Google PCA Practice Test 47 — 15 Questions

Question 1

A company is migrating a monolithic e-commerce application to Google Cloud. The application has been refactored into microservices. Most services are stateless and can run on Cloud Run. However, the checkout service requires maintaining session state across multiple requests, and the session data must be available globally for low latency. The application will be deployed in multiple regions to serve a global user base. Which approach should the company take?

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

Accepted Answer

Deploy the checkout service on Cloud Run in multiple regions, and use Memorystore (Redis) with replication as the session store. Memorystore (Redis) with replication provides a globally accessible, low-latency session store that can be used by Cloud Run instances in multiple regions. Redis replication ensures data durability and high availability, while Cloud Run's stateless nature is complemented by externalizing session state to a managed caching layer. This architecture meets the requirement for global session data availability without coupling state to individual compute instances.

Answer

Run the checkout service on Compute Engine with regional managed instance groups and Cloud Filestore

Answer

Use Cloud Run with session affinity and in-memory caching within each instance

Answer

Deploy the checkout service on Google Kubernetes Engine using StatefulSets and regional persistent disks

Question 2

Refer to the exhibit. An engineer is reviewing health check logs for a target pool. The instance web-01 is still serving traffic despite two consecutive unhealthy health checks. Why is the instance not removed from the target pool?

Accepted Answer

The instance was removed after the second failure but then re-added after the next healthy check, so it is serving traffic again.. The instance was removed from the target pool after the second consecutive unhealthy health check, but it was immediately re-added after the next healthy health check. This happens when the health check's `unhealthyThreshold` is set to 2, meaning two consecutive failures trigger removal, but the instance then passes a subsequent health check, causing it to be re-added and resume serving traffic. The logs show the instance is still serving traffic because it was removed and then re-added within the same monitoring window.

Answer

The health check's timeoutSec is too high, causing delayed removal.

Answer

The forwarding rule is not using the target pool, so health checks have no effect.

Answer

The healthyThreshold is set to 2, requiring two healthy checks before re-adding, but the instance only had one healthy check.

Question 3

You are running a Cloud Run service that experiences occasional cold starts causing latency spikes. You want to minimize cold starts cost-effectively. What should you do?

Accepted Answer

Set min-instances to 1 to keep at least one instance always warm.. Setting min-instances to 1 ensures that at least one instance of your Cloud Run service is always kept warm, meaning it is initialized and ready to handle requests immediately. This eliminates cold starts for the first request after a period of inactivity, reducing latency spikes without requiring over-provisioning of resources. It is cost-effective because you only pay for the idle instance when it is not serving traffic, and you avoid the higher costs of larger machine types or excessive concurrent instances.

Answer

Increase the max-instances setting to allow more concurrent requests.

Answer

Set concurrency to 1 to ensure each instance handles one request at a time.

Answer

Use a larger machine type (e.g., 2 vCPU) to reduce startup time.

Question 4

Your company runs a global e-commerce platform on Google Cloud. The application is deployed across multiple regions for low latency. You use Cloud SQL for transactional data and Cloud Spanner for global consistency of inventory. Recently, the operations team reported that the application is experiencing increased latency during peak hours, and the monthly cloud bill has risen significantly. Upon investigation, you find that the Cloud SQL instance is underutilized (CPU < 20%) while Cloud Spanner split utilization is over 80%. The application instances are fronted by a global external HTTPS load balancer. Network egress costs are high. Which course of action would best address both the latency and cost issues?

Accepted Answer

Increase the number of splits in Cloud Spanner to reduce hot spots, and configure Cloud CDN in front of the load balancer to cache static content.. The primary performance issue is Cloud Spanner split utilization over 80%, indicating hot spots that cause increased latency. Increasing the number of splits redistributes load across more nodes, reducing contention. Additionally, configuring Cloud CDN caches static content at edge locations, reducing network egress costs and latency by serving content closer to users.

Answer

Reduce the Cloud SQL instance tier to a lower machine type to save costs, and add read replicas in other regions for failover.

Answer

Add more nodes to the Cloud SQL instance and enable automatic storage increase to handle peak loads.

Answer

Move the transactional database to Cloud Spanner and decommission Cloud SQL to reduce complexity.

Question 5

A company uses Terraform to manage Google Cloud infrastructure. They want to store the Terraform state file in a remote backend with state locking to prevent concurrent modifications. Which Google Cloud service supports this natively?

Accepted Answer

Cloud Storage. Google Cloud Storage (GCS) is the only option that natively supports Terraform's remote state backend with state locking. Terraform uses GCS's object versioning and a write-lock mechanism via a separate lock file (e.g., `default.tflock`) stored in the same bucket, leveraging GCS's strong consistency for atomic operations. This prevents concurrent `terraform apply` commands from corrupting the state.

Answer

Cloud Firestore

Answer

Cloud Spanner

Answer

Bigtable

Answer

Cloud SQL

Question 6

A company has two VPC networks in the same project: 'vpc-prod' and 'vpc-dev'. They want to allow communication between instances in both VPCs. What is the simplest method?

Accepted Answer

Create a VPC Network Peering connection between them. VPC Network Peering is the simplest method because it directly connects two VPCs using Google's internal infrastructure, allowing private RFC 1918 IP communication across the networks without requiring external gateways, VPN tunnels, or additional bandwidth costs. It requires no routes to be manually configured—Google automatically adds the necessary routes for each peered VPC's subnets—and only a single firewall rule to permit traffic between the instances.

Answer

Set up a Cloud VPN tunnel between the two VPCs

Answer

Configure a custom route in each VPC pointing to the other's subnet

Answer

Add firewall rules allowing traffic between the VPCs

Question 7

Refer to the exhibit. A team wants to grant the ability to run queries (but not modify) on BigQuery datasets to a new set of users who have email addresses in the 'example.com' domain. What is the simplest way to achieve this?

Accepted Answer

No action needed; new users with 'example.com' accounts already have the dataViewer role through the existing domain membership. The exhibit shows an existing IAM policy binding that grants the 'roles/bigquery.dataViewer' role to 'domain:example.com'. IAM domain-based bindings automatically apply to all authenticated users with email addresses in that domain. When new users with 'example.com' accounts are added, they inherit this role without any additional configuration, as IAM evaluates domain membership at access time.

Answer

Create a new binding with role 'roles/bigquery.dataViewer' and include the new users as members

Answer

Remove the domain binding and only grant access to individual users

Answer

Add the new users to the existing 'domain:example.com' member list

Question 8

Which TWO statements about Google Cloud VPC networks are true? (Choose two.)

Accepted Answer

Subnets are regional resources.. Subnets in Google Cloud VPC are regional resources. When you create a subnet, you specify a region and a CIDR block, and the subnet spans all zones within that region. This allows resources in different zones of the same region to use the same subnet without additional configuration.

Answer

VPC networks are project-level resources.

Answer

Firewall rules are regional.

Answer

Subnets are zonal resources.

Question 9

A company uses BigQuery for large-scale analytics. They have a fixed monthly budget and want to ensure predictable costs for query processing, even when many users run concurrent queries. Which BigQuery pricing model should they use?

Accepted Answer

Flat-rate pricing with slot reservations. BigQuery slot reservations (fixed capacity) provide predictable pricing by purchasing a set number of slots. On-demand pricing charges per byte scanned, which can vary with usage. Flat-rate pricing via reservations is best for predictable costs. Committed use discounts apply to slot reservations, but the core model is reservations.

Answer

On-demand pricing with flat-rate discounts

Answer

Autoscaling slot reservations

Answer

On-demand pricing with committed use discounts

Question 10

A company wants to reduce Cloud Storage costs for archival data that is accessed less than once a year. They have data stored in Standard storage class. Which TWO actions should they take? (Choose TWO.)

Accepted Answer

Use a lifecycle rule to delete objects that are older than 365 days. Using lifecycle policies to transition from Standard to Archive storage class reduces cost for rarely accessed data. Also, deleting unnecessary objects reduces storage. Coldline is for 90 days, Nearline for 30 days; Archive is for long-term archival. Changing object names does not reduce cost.

Answer

Set a lifecycle rule to transition objects older than 90 days to Coldline storage class

Answer

Rename objects to start with 'archive/' to reduce costs

Answer

Set a lifecycle rule to transition objects older than 30 days to Nearline storage class

Question 11

A company uses Cloud CDN to serve static content from a Cloud Storage bucket. They update the content every 2 hours, but users sometimes see stale content for up to 24 hours. They need users to see the latest content within 5 minutes of an update. Which action should they take?

Accepted Answer

Set a short max-age TTL (e.g., 5 minutes) on the objects. To ensure content is served fresh, they should set a short max-age (cache TTL) on the object metadata or via the bucket's default object holding. Invalidating the cache is a reactive measure, but setting a short TTL ensures that cache entries expire quickly. Cache modes affect query string handling, not TTL. Cache keys affect how content is cached but not the freshness duration.

Answer

Use cache invalidation after each update

Answer

Use custom cache keys

Answer

Change cache mode to CACHE_ALL_STATIC

Question 12

A company runs batch analytics workloads on Compute Engine that can tolerate interruptions. They want to reduce compute costs by up to 60-90%. Which compute option is the most cost-effective?

Accepted Answer

Preemptible VMs. Preemptible VMs offer significant discounts (60-90%) for interruptible workloads. Spot VMs are the newer equivalent with same discounts but no max runtime limit.

Answer

On-demand VMs

Answer

Committed use discounts (1-year)

Answer

Sustained use discounts

Question 13

A company wants to implement blameless postmortems as part of their SRE practices. Which THREE principles should they follow?

Accepted Answer

Create actionable recommendations to prevent recurrence. Blameless postmortems focus on learning from incidents without assigning blame. They should focus on system improvements, include everyone involved, and document findings. Assigning blame, only involving management, and ignoring fixes are contrary to blameless culture.

Answer

Only involve senior management in the review

Answer

Assign responsibility to the team that caused the incident

Question 14

Which Google Cloud service automatically computes the optimal size or tier for underutilized Compute Engine instances and generates recommendations to reduce cost?

Accepted Answer

Recommender (Active Assist). Active Assist includes rightsizing recommendations that analyze VM utilization and suggest appropriate machine types or tiers to reduce cost without sacrificing performance.

Answer

Cloud Monitoring

Answer

Cloud Profiler

Answer

Cost Management

Question 15

A company wants to reduce costs for their BigQuery workloads. Which TWO actions will help achieve this?

Accepted Answer

Use clustered and partitioned tables to limit data scanned. Using flat-rate pricing with slot reservations provides predictable costs, and using clustered and partitioned tables reduces the amount of data scanned per query, lowering on-demand costs. Materialized views can reduce costs but are not always needed. Deleting datasets and compression are not relevant.

Answer

Enable compression on exported data

Answer

Delete unnecessary datasets

Answer

Use materialized views to pre-compute results