cert
cert-1
cert-2

Pass Google Associate Cloud Engineer Exam in First Attempt Guaranteed!

Get 100% Latest Exam Questions, Accurate & Verified Answers to Pass the Actual Exam!
30 Days Free Updates, Instant Download!

cert-5
cert-6
Associate Cloud Engineer Exam - Verified By Experts
Associate Cloud Engineer Premium Bundle
$29.98

Associate Cloud Engineer Premium Bundle

$79.99
$109.97
  • Premium File 391 Questions & Answers. Last update: Nov 10, 2026
  • Training Course 234 Video Lectures
  • Study Guide 849 Pages
 
$109.97
$79.99
accept 42 downloads in last 7 days
block-screenshots
Purchase Google Associate Cloud Engineer Exam Training Products Individually
Google Associate Cloud Engineer Premium File
Premium File 391 Q&A
$76.99$69.99
Google Associate Cloud Engineer Video Training Course
Training Course 234 Lectures
$21.99 $19.99
Google Associate Cloud Engineer PDF Study Guide
Study Guide 849 Pages
$21.99 $19.99
Google Associate Cloud Engineer Practice Test 23

Question 220

Cloud Run revision B begins returning HTTP 5xx immediately after a release, while revision A remains healthy. The team wants to restore service rapidly without rebuilding a container image. What should it do first?

  1. Increase revision B's CPU allocation and continue sending it all traffic without reviewing the errors.
  2. Delete revision A so every request is forced onto the new failing revision.
  3. Change the Cloud Run service maximum instance limit; that automatically restores an older working revision.
  4. Shift the service traffic allocation back to the previously healthy revision A and investigate revision B off the serving path.

Explanation

Correct Answer

D:

Shift the service traffic allocation back to the previously healthy revision A and investigate revision B off the serving path. Cloud Run supports traffic migration among revisions. Moving production traffic to the healthy revision restores service without a new image build while the faulty revision is investigated.

Incorrect Answers

A:

Increase revision B's CPU allocation and continue sending it all traffic without reviewing the errors. A generic CPU increase is not justified by the given evidence; it keeps the known failing revision exposed to users and delays safe rollback.

B:

Delete revision A so every request is forced onto the new failing revision. Removing the healthy revision would eliminate the known-good rollback target and cannot resolve the newly introduced 5xx regression.

C:

Change the Cloud Run service maximum instance limit; that automatically restores an older working revision. Instance limits control autoscaling, not the code or configuration selected by the service's traffic split. Changing a limit does not roll back to A.

Question 221

A Cloud Run deployment must keep 1 warm instance to reduce cold-start delays for expected low traffic. What configuration is appropriate?

  1. Increase request timeout; a longer timeout guarantees that an idle instance stays allocated.
  2. Set an appropriate Cloud Run minimum instance count of one, accounting for idle billing and best-effort warm capacity.
  3. Set maximum instances to one with no minimum; this always reserves a warm instance.
  4. Deploy a second revision with no traffic and assume that creates a warm stable instance automatically.

Explanation

Correct Answer

B:

Service-level minimum instances can keep at least one instance ready during low traffic, reducing cold-start frequency. It incurs cost for idle capacity and is a best-effort warm target rather than an absolute no-cold-start guarantee.

Incorrect Answers

A:

Request timeout affects the duration of individual requests, not idle instance retention.

C:

A maximum caps scale-out but does not prevent scale-to-zero when the minimum remains zero.

D:

A zero-traffic revision is not by itself a guarantee of warm serving capacity for the production revision.

Question 222

A Cloud Run service has a steady ten-instance footprint. Each instance opens a pool of 50 idle connections to a Cloud SQL for PostgreSQL database, leaving almost no connection capacity even before traffic peaks. What client-side change directly addresses that idle-pool overhead without lowering the instance limit?

  1. Reduce per-instance idle connection pool size and use appropriately bounded connection pooling for the application workload.
  2. Increase the number of idle connections opened by every Cloud Run instance so fewer requests create new connections.
  3. Increase the Cloud Run maximum instance count while keeping the 50-idle-connection pool per instance.
  4. Disable the Cloud SQL connection limit check because Cloud Run instance management already guarantees database safety.

Explanation

Correct Answer

A:

Reduce per-instance idle connection pool size and use appropriately bounded connection pooling for the application workload. With ten instances each holding 50 idle connections, the baseline consumes approximately 500 database connections before workload demand. Smaller pools reduce this client-side footprint directly.

Incorrect Answers

B:

Increase the number of idle connections opened by every Cloud Run instance so fewer requests create new connections. Increasing the pool makes the existing per-instance connection multiplication worse, not better. It uses still more connections even at zero request load.

C:

Increase the Cloud Run maximum instance count while keeping the 50-idle-connection pool per instance. Adding more instances at the same pool size increases the maximum aggregate connections. The stem explicitly asks for a client-side change without reducing the instance ceiling.

D:

Disable the Cloud SQL connection limit check because Cloud Run instance management already guarantees database safety. Cloud Run autoscaling does not nullify Cloud SQL's connection limits. Exceeding the server limit remains possible when many instances maintain independent pools.

Question 223

A container serves long-running requests but is terminated before its processing finishes because the configured Cloud Run request timeout is too short. What should change?

  1. Set HTTP keep-alive headers indefinitely; Cloud Run has no request duration limit.
  2. Increase the configured timeout within current platform limits, or redesign lengthy work for asynchronous execution.
  3. Increase the client’s DNS TTL so Cloud Run never closes a request.
  4. Disable all concurrency limits because concurrency determines the allowed duration of a request.

Explanation

Correct Answer

B:

Set a supported request timeout if the work is legitimately synchronous. For tasks exceeding request limits or requiring durable completion, enqueue work and use an appropriate background job or worker pattern with retries.

Incorrect Answers

A:

HTTP headers cannot override the service’s configured server-side request timeout.

C:

DNS caching changes how long a name lookup is reused; it does not extend a running HTTP request.

D:

Concurrency governs simultaneous requests; it is not a substitute for a duration budget.

Question 224

A public Cloud Run handler has an unauthenticated invoker binding despite a private business requirement. What immediate security change is appropriate?

  1. Keep allUsers Invoker but add a note in the service description that it is private.
  2. Enable Cloud Logging and keep unauthenticated invocations because logs enforce access control.
  3. Hide the Cloud Run URL in an internal wiki while preserving public invocation privileges.
  4. Remove unauthorized public invocation permission and require approved authenticated callers.

Explanation

Correct Answer

D:

Review Cloud Run IAM and invoker settings to remove public access, require authenticated identities, and validate the intended ingress mode. Test both allowed and denied requests after the change.

Incorrect Answers

A:

A descriptive label has no effect on an anonymous caller’s invocation permission.

B:

Logging records events but does not replace authentication and authorization.

C:

A URL that is not advertised can still be discovered or reused when public invocation remains enabled.

Question 225

A Cloud Run service cannot reach a private Cloud SQL database because its egress does not enter the approved VPC. Which connection design should be evaluated?

  1. Make the Cloud SQL database public to avoid private VPC routing.
  2. Increase Cloud Run concurrency; private network connectivity is proportional to container count.
  3. Assign Billing Account Admin to Cloud Run so it can discover the database subnet.
  4. Configure supported Direct VPC egress or a connector to the private network, then verify database authorization.

Explanation

Correct Answer

D:

Cloud Run needs an approved network egress path to the VPC containing the private database. Configure supported egress, routes, firewall controls and the database/client identity or credentials separately, then test connectivity.

Incorrect Answers

A:

Publicly exposing a private database contradicts the approved network boundary and is not needed for serverless connectivity.

B:

Concurrency changes throughput behavior, not the available route into a VPC.

C:

Billing permissions have no authority over private IP routing or database authentication.

Question 226

A Pub/Sub-based cloud service receives duplicate events, and the handler performs irreversible side effects twice. What app-level design should be added?

  1. Disable Pub/Sub retries globally; all messages will now be delivered exactly once.
  2. Store only the message body in logs; writing logs prevents side effects from repeating.
  3. Increase subscriber concurrency; concurrent delivery makes duplicate mutations less likely.
  4. Make the handler idempotent using event IDs, deduplication state or transactional safeguards.

Explanation

Correct Answer

D:

Use stable message identifiers and an atomic check/write or transaction so repeated processing has the same business outcome. Acknowledge only after the required operation succeeds, and test duplicate and retry scenarios.

Incorrect Answers

A:

At-least-once delivery and redelivery can occur; eliminating retries can also lose failed work.

B:

Logging is useful for investigation but does not block a second financial or database mutation.

C:

Higher concurrency can increase races and does not guarantee unique delivery.

Question 227

An app needs to process new Cloud Storage objects using an event rather than scanning every bucket each minute. Which trigger model fits?

  1. Poll every bucket object from a permanently running VM and assume the scans are event-driven.
  2. Create a scheduled cost-report export to trigger application processing for each object.
  3. Route supported Cloud Storage object-finalization events through Eventarc to an authorized Cloud Run handler.
  4. Enable Cloud DNS caching, which automatically starts application code when a file appears.

Explanation

Correct Answer

C:

Eventarc can deliver supported object change events to a Cloud Run service. Configure a suitable bucket event filter, trigger location, delivery identity and idempotent handling, then test with a new object.

Incorrect Answers

A:

Periodic full scans waste work and do not provide an object-triggered handler.

B:

A billing export does not describe new object events or invoke the application.

D:

DNS caching is unrelated to storage event notification or compute activation.

Question 228

A team has a Python conversational agent and needs a managed remotely callable runtime on Google Cloud without administering its own GKE cluster. Which supported product capability meets the deployment requirement?

  1. Store the Python source only in a Cloud Storage bucket; object storage creates a persistent agent RPC endpoint automatically.
  2. Deploy the supported agent with Gemini Enterprise Agent Platform Agent Runtime using the SDK or an eligible source deployment method.
  3. Use a Cloud Monitoring dashboard as the agent execution service; dashboards execute Python agents when queried.
  4. Create a Cloud SQL instance and upload the agent code as a table, which makes it a managed HTTP agent endpoint.

Explanation

Correct Answer

B:

Deploy the supported agent with Gemini Enterprise Agent Platform Agent Runtime using the SDK or an eligible source deployment method. Agent Runtime hosts supported deployed agents and exposes remote operation endpoints, eliminating the need for a team-managed Kubernetes node pool.

Incorrect Answers

A:

Store the Python source only in a Cloud Storage bucket; object storage creates a persistent agent RPC endpoint automatically. Cloud Storage holds objects but does not execute code or expose the agent operation contract just by receiving a source file.

C:

Use a Cloud Monitoring dashboard as the agent execution service; dashboards execute Python agents when queried. Monitoring dashboards visualize telemetry; they do not run the agent code or serve its application API.

D:

Create a Cloud SQL instance and upload the agent code as a table, which makes it a managed HTTP agent endpoint. Cloud SQL stores relational data but does not execute and publish arbitrary Python agents as its core managed service.

Question 229

An Agent Runtime deployment must come directly from a linked Git repository through Developer Connect so the team can trace the exact reviewed commit. The agent is written in Python. Which workflow is documented?

  1. Paste generated agent source into a Cloud Logging filter; Logging exports create remote Agent Runtime deployments from commits.
  2. Link the repository with Developer Connect and use Agent Runtime's supported Developer Connect Git deployment path.
  3. Deploy from a plain Cloud Billing budget; billing account association includes the linked Git revision.
  4. Use a static Artifact Registry image alone and assert it proves deployment from the currently linked Git commit without any repository connection.

Explanation

Correct Answer

B:

Link the repository with Developer Connect and use Agent Runtime's supported Developer Connect Git deployment path. The documented Developer Connect path deploys an eligible Python agent from a connected repository while preserving a version-controlled source workflow.

Incorrect Answers

A:

Paste generated agent source into a Cloud Logging filter; Logging exports create remote Agent Runtime deployments from commits. Cloud Logging query filters do not compile or deploy agent source code and cannot establish Git provenance.

C:

Deploy from a plain Cloud Billing budget; billing account association includes the linked Git revision. Billing configuration controls payment responsibility, not deployment source integration or Git commit traceability.

D:

Use a static Artifact Registry image alone and assert it proves deployment from the currently linked Git commit without any repository connection. A container image could be deployed by a different runtime method but does not itself implement the required direct Developer Connect Git workflow.

Question 230

A non-Python agent has already been built and tested as an OCI container, stored in Artifact Registry, and implements Agent Runtime's server contract. Which deployment method retains that image?

  1. Use only the Python-local-object SDK execution path, which accepts arbitrary non-Python images without a container contract.
  2. Upload the image digest into Secret Manager and rely on Secret Manager to launch the agent service.
  3. Deploy the Artifact Registry container image through Agent Runtime's supported container-image configuration.
  4. Use Developer Connect source deployment, which is the documented non-Python entrypoint for any compiled language.

Explanation

Correct Answer

C:

Deploy the Artifact Registry container image through Agent Runtime's supported container-image configuration. Agent Runtime supports Artifact Registry container images built with any language when the image satisfies its runtime contract.

Incorrect Answers

A:

Use only the Python-local-object SDK execution path, which accepts arbitrary non-Python images without a container contract. The in-memory object SDK path targets supported Python agent objects and is not a generic hosting mechanism for arbitrary OCI images.

B:

Upload the image digest into Secret Manager and rely on Secret Manager to launch the agent service. Secret Manager holds sensitive configuration and does not execute containers or provide a managed agent service endpoint.

D:

Use Developer Connect source deployment, which is the documented non-Python entrypoint for any compiled language. The Developer Connect source path is Python-only in the documented deployment matrix; the already-built non-Python container fits the image method.

Associate Cloud Engineer Practice Tests