Fortinet FCP_FSM_AN-7.2 (FCP - FortiSIEM 7.2 Analyst) Practice Test 17

Question 160

For Cloud Services, an analyst must choose where a FortiSIEM ML job executes. Which pair of modes separates local execution from an AWS-backed execution option?

  1. Email and SMS aggregation modes
  2. Local and AWS machine-learning run modes
  3. Real-time Group By and Nested Lookup modes
  4. CMDB and Incident Resolution modes

Explanation

Correct Answer

B:

FortiSIEM provides run-mode selections such as Local or AWS, with Auto variants. They determine execution location and automation characteristics for the daily unique destination counts task.

Incorrect Answers

A:

In machine-learning analysis at Cloud Services: Email and SMS are notification transports; neither selects local or AWS compute for the learner.

C:

Considering machine-learning analysis at Cloud Services: These are analytics search features, not the named local-versus-cloud execution modes for a FortiSIEM ML job.

D:

Under machine-learning analysis at Cloud Services: CMDB inventory and incident verdict fields are not the documented run-mode choices for executing machine-learning work.

Question 161

Manufacturing is correlating svc.backup suspicious login activity with a high-value host-66 object. What additional data should remain available to improve prioritization?

  1. An inactive remediation script disconnected from the incident
  2. Host criticality from CMDB together with relevant user-event context
  3. Only the number of characters in the hostname
  4. Use only event timestamp as the user/entity correlation key

Explanation

Correct Answer

B:

For entity-behavior risk evaluation at Manufacturing for host-66: Asset criticality increases the meaning of risky activity. Combining CMDB importance with user and host observations helps analysts prioritize the potentially impactful access.

Incorrect Answers

A:

An unused script does not enrich risk signals or explain the actual impact of an access attempt against host-66.

C:

Hostname length is an arbitrary string property and does not reflect the business importance of host-66.

D:

Under entity-behavior risk evaluation at Manufacturing for host-66: Timestamps cannot establish stable identity across VPN, identity-provider and other event sources for a UEBA investigation.

Question 162

A Manufacturing rule tracks elevated risk for svc.backup, and the SOC wants a dashboard filter showing only the tagged subset. What should be preserved in the query?

  1. The UEBA tag/entity attribute and the filter that selects the intended population
  2. A notification recipient email as the aggregation key
  3. An unconditional count of all tenants’ unrelated events
  4. Only a report title that contains the word risk

Explanation

Correct Answer

A:

A useful dashboard needs the same tag or entity identity carried through the rule/event model. Filtering by that attribute isolates the relevant svc.backup activity.

Incorrect Answers

B:

For entity-behavior risk evaluation at Manufacturing: Notification recipients identify delivery targets, not the entity attribute used to filter UEBA telemetry.

C:

Broad counts mix unrelated observations and obscure the selected Manufacturing tagged behavior.

D:

Under entity-behavior risk evaluation at Manufacturing: Human-readable text does not itself filter data by the underlying user tag or risk attribute.

Question 163

East Retail is building a scheduled analytics rule for rejected request. A designer tries adding two distinct subpatterns to its condition. Which FortiSIEM 7.2 restriction explains the rejection?

  1. Scheduled rules must use only raw-text keywords as filters
  2. All rules are limited to one event attribute globally
  3. Scheduled rules support only one subpattern
  4. Scheduled rules cannot use any event filters

Explanation

Correct Answer

C:

Considering correlation-rule investigation at East Retail with the specified 7.2 value: FortiSIEM explicitly restricts scheduled rule conditions to a single subpattern. Multi-subpattern temporal correlation belongs to applicable real-time rule designs.

Incorrect Answers

A:

In correlation-rule investigation at East Retail with the specified 7.2 value: Normalized Event Attribute conditions are supported in rule filtering; the problem is the second subpattern rather than parser field use.

B:

For correlation-rule investigation at East Retail with the specified 7.2 value: A subpattern may use multiple filter attributes. The documented scheduled restriction is a structural subpattern count.

D:

Under correlation-rule investigation at East Retail with the specified 7.2 value: Scheduled rules can specify the one subpattern’s filters; the limitation concerns the number of subpatterns, not whether rejected request can be selected.

Question 164

A West Operations report groups rejected request events by User and uses COUNT(Matched Events). The analyst wants only groups exceeding 7 events to remain in the result, not to remove the underlying records before aggregation. What should be configured?

  1. Group By COUNT(Matched Events) instead of User
  2. An event filter that rejects all records until their count exceeds 7
  3. A Display Condition of COUNT(Matched Events) > 7
  4. A time range of exactly 7 seconds

Explanation

Correct Answer

C:

For A West Operations report groups rejected request events by User and uses COUNT(Matched Events). The analyst wants only: Considering aggregate-query calculation at West Operations with the specified 7 value: FortiSIEM display conditions filter aggregates after grouping, analogous to HAVING. Groups must first be counted, then those above 7 remain visible.

Incorrect Answers

A:

For A West Operations report groups rejected request events by User and uses COUNT(Matched Events). The analyst wants only: In aggregate-query calculation at West Operations with the specified 7 value: COUNT is the measure to test, while User defines the groups whose counts are compared; swapping these loses the per-source view.

B:

An event filter evaluates record attributes before aggregation. Individual cloud API audit events events do not carry the completed group count to filter on.

D:

The literal threshold is a count of records, not a query time window. Shortening time would alter which rejected request events are included.

Question 165

Partner Services has created a rule to detect suspicious launch and populated its conditions, but wants to minimize false alerts before production. What is the appropriate immediate next workflow?

  1. Activate the rule without testing because Save guarantees accuracy
  2. Test the saved inactive rule against relevant events, then activate it
  3. Delete all prior log events so test output is empty
  4. Replace the rule with a notification policy that has no conditions

Explanation

Correct Answer

B:

FortiSIEM saves new rules inactive. Testing the proposed suspicious launch conditions against observed events exposes overly broad filters or incorrect thresholds before live incidents appear.

Incorrect Answers

A:

For Partner Services has created a rule to detect suspicious launch and populated its conditions, but wants to minimize false: In rule deployment or exception at Partner Services: Saving validates configuration structure, not the operational meaning of the detection for Partner Services event traffic.

C:

Erasing evidence prevents meaningful verification of the detection’s ability to match legitimate suspicious launch activity.

D:

For Partner Services has created a rule to detect suspicious launch and populated its conditions, but wants to minimize false: Under rule deployment or exception at Partner Services: Automation policies act after incidents; they cannot independently perform the rule’s required event correlation.

Question 166

For the response to an incident affecting host-78, cloud Services wants the same vetted action, isolate an infected endpoint, to execute when its verified connection incident rule fires, with no per-incident analyst click. Which configuration automates response?

  1. A historical event export scheduled daily
  2. A dashboard Group By on incident time
  3. An incident comment template without an action
  4. An enabled Automation Policy with Run Remediation/Script

Explanation

Correct Answer

D:

Under correlation-rule investigation at the SOC for host-78: The policy can invoke an approved remediation script as the rule generates a matching incident. Preconditions and device scope must be validated before unattended action.

Incorrect Answers

A:

In correlation-rule investigation at the SOC for host-78: Export creates copies of evidence, not a near-real-time response on incident generation.

B:

For correlation-rule investigation at the SOC for host-78: Dashboard aggregation shows incident patterns but is not the action engine for the Cloud Services response.

C:

Considering correlation-rule investigation at the SOC for host-78: Comments convey instructions but do not execute isolate an infected endpoint automatically on the selected systems.

Question 167

Corporate HQ wants to use the same nightly database write rate training job for host-68 but discovers that the training dataset contains almost no valid records. What is the priority before production inference?

  1. Increase the notification recipient count
  2. Repair the event/report source and confirm adequate representative training samples
  3. Change the ZTNA enforcement port for unrelated traffic
  4. Mark all nonexistent model findings as resolved

Explanation

Correct Answer

B:

Insufficient or malformed nightly database write rate data prevents reliable learning. Verify the report filters, time span, field types and job inputs before interpreting model output.

Incorrect Answers

A:

Alert recipients affect message delivery after a finding, not whether enough valid nightly database write rate samples exist for training.

C:

Considering historical or real-time event analysis at Corporate HQ for host-68: External access-policy ports do not address missing training observations in the ML pipeline.

D:

Under historical or real-time event analysis at Corporate HQ for host-68: Incident statuses cannot compensate for an empty or nonrepresentative training dataset.

Question 168

A Research correlation rule must operate per (User, Destination Domain) pair instead of per User alone, because the same actor contacts several unrelated target systems. What change prevents cross-target aggregation?

  1. Group By only User
  2. Remove all aggregate constraints
  3. Group By only Destination Domain
  4. Group By both User and Destination Domain

Explanation

Correct Answer

D:

A composite key partitions the evidence by actor-and-target combination, so separate Destination Domain systems do not contribute to the same rule threshold.

Incorrect Answers

A:

This still merges all Destination Domain targets reached by that actor and can create a combined count unsupported by any single pair.

B:

Without thresholds the rule cannot evaluate the intended per-pair policy denial frequency, regardless of the chosen grouping keys.

C:

Considering subpattern correlation at Research: This combines different attacking sources against one target and loses the required actor-specific relationship.

Question 169

East Retail searches service health records and sees the right event count but needs the Service value beside each Host in unaggregated rows. Which change addresses the output without excluding matches?

  1. Add Service to the displayed event fields
  2. Filter to events where Service equals a single value
  3. Shorten the historical time range until one row remains
  4. Group the results only by Host

Explanation

Correct Answer

A:

Display Fields controls the attributes shown for matching events. Including Service preserves the filter and makes the requested evidence visible alongside Host.

Incorrect Answers

B:

A new equality filter would discard legitimate service health records records with other Service values, changing the evidence set rather than its columns.

C:

Considering historical or real-time event analysis at East Retail using service health records: Changing time arbitrarily alters the count and risks removing relevant events; the problem is display columns, not volume.

D:

A group summary collapses event rows and can hide each individual Service relationship sought by the investigator.

Question 170

North Finance knows a legitimate scheduled scanner will produce suspicious launch events on host-61 during a maintenance window. The detection should remain active for all other assets and times. Which rule feature targets the exception?

  1. Disable the entire rule permanently
  2. Delete the parser for the log source
  3. Change the incident severity to low with no exception
  4. A rule exception constrained by relevant attributes and an effective period

Explanation

Correct Answer

D:

Rule exceptions suppress incidents for matching host-61 conditions during a defined single or recurring interval, preserving detection elsewhere.

Incorrect Answers

A:

Global disablement discards all suspicious launch detections beyond the approved host-61 maintenance exception.

B:

For correlation-rule investigation at North Finance for host-61: Removing event parsing would harm all observations from endpoint process alerts and is not a precise, time-bound exception.

C:

Considering correlation-rule investigation at North Finance for host-61: Lowering priority still creates incidents for the expected scan and does not delimit the approved maintenance period.

FCP - FortiSIEM 7.2 Analyst (FCP_FSM_AN-7.2) Practice Tests