Google PCA Designing and Planning a Cloud Solution Architecture — 15 Questions

Question 1

A company wants to migrate its on-premises monolithic application to Google Cloud with minimal changes. They plan to run it on a virtual machine with a predictable workload that runs 24/7 for a one-year commitment. Which compute option is MOST cost-effective?

Accepted Answer

Committed use discounts. Committed use discounts provide significant cost savings for predictable, always-on workloads over a 1-year (or 3-year) term. While preemptible VMs are cheaper, they can be terminated at any time, making them unsuitable for a 24/7 workload. On-demand VMs are more expensive than committed use. Spot VMs are similar to preemptible but still not guaranteed.

Answer

Spot VMs

Answer

On-demand VMs

Answer

Preemptible VMs

Question 2

A company is building a microservices architecture on Google Kubernetes Engine (GKE) and needs to ensure each microservice can only access specific Cloud Storage buckets. IAM permissions should be assigned at the pod level, not at the node level. What is the recommended approach?

Accepted Answer

Use Workload Identity to bind a Kubernetes service account to a GCP service account. Workload Identity is the recommended approach because it allows you to bind a Kubernetes service account (KSA) to a Google Cloud service account (GSA), enabling pods to authenticate to Google Cloud APIs without managing or storing service account keys. This satisfies the requirement for pod-level IAM permissions, as each pod can inherit the GSA's roles (e.g., roles/storage.objectViewer) for specific Cloud Storage buckets, while avoiding node-level assignment.

Answer

Store service account keys in Kubernetes secrets and mount them into each pod

Answer

Use a CSI driver to mount IAM credentials into pods

Answer

Assign a service account to each node pool and configure pod security policies

Question 3

A company deploys a global application on Cloud Run and uses Cloud SQL for user data. They need sub-10ms read latency for frequently accessed user profiles across regions. Traffic is unpredictable but consistently high. The database must support strong consistency. Which approach meets these requirements?

Accepted Answer

Use Cloud Spanner with multi-region configuration. Cloud Spanner with a multi-region configuration is the only option that provides strong consistency, sub-10ms read latency, and automatic scaling for unpredictable, high traffic across regions. It uses synchronous replication with Paxos-based consensus to ensure globally consistent reads without sacrificing performance.

Answer

Use Cloud Bigtable with replication across regions

Answer

Use Cloud SQL with read replicas in each region and enable synchronous replication

Answer

Use Cloud SQL with cross-region replication and Memorystore cache in each region

Question 4

A company wants to run a stateful application on Google Kubernetes Engine that requires persistent storage with high read/write performance from multiple pods simultaneously. Which storage option should they use?

Accepted Answer

Filestore. Filestore provides a fully managed NFS file share that can be mounted by multiple pods simultaneously with high performance. Persistent Disk with ReadWriteMany access mode is not supported by GKE; it supports ReadOnlyMany. Cloud Storage is object storage not suitable for POSIX-like file systems. Local SSDs are ephemeral and cannot be shared across pods.

Answer

Cloud Storage FUSE

Answer

Local SSDs on each node

Answer

Persistent Disk with ReadWriteMany access mode

Question 5

A company needs to store archival data that is accessed less than once a year and must be retained for 10 years for compliance. The data retrieval time is not critical. Which Cloud Storage class is MOST cost-effective?

Accepted Answer

Archive storage class. Archive storage is the cheapest class for long-term archival data accessed less than once a year. Coldline is for 90-day access, Nearline for 30-day, and Standard for frequent access. Archive has the lowest storage cost but higher retrieval costs and minimum storage duration of 365 days.

Answer

Coldline storage class

Answer

Standard storage class

Answer

Nearline storage class

Question 6

A company deploys a web application on Compute Engine behind a global HTTP(S) Load Balancer. Users report intermittent 502 errors. Investigation shows backend instances are healthy. What is the MOST likely cause?

Accepted Answer

The backend service's timeout setting is too low. HTTP 502 errors from a load balancer often indicate that the backend is closing the connection or not responding in time. The backend connection timeout or request timeout might be too low. Instance health checks passing means the instance is responsive to health check probes, but the actual application requests may be timing out. SSL certificate errors usually cause 400-level errors. Insufficient capacity would cause 503 or 504 errors. Incorrect firewall rules would prevent health checks from passing.

Answer

The SSL certificate has expired

Answer

The firewall rules are blocking traffic from the load balancer

Answer

The backend instances are running out of memory

Question 7

A company is migrating its on-premises data center to Google Cloud. They need a dedicated, low-latency, high-bandwidth connection between their on-premises network and VPC. They anticipate consistent traffic above 10 Gbps. Which connectivity option should they choose?

Accepted Answer

Dedicated Interconnect. Dedicated Interconnect provides a direct physical connection with bandwidth up to 100 Gbps per circuit, ideal for high-throughput, low-latency requirements. Partner Interconnect relies on a service provider and typically offers lower bandwidth. HA VPN is over the public internet and may not meet high bandwidth or low latency. Classic VPN is a single tunnel without high availability.

Answer

Partner Interconnect

Answer

HA VPN

Answer

Classic VPN

Question 8

A company wants to run a containerized application that scales down to zero when not in use and only incurs costs when requests are being processed. They do not want to manage infrastructure. Which compute service should they use?

Accepted Answer

Cloud Run. Cloud Run is a fully managed serverless container platform that scales to zero and charges only for resources used during request processing. GKE Autopilot manages infrastructure but does not scale to zero (has a minimum node). Cloud Functions scales to zero but is for functions, not containers. Compute Engine always has running VMs.

Answer

Cloud Functions

Answer

Compute Engine

Answer

Google Kubernetes Engine (GKE) Standard

Question 9

A company needs to run a batch job that processes data stored in Cloud Storage. The job runs once per day and takes about 2 hours on 8 vCPUs. Cost is the primary concern, and the job can be interrupted. Which compute option is MOST cost-effective?

Accepted Answer

Preemptible VM. Preemptible VMs (the predecessor to Spot VMs in GCP) are the most cost-effective option for batch workloads that can tolerate interruptions because they offer up to 80% discount compared to standard on-demand VMs. Since the job runs once per day for only 2 hours and can be interrupted, Preemptible VMs are ideal—they can be terminated at any time by GCP, but the job can be retried or resumed. Spot VMs are the modern equivalent, but Preemptible VMs are still available and are explicitly designed for fault-tolerant, short-lived batch processing.

Answer

Standard VM (on-demand)

Answer

Spot VM

Answer

Sole-tenant node

Question 10

A company uses Cloud SQL for MySQL and wants to achieve a Recovery Time Objective (RTO) of less than 5 minutes and a Recovery Point Objective (RPO) of less than 1 minute for a regional disaster. Which configuration should they use?

Accepted Answer

Use Cloud Spanner with multi-region configuration. Cloud Spanner with a multi-region configuration is the only option that can meet an RTO of less than 5 minutes and an RPO of less than 1 minute for a regional disaster. Cloud Spanner provides synchronous replication across multiple regions, ensuring that data is consistent with no data loss (RPO near zero) and automatic failover within seconds, achieving the required RTO. Cloud SQL, even with cross-region replicas or high availability, relies on asynchronous replication (for cross-region) or semi-synchronous replication (for regional HA), which cannot guarantee sub-minute RPO or sub-5-minute RTO during a full regional outage.

Answer

Use Cloud SQL with point-in-time recovery and automated backups

Answer

Use Cloud SQL with cross-region replica and failover

Answer

Enable high availability (regional) with a standby zone

Question 11

A company is designing a microservices architecture on Google Kubernetes Engine (GKE). They need to expose a set of internal microservices to other services within the same VPC, but not to the internet. Which GKE service type should they use?

Accepted Answer

ClusterIP. A ClusterIP service exposes the service on a cluster-internal IP, accessible only within the cluster. A NodePort service exposes on each node's IP but is not suitable for internal-only VPC access. A LoadBalancer service creates an external load balancer, which is internet-facing unless configured as internal. An Internal Load Balancer (via ingress) can be used but is more complex; the simplest internal-only service type is ClusterIP with a proxy.

Answer

LoadBalancer

Answer

NodePort

Answer

ExternalName

Question 12

A company wants to implement an event-driven architecture where uploads to a Cloud Storage bucket trigger processing in a serverless function. The function must process each object within a few seconds and handle bursts of thousands of uploads. Which service should they use?

Accepted Answer

Cloud Functions. Cloud Functions is the correct choice because it is a fully managed, event-driven serverless compute service that natively triggers on Cloud Storage bucket events (e.g., object finalize/create). It automatically scales from zero to thousands of concurrent invocations within seconds, meeting the burst requirement, and has a maximum timeout of 9 minutes (well above the 'few seconds' requirement).

Answer

Google Kubernetes Engine

Answer

Cloud Run for Anthos

Answer

Compute Engine with autoscaling

Question 13

A company is moving a legacy application to Google Cloud. The application uses a PostgreSQL database and must be re-platformed with minimal code changes. They require high availability across zones and automatic failover within the same region. Which TWO database services meet these requirements? (Choose 2.)

Accepted Answer

AlloyDB for PostgreSQL. Cloud SQL for PostgreSQL provides regional high availability with automatic failover across zones. AlloyDB is PostgreSQL-compatible, offers high availability with a primary and standby instance, and is optimized for performance. Cloud Spanner is globally distributed but requires more code changes. Bigtable is not PostgreSQL-compatible. Firestore is NoSQL.

Answer

Firestore

Answer

Cloud Bigtable

Answer

Cloud Spanner

Question 14

A company wants to run a stateful workload on GKE that requires persistent storage with low latency and high IOPS. They need to use the fastest available storage and are willing to lose data if the pod is rescheduled. Which THREE options should they consider? (Choose 3.)

Accepted Answer

Local SSDs (ephemeral storage). Local SSDs provide the highest IOPS and lowest latency but are ephemeral. Persistent Disk SSD offers high IOPS and low latency while being persistent. Persistent Disk Balanced also provides good performance and durability, though not as high as SSD. All three are among the fastest available storage options in GKE and should be considered for the workload, with the choice depending on the tolerance for data loss and performance needs.

Answer

Memorystore for Redis

Answer

Persistent Disk (SSD)

Answer

Cloud Storage FUSE

Answer

Persistent Disk (balanced)

Question 15

A company wants to run containerized applications on Google Cloud with minimal operational overhead. They prefer to use a serverless container platform. Which TWO compute options should they consider? (Choose 2.)

Accepted Answer

GKE Autopilot. Cloud Run (fully managed) is a serverless container platform that abstracts infrastructure. GKE Autopilot is also serverless in the sense that Google manages nodes, but it still requires cluster configuration. However, GKE Autopilot is considered a 'serverless' Kubernetes offering. Cloud Functions is for functions, not containers. GKE Standard requires node management. Compute Engine is not container-focused.

Answer

Compute Engine

Answer

GKE Standard

Answer

Cloud Functions