McAfee-Secured Website

Exam Bundle

Exam Code: Associate Cloud Engineer

Exam Name Associate Cloud Engineer

Certification Provider: Google

Corresponding Certification: Associate Cloud Engineer

Google Associate Cloud Engineer Bundle $44.99

100% Google Associate Cloud Engineer Guaranteed Success With Testking Training Materials

Get Real Associate Cloud Engineer Exam Questions & Expert Verified Answers as Experienced on the Real Exam!

  • Questions & Answers

    Associate Cloud Engineer Questions & Answers

    391 Questions & Answers

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

  • Associate Cloud Engineer Video Course

    Associate Cloud Engineer Video Course

    234 Video Lectures

    Associate Cloud Engineer Video Course is developed by Google Professionals to help you pass the Associate Cloud Engineer exam.

    Description

    This course will improve your knowledge and skills required to pass Associate Cloud Engineer exam.
  • Study Guide

    Associate Cloud Engineer Study Guide

    849 PDF Pages

    Developed by industry experts, this 849-page guide spells out in painstaking detail all of the information you need to ace Associate Cloud Engineer exam.

Associate Cloud Engineer 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

    391 Questions

    $124.99
  • Associate Cloud Engineer Video Course

    Video Course

    234 Video Lectures

    $39.99
  • Study Guide

    Study Guide

    849 PDF Pages

    $29.99

Google Associate Cloud Engineer Practice Test 2

Question 10

A Cloud Monitoring request-based SLO sets 99% availability over a window with 15,000 total eligible requests. Assuming each request counts equally and no exclusions apply, how many unsuccessful requests may occur without exceeding the numerical error budget?

  1. 150 failed requests
  2. 14,850 failed requests
  3. 225 failed requests
  4. 149 failed requests

Explanation

Correct Answer

A

Allowed failures are 1.00% of 15,000, which equals 150 requests.

Incorrect Answers

B:

14,850 is the successful-request target rather than the allowed failure count.

C:

225 exceeds the SLO failure allowance of 150 in the specified technical environment and operational constraints.

D:

149 is below the maximum tolerated failures, so it is not the calculated budget.

Question 11

An organization contains 3 Google Cloud projects. Exactly 1 of them is a child of a single finance folder; the rest belong to other folders. An IAM role is bound at the finance folder. Assuming no other bindings or deny policies, how many of these projects inherit that folder-level allow?

  1. 1 project
  2. 3 projects
  3. 2 projects
  4. 4 projects

Explanation

Correct Answer

A

Only 1 project descended from that folder inherit the IAM role binding.

Incorrect Answers

B:

3 counts projects in unrelated folders that do not inherit this binding.

C:

2 counts unrelated projects instead of finance descendants in the specified technical environment and operational constraints.

D:

4 double-counts projects and violates the resource hierarchy in the specified technical environment and operational constraints.

Question 12

A cloud administrator is implementing a two-part operational checklist before a scheduled release. These two items involve different cloud resources. The first requirement is that a compromised service account key must stop authenticating. The second requirement is that a CI job should authenticate as a narrowly scoped service identity. Which ordered action pair resolves the first and second items respectively?

  1. change only the Cloud DNS zone description; then let its federated principal impersonate a dedicated deployment service account
  2. change only the Cloud DNS zone description; then reuse a human organization Owner login
  3. disable or delete the affected service account key and rotate dependencies; then let its federated principal impersonate a dedicated deployment service account
  4. disable or delete the affected service account key and rotate dependencies; then reuse a human organization Owner login

Explanation

Correct Answer

C

Revoking a key removes that credential from use. Dedicated impersonation ties delegated access to a scoped identity.

Incorrect Answers

A:

The second action is valid: Dedicated impersonation ties delegated access to a scoped identity. But the first fails: DNS description changes cannot revoke issued credentials.

B:

Both requirements remain unmet: DNS description changes cannot revoke issued credentials; Human Owner reuse weakens least privilege and auditability.

D:

The first action is valid: Revoking a key removes that credential from use. But the second fails: Human Owner reuse weakens least privilege and auditability.

Question 13

A site reliability engineer is reviewing two service requirements from the runbook. Each requirement refers to its own target workload. First, a production project must reside beneath the finance division policy boundary. Also, a principal needs least-privilege read access to project resources. Which option correctly addresses requirement one and then requirement two?

  1. assign finance tags only to one virtual machine; then grant basic Owner role at organization scope
  2. move or create the project under the finance folder in the resource hierarchy; then grant an appropriate read-only predefined IAM role at project scope
  3. move or create the project under the finance folder in the resource hierarchy; then grant basic Owner role at organization scope
  4. assign finance tags only to one virtual machine; then grant an appropriate read-only predefined IAM role at project scope

Explanation

Correct Answer

B

Project ancestry determines inherited folder policies. Predefined read roles can limit privileges and scope.

Incorrect Answers

A:

Both requirements remain unmet: A VM label cannot alter the project ancestry; Organization Owner grants excessive inherited permissions.

C:

The first action is valid: Project ancestry determines inherited folder policies. But the second fails: Organization Owner grants excessive inherited permissions.

D:

The second action is valid: Predefined read roles can limit privileges and scope. But the first fails: A VM label cannot alter the project ancestry.

Question 14

A team's alert-only budget sent an email after the monthly threshold, yet workloads continued to accrue charges. Management expected services to stop automatically without disrupting other projects. What should the engineer explain and configure?

  1. Delete the project automatically whenever a threshold notification is issued
  2. Assume reaching the alert threshold blocks all new API requests
  3. Review Cloud Billing Reports more often, with no approved mechanism to act on an over-budget alert
  4. Keep the budget as a notification signal and implement an explicitly approved automated cost-control workflow

Explanation

Correct Answer

D

Alert-only budgets notify; they do not impose hard usage limits, so a separate controlled response is necessary.

Incorrect Answers

A:

Project deletion is disproportionate and destructive; a budget email is not an authorization to delete workloads.

B:

Alert-only budgets do not stop API calls; charges can continue past the configured threshold.

C:

More frequent cost reporting improves awareness but cannot enforce a controlled workload response to overspending.

Question 15

A stateless web service running on one zonal managed instance group must remain available if a zone fails. The application already uses a compatible instance template and health checks. Which planned compute architecture reduces single-zone dependency?

  1. Keep a single zonal MIG and add an HTTP health check without changing where replicas run
  2. Deploy a regional managed instance group distributing instances across multiple zones in a region
  3. Increase the number of instances only within the current zonal group
  4. Create more boot disks in the same zone without creating instances

Explanation

Correct Answer

B

Regional MIGs can spread identical instances across zones while maintaining a shared autoscaling and repair configuration.

Incorrect Answers

A:

A health check detects failures, but all instances remain in the same unavailable zone during a zonal outage.

C:

Additional replicas in one zone remain vulnerable to the same zonal failure.

D:

Disks do not independently serve application requests or provide multi-zone availability.

Question 16

Before promoting an environment into production, operations must finish two tasks. Each requirement refers to its own target workload. First, an ETL workflow must transform both batch and streaming event data. Also, an analytics team must share query results without operating database servers. Which ordered pair gives an appropriate answer for task one, then task two?

  1. use Dataflow pipelines for managed data processing; then use BigQuery datasets with scoped access
  2. use Dataflow pipelines for managed data processing; then deploy an unmanaged VM-only PostgreSQL server as default
  3. use Cloud Armor as the data transformation engine; then use BigQuery datasets with scoped access
  4. use Cloud Armor as the data transformation engine; then deploy an unmanaged VM-only PostgreSQL server as default

Explanation

Correct Answer

A

Dataflow supports managed Apache Beam batch and stream processing. BigQuery is managed for analytical queries and scoped sharing.

Incorrect Answers

B:

The first action is valid: Dataflow supports managed Apache Beam batch and stream processing. But the second fails: An unmanaged relational VM introduces avoidable administration overhead.

C:

The second action is valid: BigQuery is managed for analytical queries and scoped sharing. But the first fails: Cloud Armor is web protection rather than ETL execution.

D:

Both requirements remain unmet: Cloud Armor is web protection rather than ETL execution; An unmanaged relational VM introduces avoidable administration overhead.

Question 17

Three internal application servers should receive TCP traffic only from clients on the private corporate VPC. An external endpoint would violate the deployment policy. Which load-balancing class meets this scope?

  1. Point a public Cloud DNS A record directly at a VM's ephemeral external address
  2. Create a global external Application Load Balancer with a public IP address
  3. Configure an internal Application Load Balancer for the raw unproxied TCP protocol without checking protocol support
  4. Configure an internal passthrough Network Load Balancer with eligible private backends

Explanation

Correct Answer

D

Internal passthrough load balancing provides a private regional frontend for supported TCP/UDP services.

Incorrect Answers

A:

Public DNS and external VM addresses expose the service rather than creating private load balancing.

B:

An external internet-facing frontend violates the requirement for an internal-only TCP service.

C:

Application load balancers proxy eligible HTTP(S) traffic, not an arbitrary private passthrough TCP listener.

Question 18

An enterprise platform team is preparing changes that must succeed in the same environment. Each requirement refers to its own target workload. First, a Terraform author wants consistent HCL formatting. Also, a security review needs the difference between desired and deployed Terraform. Which option lists the right first configuration followed by the right second configuration?

  1. run terraform fmt on configuration files; then refresh state appropriately and review terraform plan drift
  2. run terraform fmt on configuration files; then run only a DNS name lookup on the service
  3. run gcloud auth revoke for all collaborators; then refresh state appropriately and review terraform plan drift
  4. run gcloud auth revoke for all collaborators; then run only a DNS name lookup on the service

Explanation

Correct Answer

A

Terraform fmt normalizes supported configuration syntax. Plan reveals proposed reconciliation against known state.

Incorrect Answers

B:

The first action is valid: Terraform fmt normalizes supported configuration syntax. But the second fails: DNS resolution does not inspect Terraform drift.

C:

The second action is valid: Plan reveals proposed reconciliation against known state. But the first fails: Revoking credentials does not format configuration.

D:

Both requirements remain unmet: Revoking credentials does not format configuration; DNS resolution does not inspect Terraform drift.

Question 19

A GKE Deployment has CPU-based Horizontal Pod Autoscaling configured, but utilization percentages are unavailable because the containers have no CPU requests. What should the operator specify in the pod template?

  1. Increase the container image tag length so CPU utilization appears in monitoring
  2. Set realistic container CPU resource requests and keep the CPU utilization HPA configured against them
  3. Change the Kubernetes Service type from ClusterIP to NodePort to expose CPU metrics
  4. Lower the HPA target CPU percentage while still omitting CPU requests from the Pod template

Explanation

Correct Answer

B

CPU utilization-based HPA needs CPU resource requests to calculate utilization ratios.

Incorrect Answers

A:

Image tags identify builds and do not define a CPU utilization denominator.

C:

Service exposure does not supply missing container CPU requests in the HPA metric calculation.

D:

A different target does not supply the request-based denominator needed for CPU utilization-based HPA decisions.

Question 20

An engineer overwrote a Cloud Storage object that contained the prior configuration of a batch application. The team wants future overwrites to keep earlier object generations available for supported recovery. Which bucket feature addresses that requirement?

  1. Increase the project's VM instance count whenever an object changes
  2. Apply a bucket retention policy without enabling versioning and expect every overwrite to create a retrievable prior generation
  3. Enable Object Versioning and apply appropriate noncurrent-version lifecycle management
  4. Set the bucket DNS TTL to a longer interval before every object overwrite

Explanation

Correct Answer

C

Versioning retains previous generations when overwritten so recovery can select an earlier object generation.

Incorrect Answers

A:

Additional compute capacity does not maintain object history in Cloud Storage.

B:

Retention can constrain replacement but does not itself create an ordinary history of older live object generations.

D:

DNS caching controls name resolution, not preservation of earlier Cloud Storage object generations.