Practice Exams:

Fortinet FCP_FGT_AD-7.6 Dumps

EXAM: FCP - FortiGate 7.6 Administrator
Verified By Experts
Premium Bundle
$49.99

FCP_FGT_AD-7.6 Premium ETE File


  • Premium File 128 Questions & Answers
Whats Included:
  • Latest Questions
  • 100% Accurate Answers
  • Fast Exam Updates

Fortinet FCP_FGT_AD-7.6 Practice Test 8

Question 70

A ticketing company is testing whether supported live sessions stay intact during a planned cluster maintenance window. Which combined implementation and verification plan most reliably resolves this specific FortiOS 7.6 issue?

  1. For the ticketing company deployment: Implement: Check FortiAnalyzer registration authorization, configured destination and log transport when forwarded events stop appearing. Verify: Compare a locally observed event with the FortiAnalyzer device status and ingestion view, and verify a fresh timestamp after reconnection.
  2. For the ticketing company deployment: Implement: Enable applicable session pickup and verify that synchronized eligible sessions survive a planned active-passive FGCP failover. Verify: Maintain representative long-lived TCP sessions, trigger an approved failover, and inspect session and HA status before and after.
  3. For the ticketing company deployment: Implement: Configure matching FGCP mode, group settings and dedicated heartbeat links on supported compatible FortiGate peers before joining them. Verify: Maintain representative long-lived TCP sessions, trigger an approved failover, and inspect session and HA status before and after.
  4. For the ticketing company deployment: Implement: Enable applicable session pickup and verify that synchronized eligible sessions survive a planned active-passive FGCP failover. Verify: Reach each peer's unique management address and inspect role-specific health while production traffic continues through the cluster.

Answer & Explanations

Correct answer: B

A.

The situation is that a ticketing company is testing whether supported live sessions stay intact during a planned cluster maintenance window. This action and check address the distinct objective “Repair FortiAnalyzer registration and logging path” instead of the stated acceptance requirement. The matching action is to enable applicable session pickup and verify that synchronized eligible sessions survive a planned active-passive FGCP failover. Its confirming test is to maintain representative long-lived TCP sessions, trigger an approved failover, and inspect session and HA status before and after.

B.

The situation is that a ticketing company is testing whether supported live sessions stay intact during a planned cluster maintenance window. Implement: Enable applicable session pickup and verify that synchronized eligible sessions survive a planned active-passive FGCP failover. Verify: Maintain representative long-lived TCP sessions, trigger an approved failover, and inspect session and HA status before and after. This pairing is justified because HA configuration synchronization does not inherently preserve every live connection; session-pickup behavior and protocol support determine failover continuity.

C.

This action and check address the alternate operation “For the ticketing company deployment: Implement: Configure matching FGCP mode, group settings and dedicated heartbeat links on supported compatible…” instead of the stated acceptance requirement. The matching action is to enable applicable session pickup and verify that synchronized eligible sessions survive a planned active-passive FGCP failover. Its confirming test is to maintain representative long-lived TCP sessions, trigger an approved failover, and inspect session and HA status before and after.

D.

This action and check address the alternate operation “For the ticketing company deployment: Implement: Enable applicable session pickup and verify that synchronized eligible sessions survive a planned…” instead of the stated acceptance requirement. The matching action is to enable applicable session pickup and verify that synchronized eligible sessions survive a planned active-passive FGCP failover. Its confirming test is to maintain representative long-lived TCP sessions, trigger an approved failover, and inspect session and HA status before and after.

Question 71

A public transit operator needs to check both FortiGate peers directly during software maintenance without disturbing the shared gateway. Which FortiOS 7.6 administrative change best meets this requirement without weakening unrelated controls?

  1. For the public transit operator deployment: Use bounded packet sniffing and diagnose debug flow for a specific source, destination and port to identify routing, policy or session decisions.
  2. For the public transit operator deployment: Enable applicable session pickup and verify that synchronized eligible sessions survive a planned active-passive FGCP failover.
  3. For the public transit operator deployment: Configure a reserved HA management interface and appropriate management routes where individual cluster members require independent access.
  4. For the public transit operator deployment: Configure matching FGCP mode, group settings and dedicated heartbeat links on supported compatible FortiGate peers before joining them.

Answer & Explanations

Correct answer: C

A.

This option addresses the distinct objective “Trace a denied flow at the correct forwarding stage”, not the immediate requirement. Packet arrival alone does not prove policy acceptance; a constrained flow trace can reveal the stage where FortiOS rejects or forwards the traffic. For the stated case, dedicated HA management lets operators address each member separately without confusing the cluster data-plane identity with member-specific diagnostic reachability.

B.

This option addresses the distinct objective “Synchronize eligible sessions for HA failover”, not the immediate requirement. HA configuration synchronization does not inherently preserve every live connection; session-pickup behavior and protocol support determine failover continuity. For the stated case, dedicated HA management lets operators address each member separately without confusing the cluster data-plane identity with member-specific diagnostic reachability.

C.

Configure a reserved HA management interface and appropriate management routes where individual cluster members require independent access. This resolves the issue because dedicated HA management lets operators address each member separately without confusing the cluster data-plane identity with member-specific diagnostic reachability. To verify, reach each peer's unique management address and inspect role-specific health while production traffic continues through the cluster.

D.

This option addresses the distinct objective “Build compatible FGCP heartbeat and cluster membership”, not the immediate requirement. FGCP members negotiate HA using compatible hardware and cluster parameters, while heartbeat connectivity drives membership and primary selection. For the stated case, dedicated HA management lets operators address each member separately without confusing the cluster data-plane identity with member-specific diagnostic reachability.

Question 72

A public transit operator needs to check both FortiGate peers directly during software maintenance without disturbing the shared gateway. The engineer proposes this change: Configure a reserved HA management interface and appropriate management routes where individual cluster members require independent access Which explanation most accurately establishes why that change meets the operational requirement?

  1. For the public transit operator deployment: HA configuration synchronization does not inherently preserve every live connection; session-pickup behavior and protocol support determine failover continuity.
  2. For the public transit operator deployment: FGCP members negotiate HA using compatible hardware and cluster parameters, while heartbeat connectivity drives membership and primary selection.
  3. For the public transit operator deployment: Packet arrival alone does not prove policy acceptance; a constrained flow trace can reveal the stage where FortiOS rejects or forwards the traffic.
  4. For the public transit operator deployment: Dedicated HA management lets operators address each member separately without confusing the cluster data-plane identity with member-specific diagnostic reachability.

Answer & Explanations

Correct answer: D

A.

That statement explains the distinct objective “Synchronize eligible sessions for HA failover” rather than the cause in this case. The deciding mechanism here is that dedicated HA management lets operators address each member separately without confusing the cluster data-plane identity with member-specific diagnostic reachability.

B.

That statement explains the distinct objective “Build compatible FGCP heartbeat and cluster membership” rather than the cause in this case. The deciding mechanism here is that dedicated HA management lets operators address each member separately without confusing the cluster data-plane identity with member-specific diagnostic reachability.

C.

That statement explains the distinct objective “Trace a denied flow at the correct forwarding stage” rather than the cause in this case. The deciding mechanism here is that dedicated HA management lets operators address each member separately without confusing the cluster data-plane identity with member-specific diagnostic reachability.

D.

Dedicated HA management lets operators address each member separately without confusing the cluster data-plane identity with member-specific diagnostic reachability. It supports this action: Configure a reserved HA management interface and appropriate management routes where individual cluster members require independent access. To confirm, reach each peer's unique management address and inspect role-specific health while production traffic continues through the cluster.

Question 73

A public transit operator needs to check both FortiGate peers directly during software maintenance without disturbing the shared gateway. After implementing this change—Configure a reserved HA management interface and appropriate management routes where individual cluster members require independent access—which test best establishes that the intended outcome is actually working?

  1. For the public transit operator deployment: Reach each peer's unique management address and inspect role-specific health while production traffic continues through the cluster.
  2. For the public transit operator deployment: Maintain representative long-lived TCP sessions, trigger an approved failover, and inspect session and HA status before and after.
  3. For the public transit operator deployment: Inspect HA status on both peers, confirm their cluster roles and heartbeat link health, and simulate a controlled member restart.
  4. For the public transit operator deployment: Reproduce one test connection and correlate ingress packet capture, trace ID, policy lookup and egress behavior.

Answer & Explanations

Correct answer: A

A.

Reach each peer's unique management address and inspect role-specific health while production traffic continues through the cluster. This is the relevant test because dedicated HA management lets operators address each member separately without confusing the cluster data-plane identity with member-specific diagnostic reachability.

B.

This check would help evaluate the distinct objective “Synchronize eligible sessions for HA failover”, but not the required result. For this case, reach each peer's unique management address and inspect role-specific health while production traffic continues through the cluster.

C.

This check would help evaluate the distinct objective “Build compatible FGCP heartbeat and cluster membership”, but not the required result. For this case, reach each peer's unique management address and inspect role-specific health while production traffic continues through the cluster.

D.

This check would help evaluate the distinct objective “Trace a denied flow at the correct forwarding stage”, but not the required result. For this case, reach each peer's unique management address and inspect role-specific health while production traffic continues through the cluster.

Question 74

A public transit operator needs to check both FortiGate peers directly during software maintenance without disturbing the shared gateway. Which proposed shortcut would most directly cause loss of reliable per-node health visibility during failover diagnosis, and therefore must be avoided for this requirement?

  1. For the public transit operator deployment: Assume identical firewall policies guarantee all in-flight sessions remain established after any failover.
  2. For the public transit operator deployment: Assign the same management IP to both nodes and troubleshoot individual members through one ambiguous endpoint.
  3. For the public transit operator deployment: Connect two mismatched units and assume any pair automatically forms a synchronized FGCP cluster.
  4. For the public transit operator deployment: Enable unrestricted debug tracing during peak traffic and infer acceptance only from ingress packet capture.

Answer & Explanations

Correct answer: B

A.

That shortcut would relate to the distinct objective “Synchronize eligible sessions for HA failover”, rather than the specific configuration or use described. The unsafe step in this case is to assign the same management IP to both nodes and troubleshoot individual members through one ambiguous endpoint.

B.

Assign the same management IP to both nodes and troubleshoot individual members through one ambiguous endpoint. The proposed shortcut bypasses the needed control: Dedicated HA management lets operators address each member separately without confusing the cluster data-plane identity with member-specific diagnostic reachability. The appropriate response is to configure a reserved HA management interface and appropriate management routes where individual cluster members require independent access.

C.

That shortcut would relate to the distinct objective “Build compatible FGCP heartbeat and cluster membership”, rather than the specific configuration or use described. The unsafe step in this case is to assign the same management IP to both nodes and troubleshoot individual members through one ambiguous endpoint.

D.

That shortcut would relate to the distinct objective “Trace a denied flow at the correct forwarding stage”, rather than the specific configuration or use described. The unsafe step in this case is to assign the same management IP to both nodes and troubleshoot individual members through one ambiguous endpoint.

Question 75

A public transit operator needs to check both FortiGate peers directly during software maintenance without disturbing the shared gateway. Which combined implementation and verification plan most reliably resolves this specific FortiOS 7.6 issue?

  1. For the public transit operator deployment: Implement: Enable applicable session pickup and verify that synchronized eligible sessions survive a planned active-passive FGCP failover. Verify: Reach each peer's unique management address and inspect role-specific health while production traffic continues through the cluster.
  2. For the public transit operator deployment: Implement: Configure a reserved HA management interface and appropriate management routes where individual cluster members require independent access. Verify: Inspect HA status on both peers, confirm their cluster roles and heartbeat link health, and simulate a controlled member restart.
  3. For the public transit operator deployment: Implement: Configure a reserved HA management interface and appropriate management routes where individual cluster members require independent access. Verify: Reach each peer's unique management address and inspect role-specific health while production traffic continues through the cluster.
  4. For the public transit operator deployment: Implement: Use bounded packet sniffing and diagnose debug flow for a specific source, destination and port to identify routing, policy or session decisions. Verify: Reproduce one test connection and correlate ingress packet capture, trace ID, policy lookup and egress behavior.

Answer & Explanations

Correct answer: C

A.

This action and check address the alternate operation “For the public transit operator deployment: Implement: Enable applicable session pickup and verify that synchronized eligible sessions survive a…” instead of the stated acceptance requirement. The matching action is to configure a reserved HA management interface and appropriate management routes where individual cluster members require independent access. Its confirming test is to reach each peer's unique management address and inspect role-specific health while production traffic continues through the cluster.

B.

This action and check address the alternate operation “For the public transit operator deployment: Implement: Configure a reserved HA management interface and appropriate management routes where individual…” instead of the stated acceptance requirement. The matching action is to configure a reserved HA management interface and appropriate management routes where individual cluster members require independent access. Its confirming test is to reach each peer's unique management address and inspect role-specific health while production traffic continues through the cluster.

C.

Implement: Configure a reserved HA management interface and appropriate management routes where individual cluster members require independent access. Verify: Reach each peer's unique management address and inspect role-specific health while production traffic continues through the cluster. This pairing is justified because dedicated HA management lets operators address each member separately without confusing the cluster data-plane identity with member-specific diagnostic reachability.

D.

This action and check address the distinct objective “Trace a denied flow at the correct forwarding stage” instead of the stated acceptance requirement. The matching action is to configure a reserved HA management interface and appropriate management routes where individual cluster members require independent access. Its confirming test is to reach each peer's unique management address and inspect role-specific health while production traffic continues through the cluster.

Question 76

A factory NOC must gather hardware health statistics from the standby cluster member before a planned failover. Which FortiOS 7.6 administrative change best meets this requirement without weakening unrelated controls?

  1. For the factory NOC deployment: Configure matching FGCP mode, group settings and dedicated heartbeat links on supported compatible FortiGate peers before joining them.
  2. For the factory NOC deployment: Use bounded packet sniffing and diagnose debug flow for a specific source, destination and port to identify routing, policy or session decisions.
  3. For the factory NOC deployment: Enable applicable session pickup and verify that synchronized eligible sessions survive a planned active-passive FGCP failover.
  4. For the factory NOC deployment: Configure a reserved HA management interface and appropriate management routes where individual cluster members require independent access.

Answer & Explanations

Correct answer: D

A.

The situation is that a factory NOC must gather hardware health statistics from the standby cluster member before a planned failover. This option addresses the distinct objective “Build compatible FGCP heartbeat and cluster membership”, not the immediate requirement. FGCP members negotiate HA using compatible hardware and cluster parameters, while heartbeat connectivity drives membership and primary selection. For the stated case, dedicated HA management lets operators address each member separately without confusing the cluster data-plane identity with member-specific diagnostic reachability.

B.

The situation is that a factory NOC must gather hardware health statistics from the standby cluster member before a planned failover. This option addresses the distinct objective “Trace a denied flow at the correct forwarding stage”, not the immediate requirement. Packet arrival alone does not prove policy acceptance; a constrained flow trace can reveal the stage where FortiOS rejects or forwards the traffic. For the stated case, dedicated HA management lets operators address each member separately without confusing the cluster data-plane identity with member-specific diagnostic reachability.

C.

The situation is that a factory NOC must gather hardware health statistics from the standby cluster member before a planned failover. This option addresses the distinct objective “Synchronize eligible sessions for HA failover”, not the immediate requirement. HA configuration synchronization does not inherently preserve every live connection; session-pickup behavior and protocol support determine failover continuity. For the stated case, dedicated HA management lets operators address each member separately without confusing the cluster data-plane identity with member-specific diagnostic reachability.

D.

The situation is that a factory NOC must gather hardware health statistics from the standby cluster member before a planned failover. Configure a reserved HA management interface and appropriate management routes where individual cluster members require independent access. This resolves the issue because dedicated HA management lets operators address each member separately without confusing the cluster data-plane identity with member-specific diagnostic reachability. To verify, reach each peer's unique management address and inspect role-specific health while production traffic continues through the cluster.

Question 77

A factory NOC must gather hardware health statistics from the standby cluster member before a planned failover. The engineer proposes this change: Configure a reserved HA management interface and appropriate management routes where individual cluster members require independent access Which explanation most accurately establishes why that change meets the operational requirement?

  1. For the factory NOC deployment: Dedicated HA management lets operators address each member separately without confusing the cluster data-plane identity with member-specific diagnostic reachability.
  2. For the factory NOC deployment: FGCP members negotiate HA using compatible hardware and cluster parameters, while heartbeat connectivity drives membership and primary selection.
  3. For the factory NOC deployment: Packet arrival alone does not prove policy acceptance; a constrained flow trace can reveal the stage where FortiOS rejects or forwards the traffic.
  4. For the factory NOC deployment: HA configuration synchronization does not inherently preserve every live connection; session-pickup behavior and protocol support determine failover continuity.

Answer & Explanations

Correct answer: A

A.

The situation is that a factory NOC must gather hardware health statistics from the standby cluster member before a planned failover. Dedicated HA management lets operators address each member separately without confusing the cluster data-plane identity with member-specific diagnostic reachability. It supports this action: Configure a reserved HA management interface and appropriate management routes where individual cluster members require independent access. To confirm, reach each peer's unique management address and inspect role-specific health while production traffic continues through the cluster.

B.

The situation is that a factory NOC must gather hardware health statistics from the standby cluster member before a planned failover. That statement explains the distinct objective “Build compatible FGCP heartbeat and cluster membership” rather than the cause in this case. The deciding mechanism here is that dedicated HA management lets operators address each member separately without confusing the cluster data-plane identity with member-specific diagnostic reachability.

C.

The situation is that a factory NOC must gather hardware health statistics from the standby cluster member before a planned failover. That statement explains the distinct objective “Trace a denied flow at the correct forwarding stage” rather than the cause in this case. The deciding mechanism here is that dedicated HA management lets operators address each member separately without confusing the cluster data-plane identity with member-specific diagnostic reachability.

D.

The situation is that a factory NOC must gather hardware health statistics from the standby cluster member before a planned failover. That statement explains the distinct objective “Synchronize eligible sessions for HA failover” rather than the cause in this case. The deciding mechanism here is that dedicated HA management lets operators address each member separately without confusing the cluster data-plane identity with member-specific diagnostic reachability.

Question 78

A factory NOC must gather hardware health statistics from the standby cluster member before a planned failover. After implementing this change—Configure a reserved HA management interface and appropriate management routes where individual cluster members require independent access—which test best establishes that the intended outcome is actually working?

  1. For the factory NOC deployment: Inspect HA status on both peers, confirm their cluster roles and heartbeat link health, and simulate a controlled member restart.
  2. For the factory NOC deployment: Reach each peer's unique management address and inspect role-specific health while production traffic continues through the cluster.
  3. For the factory NOC deployment: Reproduce one test connection and correlate ingress packet capture, trace ID, policy lookup and egress behavior.
  4. For the factory NOC deployment: Maintain representative long-lived TCP sessions, trigger an approved failover, and inspect session and HA status before and after.

Answer & Explanations

Correct answer: B

A.

The situation is that a factory NOC must gather hardware health statistics from the standby cluster member before a planned failover. This check would help evaluate the distinct objective “Build compatible FGCP heartbeat and cluster membership”, but not the required result. For this case, reach each peer's unique management address and inspect role-specific health while production traffic continues through the cluster.

B.

The situation is that a factory NOC must gather hardware health statistics from the standby cluster member before a planned failover. Reach each peer's unique management address and inspect role-specific health while production traffic continues through the cluster. This is the relevant test because dedicated HA management lets operators address each member separately without confusing the cluster data-plane identity with member-specific diagnostic reachability.

C.

The situation is that a factory NOC must gather hardware health statistics from the standby cluster member before a planned failover. This check would help evaluate the distinct objective “Trace a denied flow at the correct forwarding stage”, but not the required result. For this case, reach each peer's unique management address and inspect role-specific health while production traffic continues through the cluster.

D.

The situation is that a factory NOC must gather hardware health statistics from the standby cluster member before a planned failover. This check would help evaluate the distinct objective “Synchronize eligible sessions for HA failover”, but not the required result. For this case, reach each peer's unique management address and inspect role-specific health while production traffic continues through the cluster.

Question 79

A factory NOC must gather hardware health statistics from the standby cluster member before a planned failover. Which proposed shortcut would most directly cause loss of reliable per-node health visibility during failover diagnosis, and therefore must be avoided for this requirement?

  1. For the factory NOC deployment: Connect two mismatched units and assume any pair automatically forms a synchronized FGCP cluster.
  2. For the factory NOC deployment: Enable unrestricted debug tracing during peak traffic and infer acceptance only from ingress packet capture.
  3. For the factory NOC deployment: Assign the same management IP to both nodes and troubleshoot individual members through one ambiguous endpoint.
  4. For the factory NOC deployment: Assume identical firewall policies guarantee all in-flight sessions remain established after any failover.

Answer & Explanations

Correct answer: C

A.

The situation is that a factory NOC must gather hardware health statistics from the standby cluster member before a planned failover. That shortcut would relate to the distinct objective “Build compatible FGCP heartbeat and cluster membership”, rather than the specific configuration or use described. The unsafe step in this case is to assign the same management IP to both nodes and troubleshoot individual members through one ambiguous endpoint.

B.

The situation is that a factory NOC must gather hardware health statistics from the standby cluster member before a planned failover. That shortcut would relate to the distinct objective “Trace a denied flow at the correct forwarding stage”, rather than the specific configuration or use described. The unsafe step in this case is to assign the same management IP to both nodes and troubleshoot individual members through one ambiguous endpoint.

C.

The situation is that a factory NOC must gather hardware health statistics from the standby cluster member before a planned failover. Assign the same management IP to both nodes and troubleshoot individual members through one ambiguous endpoint. The proposed shortcut bypasses the needed control: Dedicated HA management lets operators address each member separately without confusing the cluster data-plane identity with member-specific diagnostic reachability. The appropriate response is to configure a reserved HA management interface and appropriate management routes where individual cluster members require independent access.

D.

The situation is that a factory NOC must gather hardware health statistics from the standby cluster member before a planned failover. That shortcut would relate to the distinct objective “Synchronize eligible sessions for HA failover”, rather than the specific configuration or use described. The unsafe step in this case is to assign the same management IP to both nodes and troubleshoot individual members through one ambiguous endpoint.

Question 80

A factory NOC must gather hardware health statistics from the standby cluster member before a planned failover. Which combined implementation and verification plan most reliably resolves this specific FortiOS 7.6 issue?

  1. For the factory NOC deployment: Implement: Use bounded packet sniffing and diagnose debug flow for a specific source, destination and port to identify routing, policy or session decisions. Verify: Reproduce one test connection and correlate ingress packet capture, trace ID, policy lookup and egress behavior.
  2. For the factory NOC deployment: Implement: Enable applicable session pickup and verify that synchronized eligible sessions survive a planned active-passive FGCP failover. Verify: Reach each peer's unique management address and inspect role-specific health while production traffic continues through the cluster.
  3. For the factory NOC deployment: Implement: Configure a reserved HA management interface and appropriate management routes where individual cluster members require independent access. Verify: Inspect HA status on both peers, confirm their cluster roles and heartbeat link health, and simulate a controlled member restart.
  4. For the factory NOC deployment: Implement: Configure a reserved HA management interface and appropriate management routes where individual cluster members require independent access. Verify: Reach each peer's unique management address and inspect role-specific health while production traffic continues through the cluster.

Answer & Explanations

Correct answer: D

A.

The situation is that a factory NOC must gather hardware health statistics from the standby cluster member before a planned failover. This action and check address the distinct objective “Trace a denied flow at the correct forwarding stage” instead of the stated acceptance requirement. The matching action is to configure a reserved HA management interface and appropriate management routes where individual cluster members require independent access. Its confirming test is to reach each peer's unique management address and inspect role-specific health while production traffic continues through the cluster.

B.

This action and check address the alternate operation “For the factory NOC deployment: Implement: Enable applicable session pickup and verify that synchronized eligible sessions survive a planned…” instead of the stated acceptance requirement. The matching action is to configure a reserved HA management interface and appropriate management routes where individual cluster members require independent access. Its confirming test is to reach each peer's unique management address and inspect role-specific health while production traffic continues through the cluster.

C.

This action and check address the alternate operation “For the factory NOC deployment: Implement: Configure a reserved HA management interface and appropriate management routes where individual cluster…” instead of the stated acceptance requirement. The matching action is to configure a reserved HA management interface and appropriate management routes where individual cluster members require independent access. Its confirming test is to reach each peer's unique management address and inspect role-specific health while production traffic continues through the cluster.

D.

The situation is that a factory NOC must gather hardware health statistics from the standby cluster member before a planned failover. Implement: Configure a reserved HA management interface and appropriate management routes where individual cluster members require independent access. Verify: Reach each peer's unique management address and inspect role-specific health while production traffic continues through the cluster. This pairing is justified because dedicated HA management lets operators address each member separately without confusing the cluster data-plane identity with member-specific diagnostic reachability.

FCP - FortiGate 7.6 Administrator (FCP_FGT_AD-7.6) Practice Tests