Sign in
Home/CrowdStrike/CrowdStrike Certified Falcon Administrator (CCFA)/Free questions

CrowdStrike Certified Falcon Administrator (CCFA) — Free Practice Questions

10 free sample questions from a bank of 279, with the correct answers and explanations. No signup required — start practising right now.

1What is the function of a single asterisk (*) in an ML exclusion pattern?
  • The single asterisk will match any number of characters, including none. It does include separator characters, such as \ or /, which separate portions of a file path
  • The single asterisk will match any number of characters, including none. It does not include separator characters, such as \ or /, which separate portions of a file path
  • The single asterisk is the insertion point for the variable list that follows the path
  • The single asterisk is only used to start an expression, and it represents the drive letter
Answer: B

The short version

B — Single asterisk matches any characters except path separators. In Falcon ML exclusion globs, one star stays within a path segment while double star recurses across directories. The settled key is B because that separator rule is the core distinction tested here.

Key concepts in this question

  • ML exclusions: paths that suppress machine-learning detections where trusted software would otherwise alert.
  • Glob single star: wildcard for zero or more characters inside one directory level, never crossing slash or backslash.
  • Glob double star: recursive wildcard that spans zero or more directories beneath the current location.
  • Separators: slash and backslash that delimit folders, used by the engine to limit single-star scope.
  • Variable insertion: separate syntax for lists that follow a path, not the role of a lone star.

Why B is correct

B is correct because CrowdStrike documents single star as match-anything-within-segment, explicitly excluding separators. That lets admins allow files in one folder without blanketing subfolders, while double star is reserved for recursive coverage. The vendor prose and glob standard both confirm that boundary behavior, so B alone states the function precisely.

Why the others are wrong

  • A. Including separators describes double-star recursion, not single-star behavior, so it reverses the tested rule.
  • C. Insertion point for a variable list is a different placeholder concept, not the meaning of a lone star.
  • D. Drive-letter starter is unrelated; single star may appear anywhere in a pattern, not only at the start.

CCFA exam tip

For exclusion globs remember one star stays in its folder and two stars walk subfolders; separators decide the scope.

2One of your development teams is working on code for a new enterprise application but Falcon continually flags the execution as a detection during testing. All development work is required to be stored on a file share in a folder called "devcode." What setting can you use to reduce false positives on this file path?
  • USB Device Policy
  • Firewall Rule Group
  • Containment Policy
  • Machine Learning Exclusions
Answer: D

The short version

D — Machine Learning Exclusions calm Falcon flags on the devcode share. Development builds look like suspicious behavior to ML models, and all code must live under the devcode folder. The settled key is D because only ML exclusions tune that detection layer by path.

Key concepts in this question

  • Falcon detections: ML plus behavioral signals that fire on unusual compile, injection or script activity common in dev testing.
  • False positives: benign dev binaries flagged because their actions resemble malware techniques during testing.
  • ML exclusions: policy entries that suppress ML scoring for defined paths, hashes or processes without disabling the sensor.
  • USB policy: controls removable media use, unrelated to network share build folders.
  • Firewall and containment: network rules and host isolation, not tuning for local execution scoring.

Why D is correct

D is correct because the need is fewer ML detections on a known file path while keeping protection elsewhere. Creating an ML exclusion for the devcode share tells Falcon to trust executions from that location during testing. That matches CrowdStrike guidance to scope exclusions narrowly to the required dev path, preserving visibility elsewhere instead of lowering global aggressiveness.

Why the others are wrong

  • A. USB Device Policy governs removable storage access, so it cannot quiet execution scoring on a file share.
  • B. Firewall Rule Groups manage network connections, not local ML verdicts on newly built binaries.
  • C. Containment Policy isolates hosts from the network and is far too blunt for dev-time false-positive tuning.

CCFA exam tip

When dev builds on a fixed path keep tripping Falcon, think Machine Learning Exclusions scoped to that path.

3When a host belongs to more than one host group, how is sensor update precedence determined?
  • Groups have no impact on sensor update policies
  • Sensors of hosts that belong to more than one group must be manually updated
  • The highest precedence policy from the most important group is applied to the host
  • All of the host's groups are examined in aggregate and the policy with highest precedence is applied to the host
Answer: D

The short version

D — Hosts in several groups get the highest-precedence policy across all their groups. Falcon does not pick one favorite group; it aggregates every membership and applies the winner. The settled key is D because aggregate evaluation is the documented precedence logic.

Key concepts in this question

  • Host groups: static or dynamic collections used to assign sensor update and other policies at scale.
  • Sensor update policy: controls version, scheduling and precedence for Falcon sensor upgrades.
  • Precedence: numeric priority that decides which policy wins when assignments overlap.
  • Aggregate evaluation: engine looks at every group a host belongs to, then selects the highest-precedence policy among them.
  • Manual update: fallback for unmanaged hosts, not the automatic outcome for multi-group membership.

Why D is correct

D is correct because CrowdStrike evaluates all groups together rather than choosing a single most important group container. If a host sits in three groups with different update policies, Falcon compares their precedence values and enforces the top one. That ensures predictable upgrades without forcing admins to keep group membership exclusive.

Why the others are wrong

  • A. Groups directly drive update assignment, so claiming no impact contradicts Falcon grouping design.
  • B. Manual updates are unnecessary; the platform resolves overlap automatically through precedence.
  • C. Highest policy from the most important group is incomplete because Falcon looks across all memberships, not one group.

CCFA exam tip

For multi-group hosts remember aggregate then highest precedence wins; never assume one group alone decides.

4What may prevent a user from logging into Falcon via single sign-on (SSO)?
  • The SSO username doesn't match their email address in Falcon
  • The maintenance token has expired
  • Falcon is in reduced functionality mode
  • The user never configured their security questions
Answer: A

The short version

A — SSO login fails when the SSO username does not match the Falcon email address. Falcon links the identity provider assertion to the console user record by email. The settled key is A because identity mismatch blocks the match while the other options affect different flows.

Key concepts in this question

  • Falcon SSO: SAML or OIDC login that passes a username assertion from the identity provider to CrowdStrike.
  • Email mapping: Falcon matches the assertion to the Users record by email address as the unique key.
  • Maintenance token: one-time code for local sensor troubleshooting, unrelated to console SSO auth.
  • Reduced functionality mode: sensor state with limited telemetry, not a console login gate.
  • Security questions: legacy consumer auth step, not part of Falcon SSO user provisioning.

Why A is correct

A is correct because a mismatch means Falcon cannot find the corresponding user object for the asserted identity. Even with a valid provider token, the console rejects login when the SSO username differs from the stored email. Aligning the IdP NameID or username attribute to the exact Falcon email restores access, which is the documented SSO prerequisite.

Why the others are wrong

  • B. Expired maintenance token blocks local sensor maintenance actions, not browser SSO to the console.
  • C. Reduced functionality mode limits sensor capability on an endpoint, but it does not gate Falcon console SSO.
  • D. Missing security questions have no bearing on enterprise SSO, which relies on provider assertion plus email match.

CCFA exam tip

For Falcon SSO failures first compare IdP username to Falcon email character for character; mismatch is the classic blocker.

5The Customer ID (CID) is important in which of the following scenarios?
  • When adding a user to the Falcon console under the Users application
  • When performing the sensor installation process
  • When setting up API keys
  • When performing a Host Search
Answer: B

The short version

B — Customer ID is essential during sensor installation. The CID binds each installer to the correct Falcon tenant so telemetry lands in the right cloud. The settled key is B because install-time tenancy cannot be inferred from user or API context.

Key concepts in this question

  • Customer ID: unique tenant identifier shown in Falcon console that directs sensor registration.
  • Sensor install: package deployment that requires CID plus token or policy to authenticate the endpoint to its cloud.
  • Users app: console identity management that uses email and roles, not CID, to create analysts.
  • API keys: OAuth credentials scoped to a tenant already, so CID is not entered separately during their setup.
  • Host search: post-install investigation by hostname or attributes, which assumes sensors already report with CID.

Why B is correct

B is correct because the installer command line or deployment tool must include the CID for the sensor to check in to the proper environment. Without it, the sensor cannot determine which CrowdStrike cloud tenant owns the endpoint. User creation, API setup and host search all happen after that tenancy binding, so only installation depends on manually supplying CID.

Why the others are wrong

  • A. Adding a console user relies on email plus role assignment, with tenancy already set by the logged-in admin session.
  • C. API key setup uses client ID plus secret within the tenant, not a separate CID entry step.
  • D. Host search queries data from enrolled sensors, so CID was already consumed at install and is not entered then.

CCFA exam tip

When you see CID think sensor install command line; console users and API keys live inside the tenant already.

6Which statement describes what is recommended for the Default Sensor Update policy?
  • The Default Sensor Update policy should align to an organization's overall sensor updating practice while leveraging Auto N-1 and Auto N-2 configurations where possible
  • The Default Sensor Update should be configured to always automatically upgrade to the latest sensor version
  • Since the Default Sensor Update policy is pre-configured with recommend settings out of the box, configuration of the Default Sensor Update policy is not required
  • No configuration is required. Once a Custom Sensor Update policy is created the Default Sensor Update policy is disabled
Answer: A

The short version

A — Default Sensor Update policy should mirror org rollout practice while using auto upgrade. The default catches every host without a custom policy, so it must be intentional, not left untouched or forced to latest. The settled key is A because alignment plus controlled automation is the recommended posture.

Key concepts in this question

  • Default policy: catch-all sensor update definition that applies where no custom assignment matches.
  • Custom policies: targeted update rings by group, OS or criticality for phased testing before broad rollout.
  • Auto upgrade: Falcon feature that moves sensors forward within constraints, balancing freshness and stability.
  • Always latest: aggressive stance that skips testing and risks incompatibility with business apps.
  • Disabled default: misconception; the default stays active as fallback even after customs are created.

Why A is correct

A is correct because CrowdStrike advises treating the default as part of the overall updating strategy. Admins align its version, schedule and auto behavior to org change windows, then use custom policies for pilot rings. Leveraging automation with guardrails keeps the fallback population current without forcing untested builds everywhere, which matches vendor best practice.

Why the others are wrong

  • B. Always auto to latest removes testing and can push incompatible sensors to production without control.
  • C. Preconfigured out-of-box settings are a starting point, not a recommendation to leave the default unreviewed.
  • D. Creating customs does not disable the default; it remains the fallback for unmatched hosts, so config is still required.

CCFA exam tip

Treat the default update policy as a deliberate fallback ring that mirrors org practice with controlled automation.

7You need to have the ability to monitor suspicious VBA macros. Which Sensor Visibility setting should be turned on within the Prevention policy settings?
  • Script-based Execution Monitoring
  • Interpreter-Only
  • Additional User Mode Data
  • Engine (Full Visibility)
Answer: A

The short version

A — Script-based Execution Monitoring gives visibility into suspicious VBA macros. Falcon prevention policies separate basic interpreter telemetry from deeper script-content inspection. The settled key is A because only that toggle monitors macro contents without blocking them outright.

Key concepts in this question

  • VBA macros: Office automation often abused for downloaders and payload execution, needing content-level visibility.
  • Script-based Execution Monitoring: prevention setting that inspects scripts and shells popular for malicious code.
  • Interpreter-Only: narrower mode focused on interpreter processes, missing broader macro and shell context.
  • Additional User Mode Data: extra telemetry for advanced triage, not the primary macro-content switch.
  • Engine Full Visibility: ML engine depth control, distinct from the script-content monitoring toggle.

Why A is correct

A is correct because Falcon documentation maps VBA macros, VBScript, JScript and Excel 4.0 macros to Script-based Execution Monitoring. Enabling it lets the sensor examine macro contents and related shell activity for detections and hunting. The setting monitors without killing scripts by itself, which matches the need to watch suspicious VBA while tuning response separately.

Why the others are wrong

  • B. Interpreter-Only narrows to interpreter launches and lacks the documented macro-content coverage of the correct toggle.
  • C. Additional User Mode Data enriches context but is not the designated switch for VBA macro monitoring.
  • D. Engine Full Visibility adjusts engine depth, not the script-content inspection path for macros.

CCFA exam tip

For VBA plus Excel 4.0 plus JScript visibility choose Script-based Execution Monitoring in the prevention policy.

8What is the purpose of the Machine-Learning Prevention Monitoring Report?
  • It is designed to give an administrator a quick overview of machine-learning aggressiveness settings as well as the numbers of items actually quarantined
  • It is the dashboard used by an analyst to view all items quarantined and to release any items deemed non-malicious
  • It is the dashboard used to see machine-learning preventions, and it is used to identify spikes in activity and possible targeted attacks
  • It is designed to show malware that would have been blocked in your environment based on different Machine-Learning Prevention settings
Answer: D

The short version

D — The report shows what would have blocked under different ML aggressiveness levels. It is a what-if view for tuning, not a quarantine console or live prevention dashboard. The settled key is D because modeling blocked malware is its stated purpose.

Key concepts in this question

  • ML aggressiveness: sensitivity slider balancing detection rate against false positives in Falcon prevention.
  • Monitoring report: read-only modeling view that replays recent files against candidate ML thresholds.
  • Quarantine manager: separate workflow for releasing items already held, not for threshold modeling.
  • Prevention dashboard: live view of actual blocks and spikes, distinct from hypothetical would-have-blocked analysis.
  • Tuning use: admins compare settings to pick a level before enforcing it broadly.

Why D is correct

D is correct because the Machine-Learning Prevention Monitoring Report is designed to preview malware that would have been stopped at each aggressiveness tier. That lets teams gauge coverage versus noise in their own environment before changing enforcement. Vendor docs and console behavior confirm this modeling role rather than quarantine release or live-spike triage.

Why the others are wrong

  • A. Overview of settings plus counts is too generic; the value is per-threshold would-have-blocked samples, not just counts.
  • B. Quarantine review and release happens in a different containment view, not in this monitoring report.
  • C. Live preventions and spike hunting belong to prevention dashboards, while this report models alternate thresholds.

CCFA exam tip

When the stem says would have been blocked based on different ML levels, choose the ML Prevention Monitoring Report.

9The Remote Access Graph in Visibility Reports displays:
  • a bar chart where a bar represents a daily count of remote connections
  • a geographical chart showing the geo-location of remote IP address
  • a graph showing connections between hosts and users
  • a pie chart showing a count per remote logon type
Answer: C

The short version

C — Remote Access Graph links hosts and users with remote logons. It is a relationship graph for RDP plus network logons, not a bar, geo or pie summary. The settled key is C because host-user connections are its defined display.

Key concepts in this question

  • Visibility reports: Falcon hunting views that summarize logon, network and execution patterns across endpoints.
  • Remote Access Graph: node-link view of which users remotely accessed which systems and by which logon type.
  • Logon types: Type 10 terminal plus Type 3 network logons commonly pivoted for lateral-movement review.
  • Bar charts: daily counts of connections, useful for volume but not relationship mapping.
  • Geo and pie views: location or type breakdowns that omit the host-to-user linkage the question seeks.

Why C is correct

C is correct because CrowdStrike defines this graph as a visual map between users and remotely accessed systems. Analysts filter by RDP or network logon to find unusual access paths and lateral movement. That host-user edge display is exactly what distinguishes the graph from count-based or location-based charts.

Why the others are wrong

  • A. Daily-count bars show volume over time but do not reveal which user touched which host.
  • B. Geographical charts show remote IP location, not the host-user relationship graph.
  • D. Pie counts per logon type summarize mix without mapping connections between entities.

CCFA exam tip

For host plus user remote-access relationships choose the graph view; for volumes choose bars and for mix choose pie.

10What internet domain needs to be added to any required allowlists to allow sensors to communicate with the CrowdStrike Cloud?
  • falconcloud.net
  • cloudprotect-cs.net
  • cloudsink.net
  • csfalcon.net
Answer: C

The short version

C — Allowlist cloudsink.net so sensors can reach the CrowdStrike cloud. Sensors beacon to that domain for check-in, policy and telemetry. The settled key is C because it is the documented sensor-to-cloud allowlist entry.

Key concepts in this question

  • Sensor-to-cloud: persistent TLS from endpoint to CrowdStrike for registration, commands and event upload.
  • Allowlist: firewall plus proxy exception that permits required domains while denying other egress.
  • cloudsink.net: published domain for sensor communication that must remain reachable.
  • Lookalike domains: plausible but incorrect Falcon-style names that would leave sensors disconnected if used.
  • Connectivity symptom: sensors offline or not checking in when the correct domain is blocked.

Why C is correct

C is correct because CrowdStrike publishes cloudsink.net as the sensor communication domain to permit through egress controls. Adding it to firewalls and proxies lets installs register and keeps heartbeats flowing. The other domains do not serve that sensor channel, so only C restores expected cloud connectivity per docs.

Why the others are wrong

  • A. falconcloud.net is not the documented sensor allowlist domain, so it would not restore connectivity.
  • B. cloudprotect-cs.net is not the published sensor channel and would leave beacons blocked.
  • D. csfalcon.net is likewise not the required allowlist entry for sensor-to-cloud traffic.

CCFA exam tip

For sensor-to-cloud allowlists memorize cloudsink.net; test connectivity when sensors fail to check in.

Want the full bank of 279 questions for CrowdStrike Certified Falcon Administrator (CCFA)? See all practice exams.