Google PCA Analyze and optimize technical and business processes Practice Test 4 — 15 Questions

Question 1

A company is using Cloud Spanner to serve a global gaming application. They have a single instance in us-central1. Players in Asia experience high latency. The application reads and writes player profiles. The team wants to reduce latency for Asian players while keeping write latency low for global consistency. They need a solution that minimizes operational overhead and uses native Spanner capabilities. What should they do?

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

Accepted Answer

Configure a multi-region instance configuration that includes us-central1 and an Asian region.. A multi-region instance configuration in Cloud Spanner is the correct solution because it provides a single writable instance that spans multiple geographic regions, allowing reads and writes to be served locally in each region while maintaining strong global consistency. This minimizes latency for Asian players by enabling local reads and writes, and it uses native Spanner capabilities without additional operational overhead. Option A directly addresses the requirement for low write latency and global consistency by leveraging Spanner's built-in multi-region replication.

Answer

Add read replicas in Asia using Spanner's read-only replicas.

Answer

Use Cloud CDN to cache player profiles at the edge.

Answer

Create a new instance in asia-east1 and use Directed Read options to route reads from Asia.

Question 2

A large enterprise is migrating its on-premises data warehouse to BigQuery. The current warehouse is 100 TB and uses complex ETL jobs that run on SQL Server Integration Services (SSIS). The team wants to minimize migration effort and maintain the same SQL logic for transformations. They plan to use BigQuery's standard SQL. They also need to schedule transformations and load data from multiple on-premises sources. Which approach should they take?

Accepted Answer

Use Cloud Composer (Airflow) to orchestrate SQL statements in BigQuery, and use Data Transfer Service for scheduled loads.. Cloud Composer (Airflow) can orchestrate BigQuery SQL statements, and Data Transfer Service handles scheduled loads from on-premises sources, minimizing code changes. Option A is incorrect because running SSIS on Compute Engine still requires managing virtual machines and does not leverage BigQuery's full capabilities. Option B is incorrect because rewriting SSIS packages as Dataflow pipelines using Apache Beam would require significant recoding. Option C is incorrect because Data Fusion is a data integration service for replicating and transforming data, but it does not directly replicate SSIS package logic.

Answer

Connect SSIS to BigQuery using ODBC and run SSIS packages in a VM on Compute Engine.

Answer

Rewrite SSIS packages as Dataflow pipelines using Apache Beam.

Answer

Use Data Fusion to replicate SSIS packages.

Question 3

A company runs a batch job every night that processes large CSV files stored in Cloud Storage. The job runs on a single Compute Engine VM and takes 4 hours to complete. The team wants to reduce execution time without increasing cost substantially. The job is CPU-intensive and can be parallelized. What should they do?

Accepted Answer

Split the input files into smaller chunks and use multiple VMs in a managed instance group with a job scheduler.. Splitting the input files into smaller chunks and using multiple VMs in a managed instance group with a job scheduler effectively parallelizes the CPU-intensive batch job without significantly increasing cost. This approach leverages existing Compute Engine resources and scales horizontally using multiple smaller, cost-effective VMs. Option A (Dataproc with Spark) introduces managed service overhead and potential cluster costs, which may increase expenses. Option B (Cloud Batch) provides autoscaling but still requires the user to design parallelization and may not be as cost-efficient for this specific workload. Option C (using a VM with more vCPUs) increases cost due to vertical scaling and may not provide proportional performance gains. Thus, D is the most suitable choice.

Answer

Migrate the job to Dataproc and run it as a Spark job.

Answer

Use Cloud Batch to run the job as a batch job that automatically scales.

Answer

Use a VM with more vCPUs and a higher CPU platform.

Question 4

A retail company runs a Java-based inventory service on Compute Engine VMs behind an HTTP(S) load balancer. The service has unpredictable traffic but must scale automatically. The operations team wants to minimize administrative overhead and reduce cost by using committed use discounts (CUDs) for the baseline load. They plan to move the service to a managed compute platform. Which Google Cloud solution should the team choose?

Accepted Answer

Migrate the service to App Engine flexible environment and purchase 1-year committed use discounts for the underlying Compute Engine instances.. App Engine flexible environment is a managed platform that automatically scales and runs on Compute Engine VMs, allowing the purchase of committed use discounts for the underlying instances. This meets the requirements of automatic scaling, reduced administrative overhead, and cost optimization through CUDs. The other options either do not support CUDs or require more management effort.

Answer

Migrate the service to Google Kubernetes Engine (GKE) Standard and purchase 1-year committed use discounts for the underlying Compute Engine nodes.

Answer

Migrate the service to App Engine standard environment and purchase 1-year committed use discounts for the App Engine instances.

Answer

Migrate the service to Cloud Run and purchase 1-year committed use discounts for the Cloud Run instances.

Question 5

A healthcare analytics company runs a critical patient-data pipeline on Google Cloud. The pipeline processes data in Cloud Storage, transforms it in Dataflow, and loads it into BigQuery. The finance team wants to reduce BigQuery costs without affecting query performance or data freshness. The data is partitioned by day and the queries filter on the partition column. Which action should the architect take first?

Accepted Answer

Enable BigQuery partitioning and clustering on the fact tables.. Partitioning and clustering are core BigQuery cost-optimization techniques that reduce the amount of data scanned by queries. Because the data is already partitioned by day and queries filter on that column, enabling partitioning and clustering ensures that only relevant data is processed, directly lowering query costs without impacting performance or freshness. Other options either limit usage, change the analytics engine, or optimize a different service.

Answer

Switch the Dataflow pipeline to use Streaming Engine to reduce BigQuery load costs.

Answer

Set a custom quota on the number of bytes billed per day for the project.

Answer

Move the transformed data from BigQuery to Cloud Bigtable for analytical queries.

Question 6

A media company runs a video transcoding workload on Google Cloud. Jobs are submitted to a Pub/Sub topic and processed by a managed instance group of Compute Engine VMs that poll the subscription. The team wants to reduce the operational overhead of managing the VM fleet while keeping the ability to run custom transcoding binaries that require a specific Linux kernel module. They also need the ability to scale down to zero when no jobs are pending. Which approach should the architect recommend?

Accepted Answer

Use a Managed Instance Group with autoscaling based on Pub/Sub queue depth, and bake the kernel module into a custom image.. The requirement for a custom Linux kernel module rules out fully managed serverless platforms such as Cloud Run functions and GKE Autopilot, which do not expose node-level kernel customization. A managed instance group with a custom image preserves the kernel module and custom binary, and autoscaling on Pub/Sub queue depth allows the fleet to scale down to zero when idle, satisfying both the technical and cost requirements.

Answer

Deploy the transcoding workload as a container on a GKE Autopilot cluster with a node auto-provisioning policy and Pub/Sub-triggered Horizontal Pod Autoscaling.

Answer

Migrate the transcoding logic to Cloud Run functions triggered by Pub/Sub push subscriptions.

Answer

Move the workload to a Dataproc cluster with a custom initialization action that installs the kernel module.

Question 7

A media company uses a Dataflow streaming pipeline to process video metadata. The pipeline reads from Pub/Sub, performs transformations, and writes to BigQuery. The company wants to optimize the pipeline to reduce cost and latency. They notice that the pipeline sometimes lags during peak hours. Which action should they take?

Accepted Answer

Increase the number of workers by setting a higher maximum number of workers.. Increasing the maximum number of workers enables Dataflow to scale out during peak hours, reducing lag and improving latency. This directly addresses the pipeline's performance issue. While it may increase cost, the goal is to optimize for latency during peaks, and autoscaling will scale down when demand decreases.

Answer

Use a larger machine type for the workers to increase per-worker throughput.

Answer

Enable autoscaling and set a lower maximum number of workers to reduce cost.

Answer

Switch the pipeline to use batch processing instead of streaming.

Question 8

A media company runs a 12-hour nightly transcoding batch on a managed instance group of 200 Compute Engine VMs. The pipeline completes in about 6 hours, but the team notices that many VMs sit idle during the final hour while a few stragglers finish, and the group stays at full size until every task completes. They want to cut compute spend without slowing down the nightly deadline. What should they do?

Accepted Answer

Convert the workload to use Spot VMs with a termination action of STOP and enable autoscaling based on a queue-depth signal.. The waste comes from provisioning for peak and holding capacity until the slowest task finishes. Using Spot VMs lowers the rate paid per hour, and driving autoscaling from queue depth rather than elapsed time lets the group shrink as work drains, so idle workers during the tail disappear. The STOP termination action keeps interrupted workers resumable, protecting the nightly deadline while reducing spend.

Answer

Purchase a 3-year committed use discount for the full 200-VM footprint to lock in the lowest hourly rate.

Answer

Move the transcoding job to a Cloud Run service configured with a minimum instance count equal to the peak concurrency.

Answer

Replace the managed instance group with a single large n2-standard-96 VM and run all transcoding tasks in parallel on it.

Question 9

A retail company runs its order management system on a single Compute Engine instance with a local SSD. The system is business-critical and the company wants to improve its disaster recovery posture. The recovery point objective (RPO) is 5 minutes and the recovery time objective (RTO) is 30 minutes. The application cannot be easily modified to support active-active replication. Which approach should the architect recommend?

Accepted Answer

Configure the instance to use a regional persistent disk and replicate the application to a second instance in another zone with a startup script that attaches the replica disk.. A regional persistent disk replicates data synchronously across two zones, achieving an RPO of effectively zero and enabling a standby instance to attach the replica and start within the 30-minute RTO. This avoids application changes for active-active replication. Daily or hourly snapshots cannot meet the 5-minute RPO, and a managed instance group does not provide shared data replication for a stateful single-instance application.

Answer

Use a managed instance group with autoscaling across multiple zones and a load balancer.

Answer

Schedule hourly disk snapshots and use a Cloud Function to restore the instance in another region.

Answer

Take daily snapshots of the persistent disk and store them in a multi-region bucket.

Question 10

A retail company operates a customer-facing API on Cloud Run that reads from Cloud SQL for PostgreSQL. During seasonal peaks, database connection errors appear and the service occasionally fails to start new revisions. The team wants to reduce connection pressure and improve resilience without changing the database engine. (Choose two.)

Accepted Answer

Configure Cloud SQL for PostgreSQL with a higher maximum connections flag and apply it with a maintenance window.. Serverless instances create and drop connections rapidly, and each one consumes a PostgreSQL backend slot. Inserting a pooler between Cloud Run and the database multiplexes those clients onto fewer backends, and raising the max_connections flag provides additional headroom so revisions can start during peaks. Together they relieve the pressure without changing the engine.

Answer

Deploy the Cloud SQL Auth Proxy as a sidecar container alongside the Cloud Run service and connect through it.

Answer

Increase the Cloud Run maximum instance count so more revisions can start during peak traffic.

Answer

Enable the Cloud SQL connector's built-in connection pooling by setting the max pool size environment variable on the Cloud Run service.

Question 11

A retail company runs a three-tier web application on a single Compute Engine instance in us-central1. The instance has a persistent disk (pd-ssd) that is 500 GB and is currently 60% full. The company wants to improve the resilience of the application and reduce recovery time after a zone failure. They have a recovery point objective (RPO) of 5 minutes and a recovery time objective (RTO) of 15 minutes. The application writes to the disk frequently. Which approach should they take?

Accepted Answer

Convert the persistent disk to a regional persistent disk and enable automatic replication across two zones in us-central1.. The requirement for an RPO of 5 minutes and RTO of 15 minutes with frequent writes points to a solution that provides synchronous replication and automatic failover. A regional persistent disk replicates data across two zones in the same region with zero data loss and allows rapid recovery. Other options either do not meet the RPO, lack performance, or require manual intervention.

Answer

Create a snapshot schedule for the persistent disk with a 5-minute interval and store snapshots in a multi-region bucket.

Answer

Create a second instance in another zone and set up asynchronous replication using rsync every 10 minutes.

Answer

Use a Cloud Storage bucket with versioning and mount it as a file system on the instance using Cloud Storage FUSE.

Question 12

A financial services company is migrating a legacy on-premises Oracle database to Google Cloud. The database supports a trading application with strict consistency requirements and frequent complex joins. The company wants to minimize refactoring effort and operational overhead. They are evaluating Cloud SQL for PostgreSQL and Cloud Spanner. Which two factors should the architect consider when choosing between these services? (Choose two.)

Accepted Answer

Cloud SQL for PostgreSQL supports the PostgreSQL wire protocol, which may ease migration from Oracle with tools like ora2pg, while Cloud Spanner requires a different SQL dialect and client libraries.. The key differentiators for this migration are compatibility with existing Oracle schemas and tooling, where Cloud SQL for PostgreSQL's PostgreSQL wire protocol reduces refactoring, and scalability, where Cloud Spanner provides horizontal scaling with strong consistency for high-volume trading. These two factors directly address the company's goals of minimizing refactoring and operational overhead while meeting strict consistency requirements.

Answer

Cloud Spanner is a fully managed relational database that supports ACID transactions, while Cloud SQL for PostgreSQL does not support ACID transactions.

Answer

Cloud SQL for PostgreSQL automatically shards data across multiple nodes, while Cloud Spanner requires manual sharding.

Answer

Cloud SQL for PostgreSQL offers a 99.999% SLA, while Cloud Spanner offers a 99.99% SLA.

Question 13

A healthcare company runs a patient portal on Google Kubernetes Engine (GKE). They need to comply with HIPAA and ensure that protected health information (PHI) is encrypted at rest and in transit. They also want to minimize operational overhead. Which two actions should they take? (Choose two.)

Accepted Answer

Enable Application-layer Secrets Encryption in GKE using a customer-managed encryption key (CMEK) from Cloud KMS.. To meet HIPAA requirements for encryption at rest and in transit while minimizing operational overhead, the company should enable Application-layer Secrets Encryption with CMEK to encrypt Kubernetes Secrets at rest, and implement a service mesh with mTLS to encrypt service-to-service traffic. These actions provide encryption for sensitive data and are manageable within GKE.

Answer

Use Google-managed encryption keys for all persistent disks and enable HTTPS for all ingress traffic.

Answer

Store all PHI in a Cloud SQL database with encryption at rest enabled and use Cloud VPN for all client connections.

Answer

Use Cloud Armor to protect the portal from web attacks and enable VPC Service Controls to restrict data access.

Question 14

A media company uses Cloud Storage to store video files that are accessed frequently for the first 30 days and then rarely after that. They want to minimize storage costs while ensuring that files are immediately accessible when needed. What should they do?

Accepted Answer

Set a lifecycle rule to transition objects to Nearline Storage after 30 days and then to Coldline Storage after 90 days.. The media company needs to minimize costs for data that is frequently accessed for 30 days and rarely thereafter, while maintaining immediate accessibility. A lifecycle policy that transitions to Nearline Storage after 30 days and then to Coldline Storage after 90 days aligns with the access patterns and avoids early deletion fees, as Nearline has a 30-day minimum and Coldline has a 90-day minimum.

Answer

Set a lifecycle rule to transition objects to Archive Storage after 30 days.

Answer

Set a lifecycle rule to transition objects to Coldline Storage after 30 days.

Answer

Set a lifecycle rule to transition objects to Nearline Storage after 30 days.

Question 15

A logistics company runs a real-time tracking application on Google Cloud. The application uses Pub/Sub to ingest events from thousands of vehicles and Dataflow to process them, writing results to BigQuery. During peak hours, they notice that the Dataflow pipeline is lagging, and the Pub/Sub backlog is growing. They want to optimize the pipeline to handle the increased load without increasing costs significantly. What should they do?

Accepted Answer

Use Dataflow's streaming engine and optimize the pipeline by reducing the number of shuffles and using combiner functions.. To optimize the Dataflow pipeline for increased load without significant cost increases, the best approach is to use Dataflow's streaming engine and optimize the pipeline code by reducing shuffles and using combiner functions. These changes improve efficiency and throughput, allowing the pipeline to handle more data with the same or fewer resources, thus controlling costs.

Answer

Switch from Pub/Sub to a different messaging service like Kafka on Compute Engine to handle higher throughput.

Answer

Enable autoscaling in Dataflow and set a maximum number of workers based on the expected peak load.

Answer

Increase the number of workers in the Dataflow pipeline by using a larger machine type and more maximum workers.