100% Latest & Updated Cisco 350-901 Practice Test Questions, Exam Dumps & Verified Answers!
30 Days Free Updates, Instant Download!
Question 50
A Terraform run fails after a provider upgrade silently changes a resource attribute. Independently, reviewers flag this potentially unsafe practice for assessment: 'Approve a provider upgrade from a successful download alone'. Which verification addresses that practice without assuming it caused the incident?
- Allow unrestricted automatic upgrades in production CI
- Review variable injection and credential exposure in state/logs
- Compare locked provider hashes across runners
- Inspect init, validate and plan outputs against staging baseline
Explanation
Correct Answer
D:Installation success does not prove safe plan results. The intended behavior is: A new provider needs schema and behavior verification.
Incorrect Answers
A:Untested upgrades can change schema or behavior.
B:This checks separate secret settings, not the named failure. Credentials must not be embedded in Terraform source.
C:This checks preserve dependency lock, not the named failure. The lockfile records selected provider versions and hashes.
Question 51
An enterprise stores Terraform state for a fleet of network VLANs in a shared backend and needs reliable concurrency. The specific implementation objective is to centralize authorized state. Which configuration best meets the stated requirement?
- Run forced replacement on all resources whenever a drift alert appears
- Use a controlled remote backend appropriate to the provider and repository
- Publish the entire Terraform state as a CI artifact for anonymous users
- Launch two unrestricted apply jobs against the same state
Explanation
Correct Answer
B:Shared state prevents inconsistent local views.
Incorrect Answers
A:Destructive blanket refresh is not a safe reconciliation plan.
C:State exposure can disclose network and credential details.
D:Simultaneous applies race to modify the same infrastructure.
Question 52
An enterprise stores Terraform state for a fleet of network VLANs in a shared backend and needs reliable concurrency. The specific implementation objective is to prevent parallel mutations. Which design action is most appropriate for this case?
- Enable supported state locking or serialized change execution
- Keep writable local state files on every engineer laptop
- Publish the entire Terraform state as a CI artifact for anonymous users
- Run forced replacement on all resources whenever a drift alert appears
Explanation
Correct Answer
A:Concurrent writes can corrupt tracked resource relationships.
Incorrect Answers
B:Divergent snapshots can overwrite one another.
C:State exposure can disclose network and credential details.
D:Destructive blanket refresh is not a safe reconciliation plan.
Question 53
An enterprise stores Terraform state for a fleet of network VLANs in a shared backend and needs reliable concurrency. The specific implementation objective is to protect sensitive state. What should the engineer implement for this specific objective?
- Launch two unrestricted apply jobs against the same state
- Run forced replacement on all resources whenever a drift alert appears
- Keep writable local state files on every engineer laptop
- Encrypt backend storage and restrict read/write access to the minimum roles
Explanation
Correct Answer
D:State may contain endpoint identifiers and secrets.
Incorrect Answers
A:Simultaneous applies race to modify the same infrastructure.
B:Destructive blanket refresh is not a safe reconciliation plan.
C:Divergent snapshots can overwrite one another.
Question 54
An enterprise stores Terraform state for a fleet of network VLANs in a shared backend and needs reliable concurrency. The specific implementation objective is to reconcile actual versus tracked state. Which adjustment would satisfy this particular constraint?
- Launch two unrestricted apply jobs against the same state
- Use plan and state inspection to resolve drift before the next apply
- Publish the entire Terraform state as a CI artifact for anonymous users
- Keep writable local state files on every engineer laptop
Explanation
Correct Answer
B:State is not proof the live controller still matches it.
Incorrect Answers
A:Simultaneous applies race to modify the same infrastructure.
C:State exposure can disclose network and credential details.
D:Divergent snapshots can overwrite one another.
Question 55
An enterprise stores Terraform state for a fleet of network VLANs in a shared backend and needs reliable concurrency. The deployment plan must both (1) prevent parallel mutations and (2) reconcile actual versus tracked state. Which TWO actions together satisfy those requirements? Choose TWO.
- Enable supported state locking or serialized change execution
- Run forced replacement on all resources whenever a drift alert appears
- Publish the entire Terraform state as a CI artifact for anonymous users
- Launch two unrestricted apply jobs against the same state
- Use plan and state inspection to resolve drift before the next apply
Explanation
Correct Answers
A:Concurrent writes can corrupt tracked resource relationships.
E:State is not proof the live controller still matches it.
Incorrect Answers
B:Destructive blanket refresh is not a safe reconciliation plan.
C:State exposure can disclose network and credential details.
D:Simultaneous applies race to modify the same infrastructure.
Question 56
An enterprise stores Terraform state for a fleet of network VLANs in a shared backend and needs reliable concurrency. Three conditions must be met: (1) prevent parallel mutations; (2) protect sensitive state; and (3) reconcile actual versus tracked state. Which THREE actions form the narrowest complete plan? Choose THREE.
- Run forced replacement on all resources whenever a drift alert appears
- Launch two unrestricted apply jobs against the same state
- Keep writable local state files on every engineer laptop
- Use plan and state inspection to resolve drift before the next apply
- Enable supported state locking or serialized change execution
- Encrypt backend storage and restrict read/write access to the minimum roles
Explanation
Correct Answers
D:State is not proof the live controller still matches it.
E:Concurrent writes can corrupt tracked resource relationships.
F:State may contain endpoint identifiers and secrets.
Incorrect Answers
A:Destructive blanket refresh is not a safe reconciliation plan.
B:Simultaneous applies race to modify the same infrastructure.
C:Divergent snapshots can overwrite one another.
Question 57
Two controllers show conflicting VLAN settings after simultaneous pipelines wrote infrastructure state. Independently, reviewers flag this potentially unsafe practice for assessment: 'Keep writable local state files on every engineer laptop'. Which inspection most directly evaluates that practice?
- Inspect actual backend configuration and state lineage
- Launch two unrestricted apply jobs against the same state
- Compare plan diff and controller readback for suspected drift
- Audit backend encryption and access control
Explanation
Correct Answer
A:Divergent snapshots can overwrite one another. The intended behavior is: Shared state prevents inconsistent local views.
Incorrect Answers
B:Simultaneous applies race to modify the same infrastructure.
C:This checks reconcile actual versus tracked state, not the named failure. State is not proof the live controller still matches it.
D:This checks protect sensitive state, not the named failure. State may contain endpoint identifiers and secrets.
Question 58
Two controllers show conflicting VLAN settings after simultaneous pipelines wrote infrastructure state. Independently, reviewers flag this potentially unsafe practice for assessment: 'Launch two unrestricted apply jobs against the same state'. What diagnostic evidence should first be collected to assess that practice?
- Review lock records and pipeline serialization
- Compare plan diff and controller readback for suspected drift
- Keep writable local state files on every engineer laptop
- Audit backend encryption and access control
Explanation
Correct Answer
A:Simultaneous applies race to modify the same infrastructure. The intended behavior is: Concurrent writes can corrupt tracked resource relationships.
Incorrect Answers
B:This checks reconcile actual versus tracked state, not the named failure. State is not proof the live controller still matches it.
C:Divergent snapshots can overwrite one another.
D:This checks protect sensitive state, not the named failure. State may contain endpoint identifiers and secrets.
Question 59
Two controllers show conflicting VLAN settings after simultaneous pipelines wrote infrastructure state. Independently, reviewers flag this potentially unsafe practice for assessment: 'Publish the entire Terraform state as a CI artifact for anonymous users'. Which focused check best tests this specific assumption?
- Compare plan diff and controller readback for suspected drift
- Audit backend encryption and access control
- Review lock records and pipeline serialization
- Keep writable local state files on every engineer laptop
Explanation
Correct Answer
B:State exposure can disclose network and credential details. The intended behavior is: State may contain endpoint identifiers and secrets.
Incorrect Answers
A:This checks reconcile actual versus tracked state, not the named failure. State is not proof the live controller still matches it.
C:This checks prevent parallel mutations, not the named failure. Concurrent writes can corrupt tracked resource relationships.
D:Divergent snapshots can overwrite one another.
Question 60
Two controllers show conflicting VLAN settings after simultaneous pipelines wrote infrastructure state. Independently, reviewers flag this potentially unsafe practice for assessment: 'Run forced replacement on all resources whenever a drift alert appears'. Which verification addresses that practice without assuming it caused the incident?
- Review lock records and pipeline serialization
- Compare plan diff and controller readback for suspected drift
- Keep writable local state files on every engineer laptop
- Audit backend encryption and access control
Explanation
Correct Answer
B:Destructive blanket refresh is not a safe reconciliation plan. The intended behavior is: State is not proof the live controller still matches it.
Incorrect Answers
A:This checks prevent parallel mutations, not the named failure. Concurrent writes can corrupt tracked resource relationships.
C:Divergent snapshots can overwrite one another.
D:This checks protect sensitive state, not the named failure. State may contain endpoint identifiers and secrets.
Designing, Deploying, and Managing Network Automation Systems (350-901) Practice Tests