Courseiva

Google Cloud Digital Leader (GCDL) — Questions 301–375

848 questions total · 12 pages · All types, answers revealed

Page 4

Page 5 of 12

Page 6
301
MCQmedium

A company wants to use machine learning models but has no in-house data science team. They need a service that allows them to train custom models using their own data without managing infrastructure. Which Google Cloud service should they use?

A.AI Platform Notebooks
B.BigQuery ML
C.Cloud TPUs
D.Vertex AI (including AutoML)
AnswerD

Vertex AI, including its AutoML capabilities, is a unified managed machine learning platform that addresses the need for ML without in-house expertise. AutoML automates the entire model development pipeline: it handles data validation, feature engineering, architecture search, and hyperparameter tuning, then automatically deploys the trained model to a scalable serving endpoint. Users simply upload labeled data and specify the objective, and the platform manages the infrastructure. This makes Vertex AI the correct choice for a company that wants to leverage ML models with minimal manual involvement and no dedicated data science team.

Why this answer

Vertex AI provides a unified platform for ML, including AutoML for custom model training without managing infrastructure. BigQuery ML is for SQL-based ML, AI Platform Notebooks require manual setup, and Cloud TPUs are hardware accelerators, not a managed service.

302
MCQmedium

A government agency is evaluating whether to move citizen services to the cloud. Officials are concerned about vendor lock-in — specifically that they might become entirely dependent on one provider. Which approach best mitigates this risk while still allowing the agency to benefit from cloud services?

A.Avoiding cloud entirely and keeping all services on-premises to maintain full control
B.Using only one cloud provider's most specialized proprietary services for all workloads to maximize integration
C.Adopting open standards, containerized workloads, and a multi-cloud or hybrid architecture to preserve portability while benefiting from cloud services
D.Negotiating a contract with the cloud provider that forbids them from changing their service APIs
AnswerC

Adopting open standards such as Kubernetes for orchestration, containerized workloads for packaging dependencies, and a multi-cloud or hybrid architecture to distribute risk directly addresses the root cause of cloud lock-in: proprietary APIs and data gravity. Containers run consistently across any CNCF-compliant cluster, while SQL-compatible databases and open protocols (e.g., OIDC, S3) let teams migrate components between providers without rewriting code. This preserves the ability to consume high-value cloud services (managed AI, analytics, serverless) while retaining the architectural freedom to move workloads when business, cost, or compliance needs change.

Why this answer

Adopting open standards (e.g., OCI container images, Kubernetes APIs), containerized workloads, and a multi-cloud or hybrid architecture ensures workload portability across providers. This approach prevents vendor lock-in by allowing the agency to migrate services between cloud platforms or back to on-premises without rewriting applications, while still leveraging cloud benefits like scalability and managed services.

Exam trap

Google Cloud often tests the misconception that avoiding cloud entirely or using a single provider's proprietary services is safer, but the correct answer emphasizes architectural portability through open standards and containerization, not contractual or avoidance-based solutions.

How to eliminate wrong answers

Option A is wrong because avoiding the cloud entirely forfeits scalability, cost efficiency, and operational benefits, and does not address vendor lock-in—it simply replaces it with hardware vendor lock-in. Option B is wrong because using only one provider's proprietary services (e.g., AWS Lambda, Azure Functions) maximizes integration but creates deep dependency on that provider's APIs and runtime, making migration nearly impossible without significant rework. Option D is wrong because negotiating a contract that forbids API changes is impractical—cloud providers continuously evolve APIs for security, performance, and features; such a clause would be unenforceable and would prevent the provider from delivering updates, effectively freezing the platform.

303
Matchingmedium

Match each Google Cloud migration term to its description.

Drag a concept onto its matching description — or click a concept then click the description.

Concepts
Matches

Move workloads without modification

Tool to migrate VMs to GCP

Physical device for large offline data transfers

Online data transfer from other cloud or on-prem

Automated data import into BigQuery

Why these pairings

Correct matches: Lift and Shift moves apps without changes; Replatform optimizes; Refactor re-architects; Migrate for Compute Engine is the VM migration tool. Common confusions include mixing tool names and migration strategies.

304
MCQmedium

A company uses Cloud Storage to store backup files. The files are accessed on average once per year. To minimize storage costs while complying with a 365-day retention policy, which storage class should they use?

A.Coldline
B.Archive
C.Nearline
D.Standard
AnswerB

Archive is the lowest-cost storage class in Cloud Storage, purpose-built for long-term archival where data is accessed less than once a year. Its 365-day minimum storage duration aligns exactly with the retention policy; because the data is kept for 365 days, there is no early-deletion penalty. While retrieval costs are higher, the once-per-year access pattern makes these costs negligible, resulting in the lowest total cost of ownership.

Why this answer

Archive storage class is the correct choice because it offers the lowest storage cost for data that is accessed less than once per year, while still meeting the 365-day retention policy. Archive has a 365-day minimum storage duration, which aligns perfectly with the retention requirement, and its retrieval costs are acceptable given the infrequent access pattern.

Exam trap

Google Cloud often tests the misconception that Archive is only for data that is never accessed, but it actually allows retrieval with higher latency and costs, making it suitable for data accessed as infrequently as once per year with a 365-day retention policy.

How to eliminate wrong answers

Option A (Coldline) is wrong because Coldline has a 90-day minimum storage duration and higher storage cost than Archive, making it less cost-effective for data accessed once per year. Option C (Nearline) is wrong because Nearline has a 30-day minimum storage duration and higher storage cost than Archive, and is designed for data accessed less than once per month, not once per year. Option D (Standard) is wrong because Standard has no minimum storage duration but has the highest storage cost, making it the most expensive option for long-term, rarely accessed data.

305
MCQhard

A company must meet regulatory requirements that restrict where data can be stored and processed. They need to ensure that Google Cloud personnel have limited and audited access to their data. Which combination of services should they use?

A.Access Transparency and VPC Service Controls
B.Assured Workloads and Access Transparency
C.Cloud KMS and Cloud Audit Logs
D.VPC Service Controls and Cloud Audit Logs
AnswerB

Assured Workloads is the correct foundation because it enforces data residency by pinning resources to a selected region and imposes access restrictions such as preventing Google personnel from accessing customer data without explicit approval. Access Transparency complements it by providing detailed, audit-ready logs of any Google employee access actions, satisfying the regulatory need for both enforcement and accountability. This pairing directly addresses location restrictions and personnel access tracking, which are the core requirements.

Why this answer

Assured Workloads provides regulatory compliance controls and access restrictions for specific regions. Access Transparency logs Google personnel access. Cloud Audit Logs track user activity.

The question asks for a combination that restricts personnel access and provides audit logs.

306
MCQeasy

A data engineering team needs to build a pipeline that reads event data from Pub/Sub in real time, applies transformations and aggregations, and writes results to BigQuery — all without managing any infrastructure. Which Google Cloud product is designed for this serverless stream and batch data processing use case?

A.Cloud Dataflow, Google Cloud's serverless stream and batch data processing service built on Apache Beam
B.Cloud Composer, Google Cloud's managed Apache Airflow service for workflow orchestration
C.Cloud Dataproc, Google Cloud's managed Spark and Hadoop service
D.BigQuery directly, using streaming inserts to load Pub/Sub data in real time
AnswerA

Dataflow is exactly right: serverless (no infrastructure management), supports both streaming (from Pub/Sub) and batch, applies transformations and aggregations, and writes natively to BigQuery. The Pub/Sub → Dataflow → BigQuery pattern is one of the most common data engineering pipelines on Google Cloud.

Why this answer

Cloud Dataflow is Google Cloud's fully managed, serverless service for both stream and batch data processing, built on Apache Beam. It directly reads from Pub/Sub, applies transformations and aggregations using the Beam SDK, and writes the results to BigQuery without requiring any infrastructure management, making it the correct choice for this use case.

Exam trap

Google Cloud often tests the distinction between a data processing service (Dataflow) and a data ingestion or orchestration service, leading candidates to mistakenly choose BigQuery streaming inserts or Cloud Composer for real-time transformations.

How to eliminate wrong answers

Option B is wrong because Cloud Composer is a workflow orchestration service (managed Apache Airflow) designed to schedule and coordinate tasks, not to perform real-time stream processing or data transformations. Option C is wrong because Cloud Dataproc is a managed Spark and Hadoop service that requires cluster management and is not serverless; it is better suited for batch processing and big data analytics on existing clusters, not for serverless stream processing. Option D is wrong because BigQuery streaming inserts can load data from Pub/Sub but do not provide built-in transformations or aggregations; they are a data ingestion method, not a full data processing pipeline, and lack the serverless stream processing capabilities of Dataflow.

307
MCQmedium

A developer needs to trigger a serverless function whenever a new file is uploaded to a Cloud Storage bucket. The function will process the file and store results in Firestore. Which Google Cloud service should they use for the function?

A.App Engine
B.Compute Engine
C.Cloud Run
D.Cloud Functions
AnswerD

Cloud Functions is a purpose-built, event-driven serverless compute platform that natively responds to GCP events, including Cloud Storage object finalization (e.g., the 'google.storage.object.finalize' trigger). When an object is uploaded, Cloud Functions automatically invokes the function with the event metadata, scaling to zero when idle and managing all infrastructure transparently. This aligns exactly with the requirement to trigger a single-purpose serverless function on a storage event, without needing a container runtime, a web server, or an intermediary like Pub/Sub.

Why this answer

Cloud Functions is an event-driven serverless compute service that can be triggered by Cloud Storage events.

308
MCQhard

An organization has multiple projects with many underutilized Compute Engine instances. They want to identify idle instances and generate recommendations for downsizing or deleting them. The finance team wants to see cost savings estimates. Which tool should they use?

A.Google Cloud Pricing Calculator
B.Billing export to BigQuery
C.Active Assist
D.Cost Management dashboard
AnswerC

Active Assist, which is built on the Google Cloud Recommender, provides prescriptive recommendations specifically for underutilized resources, including idle VM and Cloud SQL instances as well as VM rightsizing recommendations. It analyzes actual utilization metrics (such as CPU and memory over a 7- or 14-day window) alongside machine type capacities, then ranks opportunities with estimated monthly cost savings. Because Active Assist is designed to identify precisely the kind of waste described in the scenario—underutilized resources across many projects—it is the correct, ready-to-use tool for this situation.

Why this answer

Active Assist is Google Cloud's portfolio of intelligent recommendations, including the VM rightsizing recommender and idle VM recommender, which analyze usage metrics and suggest downsizing or deleting underutilized instances with estimated cost savings. It surfaces these directly in the console and via API, matching the finance team's need for savings estimates.

Exam trap

GCDL often tests the boundary between cost visibility tools (Cost Management, Billing export) and cost optimization tools (Active Assist), so candidates pick a dashboard when the question asks for recommendations.

How to eliminate wrong answers

Option A is wrong because the Pricing Calculator estimates costs for planned resources, not recommendations for existing idle instances. Option B is wrong because Billing export to BigQuery provides raw cost data for custom analysis but does not generate rightsizing or idle-instance recommendations. Option D is wrong because the Cost Management dashboard visualizes spend and trends but does not analyze instance utilization or produce downsizing recommendations.

309
MCQeasy

An engineer needs to organize resources for multiple departments and enforce organization-wide policies such as restricting VM external IP addresses. Which GCP resource hierarchy level should the policy be applied at to ensure all projects inherit it?

A.Resource
B.Folder
C.Project
D.Organization node
AnswerD

The organization node is the root of the Google Cloud resource hierarchy, encompassing all folders and projects under it. IAM policies and organizational policies set at the organization node are inherited by every child folder and project, which means you can enforce uniform access controls, billing, and audit logging across multiple departments from a single central point. This allows administrators to model the entire enterprise as a single tree and apply department-specific organization policies by creating folders underneath the node, while still having a global fallback policy. Therefore, the organization node is the correct place to define high-level, cross-departmental resource organization and governance.

Why this answer

In Google Cloud's resource hierarchy, the organization node is the root and sits above folders and projects. IAM policies and organization policies applied at the organization node are inherited by all folders, projects, and resources beneath it. To enforce a policy such as restricting external IP addresses across all projects in the company, the organization node is the correct level because it guarantees universal inheritance.

Exam trap

The trap is assuming that a folder or project-level policy is sufficient for organization-wide enforcement; candidates often forget that only the organization node guarantees inheritance to every project, including future ones.

How to eliminate wrong answers

Option A is wrong because a resource (e.g., a VM instance) is the lowest level and policies applied there affect only that single resource, not the entire organization. Option B is wrong because a folder applies only to projects and subfolders within that folder, so it would not cover departments outside that folder. Option C is wrong because a project-level policy applies only to resources within that one project, not to all projects in the organization.

310
MCQmedium

A manufacturing company deploys sensors in its factories that send data to cloud platforms for real-time analysis. The cloud-based system predicts equipment failures 48 hours in advance, enabling maintenance before failures occur. What operational model shift does this represent?

A.The company is automating its accounting system using cloud software.
B.A shift from reactive (break-fix) maintenance to predictive maintenance, enabled by IoT sensor data and cloud AI/ML.
C.The company is replacing human maintenance workers with robots.
D.The factory is migrating its ERP system to the cloud to improve supply chain visibility.
AnswerB

The factory is shifting from run-to-failure (break-fix) maintenance to predictive maintenance, where IoT sensors—such as vibration, temperature, and acoustic monitors—continuously stream telemetry from equipment into the cloud. Cloud-based machine learning models analyze this data for anomaly detection and remaining useful life (RUL) estimation, allowing maintenance to be scheduled proactively before an actual breakdown occurs. This transformation is the archetypal Industry 4.0 use case combining edge sensing, cloud AI/ML, and operational decision-making, directly reducing unplanned downtime and maintenance costs.

Why this answer

The scenario describes a shift from reactive maintenance (fixing equipment after it fails) to predictive maintenance, where IoT sensors collect real-time data and cloud-based AI/ML models analyze it to forecast failures 48 hours in advance. This transformation leverages cloud computing's scalability and advanced analytics to prevent downtime, rather than simply automating existing processes or replacing human roles.

Exam trap

Google Cloud often tests the distinction between operational model shifts (e.g., reactive to predictive) and simple technology replacements (e.g., automating accounting or migrating ERP), so candidates mistakenly choose options that describe a different cloud benefit (like cost savings or scalability) rather than the specific shift in maintenance strategy.

How to eliminate wrong answers

Option A is wrong because it misrepresents the operational shift as accounting automation, while the question focuses on equipment maintenance and failure prediction, not financial processes. Option C is wrong because it incorrectly suggests replacing human workers with robots, whereas the scenario uses cloud AI/ML to augment human decision-making for maintenance scheduling, not to eliminate workers. Option D is wrong because it confuses the shift with ERP migration for supply chain visibility, but the core change is about maintenance strategy (predictive vs. reactive), not enterprise resource planning or supply chain management.

311
MCQeasy

An organization needs to store API keys, database passwords, and certificates securely, with automatic rotation and audit logging. Which Google Cloud service should they use?

A.Cloud Storage with encryption
B.Cloud KMS
C.Cloud HSM
D.Secret Manager
AnswerD

Secret Manager is the purpose-built Google Cloud service for storing API keys, database passwords, certificates, and other sensitive configuration data. It maintains immutable secret versions, supports granular IAM roles and conditions (e.g., accessor, giver, view access to a specific secret version), and is natively integrated with Cloud Audit Logs so every access to a secret's plaintext is recorded. You can automate rotation by using Cloud Functions or Cloud Scheduler to regenerate the secret value and add a new version, and then grant applications access only to the active version. It is the only option here that directly satisfies the requirements for secure storage, rotation, and auditing of manage to manage runtime secrets.

Why this answer

Secret Manager is purpose-built for storing API keys, passwords, and certificates with versioning, automatic rotation, IAM-based access control, and audit logging via Cloud Audit Logs. It directly addresses the requirement for secure storage plus rotation and auditing.

Exam trap

The trap is confusing Cloud KMS (encryption keys) with Secret Manager (application secrets); candidates often pick KMS because it sounds more 'secure,' but KMS does not store or rotate API keys and passwords.

How to eliminate wrong answers

Option A is wrong because Cloud Storage with encryption only encrypts objects at rest; it does not provide secret versioning, rotation, or fine-grained secret access auditing. Option B is wrong because Cloud KMS manages encryption keys, not application secrets like API keys and passwords, and does not rotate secrets themselves. Option C is wrong because Cloud HSM provides hardware-backed key storage for cryptographic keys, not a general secret store with rotation for API credentials.

312
MCQhard

A logistics company collects GPS data from 50,000 trucks every 30 seconds. Previously they sampled only 1% of this data due to storage costs. In the cloud, they store and analyze 100% of the data and discover route optimization patterns that reduce fuel costs by 12%. Which concept does this illustrate about cloud and data?

A.Cloud storage is cheaper per GB than on-premises — the only benefit is cost reduction.
B.Cloud scale removes data storage and analysis constraints, enabling companies to derive business insights from complete datasets that were previously too costly to collect and process.
C.GPS data is only useful when analyzed by Google's AI — the company's own analytics wouldn't find patterns.
D.The company should have used data sampling more aggressively to reduce cloud costs further.
AnswerB

Correct: on-premises infrastructure imposes fixed capacity and cost ceilings, which typically force data retention policies like time-windowing or sampling. Cloud's elastic storage and distributed processing lower the marginal cost of holding and analyzing every telemetry point, so organizations can run full-dataset analytics to detect subtle correlations—such as the link between route variations and fuel efficiency. This shift from sampling to comprehensive data analysis is what unlocks the 12% fuel-saving insight, proving that cloud's real advantage is enabling a whole-dataset decision-making paradigm.

Why this answer

It directly captures the core transformation that cloud computing enables: the removal of data storage and processing constraints. By moving from sampling 1% of GPS data to analyzing 100%, the company could identify route optimization patterns that were invisible in the sampled subset, leading to a 12% fuel cost reduction. This illustrates how cloud elasticity and pay-as-you-go pricing allow businesses to process complete datasets, unlocking insights that were previously economically infeasible with on-premises or limited storage.

Exam trap

The GCDL exam often tests the misconception that cloud benefits are purely about cost savings (Option A) or that specialized AI is required (Option C), when the real transformative value is the ability to process complete datasets at scale, removing prior constraints on data volume and analysis.

How to eliminate wrong answers

Option A is wrong because it incorrectly claims that cost reduction is the only benefit of cloud storage; in reality, the primary advantage demonstrated here is the ability to store and analyze 100% of data, which enables new business insights (route optimization) that were impossible with sampling. Option C is wrong because it falsely asserts that GPS data is only useful when analyzed by Google's AI; the company's own analytics (or any cloud-based analytics service) can find patterns, and the question does not specify any dependency on Google's proprietary AI. Option D is wrong because it suggests more aggressive sampling to reduce costs, which would directly contradict the lesson learned: sampling prevented discovery of the optimization pattern, and the cloud's value is in eliminating the need for sampling, not increasing it.

313
MCQmedium

A security team needs to protect a web application behind an HTTP(S) Load Balancer from SQL injection and cross-site scripting (XSS) attacks. Which Google Cloud service provides these protections?

A.Cloud CDN
B.VPC firewall rules
C.Cloud DNS
D.Cloud Armor
AnswerD

Cloud Armor is the correct answer because it is Google Cloud's managed Web Application Firewall (WAF) service that provides application-layer (L7) protection for services behind Cloud Load Balancing. It includes pre-configured rules from the OWASP ModSecurity Core Rule Set, which specifically detects and blocks SQL injection and cross-site scripting patterns in HTTP headers, query parameters, and request bodies. Cloud Armor also offers adaptive protection, rate limiting, and geo-based access controls, making it the appropriate tool for defending a web application against these common web exploits.

Why this answer

Cloud Armor is Google Cloud's edge security service that attaches to HTTP(S) Load Balancers and provides WAF capabilities, including preconfigured rules for SQL injection and XSS (the OWASP ModSecurity core rule set). It inspects requests at the load balancer before they reach the backend, blocking malicious payloads.

Exam trap

GCDL often tests whether candidates know Cloud Armor requires an HTTP(S) Load Balancer and provides WAF (SQLi/XSS) protection, versus VPC firewall rules which only filter L3/L4 traffic.

How to eliminate wrong answers

Option A is wrong because Cloud CDN caches and accelerates content delivery; it does not inspect or block application-layer attacks like SQLi or XSS. Option B is wrong because VPC firewall rules operate at L3/L4 (IP, port, protocol) and cannot inspect HTTP payloads for injection attacks. Option C is wrong because Cloud DNS is a managed DNS service for name resolution and routing — it has no WAF or application-layer inspection capability.

314
MCQeasy

A marketing team wants to create interactive dashboards and reports for business stakeholders using data stored in BigQuery, without writing code. Which Google Cloud product is most appropriate for this self-service business intelligence requirement?

A.Looker Studio, Google's self-service data visualization tool with native BigQuery connectivity and a no-code interface
B.Cloud Dataflow, Google Cloud's stream and batch data processing service
C.BigQuery itself, where the marketing team writes SQL queries to generate reports
D.Vertex AI, Google Cloud's unified machine learning platform
AnswerA

Looker Studio is exactly designed for this use case: non-technical marketing teams creating interactive dashboards from BigQuery data through a visual interface, with no SQL or coding required. It has native BigQuery connectivity and is free to use.

Why this answer

Looker Studio (formerly Google Data Studio) is the correct choice because it is Google Cloud's self-service business intelligence tool that provides a no-code, drag-and-drop interface for creating interactive dashboards and reports. It has native, built-in connectivity to BigQuery, allowing the marketing team to visualize data without writing any code or SQL queries.

Exam trap

The trap here is that candidates may confuse BigQuery's ability to run SQL queries and generate results with a no-code self-service BI tool, overlooking that Looker Studio is the dedicated product for interactive, code-free visualization.

How to eliminate wrong answers

Option B is wrong because Cloud Dataflow is a data processing service for building stream and batch pipelines, not a self-service BI or visualization tool; it requires writing code to define data transformations. Option C is wrong because while BigQuery can generate reports via SQL queries, this requires writing code (SQL) and does not provide a no-code, interactive dashboard interface for self-service business intelligence. Option D is wrong because Vertex AI is a machine learning platform for building, deploying, and managing ML models, not a data visualization or reporting tool.

315
Multi-Selecthard

A large enterprise is evaluating moving its data analytics workloads to Google Cloud. Which TWO factors should they consider when comparing on-premises costs to cloud costs? (Choose two.)

Select 2 answers
A.Elimination of hardware maintenance labor costs
B.Increased latency due to network distance
C.Reduction in data center power and cooling expenses
D.Licensing costs for software that may not be compatible with cloud
E.Requirement to redesign applications for cloud-native services
AnswersA, C

Moving analytics workloads to the cloud eliminates the need for on-premises hardware maintenance, because the cloud provider owns and operates the physical servers, storage, and networking equipment. The enterprise no longer pays salaries, benefits, or third-party contracts for staff whose primary role is break/fix, firmware upgrades, or hardware lifecycle management. This direct reduction in operational labor overhead is a recognized cost-saving factor in a cloud migration TCO analysis.

Why this answer

Option A is correct because moving analytics workloads to Google Cloud eliminates the need to maintain physical servers, including tasks like hardware repairs, firmware updates, and replacement cycles, which removes associated labor costs from the on-premises total cost of ownership. Option C is correct because shutting down or downsizing on-premises data centers directly reduces power consumption and cooling expenses, which are significant operational costs in on-premises environments. Option B is not a cost factor but a performance consideration, and network latency does not directly represent a cost comparison item.

Option D is a potential migration or compatibility concern rather than a direct on-premises versus cloud cost factor. Option E is an application design consideration, not a cost factor for comparing on-premises and cloud expenses.

Exam trap

Google Cloud often tests the distinction between direct cost factors (like hardware maintenance and power/cooling) and indirect considerations (like latency, licensing compatibility, and application redesign) to see if candidates can separate TCO line items from migration risks or performance trade-offs.

316
MCQeasy

A company's security policy requires that all cloud-to-cloud communication between services must be encrypted in transit. An auditor asks how Google Cloud handles encryption for network traffic between Google services within its network. What is Google's default approach to encryption in transit within its infrastructure?

A.Google does not encrypt internal traffic by default; customers must configure TLS for all service-to-service communication
B.Google encrypts all traffic between its data centers and internal services by default, with no customer configuration required
C.Google only encrypts traffic that crosses the public internet; internal network traffic is unencrypted for performance
D.Encryption in transit is the customer's responsibility for all traffic, including traffic within Google's network
AnswerB

Google uses Application Layer Transport Security (ALTS) to authenticate and encrypt all traffic between Google services and between data centers by default. This is a core Google infrastructure security commitment, not an optional feature customers must enable.

Why this answer

Google Cloud encrypts all network traffic between its data centers and internal services by default, using application-layer (e.g., gRPC with TLS) and link-layer encryption (e.g., MACsec or similar). This is a foundational security measure that requires no customer configuration, ensuring data is protected in transit even within Google's own infrastructure.

Exam trap

The trap here is that candidates often assume internal cloud provider networks are unencrypted for performance reasons, but Google Cloud encrypts all inter-service traffic by default, making options that require customer action or that claim no encryption incorrect.

How to eliminate wrong answers

Option A is wrong because Google does encrypt internal traffic by default; customers do not need to configure TLS for service-to-service communication within Google's network. Option C is wrong because Google encrypts not only traffic crossing the public internet but also internal network traffic between data centers and services, so performance is not a reason to leave it unencrypted. Option D is wrong because encryption in transit within Google's network is Google's responsibility and is handled automatically, not the customer's.

317
MCQeasy

A business user asks what makes cloud storage different from simply buying a larger external hard drive for the office. Which characteristic most clearly differentiates cloud storage from local storage devices?

A.Cloud storage is faster than local storage for all types of data access
B.Cloud storage is accessible from anywhere via the internet, scales elastically without hardware purchases, and provides built-in redundancy across multiple physical locations
C.Cloud storage cannot be used for backup, while external hard drives are purpose-built for backup
D.Cloud storage requires specialized hardware on the customer's side to access data
AnswerB

These three characteristics — universal accessibility, elastic scalability, and built-in geographic redundancy — are what fundamentally differentiate cloud storage from a local external drive. A hard drive is physically local, has fixed capacity, and has no built-in redundancy.

Why this answer

Cloud storage is fundamentally different from local storage because it provides internet-based accessibility, elastic scalability without requiring hardware procurement, and built-in redundancy across geographically distributed data centers. These characteristics enable on-demand resource provisioning and high availability, which a single external hard drive cannot offer. The correct answer, B, captures these core differentiators that align with the NIST definition of cloud computing.

Exam trap

Google Cloud often tests the misconception that cloud storage is inherently faster than local storage, leading candidates to select Option A, but the real differentiator is ubiquitous access and elasticity, not raw speed.

How to eliminate wrong answers

Option A is wrong because cloud storage is not universally faster than local storage; local storage (e.g., USB 3.0 or SATA SSDs) often has lower latency and higher throughput for local access, while cloud storage performance depends on network bandwidth and latency. Option C is wrong because cloud storage is commonly used for backup (e.g., AWS S3 for backup, Azure Backup) and external hard drives are not exclusively purpose-built for backup—they are general-purpose storage devices. Option D is wrong because cloud storage does not require specialized hardware on the customer side; access is typically via standard internet protocols (HTTP/HTTPS, S3 API, NFS) using commodity devices like laptops or smartphones.

318
MCQeasy

Which Google Cloud service provides a fully managed, scalable NoSQL document database suitable for mobile and web applications with real-time data synchronization?

A.Cloud Bigtable
B.Firestore
C.Cloud SQL
D.Memorystore
AnswerB

Firestore is Google Cloud's fully managed, scalable NoSQL document database designed for mobile and web applications, offering real-time listeners, offline persistence, and automatic multi-region replication. Data is stored in documents organized into collections, with flexible schema and powerful querying, making it ideal for user profiles, chat messages, and other semi-structured data. Its serverless scaling and client SDKs are optimized for direct app-to-database access, which distinguishes it from the alternatives.

Why this answer

Firestore is Google Cloud's fully managed, scalable NoSQL document database designed for mobile and web applications, with built-in real-time synchronization across clients. It supports offline persistence and live updates, making it ideal for apps that need real-time data sync. Firestore is the direct successor to the original Cloud Datastore and is optimized for these use cases.

Exam trap

GCDL often tests the confusion between Firestore (document DB with real-time sync) and Bigtable (wide-column analytics DB), causing candidates to pick Bigtable when the question emphasizes mobile/web real-time synchronization.

How to eliminate wrong answers

Option A is wrong because Cloud Bigtable is a wide-column NoSQL database optimized for high-throughput analytics and time-series data, not for mobile/web real-time sync. Option C is wrong because Cloud SQL is a managed relational database (MySQL, PostgreSQL, SQL Server), not a NoSQL document store. Option D is wrong because Memorystore is a managed Redis/Memcached in-memory cache, not a document database with real-time sync.

319
MCQeasy

When data is transmitted between a user's browser and a Google Cloud-hosted web application over HTTPS, which security protection does this provide?

A.It prevents unauthorized users from accessing the Google Cloud Console.
B.It encrypts data in transit between the user's browser and the server, preventing eavesdropping and tampering.
C.It encrypts data stored in the server's database.
D.It authenticates the user and verifies their permissions to use the application.
AnswerB

HTTPS (TLS) establishes an encrypted tunnel between the user's browser and the web server, negotiating a session key through a handshake that authenticates the server using its certificate. This ensures that any data exchanged—such as form submissions or cookies—cannot be read (confidentiality) or altered (integrity) by an eavesdropper or man-in-the-middle. That is precisely the purpose of HTTPS on a web application: protecting data while it is in transit.

Why this answer

HTTPS (HTTP over TLS) encrypts the communication channel between the user's browser and the web server using Transport Layer Security (TLS). This ensures that any data transmitted, such as login credentials or API requests, is protected from eavesdropping and tampering while in transit. It does not protect data at rest or control access to cloud management interfaces.

Exam trap

The GCDL exam often tests the distinction between encryption in transit (HTTPS) and encryption at rest (database encryption), leading candidates to incorrectly select an option about stored data or access control.

How to eliminate wrong answers

Option A is wrong because HTTPS does not control access to the Google Cloud Console; that is managed by IAM (Identity and Access Management) policies and authentication mechanisms like OAuth 2.0. Option C is wrong because HTTPS only encrypts data in transit, not data stored in the server's database; database encryption is handled by techniques like Cloud SQL encryption at rest or customer-managed encryption keys (CMEK). Option D is wrong because HTTPS does not authenticate the user or verify their permissions; user authentication and authorization are handled by the application layer (e.g., using Firebase Authentication or IAM), not by the TLS protocol itself.

320
Multi-Selectmedium

Which TWO statements are benefits of using a multi-cloud strategy? (Choose 2)

Select 2 answers
A.Reduced overall security risk
B.Simplified operational management
C.Improved latency for all users
D.Ability to use best-of-breed services from different providers
E.Reduced vendor lock-in
AnswersD, E

With a multi-cloud strategy, you can deploy each workload on the platform that offers the strongest capability for its requirements—for example, using BigQuery for large-scale analytics, SageMaker for machine learning, and Azure's identity services in the same solution. This prevents being forced into a 'jack of all trades, master of none' situation where a single provider lacks a feature you need at the required scale or cost. The tradeoff is that best-of-breed services often expose proprietary APIs, so you must design for portability or accept some dependency at the service layer.

Why this answer

Option D is correct because a multi-cloud strategy lets an organization select each provider's strongest or most specialized service (for example, AWS Lambda for serverless, Google BigQuery for analytics, or Azure AD for identity) rather than being confined to a single vendor's catalog, directly improving capability fit. Option E is correct because distributing workloads across two or more independent providers reduces dependency on any single vendor's proprietary APIs, pricing changes, and outage risk, making it easier to migrate or rebalance workloads and thereby lowering vendor lock-in. The unmarked options do not belong: A is wrong because multi-cloud typically increases the attack surface and identity/network complexity, so it does not inherently reduce overall security risk; B is wrong because managing multiple providers, IAM models, and billing systems usually complicates rather than simplifies operational management; and C is wrong because latency depends on user location, peering, and workload placement, and multi-cloud does not guarantee improved latency for all users.

Exam trap

GCDL often tests the misconception that multi-cloud automatically reduces security risk or simplifies management, when in fact it often increases complexity and requires more sophisticated governance.

321
MCQmedium

A company's cloud costs have increased by 40% over the past quarter. The operations team wants to identify and address the root causes. Which cost optimization strategies should they investigate first?

A.Immediately upgrade all infrastructure to the latest generation hardware for better efficiency.
B.Identify idle and underutilized resources (oversized VMs, unused disks, unattached IPs), apply lifecycle policies to storage, and commit to CUDs for stable workloads.
C.Migrate all workloads to Spot VMs immediately to reduce costs by 90%.
D.Switch cloud providers to whoever has the lowest advertised list price.
AnswerB

This is the correct first step because it directly targets the largest, most actionable cost leaks in any GCP environment. Active Assist tools like Recommender identify idle VMs and oversized instances for right-sizing, while unused disks and unattached IP addresses can be immediately deleted to stop recurring charges. Storage lifecycle policies (e.g., moving from Standard to Nearline or Coldline) automatically downgrade data access tiers for rarely used data, and Committed Use Discounts (CUDs) lock in lower prices for stable, predictable baseline compute — all without sacrificing performance or availability.

Why this answer

The first step in cloud cost optimization is to identify and eliminate waste from idle or oversized resources, which is the most common source of cost inefficiency. Applying lifecycle policies to storage and committing to Committed Use Discounts (CUDs) for stable workloads are proven strategies to reduce costs without compromising performance. This approach aligns with Google Cloud's recommended FinOps practices, focusing on immediate, high-impact savings before considering architectural changes.

Exam trap

The trap here is that candidates often jump to aggressive cost-cutting measures like migrating to Spot VMs or switching providers, without first addressing the low-hanging fruit of resource waste, which is the most impactful and least risky initial step in cost optimization.

How to eliminate wrong answers

Option A is wrong because immediately upgrading to the latest generation hardware is a capital-intensive strategy that may not address the root cause of cost increases (e.g., idle resources) and could even increase costs if the new hardware is not right-sized. Option C is wrong because migrating all workloads to Spot VMs is risky for production or stateful workloads, as Spot VMs can be terminated at any time with only 30 seconds notice, leading to potential data loss or service disruption. Option D is wrong because switching cloud providers based solely on lowest advertised list price ignores hidden costs like data egress fees, network latency, and the operational overhead of migration, and does not address existing resource inefficiencies.

322
MCQeasy

A software team wants to host their container images securely within Google Cloud and integrate with Cloud Build for CI/CD pipelines and GKE for deployments. Which Google Cloud product serves as the managed repository for storing and managing container images?

A.Cloud Storage, by storing container images as objects in a bucket
B.Artifact Registry, Google Cloud's managed repository for container images and build artifacts with native CI/CD integration
C.Container Registry (gcr.io), the legacy Google container image service
D.Cloud Source Repositories, for storing source code and container build files
AnswerB

Artifact Registry is the intended service for hosting container images in Google Cloud. It supports Docker, Maven, npm, PyPI, and other formats, integrates natively with Cloud Build and GKE, and provides IAM-based access control and automatic vulnerability scanning.

Why this answer

Artifact Registry is Google Cloud's fully managed, next-generation repository for storing, managing, and securing container images and build artifacts. It natively integrates with Cloud Build for CI/CD pipelines and GKE for deployments, providing features like vulnerability scanning, access control via IAM, and support for multiple formats (Docker, Maven, npm, etc.). This makes it the correct choice for the team's requirements.

Exam trap

The GCDL exam often tests the distinction between legacy and current services, so the trap here is that candidates may pick Container Registry (gcr.io) because it is familiar and historically used for container images, but they must recognize that Artifact Registry is the modern, recommended service with broader capabilities and deeper integration.

How to eliminate wrong answers

Option A is wrong because Cloud Storage is an object storage service for arbitrary data (e.g., backups, media files), not a managed container image repository; while you could manually store a container image as a blob, it lacks native container registry features like manifest management, vulnerability scanning, and seamless integration with Cloud Build and GKE. Option C is wrong because Container Registry (gcr.io) is the legacy service that has been superseded by Artifact Registry; it lacks the multi-format support, regional repository options, and advanced security features (e.g., on-demand scanning) that Artifact Registry provides, and Google recommends migrating to Artifact Registry. Option D is wrong because Cloud Source Repositories is a Git-based source code hosting service for version control and collaboration, not a repository for storing container images or build artifacts.

323
Multi-Selectmedium

A gaming company runs a real-time multiplayer game server on Google Kubernetes Engine. They want to optimize costs while ensuring low latency for players across different regions. Which three strategies should they implement? (Choose THREE.)

Select 3 answers
A.Use committed use discounts (CUDs) for sustained resource usage.
B.Use spot VMs with a node taint and toleration for game server pods.
C.Use node auto-provisioning to automatically add nodes based on pod resource requests.
D.Deploy GKE clusters in multiple regions and use a multi-cluster ingress.
E.Use preemptible VMs for game server pods.
AnswersA, C, D

Committed use discounts (CUDs) are the right choice for the always-on, predictable portion of a real-time multiplayer game's compute capacity. By committing to 1- or 3-year usage of vCPUs and memory, you can save up to 57% compared with on-demand pricing, and the commitment has no effect on node preemptibility or workload availability. Because the game servers run continuously to support players, this steady baseline resource usage is an ideal fit for CUDs, directly lowering infrastructure cost without sacrificing reliability.

Why this answer

Option A is correct because committed use discounts (CUDs) provide significant discounts (up to 57% for 3-year commitments) on sustained GKE compute usage, which is ideal for a real-time game server that runs continuously and predictably. Option C is correct because node auto-provisioning dynamically creates node pools based on pending pod resource requests, avoiding over-provisioning and reducing costs while still meeting the latency needs of game server pods. Option D is correct because deploying GKE clusters in multiple regions and using Multi Cluster Ingress (MCI) places players on the nearest healthy cluster, minimizing latency across regions while also improving availability.

Option B is not appropriate because spot VMs can be preempted at any time, which would abruptly terminate real-time game sessions and harm player experience. Option E is likewise incorrect because preemptible VMs are short-lived (max 24 hours) and can be reclaimed with only 30 seconds' notice, making them unsuitable for latency-sensitive, stateful multiplayer game servers.

Exam trap

Google Cloud often tests the misconception that spot/preemptible VMs are acceptable for stateful, latency-sensitive workloads because they are cheaper, but the exam expects you to recognize that their unpredictable termination makes them unsuitable for real-time multiplayer game servers.

324
MCQmedium

A global financial services firm is migrating its risk analysis workloads to Google Cloud to accelerate new model deployments. Which cloud benefit most directly supports faster time-to-market?

A.Security compliance
B.Rapid provisioning and deployment
C.Pay-as-you-go pricing
D.Global infrastructure
AnswerB

Rapid provisioning and deployment are core cloud benefits because infrastructure is delivered via software-defined APIs and orchestration templates. Instead of waiting weeks for physical hardware to be racked and configured, developers can spin up compute, storage, and networking resources in minutes. This dramatically shortens the time from concept to running workload, enabling faster iteration for risk models and analytics.

Why this answer

Rapid provisioning and deployment (B) directly accelerates time-to-market because it allows the firm to spin up risk analysis environments in minutes using Infrastructure as Code (IaC) tools like Terraform or Deployment Manager, rather than waiting weeks for hardware procurement. This speed enables data scientists to iterate on models faster, deploy new versions immediately, and respond to market changes without infrastructure bottlenecks.

Exam trap

Google Cloud often tests the misconception that 'global infrastructure' (D) is the key to faster deployments, but the trap here is that global reach improves latency and redundancy, not the speed of provisioning new resources.

How to eliminate wrong answers

Option A is wrong because security compliance, while critical for financial services, is a risk mitigation requirement that often slows down deployments due to audits and policy checks; it does not directly reduce time-to-market. Option C is wrong because pay-as-you-go pricing is a cost optimization model that shifts expenses from CapEx to OpEx, but it has no direct impact on how quickly workloads can be provisioned or models deployed. Option D is wrong because global infrastructure provides geographic reach and low-latency access, but it does not inherently speed up the deployment lifecycle; provisioning speed is determined by automation and resource availability, not just the number of regions.

325
MCQeasy

A multinational corporation wants to serve content to users worldwide with low latency. They plan to use Google Cloud's content delivery network and points of presence. Which cloud benefit enables this?

A.Cost optimisation
B.Scalability
C.Agility
D.Global reach
AnswerD

Global reach is the Google Cloud architectural pillar that specifically ensures content and services are delivered with low latency and high availability wherever users are located. By using a content delivery network (CDN) like Cloud CDN, content is cached at edge points of presence (PoPs) around the world so that users are served from a nearby node rather than a distant origin server. This reduces round-trip time and improves the digital experience for a multinational corporation's global user base. Therefore, global reach directly matches the requirement of serving content to users worldwide with low latency.

Why this answer

Global reach is the correct benefit because Google Cloud's CDN and global points of presence (PoPs) place content close to users worldwide, reducing latency by serving from edge locations. This directly enables low-latency content delivery across multiple continents. Cost optimisation, scalability, and agility are valid cloud benefits but do not specifically describe the geographic distribution enabled by CDN PoPs.

Exam trap

GCDL often tests the distinction between general cloud benefits (cost, scalability, agility) and the specific benefit of global PoPs, which is global reach.

How to eliminate wrong answers

Option A is wrong because cost optimisation refers to reducing infrastructure spend through pay-as-you-go and economies of scale, not to geographic distribution of content. Option B is wrong because scalability refers to the ability to handle growing workloads by adding resources, not to serving users from multiple global locations. Option C is wrong because agility refers to the speed of deploying and iterating on resources, not to the physical distribution of PoPs for low-latency delivery.

326
MCQmedium

A company's security architect wants to implement 'privacy by design' principles when building a new customer data platform on Google Cloud. What does privacy by design mean in this context?

A.Privacy by design means the platform must refuse to collect any personal data from customers.
B.Privacy by design means privacy protections (encryption, data minimization, access controls, retention policies) are architected into the system from the start, not added after deployment.
C.Privacy by design is a legal requirement that mandates using only on-premises systems for customer data.
D.Privacy by design means storing all data in an encrypted format and using a VPN for all access.
AnswerB

Privacy by design makes privacy a foundational design principle: choosing which data to collect, how to protect it, who can access it, and when to delete it are designed before the first line of code — not discovered at audit time.

Why this answer

Privacy by design is a foundational principle that requires embedding privacy controls—such as encryption, data minimization, access controls, and retention policies—into the architecture of a system from the initial design phase, rather than retrofitting them after deployment. In the context of Google Cloud, this means using services like Cloud KMS for encryption, IAM for fine-grained access control, and data lifecycle policies to minimize data collection and enforce retention limits from the start. Option B correctly captures this proactive, integrated approach.

Exam trap

The trap here is that candidates often confuse privacy by design with a single technical control (like encryption or VPNs) or assume it prohibits data collection entirely, when in fact it is a holistic architectural approach that integrates multiple privacy controls from the outset.

How to eliminate wrong answers

Option A is wrong because privacy by design does not mandate refusing to collect any personal data; rather, it emphasizes collecting only the minimum necessary data (data minimization) and implementing protections around it. Option C is wrong because privacy by design is not a legal requirement that mandates on-premises systems; Google Cloud supports privacy by design through cloud-native services like Confidential VMs and Data Loss Prevention (DLP) that comply with regulations like GDPR. Option D is wrong because privacy by design is broader than just encryption and VPNs; it encompasses data minimization, purpose limitation, retention policies, and access controls, not just technical safeguards.

327
MCQmedium

A company has a folder for each department. The Finance team needs to prevent all projects under its folder from creating external IP addresses. What is the most efficient way to enforce this restriction?

A.Configure a service perimeter in VPC Service Controls
B.Use a deny IAM policy at the folder level
C.Apply an organization policy constraint at the folder level
D.Set a VPC firewall rule in each project
AnswerC

The correct answer is to apply an organization policy constraint such as constraints/compute.vmExternalIpAccess at the folder level. This list constraint denies or permits the assignment of external IP addresses to VM instances, and because organization policies are hierarchical, the folder-level setting is inherited by every project and resource within that folder, enforcing the guardrail centrally.

Why this answer

Organization policies can be defined at any level of the resource hierarchy and are inherited by child resources. Setting a constraint at the folder level applies to all projects under that folder.

328
MCQmedium

A data engineer needs to process a continuous stream of events from a global user base, perform real-time transformations, and write the results to both Cloud Storage and BigQuery. The solution must handle sudden traffic spikes and be fully managed (no server management). Which combination of services should the engineer use?

A.Pub/Sub, Cloud Functions, Cloud Storage
B.Pub/Sub, Dataflow, Cloud Functions
C.Cloud Scheduler, Cloud Functions, BigQuery
D.Pub/Sub, Dataflow, Cloud Storage, BigQuery
AnswerD

This pipeline uses Pub/Sub for asynchronous ingestion, then Dataflow (the fully managed Apache Beam runner) to read the unbounded stream, apply transforms, and write to two complementary sinks: Cloud Storage for durable raw data or archives, and BigQuery for interactive analytics. Dataflow handles the challenges of streaming—windowing, triggering, exactly-once processing, and auto-scaling—so all services are purpose-built for their roles and form a complete, production-ready architecture.

Why this answer

Pub/Sub ingests the continuous global event stream, Dataflow performs the real-time transformations in a fully managed, autoscaling way, and the pipeline can write results to both Cloud Storage and BigQuery as sinks. This combination meets every requirement: global ingestion, real-time processing, dual output, spike handling via autoscaling, and no server management. Dataflow is the only option that natively supports streaming transformations with exactly-once semantics and writes to multiple sinks.

Exam trap

The trap is picking Cloud Functions as the transformation engine because it is serverless and easy, but the exam expects you to know Dataflow is the managed service for continuous stream processing with multiple sinks.

How to eliminate wrong answers

Option A is wrong because Cloud Functions is event-driven and not designed for continuous stream transformations with windowing and autoscaling; it also lacks a native BigQuery sink in this pipeline. Option B is wrong because Cloud Functions cannot serve as the transformation engine for a continuous stream and does not write to both Cloud Storage and BigQuery as required. Option C is wrong because Cloud Scheduler is for cron-based batch triggers, not continuous streaming, and Cloud Functions cannot handle the real-time transformation workload.

Option D is correct because it includes Pub/Sub for ingestion, Dataflow for transformation, and both Cloud Storage and BigQuery as sinks.

329
MCQhard

A large bank is undergoing a cloud transformation. The CTO argues that the transformation will require a 'bimodal IT' approach — running two modes of IT simultaneously. What does bimodal IT mean in this context, and what is its primary criticism?

A.Bimodal IT means using two different cloud providers for redundancy; the criticism is that it creates vendor lock-in with two providers instead of one
B.Bimodal IT runs stable core systems (Mode 1) alongside an agile innovation team (Mode 2); critics argue it creates organizational division, delays core modernization, and produces digital capabilities that eventually can't integrate with unreformed core systems
C.Bimodal IT means running both cloud and on-premises systems; the criticism is that hybrid environments are too complex to manage
D.Bimodal IT has no critics — it is universally accepted as the best approach to banking digital transformation
AnswerB

This accurately describes bimodal IT and its main criticism. The approach can be useful as a transitional strategy but is criticized for institutionalizing the divide between 'old IT' and 'digital' rather than transforming the core. Capabilities built in Mode 2 eventually need to connect to Mode 1 systems — the division doesn't disappear.

Why this answer

Bimodal IT, as described by Gartner, separates IT into Mode 1 (traditional, stable, and risk-averse systems) and Mode 2 (agile, exploratory, and fast-moving innovation teams). In a cloud transformation context, Mode 1 typically runs legacy core banking systems on-premises or in a private cloud, while Mode 2 rapidly develops cloud-native applications. The primary criticism is that this creates a permanent organizational and technical divide, where Mode 2 builds digital capabilities that cannot integrate with the unreformed, monolithic Mode 1 systems, leading to technical debt and delaying the necessary modernization of the core.

Exam trap

The GCDL exam often tests the distinction between organizational operating models (bimodal IT) and deployment architectures (hybrid cloud, multi-cloud), so the trap here is that candidates pick Option C because they mistakenly equate 'two modes' with 'two environments' (cloud and on-premises).

How to eliminate wrong answers

Option A is wrong because bimodal IT is not about using two cloud providers for redundancy; that describes a multi-cloud strategy, not the organizational separation of stable and agile IT modes. Option C is wrong because bimodal IT is not simply running cloud and on-premises systems (hybrid cloud); it is a specific organizational and process model that separates IT into two distinct modes of operation, not just a deployment architecture. Option D is wrong because bimodal IT has been heavily criticized by industry experts (e.g., Martin Fowler, Gartner's own later analysis) for creating silos, increasing integration costs, and failing to address core system modernization; it is not universally accepted.

330
MCQeasy

A retail company's IT director says: 'We need to digitize our business.' A digital transformation consultant responds that digitization and digital transformation are different things. Which statement best captures the distinction?

A.Digitization and digital transformation are synonyms; both refer to moving business operations to digital systems
B.Digitization converts analog information to digital form, while digital transformation reimagines business models, customer experiences, and operations using digital technology as the core enabler
C.Digital transformation is a subset of digitization, focused specifically on transforming customer-facing processes
D.Digitization requires cloud technology, while digital transformation can be achieved with on-premises systems alone
AnswerB

This captures the essential distinction. Digitization (scanning paper, converting spreadsheets to databases) is a prerequisite but not sufficient. Transformation means the business fundamentally changes how it creates and delivers value — not just running old processes digitally.

Why this answer

It accurately distinguishes digitization (converting analog data to digital format, e.g., scanning paper invoices into PDFs) from digital transformation (fundamentally rethinking business models, customer experiences, and operations with digital technology as the core enabler, e.g., using cloud-based analytics to personalize customer journeys). The consultant's point is that digitization is a tactical step, while digital transformation is a strategic overhaul that leverages cloud, AI, and IoT to create new value chains.

Exam trap

Google Cloud often tests the confusion between digitization and digital transformation by presenting options that conflate the two as synonyms or reverse their hierarchy, so candidates must remember that digitization is a technical conversion step, while digital transformation is a strategic business reimagination enabled by cloud and modern IT architectures.

How to eliminate wrong answers

Option A is wrong because it incorrectly treats digitization and digital transformation as synonyms, ignoring that digitization is merely converting analog to digital (e.g., digitizing a paper ledger into a spreadsheet), while digital transformation involves reimagining processes and business models using cloud, APIs, and data-driven automation. Option C is wrong because it claims digital transformation is a subset of digitization focused only on customer-facing processes; in reality, digital transformation is broader, encompassing supply chain, operations, and culture, and digitization is a foundational enabler, not the superset. Option D is wrong because it falsely ties digitization to cloud technology (digitization can be done with on-premises scanners and local databases) and suggests digital transformation can be achieved with on-premises systems alone, whereas true digital transformation often requires cloud scalability, microservices, and continuous integration/delivery pipelines to enable rapid innovation.

331
MCQhard

A global enterprise must store customer data in specific geographic regions to comply with data residency laws. They also need to be able to analyze data across all regions without centralizing it. Which approach best balances compliance and analytics?

A.Deploy identical Compute Engine instances with databases in each region.
B.Use regional Cloud Storage buckets per region and BigQuery with federated queries.
C.Store all data in a single region and use Cloud CDN for fast access.
D.Use a single Cloud Storage multi-region bucket with granular access controls.
AnswerB

Regional buckets keep each region's data physically resident, satisfying data residency law, while BigQuery federated queries read across those external buckets without copying or centralising data. This delivers cross-region analytics while preserving the compliance constraint the stem requires.

Why this answer

Option B is correct because regional Cloud Storage buckets keep each region's customer data physically within its required geography, satisfying data residency laws, while BigQuery federated queries (external tables over Cloud Storage) let analysts query that data in place across regions without copying or centralizing it. This combination preserves compliance and enables cross-regional analytics. Option A does not provide a cross-region analytics mechanism and would require custom data movement or query federation, adding complexity and risk.

Option C centralizes all data in one region, directly violating residency requirements. Option D uses a multi-region bucket, which may replicate data outside the mandated region and thus fails the residency constraint.

332
MCQhard

A company migrates its on-premises database to Cloud SQL. The security team is concerned about who is responsible for patching the underlying operating system and database engine. Under the shared responsibility model, which of the following is true?

A.The customer must patch the operating system of the Cloud SQL instances.
B.Google Cloud is responsible for patching the operating system and database engine.
C.The customer is responsible for physical security of the data centers.
D.The customer is responsible for applying database engine patches.
AnswerB

Cloud SQL is a fully managed database service where Google Cloud owns the entire infrastructure stack, including the hypervisor, virtual machine, operating system, and database engine binaries. Google's operations team automatically applies security patches for both the OS and the database engine (e.g., MySQL, PostgreSQL, or SQL Server), typically on a regular cadence with customer-controlled maintenance windows. This aligns with the shared responsibility model, where the provider handles all infrastructure layer maintenance.

Why this answer

Under the Google Cloud shared responsibility model, Google Cloud manages the underlying infrastructure, including patching the operating system and database engine for managed services like Cloud SQL. This is because Cloud SQL is a fully managed service where Google Cloud handles OS and database engine updates, ensuring security and compliance without customer intervention. The customer remains responsible for data, access management, and application-level configurations.

Exam trap

The trap here is that candidates confuse Cloud SQL (a managed service) with Compute Engine (IaaS), where the customer is responsible for OS and database patching, leading them to incorrectly select options A or D.

How to eliminate wrong answers

Option A is wrong because the customer is not responsible for patching the operating system of Cloud SQL instances; Google Cloud automatically manages OS patches as part of the managed service. Option C is wrong because physical security of data centers is solely Google Cloud's responsibility under the shared responsibility model, not the customer's. Option D is wrong because database engine patches are applied by Google Cloud for Cloud SQL, not by the customer, as the service handles version upgrades and security patches automatically.

333
MCQmedium

A healthcare company must comply with HIPAA and store all protected health information (PHI) only in the United States. They use Google Cloud and want to prevent any accidental data storage outside the US. Which two services should they implement?

A.VPC Service Controls and Organization Policies
B.Data Loss Prevention API
C.Identity-Aware Proxy
D.Cloud Armor
AnswerA

VPC Service Controls create a data security perimeter that restricts access to Google Cloud resources and prevents data exfiltration through context-aware enforcement. Organization Policies, when configured with location constraints, explicitly limit resource creation to approved geographic regions, ensuring HIPAA data residency requirements are met. Together, they enforce both the boundary and the storage location for regulated data.

Why this answer

VPC Service Controls create a security perimeter around Google Cloud resources, preventing data from being copied or moved outside allowed regions. Organization Policies allow you to set a constraint (e.g., `gcp.resourceLocations`) that restricts where resources like Cloud Storage buckets or BigQuery datasets can be created, ensuring PHI remains in the US. Together, they enforce both data exfiltration prevention and location-based resource creation restrictions.

Exam trap

Google Cloud often tests the distinction between data *protection* (DLP, IAP, Cloud Armor) and data *residency enforcement* (VPC Service Controls, Organization Policies), leading candidates to confuse content inspection or access control with geographic restriction.

How to eliminate wrong answers

Option B (Data Loss Prevention API) is wrong because it is a content inspection and redaction tool, not a data residency enforcement mechanism; it scans for sensitive data patterns but does not prevent storage outside a geographic boundary. Option C (Identity-Aware Proxy) is wrong because it controls user access to applications based on identity and context, not data location or storage restrictions. Option D (Cloud Armor) is wrong because it is a web application firewall (WAF) that protects against DDoS and OWASP Top 10 threats, with no capability to enforce data residency or prevent storage in non-US regions.

334
MCQeasy

A developer wants to deploy a containerised web application that can scale to zero when not in use, and only pay for actual request processing time. Which Google Cloud compute service should they choose?

A.App Engine Standard Environment
B.Google Kubernetes Engine (GKE)
C.Cloud Run
D.Compute Engine with Preemptible VMs
AnswerC

Cloud Run is a fully managed serverless container platform that executes stateless HTTP/S containers on demand, automatically scaling from zero to handle traffic spikes and back down to zero when requests stop. It accepts any container image that listens on HTTP/HTTPS, pulled from Artifact Registry or Container Registry, and you are billed only for the resources used during request processing (increments of 100ms) with no charge for idle instances. This makes it a direct fit for a containerized web application that needs elastic scaling and pay-per-use pricing without the need to manage clusters or nodes.

Why this answer

Cloud Run is a serverless compute platform that runs containers, scales to zero when idle, and charges per request and CPU/memory usage only during request handling. App Engine Standard can scale to zero but does not support arbitrary containers. GKE and Compute Engine are provisioned infrastructure that does not scale to zero.

335
MCQmedium

A company wants to analyze its Google Cloud spending trends by project, service, and labels. They need to run custom SQL queries on billing data and retain it for 2 years. Which approach should they use?

A.Use the Google Cloud Pricing Calculator.
B.Set up billing export to BigQuery.
C.Enable the Cost Management dashboard.
D.Configure budgets and alerts.
AnswerB

Billing export to BigQuery continuously writes detailed billing data, including usage, cost, labels, and metadata, into BigQuery tables for flexible analysis and long-term retention. This enables complex SQL queries, custom dashboards, and time-series trend analysis on actual historical spend, making it the correct and recommended method for understanding Google Cloud spending trends.

Why this answer

Billing export to BigQuery allows you to export detailed billing data into BigQuery datasets, where you can run custom SQL queries and retain data for your desired period.

336
Multi-Selectmedium

A developer wants to build a CI/CD pipeline that automatically builds a Docker container from a GitHub repository, stores the image securely, and deploys it to Cloud Run. Which THREE services should they integrate? (Select 3)

Select 3 answers
A.Cloud Build
B.Artifact Registry
C.Cloud Storage
D.Cloud Run
E.Cloud Source Repositories
AnswersA, B, D

Cloud Build ingests source code from GitHub and executes a series of build steps, including compiling code and building a Docker image using a Dockerfile. Because it is a managed service, it scales automatically and can trigger builds on commits or PRs, making it the core automation engine of this CI/CD pipeline.

Why this answer

Cloud Build (A) is the correct service to define and execute the CI/CD pipeline that pulls the source from GitHub and runs the docker build steps to produce the container image. Artifact Registry (B) is the secure, recommended place to store and manage the resulting Docker images (replacing the deprecated Container Registry), and Cloud Run can natively pull images from it. Cloud Run (D) is the deployment target where the built image is served as a fully managed serverless container, completing the pipeline.

Cloud Storage (C) is not appropriate because it stores objects/blobs rather than container images with registry semantics, and Cloud Source Repositories (E) is a Git hosting service that would duplicate GitHub rather than build or deploy anything.

Exam trap

The trap here is confusing Cloud Storage with Artifact Registry for container image storage — candidates who have not worked with GCP container workflows often assume any storage service can hold a Docker image.

337
MCQmedium

A company runs a batch job that processes data every night. The job can tolerate interruptions and currently runs on a dedicated on-premises server that is underutilized. They want to migrate to Google Cloud and minimise compute cost. Which compute option is most cost-effective?

A.Sole-tenant nodes
B.Standard (on-demand) VMs
C.VMs with GPUs
D.Preemptible VMs
AnswerD

Preemptible VMs are Compute Engine instances that use leftover capacity at a discount of as much as 60-80% compared to on-demand pricing, but they can be terminated by Google at any time with a 30-second warning. Because a nightly batch job is inherently fault-tolerant and can be designed to resume from a checkpoint, it is an ideal fit for these interruptible instances. The key is to architect the job with idempotent processing and persistent storage so that preemption does not corrupt data or lose progress, making it the correct, cost-effective choice.

Why this answer

Preemptible VMs offer significantly lower cost (up to 80% discount) but can be terminated at any time, making them ideal for fault-tolerant batch workloads. Standard VMs are more expensive, sole-tenant nodes are for isolation, and GPUs add cost.

338
Multi-Selecthard

Which TWO of the following are best practices for securing a Google Cloud environment? (Choose two.)

Select 2 answers
A.Use the same SSH key for all Compute Engine instances.
B.Grant minimal permissions to users and services using IAM roles.
C.Export service account keys and use them in on-premises applications for authentication.
D.Rotate service account keys every month.
E.Enable VPC Flow Logs for all subnets.
AnswersB, E

IAM roles provide granular, auditable control over who can perform specific actions on specific resources, and granting only the minimum permissions needed (least privilege) contains the blast radius of a compromised account or service. Predefined roles like roles/viewer or custom roles with targeted permissions allow you to align access with actual job duties, reducing the chance of accidental or malicious damage. This is a cornerstone of Google Cloud security and is far more effective than managing broad permissions or long-lived credentials.

Why this answer

Option B is correct because the principle of least privilege is a core Google Cloud security best practice: IAM roles should be scoped so users and service accounts receive only the permissions required for their tasks, reducing the blast radius of compromised credentials. Option E is correct because enabling VPC Flow Logs on all subnets provides network telemetry for auditing, anomaly detection, and incident response, which is a recommended detective control in Google Cloud. Option A is not a best practice because reusing one SSH key across all Compute Engine instances means a single key compromise grants access to every VM; keys should be unique per user/instance and managed via OS Login or metadata.

Option C is not recommended because exporting long-lived service account keys for on-premises use creates high-risk credentials that can be leaked; workload identity federation or short-lived tokens should be used instead. Option D is not among the best practices here because Google recommends avoiding service account keys altogether and, when unavoidable, rotating them regularly rather than treating a fixed monthly rotation as the primary control.

Exam trap

Google Cloud often tests the misconception that service account key rotation is a best practice, but the trap here is that the real best practice is to avoid using service account keys altogether in favor of workload identity federation or short-lived credentials.

339
MCQeasy

A company is migrating its on-premises data center to Google Cloud. They want to avoid large upfront hardware costs and only pay for the resources they consume. Which cloud benefit does this represent?

A.Security
B.Pay-as-you-go pricing
C.High availability
D.Scalability
AnswerB

Pay-as-you-go pricing directly addresses the financial model of cloud migration by replacing large capital expenditures (CapEx) on hardware with operational expenditures (OpEx) based on actual consumption. In on-premises data centers, organizations must purchase and maintain physical servers regardless of utilization, leading to sunk costs. With GCP, you are billed per second for compute resources like Compute Engine, enabling cost alignment with demand and eliminating the need for upfront procurement, which is the core benefit being asked about.

Why this answer

The scenario describes a shift from capital expenditure (CapEx) to operational expenditure (OpEx), which is the core of pay-as-you-go pricing. This model allows the company to avoid large upfront hardware costs and only pay for the compute, storage, and network resources actually consumed, aligning with Google Cloud's consumption-based billing model.

Exam trap

The trap here is that candidates often confuse 'scalability' with 'pay-as-you-go' because both involve resource adjustment, but scalability is about dynamic capacity changes, not the financial model of avoiding upfront costs.

How to eliminate wrong answers

Option A is wrong because security is a shared responsibility model in Google Cloud, not a financial model that avoids upfront costs; it does not address the consumption-based pricing described. Option C is wrong because high availability refers to redundant infrastructure and uptime guarantees (e.g., via multi-zonal deployments), not to the elimination of upfront hardware expenses. Option D is wrong because scalability is the ability to automatically adjust resources based on demand (e.g., using managed instance groups), which is a separate benefit from the pay-as-you-go pricing model that avoids initial capital outlay.

340
MCQhard

A reliability engineering team wants to proactively identify weaknesses in their distributed system by deliberately injecting failures — killing random instances, introducing network latency, and cutting off database connections — to observe how the system responds. What is this practice called?

A.Destructive testing — deliberately breaking the system to determine the breaking point.
B.Chaos engineering — deliberately injecting controlled failures to discover system weaknesses and build resilience confidence.
C.Penetration testing — simulating attacks to find security vulnerabilities.
D.Load testing — verifying the system handles expected traffic volumes.
AnswerB

Chaos engineering is the disciplined practice of introducing controlled failures (e.g., terminating a service instance, injecting latency, or simulating a network partition) into a distributed system to observe behavior against a defined steady-state hypothesis. It is a scientific method for verifying that the system can self-heal and maintain user impact within acceptable bounds, and it typically is performed with a limited blast radius and continuous experimentation. The objective is not merely to find a breaking point, but to build resilience confidence by uncovering unknown architectural weaknesses before they manifest as real outages.

Why this answer

Chaos engineering is the practice of deliberately injecting controlled failures—such as killing instances, introducing latency, or cutting database connections—into a distributed system to proactively identify weaknesses and build resilience confidence. This approach aligns with Google Cloud's reliability principles, where tools like Chaos Monkey (part of the Simian Army) or Google's internal DiRT (Disaster Recovery Testing) are used to test system behavior under failure conditions.

Exam trap

Google Cloud often tests the distinction between 'destructive testing' and 'chaos engineering' by making candidates think any deliberate failure is destructive, but the key difference is that chaos engineering is controlled, hypothesis-driven, and aims to build resilience, not just find the breaking point.

How to eliminate wrong answers

Option A is wrong because destructive testing focuses on finding the breaking point of a system by pushing it to failure, often in a non-controlled manner, and does not emphasize controlled, proactive failure injection to build resilience confidence. Option C is wrong because penetration testing specifically targets security vulnerabilities (e.g., OWASP Top 10, SQL injection) and does not cover operational failures like network latency or instance termination. Option D is wrong because load testing verifies system performance under expected or peak traffic volumes (e.g., using tools like Locust or k6), not the system's response to injected failures like database disconnections or random instance kills.

341
MCQhard

A company uses a legacy on-premises database and wants to migrate to Google Cloud with minimal changes to application code. They need a fully managed relational database that supports MySQL compatibility and automatic failover. Which service should they choose?

A.Firestore
B.Cloud Spanner
C.Cloud Bigtable
D.Cloud SQL for MySQL
AnswerD

Cloud SQL for MySQL is a fully managed service that is tightly compatible with the MySQL wire protocol, SQL dialect, and client tools, so existing schemas, queries, stored procedures, and application code can be reused with almost no modification. It provides automatic failover to a synchronous standby replica, managed backups, and read replicas, meeting high-availability needs without changing the database engine. This makes it the natural choice for migrating a legacy on-premises MySQL database with minimal code changes and low operational overhead.

Why this answer

Cloud SQL for MySQL is a fully managed relational database service that is wire-compatible with MySQL, so existing applications using MySQL drivers and SQL syntax require minimal code changes. It supports automatic failover via a regional HA configuration with a standby instance. This matches the requirement for MySQL compatibility, managed operations, and automatic failover.

Exam trap

The trap is over-engineering — candidates pick Cloud Spanner because it sounds 'more managed' or 'more scalable,' ignoring that the requirement is minimal code changes and MySQL compatibility, which only Cloud SQL provides.

How to eliminate wrong answers

Option A is wrong because Firestore is a NoSQL document database, not relational, and does not support MySQL compatibility or SQL queries. Option B is wrong because Cloud Spanner, while relational and globally distributed, uses GoogleSQL and is not MySQL-compatible without significant application changes, and it is overkill for a legacy migration. Option C is wrong because Cloud Bigtable is a wide-column NoSQL store for analytical workloads, not a relational database, and offers no MySQL compatibility.

342
MCQmedium

A retail company runs a customer-facing web application on Compute Engine instances. The security team wants to protect the application from common web exploits such as SQL injection and cross-site scripting, and also wants to mitigate distributed denial-of-service (DDoS) attacks without modifying the application code. Which Google Cloud service should the company implement to meet these requirements?

A.VPC Service Controls
B.Cloud Armor
C.Cloud Identity-Aware Proxy
D.Cloud NAT
AnswerB

Cloud Armor is Google Cloud's edge security service that works with external HTTP(S) load balancers to filter malicious traffic. It provides preconfigured WAF rules based on OWASP ModSecurity signatures that block SQL injection and cross-site scripting, plus adaptive protection and rate-based rules to absorb and mitigate DDoS attacks. Because it operates at the load balancer layer, no application code changes are required, directly satisfying both the WAF and DDoS requirements.

Why this answer

Cloud Armor is the correct choice because it is the Google Cloud edge security service that integrates with external HTTP(S) load balancers to deliver WAF filtering and DDoS protection. Its preconfigured rules defend against OWASP Top 10 threats like SQL injection and cross-site scripting, and its rate-based and adaptive protection rules absorb volumetric attacks. Because enforcement happens at the load balancer, the application itself remains unchanged, meeting every requirement in the scenario.

Exam trap

The trap here is assuming that identity-based access controls such as Identity-Aware Proxy or API perimeters such as VPC Service Controls can substitute for payload inspection and volumetric attack mitigation at the edge.

343
MCQhard

A company migrated a microservices application to Google Kubernetes Engine (GKE). They set up an internal HTTP(S) load balancer to route traffic to the services. However, some pods are not receiving traffic. What is the most likely cause?

A.The service type is NodePort instead of LoadBalancer.
B.The pods have failing readiness probes that are preventing them from being added to the load balancer's backend endpoints.
C.The cluster does not have enough nodes to schedule the pods.
D.The firewall rules are not allowing traffic from the load balancer to the nodes.
AnswerB

Readiness probes are executed by the kubelet on a recurring basis; if the probe fails, the pod is marked NotReady and its IP is removed from the Service’s Endpoints object. With no ready pods, the EndpointSlices are empty, so the internal load balancer's backend set has zero healthy instances and will not forward any requests. This perfectly matches the symptom of a pod that is Running (it exists) but receives no traffic, making this the correct diagnosis.

Why this answer

GKE internal HTTP(S) load balancers use network endpoint groups (NEGs) that only include pod IPs whose readiness probes are passing. If a pod's readiness probe fails, the pod is removed from the backend endpoints and receives no traffic, even though it is Running. This is the most common cause of 'some pods not receiving traffic' in GKE.

Exam trap

GCDL often tests the distinction between liveness and readiness probes — candidates incorrectly assume a Running pod always receives traffic, forgetting that readiness gates endpoint membership.

How to eliminate wrong answers

Option A is wrong because an internal HTTP(S) load balancer in GKE is provisioned via a Service of type LoadBalancer with the internal annotation or via Ingress/Gateway — the Service type itself is not NodePort, and NodePort would still route traffic via kube-proxy rather than explaining selective pod exclusion. Option C is wrong because insufficient nodes would cause pods to remain Pending, not Running-but-unreachable; the symptom described is selective traffic loss, not scheduling failure. Option D is wrong because firewall rules blocking LB-to-node traffic would affect all pods uniformly, not just 'some pods'.

344
MCQeasy

A startup wants to deploy a containerized web application that can scale to zero during periods of no traffic, and they want to minimize operational overhead. Which Google Cloud compute service is the BEST fit?

A.App Engine Flexible Environment
B.Google Kubernetes Engine (GKE)
C.Cloud Run
D.Compute Engine with managed instance groups
AnswerC

Cloud Run runs stateless containers on a fully managed, serverless platform that automatically scales from zero to thousands of instances based on incoming traffic, and you pay only for the processing time during active requests. It eliminates infrastructure management entirely—no cluster, VMs, or capacity planning—and supports fine-grained revisioning and traffic splitting. This makes it the ideal choice for a containerized web app that must scale to zero during idle periods.

Why this answer

Cloud Run is a serverless container platform that automatically scales your containers up and down, including scaling to zero when there is no traffic. It abstracts away infrastructure management, making it ideal for this use case.

345
MCQeasy

What is a key benefit of using a Platform as a Service (PaaS) offering like App Engine compared to Infrastructure as a Service (IaaS) like Compute Engine?

A.Lower cost for always-on workloads
B.No need to manage the underlying OS or runtime environment
C.Full control over the operating system and hardware
D.Ability to install custom software on the server
AnswerB

The defining characteristic of PaaS is that the vendor owns and operates the entire stack below your application code, including the operating system, runtime interpreter, and middleware libraries. You are responsible only for uploading and deploying your application, while the provider handles patching, security, and infrastructure scaling. This abstraction eliminates the operational burden of OS administration, which is why this is considered a key benefit.

Why this answer

PaaS platforms like App Engine abstract away the operating system, runtime patches, scaling, and infrastructure management, so developers only ship application code. IaaS like Compute Engine gives you a VM where you remain responsible for OS patching, runtime installation, and capacity planning. The core PaaS benefit is therefore the elimination of OS/runtime management overhead.

Exam trap

The trap here is conflating 'managed service' with 'cheaper' or 'more control' — GCDL frequently tests the direction of the abstraction trade-off: PaaS removes management burden but also removes control, while IaaS is the opposite.

How to eliminate wrong answers

Option A is wrong because PaaS is not inherently cheaper for always-on workloads — App Engine Standard's instance-hour pricing can exceed a committed-use Compute Engine VM for steady-state traffic, and cost depends heavily on scaling pattern. Option C is wrong because full control over the OS and hardware is a hallmark of IaaS (Compute Engine, bare metal), not PaaS; App Engine sandboxes the runtime and forbids OS-level access. Option D is wrong because installing arbitrary custom software on the server is only possible with IaaS VMs or containers you manage — App Engine restricts you to supported runtimes and libraries, which is precisely the trade-off PaaS makes.

346
MCQmedium

A data analytics team processes large datasets using Apache Spark. They want a fully managed service that allows them to run Spark jobs without provisioning clusters. Which Google Cloud service should they use?

A.BigQuery
B.Dataflow
C.Dataproc
D.Cloud Dataprep
AnswerC

Dataproc is Google Cloud's managed service for running Apache Spark and Hadoop clusters, providing native Spark support with custom cluster configurations, autoscaling, and integration with cloud storage. Dataproc Serverless further allows teams to run Spark jobs without provisioning or managing any cluster infrastructure at all. This makes Dataproc the straightforward choice for a team already using Spark.

Why this answer

Dataproc is Google Cloud's fully managed Apache Spark and Hadoop service, letting teams submit Spark jobs to ephemeral or long-running clusters without provisioning infrastructure themselves. It supports Spark, Spark SQL, PySpark, Hive, and Pig, and integrates with Cloud Storage and BigQuery as data sources. That matches the requirement exactly.

Exam trap

GCDL often tests the difference between 'managed Spark' (Dataproc) and 'managed Beam' (Dataflow) — candidates who see 'large datasets' and 'processing' often jump to Dataflow or BigQuery without noticing the explicit Apache Spark requirement.

How to eliminate wrong answers

Option A is wrong because BigQuery is a serverless data warehouse with its own SQL engine — it does not run arbitrary Apache Spark jobs, even though it can query external data and supports Spark via the BigQuery connector. Option B is wrong because Dataflow is a managed Apache Beam runner for stream and batch pipelines; it executes Beam pipelines, not native Spark jobs, and requires rewriting Spark code into Beam. Option D is wrong because Cloud Dataprep is a UI-driven data preparation tool (powered by Trifacta) for visual cleansing and transformation, not a Spark job execution platform.

347
MCQeasy

Which Google Cloud resource serves as the root node in the resource hierarchy and is linked to a Google Workspace or Cloud Identity account?

A.Organization node
B.Billing account
C.Folder
D.Project
AnswerA

The organization node is the top-level (root) node in the Google Cloud resource hierarchy. It represents the entire enterprise and is typically created automatically when a Google Cloud account is established or a domain is claimed. All other resources—folders, projects, and billing associations—are ultimately attached beneath it, allowing organization-wide IAM policies, organization policies, and access governance to be applied uniformly.

Why this answer

The Organization node is the root node in the Google Cloud resource hierarchy and is automatically created when a Google Workspace or Cloud Identity account is linked to Google Cloud. It sits above folders and projects, and serves as the root for applying organization-level IAM policies and constraints.

Exam trap

GCDL often tests the confusion between the organization node and billing account — candidates may think the billing account is the root, but it is not part of the resource hierarchy.

How to eliminate wrong answers

Option B is wrong because a billing account is used for payment and is not part of the resource hierarchy; it can be linked to multiple projects but does not serve as a root node. Option C is wrong because folders are optional containers within an organization that group projects and other folders, but they are not the root node. Option D is wrong because a project is a resource container that holds resources like VMs and buckets, but it is a child of folders or the organization, not the root.

348
MCQeasy

A company's security team wants to ensure that only approved corporate devices can access Google Cloud resources, regardless of whether the user has valid credentials. Which Google Cloud security capability enforces device-level access requirements?

A.Cloud Armor, which filters incoming requests based on IP allowlists and denylists
B.Access Context Manager, which enforces device-level access requirements as part of context-aware access control policies
C.Identity-Aware Proxy (IAP), which provides application-level authentication but without device checks
D.VPC Service Controls, which restrict access to Google APIs based on network perimeter membership
AnswerB

Access Context Manager is precisely the service for this. It allows security teams to define access levels (policies) that include device attribute requirements — managed/enrolled devices, disk encryption, screen lock. These conditions must be met in addition to valid credentials for access to be granted.

Why this answer

Access Context Manager is the correct choice because it allows security teams to define context-aware access policies that include device-level attributes such as device OS type, device ID, and whether the device is managed (e.g., via endpoint verification or third-party EMM). This enforces device-level access requirements even if the user has valid credentials, directly addressing the scenario.

Exam trap

The trap here is that candidates confuse IAP's role in user authentication with device-level enforcement, not realizing that IAP delegates device context checks to Access Context Manager via access levels.

How to eliminate wrong answers

Option A is wrong because Cloud Armor filters traffic based on IP addresses (allowlists/denylists) and other network-layer attributes, not device-level identity or management status. Option C is wrong because Identity-Aware Proxy (IAP) provides authentication and authorization at the application layer but does not natively enforce device-level checks; it relies on Access Context Manager for such context. Option D is wrong because VPC Service Controls restrict access to Google APIs based on network perimeter (e.g., VPC, IP ranges) and do not evaluate device-level attributes.

349
Multi-Selectmedium

A company wants to implement a serverless event-driven architecture where object uploads to Cloud Storage trigger a function that processes the file and stores results in Firestore. The function needs to be written in Python. Which three Google Cloud services are required?

Select 3 answers
A.Pub/Sub
B.Cloud Functions
C.Cloud Build
D.Cloud Storage
E.Firestore
AnswersB, D, E

Cloud Functions is the serverless compute layer that runs the Python code when a file is uploaded; the Cloud Storage event (object.finalize) triggers the function with event metadata like bucket and object name. It automatically scales to zero when idle, handles the processing synchronously or asynchronously, and integrates seamlessly with GCP services. Since the requirement specifies running Python code in an event-driven way, Cloud Functions is the correct service for executing the logic.

Why this answer

Cloud Functions (B) is correct because it is the serverless compute service that runs the Python function in response to the event, and it natively supports Python runtimes. Cloud Storage (D) is correct because it is the source of the event: object uploads (finalize events) to a bucket trigger the function, typically via an Eventarc or Cloud Storage trigger. Firestore (E) is correct because it is the destination datastore where the function writes the processing results, as required by the scenario.

Pub/Sub (A) is not required because Cloud Storage events can trigger Cloud Functions directly without an explicit Pub/Sub topic, and Cloud Build (C) is a CI/CD service for building and deploying code, not a runtime component of this event-driven flow.

Exam trap

GCDL often tests whether candidates add Pub/Sub reflexively to any event-driven design; the trap is that Cloud Storage can trigger Cloud Functions directly, so Pub/Sub is not required unless multiple subscribers or decoupling is explicitly needed.

350
MCQmedium

A company wants to ensure that even if an attacker compromises an employee's password and passes MFA, the attacker cannot access sensitive Google Cloud resources from an unmanaged personal laptop. Which Google security feature enforces device trust as part of access decisions?

A.Cloud Armor — it inspects device fingerprints on incoming requests.
B.Access Context Manager with device policy conditions requiring managed, compliant devices.
C.Cloud Firewall rules that allow only corporate office IP ranges.
D.Two-step verification — the second factor proves the device is trusted.
AnswerB

Access Context Manager lets you define access levels as a set of device policy conditions—for example, requiring the device to be enrolled in an endpoint verification service, have full disk encryption, and run a minimum OS version. These access levels can be attached to IAP or VPC Service Controls, which then enforce the policy at the application or network perimeter. This is the only option that explicitly evaluates device trust and compliance before granting access.

Why this answer

Access Context Manager allows you to define device policy conditions that require devices to be managed (e.g., via endpoint verification) and compliant with corporate security policies. When an attacker attempts to access sensitive Google Cloud resources from an unmanaged personal laptop, the access level will not be satisfied, and access is denied even if the user's password and MFA are valid. This enforces device trust as a distinct attribute in the access decision, separate from user authentication.

Exam trap

The GCDL exam often tests the distinction between authentication (MFA) and device trust, so the trap here is that candidates confuse two-step verification (MFA) with device trust, thinking that a second factor inherently proves the device is trusted, when in reality MFA only proves the user's identity, not the device's security posture.

How to eliminate wrong answers

Option A is wrong because Cloud Armor is a web application firewall that inspects HTTP/S traffic and can use device fingerprints (e.g., via reCAPTCHA or WAF rules), but it does not enforce device trust as a condition for access to Google Cloud resources; it operates at the network edge, not as an identity-aware access control. Option C is wrong because Cloud Firewall rules that allow only corporate office IP ranges restrict access based on network origin, not device trust; an attacker could still use a managed laptop from a corporate IP if they compromise it, and an unmanaged personal laptop from a corporate IP would bypass the intent. Option D is wrong because two-step verification (MFA) verifies the user's identity via a second factor (e.g., TOTP, push notification), but it does not prove the device is trusted; an attacker who has compromised the password and MFA (e.g., via session hijacking or phishing) can still access resources from any device.

351
MCQhard

A multinational manufacturing company operates thousands of IoT sensors on factory equipment. These sensors generate over 50 TB of telemetry data daily. The company wants to implement predictive maintenance to reduce unplanned downtime. Their current on-premises infrastructure is maxed out, and they have a small IT team with limited data engineering expertise. They are evaluating cloud vs. on-premises options. The data is highly sensitive and must be encrypted at rest and in transit. Additionally, they need to run machine learning models near real-time and store historical data for trend analysis. The CTO is concerned about vendor lock-in, data sovereignty, and the ability to scale globally as they open new factories. Which course of action best addresses these requirements using Google Cloud?

A.Deploy Kubernetes on-premises to orchestrate microservices, and use Cloud Storage for backup only.
B.Build a new on-premises data center with dedicated GPU servers for ML training, and hire additional data engineers.
C.Use Cloud Pub/Sub to ingest sensor data, Cloud Dataflow for stream processing, BigQuery for historical storage and analysis, and Vertex AI for predictive models, all in the desired region.
D.Store all sensor data in Cloud Storage and run ad-hoc queries using BigQuery without any streaming pipeline.
AnswerC

This fully managed pipeline uses Cloud IoT Core to securely connect and ingest sensor telemetry, then Pub/Sub provides durable, low-latency message buffering with decoupled consumers. Dataflow performs stateful, exactly-once stream processing and windowing for anomaly detection, while AI Platform trains and serves predictive models; deploying all services in the desired region satisfies data residency and compliance. Because each component is serverless and auto-scaling, the solution handles massive throughput without infrastructure management, directly addressing the company's constraints.

Why this answer

The correct answer is C. Use Cloud Pub/Sub to ingest streaming sensor data, Cloud Dataflow for processing, BigQuery for historical storage and analysis, and Vertex AI for predictive models. This provides near-real-time ML, encryption at rest and in transit, regional data sovereignty, and global scalability.

Exam trap

Google Cloud often tests the misconception that batch processing (e.g., BigQuery on stored data) can substitute for a streaming pipeline in near-real-time scenarios, leading candidates to overlook the need for continuous ingestion and processing with services like Pub/Sub and Dataflow.

How to eliminate wrong answers

Option A is wrong because deploying Kubernetes on-premises does not address the maxed-out infrastructure, requires significant data engineering expertise for management, and fails to provide the near-real-time streaming and ML capabilities needed, while Cloud Storage for backup alone ignores the primary processing requirements. Option B is wrong because building a new on-premises data center contradicts the goal of avoiding infrastructure expansion, requires hiring additional data engineers (which the small IT team cannot support), and does not address vendor lock-in or global scaling concerns. Option D is wrong because storing all data in Cloud Storage and running ad-hoc BigQuery queries lacks a streaming pipeline, making near-real-time predictive maintenance impossible, and does not provide the continuous processing needed for ML model inference on live telemetry data.

352
MCQeasy

Which term describes the process by which organizations integrate digital technology into all areas of their business, fundamentally changing how they operate and deliver value to customers?

A.IT modernization
B.Digital transformation
C.Cloud migration
D.Agile development
AnswerB

Digital transformation is the comprehensive, strategic integration of digital technology into every area of the organization, fundamentally changing operations, culture, and how value is delivered to customers. It encompasses business model reinvention, data-driven decision making, customer experience redesign, and new revenue streams, with cloud as a primary enabler providing scalable infrastructure and advanced services. Unlike narrower initiatives, it requires deep organizational change management, leadership alignment, and continuous innovation.

Why this answer

Digital transformation (B) is the correct term because it encompasses the holistic integration of digital technology across all business areas, fundamentally altering operations and customer value delivery. Unlike IT modernization, which focuses on updating legacy systems, digital transformation involves strategic changes in culture, processes, and customer engagement, often leveraging cloud computing, data analytics, and AI. This aligns with the GCDL domain's emphasis on how cloud technology enables business model innovation rather than just infrastructure upgrades.

Exam trap

The GCDL exam often tests the distinction between tactical technology upgrades (like cloud migration or IT modernization) and the strategic, business-model-changing scope of digital transformation, leading candidates to confuse a component (e.g., moving to the cloud) with the holistic process.

How to eliminate wrong answers

Option A (IT modernization) is wrong because it specifically refers to updating or replacing legacy IT systems (e.g., hardware, software) to improve efficiency, without necessarily changing business models or customer value delivery. Option C (Cloud migration) is wrong because it is a tactical move of moving applications or data to cloud infrastructure (e.g., IaaS/PaaS), which is a component of digital transformation but not the overarching strategic process. Option D (Agile development) is wrong because it is a software development methodology focused on iterative delivery and collaboration, not the enterprise-wide integration of digital technology across all business functions.

353
MCQmedium

A developer needs to store application secrets (database passwords, API keys, OAuth tokens) securely so that they are not hardcoded in source code or environment variables. The secrets should be automatically versioned and rotatable. Which Google Cloud product is designed for this secure secrets management requirement?

A.Cloud Storage, by storing secrets in an encrypted bucket with restricted IAM access
B.Secret Manager, which stores secrets encrypted at rest with automatic versioning, rotation support, and fine-grained IAM access control
C.Cloud KMS, which generates encryption keys for encrypting application data
D.Cloud SQL, by storing secrets in an encrypted database table with restricted access
AnswerB

Secret Manager is the correct answer. It provides: encrypted storage for secret values, automatic versioning (each update creates a new numbered version), rotation support via Cloud Functions, per-secret IAM bindings, and audit logs for every secret access. It directly solves the hardcoded secrets problem.

Why this answer

Secret Manager is Google Cloud's dedicated service for storing application secrets such as database passwords, API keys, and OAuth tokens. It provides encryption at rest using AES-256, automatic versioning (each new secret version is immutable and timestamped), and built-in rotation support via scheduled rotation policies. Fine-grained IAM roles (e.g., roles/secretmanager.secretAccessor) ensure least-privilege access without exposing secrets in source code or environment variables.

Exam trap

The trap here is that candidates confuse Cloud KMS (key management) with Secret Manager (secret storage), or assume that any encrypted storage service (like Cloud Storage or Cloud SQL) can substitute for a purpose-built secrets manager, ignoring the need for automatic versioning, rotation, and fine-grained per-secret access control.

How to eliminate wrong answers

Option A is wrong because Cloud Storage is an object storage service, not a secrets manager; while you can encrypt a bucket and restrict IAM, it lacks automatic versioning (object versioning is optional and not secret-aware), rotation policies, and fine-grained access control at the secret level (e.g., per-secret IAM). Option C is wrong because Cloud KMS is a key management service for creating and managing encryption keys used to encrypt data elsewhere; it does not store secrets directly, does not provide versioning of secret values, and does not support rotation of the secret itself (only key rotation). Option D is wrong because Cloud SQL is a managed relational database service; storing secrets in a database table requires custom application logic for encryption, versioning, and rotation, and it introduces additional attack surface (e.g., SQL injection) and operational overhead, making it unsuitable for secure secrets management.

354
MCQmedium

Refer to the exhibit. A user receives this error when trying to SSH into a Compute Engine instance. Which IAM role should be granted to the user?

A.roles/compute.osLogin
B.roles/compute.instanceAdmin.v1
C.roles/iam.serviceAccountUser
D.roles/compute.admin
AnswerB

roles/compute.instanceAdmin.v1 is the correct role because it grants compute.instances.get, which lets the user retrieve instance metadata required for SSH connection, and compute.instances.setMetadata, which allows updating SSH keys or enabling OS Login. This is the minimum privilege that includes the necessary API calls for ssh to succeed.

Why this answer

The error indicates the user lacks SSH access to the Compute Engine instance. The `roles/compute.instanceAdmin.v1` role includes the `compute.instances.setMetadata` permission, which allows the user to add their SSH public key to the instance's metadata, enabling SSH access. This role also provides broader instance management capabilities, making it the correct choice for resolving SSH connectivity issues.

Exam trap

Google Cloud often tests the misconception that `roles/compute.osLogin` is required for SSH access, but the trap here is that OS Login is an alternative method that must be explicitly enabled on the instance and project, whereas the default SSH access relies on metadata-based keys, which require `roles/compute.instanceAdmin.v1` to modify.

How to eliminate wrong answers

Option A is wrong because `roles/compute.osLogin` enables OS Login, which uses IAM permissions to manage SSH keys centrally, but the error suggests the user is not using OS Login or lacks the necessary metadata configuration. Option C is wrong because `roles/iam.serviceAccountUser` allows a user to impersonate a service account, but it does not grant any permissions to access Compute Engine instances or modify SSH keys. Option D is wrong because `roles/compute.admin` is a highly privileged role that includes all Compute Engine permissions, but it is overly broad for simply granting SSH access; the more specific `roles/compute.instanceAdmin.v1` is sufficient and follows the principle of least privilege.

355
Multi-Selectmedium

Which TWO of the following are key benefits of using a cloud provider like Google Cloud? (Choose exactly 2.)

Select 2 answers
A.Access to a global network infrastructure
B.Complete removal of physical security responsibilities
C.Reduced internet latency
D.Elimination of hardware costs
E.Economies of scale leading to lower costs
AnswersA, E

This is a key benefit because cloud providers operate a vast, geographically distributed network of data centers and edge nodes connected by high-capacity backbone links. Customers can deploy applications in regions closer to their end users, enabling low-latency access, high availability, and global redundancy without building their own infrastructure. This global footprint also facilitates compliance with data residency requirements and disaster recovery across regions.

Why this answer

Option A is correct because Google Cloud operates a global network of regions, zones, and edge points of presence, so customers can deploy workloads close to users and leverage Google's private backbone for reliable, high-performance connectivity. Option E is correct because cloud providers aggregate massive demand across many tenants, achieving economies of scale in hardware, power, cooling, and operations that translate into lower per-unit costs passed on to customers. Option B is not fully correct because the cloud provider secures the physical data center, but customers still retain responsibility for physical security of their own devices, offices, and access controls under the shared responsibility model.

Option C is not guaranteed because reduced internet latency depends on architecture, region selection, and last-mile connectivity, not merely on using a cloud provider. Option D is not strictly correct because while capital hardware costs are largely shifted to operational expenses, customers may still incur hardware costs for on-premises equipment, client devices, or dedicated hardware such as TPUs.

Exam trap

Google Cloud often tests the shared responsibility model to trick candidates into thinking that physical security is completely offloaded, when in fact customers still secure their own virtual assets and data.

356
MCQmedium

A company runs Compute Engine instances that experience stable, predictable usage. They want to reduce costs by committing to a 1-year term for these virtual machines. Which discount type should they use?

A.Committed use discounts
B.Preemptible VMs
C.Sole-tenant nodes
D.Sustained use discounts
AnswerA

Committed use discounts require a contractual commitment of 1 or 3 years to a specific amount of vCPUs, memory, and other resources in a region. For stable, predictable workloads that run continuously, this commitment can reduce Compute Engine costs by up to 57%, making it the most cost-effective option among the choices. Unlike sustained use discounts, CUDs lock in a lower price regardless of actual monthly usage, but only if you are willing to commit to a fixed resource level.

Why this answer

Committed use discounts (CUDs) allow customers to commit to a certain amount of resources (vCPUs, memory) for 1 or 3 years in exchange for a significant discount. This is ideal for predictable workloads.

357
MCQeasy

A startup wants to deploy a containerized web application that can scale to zero when not in use and automatically scale based on request traffic. They have limited DevOps experience and want minimal infrastructure management. Which compute service should they choose?

A.App Engine Standard
B.Google Kubernetes Engine
C.Compute Engine
D.Cloud Run
AnswerD

Cloud Run is a fully managed serverless compute platform that executes stateless containers from a Docker image, automatically scaling each instance in response to incoming HTTP requests. It scales to zero when there is no traffic, meaning you pay only for requests actually being processed, with no cluster or VM to manage. This makes it ideal for a startup's containerized web application, providing built-in HTTPS, high availability, and request-based billing without any underlying infrastructure to operate.

Why this answer

Cloud Run is a serverless container platform that scales to zero, automatically scales based on requests, and requires no cluster management, making it ideal for this scenario.

358
MCQmedium

A user attempts to create a new Compute Engine instance in the us-central1 region and receives an error indicating that the quota for 'CPUs' has been exceeded. Where should the user go to request a quota increase?

A.Cloud Console > IAM & Admin > Quotas
B.Cloud Console > Billing > Budgets
C.Google Cloud Pricing Calculator
D.Cloud Console > Support > Cases
AnswerA

The IAM & Admin > Quotas page in Cloud Console is the centralized self-service interface for viewing all project quotas, including CPU, disk, and API rate limits. When instance creation fails with a QUOTA_EXCEEDED error, you use this page to select the specific quota, edit the limit, and submit a request that routes to the appropriate approval workflow. For many quotas, the increase is automatically applied within minutes, making this the definitive path.

Why this answer

Quota increases in Google Cloud are requested through the Cloud Console under IAM & Admin > Quotas, where you can filter by service (Compute Engine), locate the specific quota (CPUs in us-central1), and submit an increase request. This is the standard, supported workflow for adjusting project quotas. The request is then reviewed and typically approved automatically or by Google support depending on the size.

Exam trap

GCDL often tests where administrative actions live in the Console, and candidates confuse quota management with billing or support, picking Billing > Budgets or Support > Cases instead of the dedicated Quotas page.

How to eliminate wrong answers

Option B is wrong because Billing > Budgets is for setting spending alerts and budget thresholds, not for requesting quota increases. Option C is wrong because the Google Cloud Pricing Calculator is a tool for estimating costs, not for managing or increasing quotas. Option D is wrong because Support > Cases is for opening support tickets about issues, not the designated path for quota increase requests—quota adjustments are handled through the Quotas page, which may then route to support if manual review is needed.

359
Multi-Selectmedium

A company is designing a disaster recovery strategy for a critical application. They need to ensure that if an entire Google Cloud region becomes unavailable, the application can still serve traffic from another region with minimal data loss. Which TWO strategies should they implement? (Choose 2.)

Select 2 answers
A.Deploy the application in multiple regions
B.Use synchronous replication for the database
C.Deploy the application in multiple zones within a single region
D.Use asynchronous replication for the database
E.Use a single load balancer
AnswersA, B

Running workloads across multiple regions means a surviving region can absorb traffic when one fails, satisfying the requirement to keep serving requests during a full regional outage. Pairing this with data replication addresses the minimal-data-loss constraint, since regional failover alone does not protect stored data.

Why this answer

Option A is correct because deploying the application in multiple regions ensures that if an entire Google Cloud region becomes unavailable, the application can continue serving traffic from another region, directly addressing the requirement for regional failover. Option B is correct because synchronous replication writes data to the standby database in another region before acknowledging the write, which minimizes data loss (near-zero RPO) as required by the scenario. Option C is incorrect because multiple zones within a single region do not protect against a full regional outage.

Option D is incorrect because asynchronous replication can result in some data loss during failover, which conflicts with the minimal data loss requirement. Option E is incorrect because a single load balancer does not provide cross-region redundancy and could itself become a single point of failure.

Exam trap

The trap is confusing zone redundancy with region redundancy — candidates often select 'multiple zones' thinking it covers regional failure, and may also pick asynchronous replication because it sounds more scalable, missing the 'minimal data loss' requirement.

360
Multi-Selectmedium

A developer is troubleshooting why a Compute Engine instance cannot start in a specific region. The error indicates a quota limit. Which TWO steps should the developer take?

Select 2 answers
A.Check the current quota for the resource in that region via the Quotas page in Cloud Console.
B.Delete unused resources in the same region to free up quota automatically.
C.Change the project to a different billing account.
D.Use a different machine type that uses fewer vCPUs.
E.Request a quota increase for the specific resource in that region.
AnswersA, E

The Quotas page in Cloud Console displays per-region limits and usage for each resource type, such as vCPUs, static IPs, and persistent disk capacity. When a Compute Engine instance creation fails due to a quota limit, the error message will explicitly indicate a quota exceeded error, and this page confirms the exact metric and current consumption. This is the correct first step because it isolates whether the failure is a quota issue or something else, such as network configuration or billing charges.

Why this answer

Option A is correct because the Quotas page in Cloud Console (IAM & Admin > Quotas) shows the current usage and limit for each resource per region, letting the developer confirm which quota is exhausted before acting. Option E is correct because when a quota is genuinely too low for the workload, the developer must submit a quota increase request for that specific resource and region, which is then reviewed and approved by Google Cloud support. Together these steps diagnose the exact quota and resolve it through the proper channel.

Option B is not appropriate because deleting resources is a destructive action and quota is not 'freed automatically' in a way that guarantees the instance can start; the developer should first identify the quota. Option C is wrong because changing the billing account does not alter regional resource quotas. Option D is wrong because switching to a smaller machine type is a workaround that may not address the actual quota (e.g., if the quota is for CPUs, in-use IP addresses, or persistent disk), and it does not resolve the underlying quota limit.

Exam trap

GCDL often tests the misconception that quotas are global or billing-related — candidates pick 'change billing account' or 'delete resources' instead of the correct check-and-request workflow for the specific regional quota.

361
Multi-Selectmedium

A company is building a data pipeline that ingests events from multiple sources, processes them in real-time, and stores the results in a data warehouse for analysis. They need a fully managed, serverless solution for stream processing. Which THREE Google Cloud services should they use?

Select 3 answers
A.Dataproc
B.Pub/Sub
C.Dataflow
D.Cloud Functions
E.BigQuery
AnswersB, C, E

Pub/Sub is the correct foundational service for ingesting streaming events because it is a scalable, durable, asynchronous messaging middleware that decouples event producers from consumers. It provides at-least-once delivery, supports push and pull subscriptions, and can buffer spikes in event volume, ensuring that events are not lost before downstream processing. As the entry point of a data pipeline, Pub/Sub is specifically designed to receive high-throughput event streams and make them available to streaming processors like Dataflow.

Why this answer

Pub/Sub (B) is correct because it is the fully managed, serverless messaging service used to ingest events from multiple sources into the pipeline with at-least-once delivery and automatic scaling. Dataflow (C) is correct because it provides fully managed, serverless stream processing with Apache Beam, handling real-time transformations, windowing, and exactly-once processing. BigQuery (E) is correct because it is the serverless data warehouse where the processed results are stored for analysis via SQL.

Dataproc (A) is not appropriate because it is a managed Hadoop/Spark service requiring cluster provisioning, not serverless stream processing. Cloud Functions (D) is event-driven compute for lightweight functions, not a stream-processing engine for continuous pipelines.

Exam trap

GCDL often tests the distinction between 'fully managed' and 'serverless' — candidates may incorrectly select Dataproc because it is a managed service, but it is not serverless, and may overlook Cloud Functions as a stream processor when it is only suitable for lightweight event handling.

362
Multi-Selectmedium

A company is adopting a multi-cloud strategy to avoid vendor lock-in and use best-of-breed services. Which TWO statements accurately describe multi-cloud? (Choose 2)

Select 2 answers
A.It connects an on-premises data centre to a single public cloud.
B.It requires a private cloud on-premises.
C.It reduces management complexity.
D.It can help avoid vendor lock-in.
E.It uses multiple public cloud providers.
AnswersD, E

By distributing workloads across multiple cloud providers, an organization avoids being dependent on a single vendor's pricing, service limitations, or strategic direction. This gives the organization negotiation leverage and makes it easier to switch workloads if one provider's offerings or terms become unfavorable. Multi-cloud also mitigates the risk of a single-vendor outage or platform failure affecting business continuity.

Why this answer

Option D is correct because multi-cloud deliberately spreads workloads across more than one provider, so dependence on any single vendor's proprietary services, pricing, and APIs is reduced, directly mitigating vendor lock-in. Option E is correct because the defining characteristic of multi-cloud is the use of two or more distinct public cloud providers (for example AWS plus Azure plus GCP), which is exactly what enables best-of-breed service selection. Option A is incorrect because connecting an on-premises data centre to a single public cloud describes a hybrid cloud, not multi-cloud.

Option B is incorrect because multi-cloud does not require an on-premises private cloud; it can consist entirely of public cloud providers. Option C is incorrect because multi-cloud typically increases management complexity, requiring separate tooling, identity, networking, and cost controls per provider rather than reducing it.

Exam trap

GCDL often tests the confusion between multi-cloud and hybrid cloud, where candidates incorrectly assume multi-cloud requires on-premises infrastructure or that it simplifies management.

363
MCQeasy

A company uses the lifecycle configuration above for archival data that is rarely accessed. What business benefit does this provide?

A.Improved data security through encryption
B.Faster data retrieval for frequently accessed files
C.Reduced storage costs by transitioning to cheaper tiers and deleting obsolete data
D.Automatic replication to another region for disaster recovery
AnswerC

Lifecycle rules automatically transition objects to colder, cheaper storage classes as they age, then expire them once obsolete. This eliminates manual intervention and ongoing costs for rarely accessed archival data, directly delivering the reduced storage expenditure the question asks about.

Why this answer

Option C is correct because a lifecycle configuration is designed to automatically transition objects to lower-cost storage classes (such as infrequent access or archive tiers) as they age and to expire/delete objects once they are no longer needed, which directly lowers storage spend for rarely accessed archival data. This matches the stated scenario of archival data that is rarely accessed, where the main business benefit is cost optimization rather than performance or security. Option A does not fit because encryption is a security control, not a lifecycle action, and lifecycle rules do not provide encryption.

Option B does not fit because lifecycle policies typically move data to slower, cheaper tiers, which can increase retrieval latency rather than speed it up. Option D does not fit because cross-region replication is a separate feature (e.g., S3 Cross-Region Replication) and is not what a lifecycle configuration accomplishes.

Exam trap

A common trap is confusing cost optimization with other benefits like security or replication. Focus on the direct effect of lifecycle policies: cost savings through tiering and deletion.

364
MCQmedium

An organization wants to ensure that all data stored in Cloud Storage is encrypted with customer-managed keys that can be rotated on demand. They also need to log every key use for audit compliance. Which combination of services should they use?

A.Cloud KMS with Customer-Supplied Encryption Keys (CSEK)
B.Google-managed encryption keys with Access Transparency
C.Cloud KMS with Customer-Managed Encryption Keys (CMEK) and Cloud Audit Logs
D.Secret Manager with encryption keys and Cloud Logging
AnswerC

CMEK allows you to create and manage key material in Cloud KMS and assign it as the key that encrypts Cloud Storage objects through a key ring and CryptoKey. You retain lifecycle control—enabling, disabling, rotating, or destroying the key—and every encryption or decryption operation under that key is recorded as a Cloud KMS operation in Cloud Audit Logs, giving you the auditability needed for security monitoring and forensic investigation.

Why this answer

Cloud KMS allows creation and management of customer-managed encryption keys (CMEK) for Cloud Storage. Cloud Audit Logs can be used to log key usage (e.g., via Data Access audit logs). CSEK is deprecated; Secret Manager is for secrets, not encryption keys for storage.

365
MCQmedium

A healthcare company runs a containerized patient-records application on Google Kubernetes Engine. The security team wants to detect and block suspicious network traffic between pods, and also wants to scan container images for known vulnerabilities before deployment. Which combination of Google Cloud services should they use to meet these requirements?

A.Cloud Armor for pod-to-pod traffic filtering and Container Registry for image scanning
B.Google Cloud Armor for pod traffic and Artifact Registry vulnerability scanning for images
C.GKE network policies with Dataplane V2 for pod traffic and Artifact Registry vulnerability scanning for images
D.VPC Service Controls for pod traffic and Binary Authorization for image scanning
AnswerC

GKE network policies, enforced by Dataplane V2 based on Cilium eBPF, let you define ingress and egress rules between pods, effectively blocking suspicious east-west traffic. Artifact Registry automatically scans container images for known vulnerabilities when they are pushed. Together they provide both runtime network segmentation and pre-deployment image scanning, exactly matching the stated requirements.

Why this answer

Protecting containerized workloads requires layered controls. GKE network policies with Dataplane V2 enforce pod-level firewall rules inside the cluster, while Artifact Registry provides automated vulnerability scanning of container images. Cloud Armor and VPC Service Controls operate at different layers and cannot inspect pod-to-pod traffic, so they do not fulfill the east-west traffic requirement.

Exam trap

The trap here is assuming that Cloud Armor or VPC Service Controls can filter internal pod-to-pod traffic, when they actually protect external edges or API perimeters.

366
MCQhard

Refer to the exhibit. A team deployed this Cloud Run service. During a load test, the service receives high traffic, but the number of container instances never exceeds 10. What is the most likely cause?

A.The maxScale annotation limits the maximum number of instances to 10.
B.The minScale of 2 forces at least two instances, but not the max.
C.The containerConcurrency of 80 limits the number of concurrent requests per instance.
D.The CPU limit of 1 vCPU is too low to handle the traffic.
AnswerA

The `autoscaling.knative.dev/maxScale` annotation directly sets the upper bound on the number of instances that Cloud Run can create. With a value of 10, the service is hard-capped at 10 concurrent instances even if traffic surges. This overrides the default maximum (which is usually 100 or unlimited), so it is the correct reason the service cannot scale beyond 10.

Why this answer

The `maxScale` annotation in Cloud Run directly caps the maximum number of container instances that can be created. When the service receives high traffic but never exceeds 10 instances, it indicates that the `maxScale` annotation is set to 10, preventing further scaling even if demand increases. This is the most direct and likely cause among the options.

Exam trap

Google Cloud often tests the distinction between scaling limits (maxScale) and performance tuning parameters (containerConcurrency, CPU limits), leading candidates to mistakenly attribute a hard instance cap to concurrency or resource constraints rather than the explicit annotation.

How to eliminate wrong answers

Option B is wrong because `minScale` of 2 only ensures a minimum of two instances are always running, but it does not impose any upper limit; the service could scale beyond 10 if `maxScale` were higher. Option C is wrong because `containerConcurrency` of 80 limits how many concurrent requests each instance can handle, but it does not cap the total number of instances; the service could still scale out to more instances to handle the load. Option D is wrong because a CPU limit of 1 vCPU per instance might cause performance bottlenecks, but it does not prevent the service from creating more than 10 instances; Cloud Run can still scale horizontally to additional instances even if each has a low CPU limit.

367
MCQeasy

A small startup can now access the same world-class AI, machine learning, and global infrastructure that previously only Fortune 500 companies with billion-dollar IT budgets could afford. Which cloud characteristic enables this competitive equalization?

A.Cloud providers charge smaller companies lower rates than enterprises.
B.Cloud's pay-per-use model and managed services give any organization access to enterprise-grade capabilities without large upfront capital investment.
C.Cloud providers assign dedicated infrastructure to small companies so they always have priority access.
D.Government regulations require cloud providers to offer equal service levels to all customers.
AnswerB

The pay-per-use model eliminates the need for large upfront capital investments in hardware and data centers, converting fixed costs into variable operational costs. Managed services further remove the engineering burden by offering enterprise-grade capabilities like AI/ML APIs, managed databases, and global load balancing as on-demand services. This combination lets any organization deploy solutions that previously required massive infrastructure budgets and specialized teams.

Why this answer

The cloud's pay-per-use model eliminates the need for large upfront capital expenditures, while managed services (e.g., AWS RDS, Azure SQL Database, Google Cloud AI Platform) abstract away the operational complexity of maintaining enterprise-grade infrastructure. This allows a small startup to leverage the same AI/ML models, GPU clusters, and global network backbones that Fortune 500 companies use, paying only for what they consume rather than provisioning for peak capacity.

Exam trap

The GCDL exam often tests the misconception that cloud providers offer 'lower rates' or 'dedicated infrastructure' to small companies, when in reality the equalization comes from the operational expenditure (OpEx) model and managed services that abstract complexity, not from preferential pricing or physical resource dedication.

How to eliminate wrong answers

Option A is wrong because cloud providers do not charge lower rates based on company size; pricing is typically based on resource consumption, commitment levels (e.g., reserved instances), and volume discounts, not on whether the customer is a startup or an enterprise. Option C is wrong because cloud providers use multi-tenant architectures (e.g., hypervisor-level isolation, VPCs) rather than assigning dedicated physical infrastructure to small companies; priority access is not guaranteed unless specific reserved capacity or dedicated hosts are purchased. Option D is wrong because no government regulations mandate equal service levels for all customers; SLAs vary by service tier and region, and providers like AWS, Azure, and GCP offer different performance guarantees based on the chosen plan (e.g., Standard vs.

Premium tiers).

368
MCQhard

A consumer goods company uses cloud-based demand sensing — analyzing real-time sales signals, social media trends, and weather data to adjust production runs dynamically. This has reduced stockouts by 35% and overstock waste by 28%. Which aspect of digital transformation does this most directly exemplify?

A.Cost reduction through infrastructure consolidation and server decommissioning
B.Operational transformation through real-time data integration and machine learning that enables continuous, signal-driven production decisions
C.Business continuity improvement through data backup and disaster recovery in the cloud
D.Employee productivity improvement through providing staff with cloud-based collaboration tools
AnswerB

This precisely describes what's happening: cloud enables the integration of diverse real-time signals (sales, social, weather) at a scale and speed that transforms how production decisions are made. The 35% stockout reduction and 28% waste reduction are measurable business outcomes of this operational transformation.

Why this answer

The scenario describes a shift from static, forecast-based production to dynamic, signal-driven decisions using real-time data integration (sales signals, social media, weather) and machine learning. This directly exemplifies operational transformation, a core pillar of digital transformation where cloud-based analytics and AI enable continuous optimization of core business processes like manufacturing.

Exam trap

Google Cloud often tests the distinction between operational transformation (changing core business processes with data and AI) and other common cloud benefits like cost savings or disaster recovery, so candidates mistakenly pick A or C when they see 'cloud' and 'reduced waste' without analyzing the process change.

How to eliminate wrong answers

Option A is wrong because it focuses narrowly on IT infrastructure cost savings (server decommissioning), whereas the question describes a business process change in production planning, not IT consolidation. Option C is wrong because it describes business continuity (backup and disaster recovery), which is about maintaining operations during disruptions, not about using real-time data to dynamically adjust production. Option D is wrong because it refers to employee productivity via collaboration tools (e.g., cloud-based email or document sharing), not to the automated, machine-learning-driven decision-making that adjusts production runs.

369
MCQeasy

An organization wants to ensure that all projects under a specific folder inherit a policy that disables the creation of external IP addresses. Which Google Cloud resource hierarchy level should the policy be applied to enforce this requirement for all child resources?

A.Folder
B.Organization node
C.Resource (e.g., VM instance)
D.Project
AnswerA

Attaching the policy at the folder level is the most operative because IAM policies are inherited down the resource hierarchy. The folder acts as a container for all projects under it, so a policy bound at this level automatically applies to every project and resource within that subtree, including any future projects created in the folder. This provides consistent enforcement for the specific business unit or department without affecting unrelated folders or projects.

Why this answer

Organization policy constraints applied at the folder level are inherited by all projects and resources within that folder, ensuring consistent enforcement across the hierarchy.

370
MCQeasy

A data analyst wants to create interactive dashboards and reports using data from BigQuery, without writing code. Which Google Cloud BI tool should they use?

A.Data Studio
B.Looker Studio
C.Dataflow
D.Looker
AnswerB

Looker Studio (formerly Data Studio) is a free, self-service BI and data visualization platform that lets analysts build interactive dashboards and reports entirely through a drag-and-drop interface, with no code required. It connects directly to data sources such as BigQuery, Google Sheets, and many other databases, and supports shared, embeddable, and scheduled reports. Because the analyst explicitly wants to avoid writing code, Looker Studio's WYSIWYG editor and pre-built connectors make it the correct choice for this requirement.

Why this answer

Looker Studio is a free, self-service BI tool that allows users to create interactive dashboards and reports from various data sources including BigQuery, with a drag-and-drop interface. Looker is a more advanced enterprise BI platform. Data Studio is the former name; it is now Looker Studio.

Dataflow is for data processing.

371
MCQeasy

Which of the following is a characteristic of the cloud according to the NIST definition?

A.Measured service
B.Manual scaling
C.Limited network access
D.Single tenancy
AnswerA

Measured service is the cloud attribute whereby the provider continuously meters, monitors, and controls resource usage (compute, storage, networking) to provide transparent, pay-per-use billing. This metering enables cost optimization by allowing customers to align spending with actual consumption, and it is a core NIST essential characteristic that distinguishes cloud from traditional IT.

Why this answer

The NIST SP 800-145 definition of cloud computing lists five essential characteristics: on-demand self-service, broad network access, resource pooling, rapid elasticity, and measured service. Measured service means cloud providers automatically monitor, control, and report resource usage (storage, processing, bandwidth, active user accounts) via metering capabilities, providing transparency to both provider and consumer. This pay-per-use metering model is a defining trait of cloud computing.

Exam trap

The trap here is confusing the five NIST essential characteristics with generic cloud benefits — candidates often pick 'manual scaling' or 'single tenancy' because they sound plausible but actually contradict elasticity and resource pooling.

How to eliminate wrong answers

Option B is wrong because manual scaling contradicts the NIST characteristic of rapid elasticity, where resources can scale automatically and elastically with demand. Option C is wrong because the NIST definition requires broad network access, meaning services are available over the network via standard mechanisms, not limited access. Option D is wrong because NIST specifies resource pooling with a multi-tenant model, where physical and virtual resources are pooled to serve multiple consumers — not single tenancy.

372
MCQmedium

A healthcare company needs to store patient data in Google Cloud and must comply with HIPAA (Health Insurance Portability and Accountability Act). Which statement correctly describes how Google Cloud helps them achieve HIPAA compliance?

A.Storing data in Google Cloud automatically makes an application HIPAA-compliant.
B.Google offers HIPAA-eligible services and signs a Business Associate Agreement (BAA), but customers must implement their own technical safeguards and access controls.
C.HIPAA compliance is impossible on public cloud; healthcare data must stay on-premises.
D.Google Cloud's automatic data encryption fully satisfies all HIPAA technical safeguard requirements.
AnswerB

Google Cloud participates in HIPAA compliance by providing a BAA and offering infrastructure that addresses physical, environmental, and certain technical safeguards. However, the Shared Responsibility Model makes the customer accountable for configuring services correctly, managing access controls, enabling audit logging, encrypting data where required, and implementing contingency plans. Without these customer-side actions, even a signed BAA does not render a workload compliant.

Why this answer

Google Cloud provides HIPAA-eligible services and offers a Business Associate Agreement (BAA) to covered entities, but compliance is a shared responsibility. Customers must configure their own technical safeguards, such as access controls, audit logging, and encryption key management, to meet HIPAA requirements. Google Cloud does not automatically make an application compliant; the customer must implement the necessary controls.

Exam trap

The GCDL exam often tests the shared responsibility model by presenting options that imply full vendor responsibility (like automatic compliance) or full customer responsibility (like impossibility), and the trap here is assuming that encryption alone satisfies all HIPAA technical safeguards, ignoring access control and audit requirements.

How to eliminate wrong answers

Option A is wrong because storing data in Google Cloud does not automatically make an application HIPAA-compliant; compliance requires the customer to implement technical safeguards and access controls, and to sign a BAA. Option C is wrong because HIPAA compliance is achievable on public cloud platforms like Google Cloud when using HIPAA-eligible services and signing a BAA, and many healthcare organizations successfully run workloads in the cloud. Option D is wrong because Google Cloud's automatic data encryption addresses only one aspect of HIPAA's technical safeguards; it does not satisfy all requirements, such as access control, audit controls, and integrity controls, which the customer must implement.

373
MCQhard

An SRE team is practicing 'chaos engineering' by simulating a zone-level failure in their staging environment. They find that their application does not automatically recover — traffic is not redirected and the service remains down. What architectural component is most likely missing?

A.The application needs more replicas in the failing zone to survive the failure
B.A load balancer with health checks across multiple zones is most likely missing — without it, there is no mechanism to detect the zone failure and automatically redirect traffic to healthy instances in surviving zones
C.The application needs a larger machine type to handle the full traffic load without the failed zone's capacity
D.Cloud Monitoring alerts need to be configured to notify the team when a zone fails, enabling manual traffic redirection
AnswerB

The load balancer is the key component. It must be configured with backend instances in multiple zones and health checks enabled. When the health check detects that zone A instances are unhealthy, it automatically removes them from the rotation and sends all traffic to healthy instances in zones B and C. Without the load balancer, clients connect directly to zone A and have no fallback.

Why this answer

In a zone-level failure, traffic cannot be redirected to healthy instances in surviving zones without a load balancer that performs health checks across multiple zones. Google Cloud's external or internal load balancers (e.g., HTTP(S) Load Balancer, TCP/UDP Network Load Balancer) use health checks to detect unhealthy instances and automatically route traffic only to healthy backends. Without this component, the application has no mechanism to detect the zone failure and reroute traffic, leaving the service down.

Exam trap

The trap here is that candidates may confuse 'scaling up' (larger machine types or more replicas) with 'resilience through load balancing', failing to recognize that without a load balancer with health checks, no amount of capacity in surviving zones will automatically redirect traffic.

How to eliminate wrong answers

Option A is wrong because adding more replicas in the failing zone does not help when the entire zone is unavailable; replicas in that zone would also be down. Option C is wrong because a larger machine type does not solve the lack of automatic traffic redirection; it only increases capacity in surviving zones, but without a load balancer, traffic is still not redirected. Option D is wrong because Cloud Monitoring alerts only notify the team of the failure; they do not automatically redirect traffic, and manual redirection is not a scalable or reliable solution for chaos engineering scenarios.

374
MCQmedium

A company wants to encrypt sensitive data stored in Cloud Storage with a key that is generated and stored on-premises using a hardware security module (HSM). They do not want Google to have access to the key. Which encryption option should they use?

A.Google-managed encryption keys
B.Customer-managed encryption keys (CMEK)
C.Cloud HSM
D.Customer-supplied encryption keys (CSEK)
AnswerD

Customer-supplied encryption keys (CSEK) are keys that the customer generates and provides directly to Google Cloud via API when writing data. Google's systems use the customer-supplied key to encrypt the data at rest and then destroy the key from its memory, ensuring that Google does not retain the key material. This aligns with the company's requirement because the customer retains exclusive control over the key, and Google cannot access the plaintext without the customer re-supplying the key, making it the correct choice.

Why this answer

Customer-Supplied Encryption Keys (CSEK) allow customers to provide their own keys; Google does not store the key and has no access to it.

375
MCQhard

A CISO is implementing a Zero Trust security architecture for the company's Google Cloud environment. Under Zero Trust, which fundamental assumption about network traffic changes compared to traditional perimeter-based security?

A.Zero Trust assumes that internal network traffic is more secure than external traffic because it has passed through the corporate firewall
B.Zero Trust assumes no traffic is trusted by default regardless of network origin — every request must be explicitly authenticated and authorized based on identity, device posture, and context
C.Zero Trust assumes all traffic is malicious and blocks all requests by default, requiring explicit allowlisting for each connection
D.Zero Trust eliminates the need for encryption since all traffic is assumed to be on secure internal networks
AnswerB

This is the core Zero Trust principle: 'never trust, always verify.' A request from inside the VPC receives the same verification scrutiny as a request from the public internet. This model is more appropriate for cloud environments where the network perimeter no longer has clear meaning — employees, services, and attackers can all be inside the 'perimeter.'

Why this answer

Zero Trust fundamentally shifts from implicit trust based on network location to explicit verification of every request. In Google Cloud, this means every API call, regardless of whether it originates from within the VPC or the internet, must be authenticated (e.g., using OAuth 2.0 or service account keys) and authorized based on identity, device posture, and context, as enforced by tools like Identity-Aware Proxy (IAP) and VPC Service Controls.

Exam trap

The trap here is that candidates often confuse Zero Trust's 'never trust, always verify' with a blanket denial of all traffic (Option C), when in reality it requires explicit verification for each request, not static allowlisting.

How to eliminate wrong answers

Option A is wrong because Zero Trust explicitly rejects the assumption that internal network traffic is more secure; it treats all traffic as untrusted, including traffic within the same VPC or subnet, and does not rely on a corporate firewall for trust. Option C is wrong because Zero Trust does not assume all traffic is malicious and block by default; it assumes no implicit trust but allows traffic after explicit authentication and authorization, not via static allowlisting. Option D is wrong because Zero Trust does not eliminate the need for encryption; in fact, it mandates encryption in transit (e.g., TLS 1.3) and at rest for all traffic, as internal networks are no longer considered trusted boundaries.

Page 4

Page 5 of 12

Page 6