McAfee-Secured Website

Exam Bundle

Exam Code: Professional Cloud Architect

Exam Name Google Cloud Certified - Professional Cloud Architect

Certification Provider: Google

Corresponding Certification: Professional Cloud Architect

Google Professional Cloud Architect Bundle $44.99

100% Google Professional Cloud Architect Guaranteed Success With Testking Training Materials

Get Real Professional Cloud Architect Exam Questions & Expert Verified Answers as Experienced on the Real Exam!

  • Questions & Answers

    Professional Cloud Architect Questions & Answers

    370 Questions & Answers

    The ultimate exam preparation tool, Questions & Answers mirrors the actual Professional Cloud Architect exam while allowing you to hone and measure every aspect of your test-taking ability.

  • Professional Cloud Architect Video Course

    Professional Cloud Architect Video Course

    63 Video Lectures

    Professional Cloud Architect Video Course is developed by Google Professionals to help you pass the Professional Cloud Architect exam.

    Description

    This course will improve your knowledge and skills required to pass Google Cloud Certified - Professional Cloud Architect exam.
  • Study Guide

    Professional Cloud Architect Study Guide

    491 PDF Pages

    Developed by industry experts, this 491-page guide spells out in painstaking detail all of the information you need to ace Professional Cloud Architect exam.

Google Cloud Certified - Professional Cloud Architect (Professional Cloud Architect) Practice Tests

Satisfaction Guaranteed

Money Back Guarantee

Testking provides hassle-free money back guarantee with our products. That is because we have 100% trust in the abilities of our professional and experience product team, and our record is a proof of that.

99.6% PASS RATE
Total Cost: $194.97
Bundle Price: $149.98

Purchase Individually

  • Questions & Answers

    Questions & Answers

    370 Questions

    $124.99
  • Professional Cloud Architect Video Course

    Video Course

    63 Video Lectures

    $39.99
  • Study Guide

    Study Guide

    491 PDF Pages

    $29.99

Google Professional Cloud Architect Practice Test 31

Question 300

Scenario: A compliance team needs tamper-resistant centralized audit evidence from many projects. What should the architect verify before declaring this requirement met?

  1. A central log destination contains only VM uptime metrics, not administrative change events
  2. Sample admin and data access events appear centrally with controlled retention
  3. Logs arrive but project developers can delete archived evidence immediately
  4. The aggregated sink captures admin activity but required data access records are missing

Explanation

Correct Answer

B:

This measures the intended outcome rather than a proxy or administrative milestone: Sample admin and data access events appear centrally with controlled retention.

Incorrect Answers

A:

Resource health is not the necessary security audit trail.

C:

Centralization without protected retention is insufficient.

D:

An incomplete event set cannot support the documented compliance review.

Question 301

Scenario: A security operations team has no visibility into risky cloud configuration drift. Which decision best closes the identified operational or design gap?

  1. Enable suitable Security Command Center findings and defined remediation ownership
  2. Depend on annual manual spreadsheet reviews to catch dangerous cloud configuration drift
  3. Collect raw logs without enabling actionable findings or assigning remediation owners
  4. Grant all application engineers Security Admin privileges so they can self-correct unseen issues

Explanation

Correct Answer

A:

Enable suitable Security Command Center findings and defined remediation ownership. The goal acceptance evidence is: An introduced misconfiguration raises a tracked finding with an assigned owner.

Incorrect Answers

B:

Periodic manual spot-checks may miss unreviewed high-risk changes.

C:

Log volume alone is not governed risk identification and response.

D:

Broad permissions can create more drift and bypass review.

Question 302

Scenario: A security operations team has no visibility into risky cloud configuration drift. Which test result would best distinguish successful implementation from progress alone?

  1. A risky public bucket is created but no finding or alert identifies it
  2. An introduced misconfiguration raises a tracked finding with an assigned owner
  3. The scanner finds a misconfiguration but no accountable team tracks remediation
  4. The dashboard lists a generic high risk score without resource-specific evidence

Explanation

Correct Answer

B:

This measures the intended outcome rather than a proxy or administrative milestone: An introduced misconfiguration raises a tracked finding with an assigned owner.

Incorrect Answers

A:

The expected security detection mechanism is ineffective.

C:

Detection without an owner does not close the control gap.

D:

An uninterpretable score alone does not demonstrate useful security findings.

Question 303

Scenario: A production API exposes authentication tokens in diagnostic logs. What change should be selected to meet the most critical requirement?

  1. Store all raw authentication headers in logs and restrict viewing to administrators
  2. Hash only the username while leaving bearer tokens and session secrets intact
  3. Redact sensitive fields and scope logging access and retention
  4. Disable all logging permanently instead of selectively removing sensitive fields

Explanation

Correct Answer

C:

Redact sensitive fields and scope logging access and retention. The goal acceptance evidence is: Red-team test tokens are absent from logs while troubleshooting metadata remains.

Incorrect Answers

A:

Even privileged logs should avoid unnecessary reusable credential material.

B:

Redacting identity labels does not remove actionable sensitive authentication tokens.

D:

Eliminating logs can make legitimate incident diagnosis and audit requirements impossible.

Question 304

Scenario: A production API exposes authentication tokens in diagnostic logs. Which measurable result best confirms the desired operating behavior?

  1. A test access token appears verbatim in diagnostic logs despite enabled redaction settings
  2. Token values disappear but all correlation and error fields are also missing
  3. Red-team test tokens are absent from logs while troubleshooting metadata remains
  4. The log bucket is access controlled, but production traces still contain usable session credentials

Explanation

Correct Answer

C:

This measures the intended outcome rather than a proxy or administrative milestone: Red-team test tokens are absent from logs while troubleshooting metadata remains.

Incorrect Answers

A:

The actual sensitive value is still exposed.

B:

Excessive scrubbing may make required troubleshooting impossible.

D:

Storage permissions alone cannot excuse leaking sensitive authentication data.

Question 305

Scenario: A company deploys custom generative AI agents on restricted customer material. Which course of action best fits the scenario and its constraints?

  1. Give the AI agent direct organization-wide access to customer storage for convenience
  2. Limit model endpoint identities, approved data paths, and tool permissions
  3. Enable all third-party tools for the agent and depend only on system prompts to prevent harmful calls
  4. Record only successful agent conversations and ignore denied tool attempts

Explanation

Correct Answer

B:

Limit model endpoint identities, approved data paths, and tool permissions. The goal acceptance evidence is: An unauthorized tool call or data retrieval is denied in sandbox tests.

Incorrect Answers

A:

Broad data privileges violate the restricted material boundary.

C:

Prompt instructions are not a security boundary for tool authorization.

D:

Missing security events prevents detection of tool misuse or data exfiltration attempts.

Question 306

Scenario: A company deploys custom generative AI agents on restricted customer material. What evidence would make the proposed resolution defensible to reviewers?

  1. An agent completes a customer request but its logs show access to unrelated customer records
  2. Unauthorized test tools can be invoked by injecting instructions into retrieved documents
  3. An unauthorized tool call or data retrieval is denied in sandbox tests
  4. Model responses look accurate while access tests reveal endpoint identities with unrestricted permissions

Explanation

Correct Answer

C:

This measures the intended outcome rather than a proxy or administrative milestone: An unauthorized tool call or data retrieval is denied in sandbox tests.

Incorrect Answers

A:

Task completion is not proof of properly scoped data retrieval.

B:

A tool path that bypasses policy fails the trusted boundary.

D:

Output quality is distinct from isolation of sensitive data and tools.

Question 307

Scenario: A public-facing chat assistant can receive instructions embedded in untrusted web pages. Which decision best addresses this requirement given the described constraints?

  1. Add a system prompt saying that all web pages are trusted instructions and let the model run their commands
  2. Apply appropriate prompt-injection and unsafe-content protections such as Model Armor where supported
  3. Log the assistant output after it is sent to external tools, then review leaked tokens weekly
  4. Disable content screening when the model sees external URLs to reduce latency

Explanation

Correct Answer

B:

Model Armor inspects prompts and responses for policy violations, including supported prompt-injection risks. Pair its template-based screening with explicit application decisions to reject or sanitize suspect retrieved instructions; the model must not blindly execute untrusted tool directives.

Incorrect Answers

A:

This explicitly promotes lower-trust retrieved text into instructions and amplifies prompt-injection risk.

C:

Post-hoc logs do not prevent execution of injected instructions or first-time sensitive-data exposure.

D:

Removing screening on untrusted retrieved content increases the attack surface under the stated threat.

Question 308

Scenario: A public-facing chat assistant can receive instructions embedded in untrusted web pages. The team has selected this architecture decision: Apply appropriate prompt-injection and unsafe-content protections such as Model Armor where supported. Which result would verify the specified requirement rather than merely report a configuration change?

  1. A poisoned web page causes the agent to email a secret before a dashboard later records the event
  2. Adversarial retrieval tests block unsafe instructions without leaking confidential data
  3. The chatbot refuses one harmless prompt while still executing instructions embedded in search results
  4. The app receives screening results but ignores the violation decision when invoking tools

Explanation

Correct Answer

B:

Adversarial samples embedded in retrieved web content are flagged and blocked by the integrated app policy; protected data are not disclosed and legitimate inputs still produce useful outputs.

Incorrect Answers

A:

Detection after exfiltration is not preventive enforcement.

C:

False positives do not demonstrate that retrieved prompt-injection paths are controlled.

D:

Unenforced scan output does not prevent malicious retrieved text from influencing tool execution.

Question 309

Scenario: A customer-support agent should not be able to invoke production billing mutations. What should the solution architect recommend to resolve this requirement?

  1. Give the customer-support agent write permission to all billing tables because it needs to read invoices
  2. Rely on prompt wording such as never change payments while exposing production update tools
  3. Isolate read-only agent tools and require explicit approval for transactional tools
  4. Allow automatic refund execution if the agent says the customer seems satisfied

Explanation

Correct Answer

C:

Isolate read-only agent tools and require explicit approval for transactional tools. The goal acceptance evidence is: A test agent can summarize invoices but cannot change payment records.

Incorrect Answers

A:

Read operations do not require broad transaction mutation rights.

B:

Text instructions cannot substitute for enforced backend tool scope.

D:

A subjective model judgment is not explicit approval for money-moving actions.

Question 310

Scenario: A customer-support agent should not be able to invoke production billing mutations. Which observable result most directly demonstrates success?

  1. The agent answers invoice questions but a test prompt also changes card payment details
  2. A test agent can summarize invoices but cannot change payment records
  3. An approval button is visible but backend APIs accept write calls without it
  4. The billing service records changes made by a generic agent identity with no approval trace

Explanation

Correct Answer

B:

This measures the intended outcome rather than a proxy or administrative milestone: A test agent can summarize invoices but cannot change payment records.

Incorrect Answers

A:

Read capability must not confer unapproved financial mutation.

C:

A UI-only gate does not enforce transactional authorization.

D:

An attributable human decision must precede privileged changes.