PCD Practice Test 50 — 15 Questions

Question 1

A retail company runs an order-processing service on Cloud Run (fully managed) in us-central1. During flash sales, order volume spikes 20x for about 10 minutes. The service must remain responsive, minimize cold starts, and keep costs predictable. The team wants to configure the service to handle these bursts. What should they do?

Choose an answer to begin — your selection is scored in the full session.

Accepted Answer

Set the minimum number of instances to a value that covers baseline traffic and configure a maximum instance limit, along with concurrency settings appropriate for the service.. Cloud Run's autoscaling can handle bursts, but to minimize cold starts and keep costs predictable, you should set a minimum instance count to keep some instances warm and a maximum instance limit to cap scaling. Tuning concurrency ensures efficient request handling per instance. This balances responsiveness and cost control for flash-sale traffic.

Answer

Configure the service to use a larger container image and increase the memory limit to reduce cold-start time.

Answer

Deploy the service to a GKE cluster with cluster autoscaler and horizontal pod autoscaler, and route traffic through an HTTP(S) load balancer.

Answer

Enable CPU always allocated and set the maximum number of instances to a high value, without setting a minimum.

Question 2

A development team runs a Node.js API on Cloud Run that reads configuration values from Secret Manager at startup. During a load test, the service scales out to 40 instances and the team observes a spike in Secret Manager API calls plus occasional HTTP 429 responses. They want to reduce the call volume and eliminate the throttling without changing the secret values or broadening IAM permissions. What should they do?

Accepted Answer

Fetch the secret once at container startup and store it in a module-level variable, and set the secret as an environment variable or mounted volume so each instance reads it locally.. Caching the secret in memory after a single startup fetch reduces Secret Manager calls to roughly one per instance, which directly removes the burst of requests that caused throttling. Delivering the value through an environment variable or mounted volume also decouples the running service from the Secret Manager API, so scaling events no longer multiply API traffic.

Answer

Enable the Secret Manager API's server-side caching by setting the --cache-ttl flag on the Cloud Run service.

Answer

Mount the secret as a volume and read it from the filesystem on every request instead of calling the API.

Answer

Increase the Cloud Run maximum instance count and set a higher CPU allocation so instances initialize faster.

Question 3

A developer needs to deploy a simple stateless web application to Google Cloud. The application is packaged as a container image and must automatically scale based on incoming request volume. The developer wants to minimize infrastructure management and pay only for resources used. Which Google Cloud service should the developer use?

Accepted Answer

Cloud Run. Cloud Run is the ideal choice for a containerized stateless web application with automatic scaling and pay-per-use billing. It abstracts away all infrastructure management and scales to zero. GKE Autopilot and Compute Engine require more management, and App Engine standard may not support the container image.

Answer

Google Kubernetes Engine (GKE) Autopilot

Answer

Compute Engine managed instance groups

Answer

App Engine standard environment

Question 4

You manage a Cloud Run service that receives traffic from users worldwide. The service is deployed in the us-central1 region. Users in Asia report high latency. You want to reduce latency for these users without changing the application code. What should you do?

Accepted Answer

Deploy the service to multiple regions and use a global external HTTP(S) load balancer with a serverless NEG backend to route users to the closest region.. Deploying the service to multiple regions and using a global external HTTP(S) load balancer with serverless NEGs allows Google's global network to route users to the nearest Cloud Run region, minimizing latency. This approach leverages Google's edge network and automatic failover, and does not require application code changes. Other options address scaling or caching but do not reduce geographic latency.

Answer

Configure the Cloud Run service to use a larger CPU and memory allocation to process requests faster.

Answer

Increase the maximum number of instances for the Cloud Run service to handle more concurrent requests.

Answer

Enable CDN caching on the Cloud Run service by setting the cache-control header in responses.

Question 5

A developer is building a Cloud Run service that reads messages from a Pub/Sub subscription. The service must process each message exactly once and acknowledge it only after successful processing. The developer wants to minimize the chance of duplicate processing. Which Pub/Sub feature should be enabled on the subscription?

Accepted Answer

Exactly-once delivery.. Exactly-once delivery is a Pub/Sub subscription setting that ensures each message is delivered and processed only once, provided the subscriber acknowledges within the deadline. It is the only option that directly targets duplicate prevention. Other features like dead-letter topics, ordering keys, and retry policies serve different purposes and do not guarantee exactly-once processing.

Answer

Message ordering with an ordering key.

Answer

Retry policy with exponential backoff.

Answer

Dead-letter topic with a maximum delivery attempt of 5.

Question 6

A developer is writing unit tests for a Cloud Function written in Node.js that publishes a message to Pub/Sub. They want to avoid making real Pub/Sub calls during tests. Which testing strategy should they use?

Accepted Answer

Use dependency injection to pass a mock Pub/Sub client to the function, and use a library like sinon to stub the publish method.. Unit tests should isolate the code under test from external dependencies. Dependency injection allows the Pub/Sub client to be replaced with a mock during tests. Using a stubbing library like sinon, the publish method can be stubbed to simulate success or failure without network calls. This approach ensures tests are fast, reliable, and do not require emulators or cloud resources.

Answer

Use the Google Cloud Client Library's built-in test mode by setting the GOOGLE_CLOUD_TEST environment variable to true.

Answer

Use the Pub/Sub emulator and set the PUBSUB_EMULATOR_HOST environment variable before running tests.

Answer

Deploy the function to a test environment and invoke it with a test message, then verify that the message appears in a test Pub/Sub topic.

Question 7

A developer is writing unit tests for a Cloud Function written in Python that stores data in Cloud Firestore. They want to isolate the function from the actual Firestore database during tests. Which approach should they use?

Accepted Answer

Use the Firestore emulator and configure the function to connect to it by setting the FIRESTORE_EMULATOR_HOST environment variable.. The Firestore emulator is the recommended tool for local testing of applications that use Firestore. It runs locally and can be used by setting the FIRESTORE_EMULATOR_HOST environment variable, which the Cloud Function's Firestore client library respects. This allows unit tests to run against a local instance, ensuring isolation from production, faster execution, and no cost. Other methods either introduce complexity or do not fully isolate the tests.

Answer

Write the tests to use the Firestore REST API directly and intercept HTTP calls with a proxy.

Answer

Create a separate Firestore database in a test project and configure the function to use it via environment variables.

Answer

Mock the Firestore client library using a Python mocking framework such as unittest.mock.

Question 8

A developer is using Cloud Build to run unit tests for a Python application. The tests require a specific version of a library that is not available in the default Cloud Build environment. The build steps are defined in cloudbuild.yaml. What should the developer do to ensure the tests run with the required library version?

Accepted Answer

Add a step that uses the python:3.9 image and runs pip install -r requirements.txt before running the tests.. Cloud Build executes each step in a separate container. To run tests with a specific library version, the developer must use a step with a Python image and install the dependencies from requirements.txt before running tests. This ensures the correct version is available in the test step. Using a custom image or the official Python image with pip install is the reliable way to control dependencies.

Answer

Modify the cloudbuild.yaml to use the gcr.io/cloud-builders/python builder, which includes all common Python libraries.

Answer

Use a Cloud Build trigger that automatically installs the required library version based on the repository's requirements.txt.

Answer

Add a step that runs pip install --user == using the default Cloud Build environment.

Question 9

A company runs a stateful application on Google Kubernetes Engine (GKE) using a StatefulSet with persistent volumes. They need to ensure that the application can survive a zone failure without data loss and maintain high availability. Which two strategies should they implement? (Choose two.)

Accepted Answer

Configure persistent volumes with a Regional Persistent Disk (Regional PD) storage class.. To survive a zone failure with a stateful application, you need both compute and storage redundancy. A regional GKE cluster spreads nodes across zones, and Regional Persistent Disks replicate data across zones. Together, they allow the application to continue running and preserve data if a zone becomes unavailable.

Answer

Enable horizontal pod autoscaling based on CPU utilization.

Answer

Use a single-zone GKE cluster with node auto-repair enabled.

Answer

Set the pod anti-affinity to ensure pods are scheduled on different nodes.

Question 10

A retail company runs a web application on Compute Engine behind an external Application Load Balancer. They want the application to remain available if an entire zone becomes unhealthy, and they want unhealthy instances removed from serving traffic automatically. The application instances are managed by a regional managed instance group (MIG) with an autoscaler. What should they do?

Accepted Answer

Configure an HTTP health check on the backend service and set the MIG to use autohealing.. A regional managed instance group spreads instances across zones in a region, and combining a load balancer health check with MIG autohealing ensures unhealthy instances are both removed from traffic and automatically recreated. This gives zone-level resilience and self-healing without the overhead of a multi-region design. Capacity changes and connection draining do not provide these behaviors.

Answer

Increase the MIG's maximum number of instances so more capacity is available during a zone failure.

Answer

Add a second regional MIG in another region and use global load balancing across both.

Answer

Enable connection draining on the backend service to allow in-flight requests to complete.

Question 11

A developer is building a microservices application on Google Kubernetes Engine (GKE). They want to ensure that each service can scale independently based on CPU utilization and that the cluster can automatically add nodes when resource demand increases. Which GKE feature should they use?

Accepted Answer

Horizontal Pod Autoscaler (HPA) combined with Cluster Autoscaler.. Horizontal Pod Autoscaler scales the number of pod replicas based on metrics like CPU utilization, ensuring each service scales independently. Cluster Autoscaler adds or removes nodes in the node pool when pods cannot be scheduled or when nodes are underutilized. This combination provides both pod-level and node-level automatic scaling, which is exactly what the scenario requires.

Answer

Istio service mesh with automatic sidecar injection.

Answer

Pod Disruption Budget (PDB) combined with node pool auto-scaling.

Answer

Vertical Pod Autoscaler (VPA) combined with node auto-provisioning.

Question 12

A company is building a highly available application on Google Cloud. They want to ensure that the application can survive a zonal failure and continue serving traffic with minimal disruption. The application runs on Compute Engine instances in a managed instance group (MIG) and uses a Cloud SQL for MySQL database. Which two actions should they take? (Choose two.)

Accepted Answer

Enable high availability (HA) on the Cloud SQL for MySQL instance to create a standby in a different zone.. To survive a zonal failure, both the compute and database tiers must be resilient. A regional managed instance group spreads instances across multiple zones, so if one zone fails, others continue serving. Cloud SQL high availability creates a standby in a different zone with automatic failover. Together, these actions ensure the application remains available during a zonal outage.

Answer

Deploy the application on a single zone with a larger machine type to handle the load.

Answer

Use a global external HTTP(S) load balancer to distribute traffic across multiple regions.

Answer

Create a read replica of the Cloud SQL instance in a different region and direct all writes to the read replica.

Question 13

A SaaS provider serves customers from a single GKE cluster in us-east1 backed by a single Cloud SQL for PostgreSQL instance. The service level objective requires the application to remain available during a regional outage. Which change best meets this objective?

Accepted Answer

Run the application in a second GKE cluster in another region, replicate the database with a cross-region replica, and direct traffic with a global load balancer with health-checked failover.. Meeting a regional-outage objective requires resources in at least two regions, not just multiple zones. A second regional GKE cluster provides compute that survives the loss of us-east1, a cross-region database replica provides a promotable copy of the data, and a global load balancer with health checks automatically routes users to the surviving region. Only a multi-region design satisfies the objective.

Answer

Enable high availability on the Cloud SQL instance and add a read replica in another zone.

Answer

Increase the Cloud SQL instance size and add more nodes to the existing GKE cluster to absorb the extra load during an outage.

Answer

Deploy the GKE cluster as a regional cluster and configure the Cloud SQL instance for high availability.

Question 14

A startup is building a new REST API on Cloud Run. The API must handle unpredictable traffic spikes and scale automatically. The startup wants to minimize costs when there is no traffic. What should they do?

Accepted Answer

Configure the Cloud Run service to allow scaling to zero by not setting a minimum instance count.. Cloud Run automatically scales to zero when there is no traffic, so you only pay for what you use. This is ideal for unpredictable traffic and minimizing costs during idle periods. Allowing the service to scale to zero by not setting a minimum instance count achieves this, while still enabling automatic scaling during spikes.

Answer

Use a Cloud Run service with a minimum of 0 instances and set a maximum of 1 instance.

Answer

Deploy the API on a Compute Engine managed instance group with autoscaling enabled.

Answer

Set the minimum number of instances to 1 and the maximum to 1000.

Question 15

A company is deploying a new microservice on Cloud Run. The service must connect to a Cloud SQL for PostgreSQL instance and a Memorystore for Redis instance. The security team requires that all connections use private IP addresses and that no public IPs are exposed. The developer needs to configure the Cloud Run service. (Choose two.)

Accepted Answer

Configure the Cloud Run service to use a Serverless VPC Access connector to connect to the VPC network.. To connect a Cloud Run service to Cloud SQL and Memorystore using private IPs, you must use a Serverless VPC Access connector to route traffic into the VPC, and ensure both Cloud SQL and Memorystore instances are configured with private IPs in the same VPC network. This keeps all traffic internal and satisfies the security requirement.

Answer

Set the Cloud SQL instance to have a public IP and configure the Cloud Run service to connect using the Cloud SQL Auth Proxy with SSL.

Answer

Enable Cloud Run's built-in Cloud SQL connection using the --add-cloudsql-instances flag and rely on that for both Cloud SQL and Memorystore connectivity.

Answer

Assign the Cloud Run service a public IP address and use VPC peering to route traffic to the private resources.