current effective settings via console.
- Review the console administrator access settings and permit a temporary unrestricted local login while leaving the current broken remote-management configuration untouched.
Explanation
Correct Answer
B:With working console access, first identify whether an interface administrative-access, trusted-host or route change caused the lockout. If correction is unsafe or a known-good backup is required, verify that its FortiOS/hardware/version and intended configuration are compatible and that a recovery copy exists before restoring. Reconnect over a restricted management path and test policy/data forwarding; do not blindly flash firmware or disable administrator authentication.
Incorrect Answers
A:Wide public management exposure introduces unnecessary risk when local console recovery exists. Restore tested trusted-source access and compatible configuration through the authorized recovery channel.
C:Rebooting usually reloads the saved configuration and does not automatically undo an erroneous committed policy. Use console to assess the exact regression and verified rollback.
D:Console recovery is the appropriate safe route when the remote policy locks out administrators, but removing authentication and retaining unrestricted access is not required. The safer procedure is an authorized console login, evidence of the policy regression, selective rollback using a compatible verified backup, and a restricted remote-management test. Preserve account security and a recovery path throughout.
Question 308
LDAPS users fail authentication with a certificate trust error after the directory server rotates its issuing CA. What should be corrected?
- Temporarily disable LDAPS server-certificate validation in the remote user connector so the new issuing CA does not interrupt logins.
- Inspect the LDAP search base DN and expand it to include the newly reorganized directory users while retaining the same trusted LDAPS issuer.
- Update the trusted issuing CA chain for the rotated LDAPS server and verify its expected identity before retesting.
- Increase LDAPS TCP connection timeout and directory bind retries to accommodate the newly rotated server certificate.
Explanation
Correct Answer
C:LDAP-over-TLS certificate validation can fail when the directory server switches to a CA that FortiGate no longer trusts. Inspect the server’s presented chain and hostname, verify the approved new issuer, import or select the appropriate trusted CA, and test LDAPS bind and group mapping. Disabling certificate verification would hide the cause while allowing impersonation risk.
Incorrect Answers
A:It may bypass the symptom but removes server identity assurance. A known issuer rotation requires validating and installing the authorized CA chain, preserving hostname/identity checks.
B:LDAP search scope determines which accounts can be found after a secure connection, but the current error is a TLS certificate trust failure immediately after the server changes its issuing CA. A wider base DN will not validate the new certificate chain. Update legitimate issuing CA trust and server identity first.
D:Retries may help network delays, but they repeat the same failed certificate-validation decision when the trusted issuer no longer matches. Establish the correct issuing chain, validity, name and required trust on FortiGate before replaying authentication or adjusting LDAP search performance.
Question 309
A blocked category can still be reached through an alternate hostname that uses a CDN. The user types the alternate domain, not the originally filtered FQDN. What should be verified?
- Inspect the alternate CDN hostname’s resolved public addresses and block their entire IP prefix at the WAN firewall.
- Widen the existing static URL wildcard to every domain containing the same word as the prohibited brand.
- Review the original blocked hostname’s FortiGuard category and apply that same rating to the CDN provider’s main domain.
- Inspect the alternate CDN hostname’s own FortiGuard rating and matching static URL rule on the effective policy; distinguish host-specific matches from broad wildcard patterns before applying a targeted correction.
Explanation
Correct Answer
D:Web filtering evaluates the destination identity and URL pattern visible in the particular session, not a conceptual corporate site name. A CDN may expose the same resource through another hostname with a different rating or unmatched static URL entry. Compare DNS/SNI/Host observations, profile match and log action, then make a supported domain-scoped rule adjustment without blocking unrelated shared CDN tenants.
Incorrect Answers
A:Resolved IPs can provide useful session correlation, but a shared CDN address or prefix may host many unrelated approved domains. An IP deny can therefore overblock while the alternate hostname remains the actual web-filter identity. Determine its FortiGuard rating and matched URL expression before choosing a bounded rule.
B:An expansive pattern could block legitimate unrelated domains and still miss other aliases. Identify actual hostnames and use tested narrow URL patterns or categories. An overly broad wildcard can block legitimate domains sharing a word; match only the prohibited host variants.
C:The CDN provider’s parent domain and an alternate customer hostname need not have the same category or static URL-filter match as the original destination. The user actually requests the alternate host. Evaluate its rating, rule and application context on the matching policy rather than inherit an unrelated broad provider classification.
Question 310
After an HA failover, FortiAnalyzer has no event for a known connection, but the former primary displays an older local log. Where should retention be checked?
- Review the new primary’s on-box historical log buffer and rely on FGCP session synchronization to recover the former primary’s missing records.
- Inspect time synchronization between FGCP members and FortiAnalyzer, then reset both appliance clocks before tracing event forwarding.
- Determine the logging HA member and validate FortiAnalyzer ingestion, device registration and time filters after failover.
- Delete both FortiAnalyzer device records and register the HA cluster again before checking local event timestamps and collector forwarding on the former primary.
Explanation
Correct Answer
C:HA failover can move session and log forwarding between physical members; inspect current and former member event stores, timestamps and FortiAnalyzer device identity. Find the current HA member that produced the test event, its timestamp and FortiAnalyzer device identity; check forwarding after failover and local log buffers on both members. A log remaining on the old primary does not prove the central collector received it.
Incorrect Answers
A:FGCP can synchronize configuration and eligible session state, not necessarily every stored historical log entry between members. The known event exists on the former primary; trace its original timestamp and forwarding target, then check FortiAnalyzer ingestion and selected device/time filters rather than assuming replicated local history.
B:Time skew can make a real event appear outside a query window, so comparing timestamps is useful. Blindly resetting clocks, however, is disruptive and does not prove that the new active member sent the event to FortiAnalyzer. Verify logging registration, destination and actual delivery after failover.
D:Premature re-registration can destroy context and does not prove whether the missing log was generated, transported or indexed. Preserve local and central evidence first.