Courseiva

Google Cloud Digital Leader (GCDL) — Questions 151–225

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

Page 2

Page 3 of 12

Page 4
151
MCQeasy

A startup wants to deploy a containerized web application without managing the underlying infrastructure. They want to only focus on code. Which Google Cloud service is most suitable?

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

Cloud Run is a fully managed serverless platform that executes stateless containers directly from an image. It abstracts all infrastructure, including servers, OS, and cluster management, and automatically scales from zero to N instances based on incoming traffic. You only pay for resources used while a request is being processed, and it brings your existing Dockerfile-based workflow without requiring Kubernetes or VM administration.

Why this answer

Cloud Run is the correct choice because it is a fully managed serverless platform that executes stateless containers in a request-driven environment, abstracting all infrastructure management. This allows the startup to focus solely on code by deploying container images directly from Artifact Registry or Container Registry, with automatic scaling down to zero when not in use.

Exam trap

The trap here is that candidates often confuse App Engine's automatic scaling with container support, overlooking that App Engine Standard requires specific language runtimes and does not accept arbitrary containers, while Cloud Run provides true container portability without infrastructure management.

How to eliminate wrong answers

Option A is wrong because Google Kubernetes Engine (GKE) requires managing a Kubernetes cluster, including node pools, networking, and scaling policies, which contradicts the requirement to avoid infrastructure management. Option B is wrong because Compute Engine provides virtual machines that demand manual provisioning, patching, and capacity planning, directly opposing the goal of focusing only on code. Option C is wrong because App Engine, while serverless, is a platform-as-a-service that restricts the runtime environment to specific supported languages and runtimes, whereas the startup explicitly wants to deploy a containerized application without such constraints.

152
MCQmedium

A global e-commerce platform runs on Compute Engine instances in multiple regions. Users report high latency when accessing the website from South America. The company wants to improve performance by delivering static content (images, CSS, JS) from the nearest edge location. Which Google Cloud service should they implement?

A.Cloud CDN
B.Cloud DNS
C.Cloud Load Balancing
D.Cloud Armor
AnswerA

Cloud CDN uses Google's globally distributed edge points of presence to cache static and dynamic content close to users. It intercepts HTTP(S) requests via the load balancing infrastructure and serves cached responses when possible, which reduces latency and cuts compute engine instance load. For a global e-commerce platform, caching product catalogs and media dramatically accelerates page rendering and improves customer experience.

Why this answer

Cloud CDN uses Google's global edge network (Points of Presence) to cache static content closer to users, reducing latency. Cloud Load Balancing distributes traffic but doesn't cache content. Cloud Armor provides security.

Cloud DNS resolves domain names.

153
MCQeasy

A company is concerned about which security responsibilities belong to Google versus which belong to them when using Google Cloud's managed database service (Cloud SQL). In the shared responsibility model, which security tasks does Google handle?

A.Google controls who can access the database and what data can be stored.
B.Google handles physical security, hardware maintenance, and OS and database software patching.
C.Google is responsible for backing up customer data and ensuring data recovery.
D.Google determines which compliance certifications the customer's application must meet.
AnswerB

This is the correct description of the shared responsibility model for managed database services like Cloud SQL. Google is responsible for data center physical security (e.g., biometric access controls, armed guards), hardware lifecycle management (failed disk replacement), and patching the underlying operating system and database engine. For example, Cloud SQL automatically applies minor version updates and security patches; customers can schedule maintenance windows for major upgrades. These infrastructure-level tasks are handled entirely by Google, which is what distinguishes a fully managed service from a self-managed Compute Engine VM.

Why this answer

In the shared responsibility model for Google Cloud services like Cloud SQL, Google is responsible for security 'of' the cloud, which includes physical security of data centers, hardware maintenance, and patching the underlying operating system and database software. This ensures the infrastructure hosting Cloud SQL instances is secure, while the customer remains responsible for securing their data, access policies, and application-level configurations.

Exam trap

The trap here is that candidates often confuse Google's responsibility for patching the database software (which Google handles) with the customer's responsibility for managing database access controls and backup configurations, leading them to incorrectly select options A or C.

How to eliminate wrong answers

Option A is wrong because Google does not control who accesses the database or what data is stored; those are customer responsibilities under IAM and data classification. Option C is wrong because while Google provides backup and recovery capabilities as a feature, the customer is responsible for configuring and enabling backups, and for verifying recovery procedures. Option D is wrong because Google does not determine which compliance certifications the customer's application must meet; the customer must assess their own compliance requirements and choose Google Cloud services that align with those certifications.

154
MCQeasy

Which Google Cloud service is a fully managed, serverless data warehouse for running SQL queries on petabyte-scale datasets?

A.Cloud SQL
B.Cloud Dataflow
C.Cloud Dataproc
D.BigQuery
AnswerD

BigQuery is a fully managed, serverless data warehouse that separates compute from storage, enabling petabyte-scale SQL analytics without provisioning infrastructure. It uses a columnar storage format and a distributed query engine that dynamically allocates slots, making it ideal for interactive BI and large data aggregations. BigQuery also includes built-in features like partitioning, clustering, and automatic recompression of data, and it supports standard SQL, which is why it is the correct answer for a serverless analytics warehouse.

Why this answer

BigQuery is the serverless data warehouse for analytics SQL queries. Dataflow is for stream processing, Dataproc for Hadoop/Spark, and Cloud SQL for OLTP.

155
MCQeasy

A CEO presents a strategic plan to 'move everything to the cloud.' The board asks what business outcome should be the primary measure of success for the cloud migration. Which answer best reflects a business-outcome-oriented approach to measuring cloud migration success?

A.The percentage of applications successfully migrated to cloud infrastructure
B.Measurable business improvements such as reduced time-to-market for new products, lower infrastructure costs, and improved customer satisfaction enabled by cloud capabilities
C.The number of cloud certifications earned by the IT team during the migration
D.Achieving 100% elimination of on-premises infrastructure by the end of the first year
AnswerB

Business-outcome-oriented success measures tie the migration directly to value creation: faster launches generate revenue, cost reduction improves margins, and customer satisfaction metrics capture whether the investment is working. These are the metrics that matter to the board.

Why this answer

It ties cloud migration success directly to measurable business outcomes, such as reduced time-to-market, lower infrastructure costs, and improved customer satisfaction. This aligns with the GCDL principle that cloud technology is a business enabler, not just an IT project. A business-outcome-oriented approach ensures that migration efforts are evaluated by their impact on strategic goals, such as agility and cost efficiency, rather than technical milestones.

Exam trap

Google Cloud often tests the distinction between technical metrics and business outcomes, trapping candidates who confuse project completion (e.g., percentage migrated) with actual business value (e.g., cost savings or agility improvements).

How to eliminate wrong answers

Option A is wrong because it measures a technical milestone (percentage of applications migrated) rather than a business outcome; simply moving applications to the cloud does not guarantee improved business performance or value. Option C is wrong because cloud certifications earned by the IT team are a measure of skill development, not a direct business outcome; certifications do not reflect whether the migration has improved operational efficiency or customer experience. Option D is wrong because achieving 100% elimination of on-premises infrastructure by a fixed deadline prioritizes a technical target over business value; a rushed migration can lead to service disruptions, cost overruns, and missed opportunities to optimize cloud-native capabilities.

156
Multi-Selecthard

A company has a multi-regional deployment of a web application on Compute Engine. They want to improve latency for users worldwide and reduce load on the origin servers. They also need to protect against SQL injection and cross-site scripting attacks. Which TWO Google Cloud services should they implement?

Select 2 answers
A.Cloud DNS
B.Cloud Armor
C.Cloud NAT
D.Cloud CDN
E.Cloud Load Balancing
AnswersB, D

Cloud Armor is Google Cloud's web application firewall (WAF) that provides configurable security policies for HTTP(S) load balancing. It includes pre-configured rules from the ModSecurity Core Rule Set to block SQL injection, cross-site scripting, and other OWASP Top 10 threats. Cloud Armor also supports IP allow/deny lists, geo-based access controls, and rate limiting, making it the precise service to protect a web application from the stated attacks.

Why this answer

Cloud Armor (B) is correct because it is Google Cloud's edge security service that provides WAF capabilities, including preconfigured rules to block SQL injection (e.g., sqli-v33-stable) and cross-site scripting (e.g., xss-v33-stable) attacks against HTTP(S) load-balanced backends. Cloud CDN (D) is correct because caching content at Google's globally distributed edge points of presence reduces latency for worldwide users and offloads requests from the Compute Engine origin servers. Cloud DNS (A) only provides authoritative DNS resolution and does not cache HTTP content or inspect application-layer attacks.

Cloud NAT (C) provides outbound internet connectivity for private instances and offers no latency or WAF benefits. Cloud Load Balancing (E) distributes traffic but does not itself cache content or provide WAF protection, so it does not satisfy both requirements.

Exam trap

GCDL often tests service-purpose pairing; candidates pick Cloud Load Balancing as an answer because it is 'needed,' but the question asks for the services that provide caching and WAF, which are Cloud CDN and Cloud Armor respectively.

157
MCQhard

A large enterprise runs a critical application on Google Cloud consisting of Compute Engine instances behind a TCP load balancer. The application experiences intermittent slow response times that last for about 10 minutes before returning to normal. This pattern has been occurring every few days at random times. The operations team has configured Cloud Monitoring alerts for CPU and memory, but no alerts have fired. They have also reviewed the load balancer logs and see no errors, but the latency spikes. The application logs show no errors during these periods. The team suspects a resource bottleneck but cannot find it. Further investigation reveals that the application makes synchronous calls to an external authentication service for each request. What is the most likely cause and corrective action?

A.The TCP load balancer is experiencing connection draining issues; switch to a proxy-based load balancer.
B.The instance group's autoscaler is configured with a cooldown period that is too long; reduce the cooldown period.
C.The application is making synchronous calls to an external authentication service that occasionally has latency spikes; implement caching and asynchronous processing.
D.The virtual machine instances are suffering from CPU throttling due to sustained use of burstable CPU; move to a machine type with more CPUs.
AnswerC

Synchronous calls to an external authentication service introduce a hard dependency where any latency spike in that service directly blocks the application's request threads, causing intermittent slowdowns that correlate with the external service's variability. Implementing caching for authentication results and switching to asynchronous processing decouples the critical path, absorbs latency spikes, and improves overall response time consistency—especially since the rest of the infrastructure is healthy.

Why this answer

The intermittent latency spikes lasting ~10 minutes, with no errors in application or load balancer logs and no CPU/memory alerts, point to an external dependency issue. The synchronous calls to the external authentication service are the likely bottleneck: if that service experiences transient latency, every request is blocked, causing the application's response time to spike. Caching authentication tokens and using asynchronous processing (e.g., a queue or background refresh) decouples the application from the external service's variability, eliminating the cascading latency.

Exam trap

Google Cloud often tests the misconception that all latency originates from internal infrastructure (load balancers, autoscalers, or CPU), when the real cause is an external dependency's synchronous call pattern that creates a hidden bottleneck without triggering resource alerts.

How to eliminate wrong answers

Option A is wrong because TCP load balancers do not have connection draining issues that cause intermittent latency spikes; connection draining is a feature for graceful shutdown, not a source of random latency, and switching to a proxy-based load balancer would not fix an external dependency problem. Option B is wrong because the autoscaler's cooldown period affects scaling decisions, not the latency of individual requests; if CPU/memory are not spiking, autoscaling is irrelevant, and a long cooldown would cause slow scaling, not 10-minute latency bursts. Option D is wrong because CPU throttling from burstable machine types would trigger CPU utilization alerts and would not produce latency spikes without CPU or memory alerts; the pattern of random 10-minute spikes with no resource alerts contradicts sustained CPU throttling.

158
MCQmedium

An organization wants to allow their finance team to view billing account cost information but prevent them from making any changes to the billing account. Which IAM role should they grant?

A.Billing Account Admin
B.Project Billing Manager
C.Billing Account User
D.Billing Account Viewer
AnswerD

Billing Account Viewer is the correct role because it provides read-only access to all billing information on the billing account, including cost breakdowns, invoices, and payment history, without allowing any modifications. This aligns with the finance team's need to view billing data and perform analysis while adhering to least-privilege security principles, reducing the risk of unintended changes.

Why this answer

The Billing Account Viewer role grants read-only access to billing account cost information, including viewing budgets, costs, and billing account details, without permission to modify the billing account. This matches the requirement to let the finance team view costs while preventing changes. It is the least-privilege role for read-only billing visibility.

Exam trap

The trap is the near-identical naming of billing roles — candidates confuse Billing Account User (which only links projects) with Billing Account Viewer (which actually reads cost data), and may over-grant by choosing Admin for a read-only requirement.

How to eliminate wrong answers

Option A is wrong because Billing Account Admin grants full control over the billing account, including modifying payment methods, budgets, and IAM policies — far more than read-only. Option B is wrong because Project Billing Manager manages billing for a specific project, not the billing account, and includes write capabilities. Option C is wrong because Billing Account User is intended to associate projects with a billing account (link projects), not to view cost data, and it does not provide the read-only cost visibility the finance team needs.

159
MCQeasy

A startup wants to secure access to its internal web applications without using a VPN. They need to enforce access based on user identity and device security posture. Which Google Cloud service should they use?

A.Identity-Aware Proxy (IAP)
B.Cloud VPN
C.VPC Service Controls
D.Cloud Armor
AnswerA

Identity-Aware Proxy (IAP) is the correct answer because it implements zero-trust access control at the application layer. When a user requests the internal web app, IAP uses OAuth-based authentication to verify their identity and validates contextual signals such as device security status, IP address, and geographic location. It then enforces fine-grained IAM roles (e.g., the 'IAP-secured Web App User' role) to allow or deny each session, providing user-centric access without exposing the entire VPC.

Why this answer

BeyondCorp Enterprise is Google's zero-trust solution that replaces VPNs by using identity and context (device, location) to grant access. Identity-Aware Proxy (IAP) is a component of BeyondCorp that enforces access at the application layer.

160
MCQeasy

Which Google Cloud service provides a serverless data warehouse for running SQL queries on petabyte-scale data with no need to manage infrastructure?

A.Cloud SQL
B.Dataflow
C.BigQuery
D.Pub/Sub
AnswerC

BigQuery is Google Cloud's serverless, fully managed analytics warehouse that runs ANSI SQL over petabyte-scale datasets using its Dremel engine and columnar storage, with no clusters to provision or tune. This directly satisfies the stem's no-infrastructure-management constraint.

Why this answer

BigQuery is a fully managed, serverless data warehouse that supports SQL queries on massive datasets.

161
MCQmedium

A security engineer needs to monitor and analyze security logs from multiple GCP projects and on-premises sources in a centralized SIEM. Which Google Cloud service is designed for log management and security analytics at scale?

A.Cloud Logging
B.Chronicle
C.Security Command Center
D.Cloud IDS
AnswerB

Chronicle is Google Cloud's cloud-native SIEM purpose-built for ingesting massive volumes of security telemetry, normalizing it, and running continuous threat detection. Unlike simpler log tools, it provides automated correlation, an entity graph, and detection rules that enable security teams to hunt for and investigate threats. It integrates with Cloud Logging and other sources, making it the correct choice for security log monitoring and analysis.

Why this answer

Chronicle is Google Cloud's SIEM platform that ingests logs from various sources and provides security analytics. Security Command Center is for vulnerability management, not a SIEM. Cloud Logging is for log storage and basics, but not a full SIEM.

162
MCQeasy

Refer to the exhibit. A user receives this error when trying to copy an object from one bucket to another. What is the most likely cause?

A.The service account used does not have the required IAM permissions to copy objects.
B.The buckets are located in different regions and cross-region copy is not allowed.
C.The destination bucket has exceeded its storage quota.
D.The source bucket name is misspelled in the request.
AnswerA

A 403 Forbidden error in Google Cloud Storage means the request was authenticated but not authorized. To copy an object, the service account must have both storage.objects.get on the source bucket (or object) and storage.objects.create on the destination bucket—roles such as roles/storage.objectViewer plus roles/storage.objectCreator, or the combined roles/storage.objectAdmin, provide these permissions. If the service account lacks them, the copy operation fails with AccessDenied regardless of whether the buckets and objects exist and are reachable. Also verify that no organization policy or IAM deny rule is blocking the specific access.

Why this answer

The error when copying an object between buckets is most likely due to insufficient IAM permissions. In Google Cloud, the service account initiating the copy must have both `storage.objects.get` (to read the source object) and `storage.objects.create` (to write to the destination bucket) permissions. Without these, the operation fails with an access denied error, even if the buckets exist and are accessible.

Exam trap

Google Cloud often tests the misconception that cross-region copy is blocked by default, but in Google Cloud Storage, cross-region copies are allowed as long as IAM permissions are correct, making permissions the primary gatekeeper.

How to eliminate wrong answers

Option B is wrong because cross-region copy is fully supported in Google Cloud Storage; objects can be copied between buckets in different regions without restriction. Option C is wrong because exceeding the storage quota would cause a quota exceeded error, not a permissions-related error, and the error message shown is typical of access issues. Option D is wrong because a misspelled bucket name would result in a 'bucket not found' error (HTTP 404), not a permissions error.

163
Matchingmedium

Match each Google Cloud networking concept to its description.

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

Concepts
Matches

Virtual Private Cloud – isolated network

Content delivery network for low-latency delivery

Distributes traffic across instances

Outbound internet for private instances

Dedicated connection between on-prem and GCP

Why these pairings

The correct matches are: VPC - virtual network, Cloud Load Balancing - traffic distribution, Cloud CDN - content caching, Cloud Interconnect - dedicated physical connection. Common confusions include mixing Cloud CDN with Load Balancing and Cloud Interconnect with Cloud VPN.

164
MCQmedium

A media company needs to deliver video content to users worldwide with low latency. They want to cache content at edge locations. Which Google Cloud service should they use?

A.Cloud CDN
B.Cloud Storage
C.Cloud Load Balancing
D.Cloud Interconnect
AnswerA

Cloud CDN caches content at Google's globally distributed edge points of presence, terminating user requests close to their location. This directly satisfies the worldwide low-latency requirement by serving cached video from the nearest edge rather than the origin, reducing round-trip time.

Why this answer

Cloud CDN uses Google's global edge network to cache content close to users, reducing latency for media delivery.

165
Multi-Selectmedium

A company is building a real-time leaderboard for an online game using Google Cloud. They need a database that can handle millions of updates per second with low latency and serve the current top scores. Which TWO services should they use together? (Choose 2)

Select 2 answers
A.Cloud SQL
B.Firestore
C.Cloud Bigtable
D.Memorystore for Redis
E.BigQuery
AnswersC, D

Cloud Bigtable is a NoSQL wide-column database engineered for high-throughput, low-latency writes of millions of rows per second when scaling with CPU and storage nodes. Its distributed storage engine appends writes to SSTables with memtable buffering, giving consistent single-digit-millisecond latencies and linear write scaling across a cluster without a single bottleneck. For a real-time leaderboard, Bigtable's row-key design (like inverted scores or user IDs) paired with its massive write capacity makes it a proven choice for globally hot update streams.

Why this answer

Cloud Bigtable (C) is correct because it is a fully managed, horizontally scalable NoSQL database designed for high-throughput, low-latency workloads, making it suitable for ingesting millions of score updates per second and durably storing the leaderboard data. Memorystore for Redis (D) is correct because it provides an in-memory data store with sub-millisecond latency, ideal for serving the current top scores in real time and handling rapid ranking operations. Together, Bigtable handles massive write throughput while Redis serves the hot leaderboard reads.

Cloud SQL (A) is not appropriate because it is a relational database that cannot scale to millions of updates per second with low latency. Firestore (B) is not ideal because, while it is a NoSQL document database, it is optimized for mobile/web app synchronization rather than extreme write throughput. BigQuery (E) is not suitable because it is an analytics data warehouse designed for large-scale queries, not real-time transactional updates.

Exam trap

GCDL often tests whether candidates recognize that Bigtable alone cannot serve sorted leaderboard queries efficiently, so answers that omit Redis (or substitute Firestore/BigQuery) fail the low-latency read requirement.

166
Multi-Selecteasy

A company wants to store archival data that is accessed less than once a year and needs the lowest storage cost. Which TWO Cloud Storage classes are most cost-effective for this use case?

Select 2 answers
A.Regional
B.Nearline
C.Standard
D.Archive
E.Coldline
AnswersD, E

Archive is the correct answer because it is Google Cloud's lowest-cost storage class, explicitly intended for data accessed less than once a year. It has the cheapest storage price but charges high retrieval fees and requires a minimum storage duration of 365 days. This matches the described access pattern of archival data, making it the most cost-effective option.

Why this answer

Coldline (E) and Archive (D) are the two lowest-cost Cloud Storage classes and are explicitly designed for data accessed less than once a year, making them the most cost-effective choices for this archival scenario. Coldline is intended for data accessed at most once per quarter and offers very low storage pricing with a 90-day minimum storage duration, while Archive is the cheapest class for data accessed less than once a year, with a 365-day minimum storage duration. Both classes charge higher retrieval and early-deletion fees, which is acceptable here because the data is rarely read.

Regional (A) and Standard (C) are hot-access classes with the highest storage prices, so they are not cost-effective for archival data. Nearline (B) is cheaper than Standard but is optimized for data accessed about once a month, so it costs more than Coldline and Archive for this less-than-annual access pattern.

Exam trap

The trap is confusing Nearline and Coldline thresholds — candidates pick Nearline for 'less than once a year' when Nearline is actually for less than once a month, making it more expensive than needed.

167
MCQeasy

A company uses Cloud Functions and notices that some functions are taking longer than expected. They want to identify which functions have the highest latency. What should they use?

A.Cloud Audit Logs
B.Error Reporting
C.Cloud Monitoring metrics
D.Cloud Logging queries
AnswerC

Cloud Monitoring collects and stores time-series metrics from Cloud Functions, including execution time, invocation count, error count, and memory usage, with built-in dashboards and alerting capabilities. The execution time metric specifically supports calculating service-level objectives and detecting latency anomalies across all invocations. This makes Cloud Monitoring the correct choice for latency analysis and performance monitoring.

Why this answer

Cloud Monitoring metrics, specifically the 'execution_time' metric for Cloud Functions, provide the precise latency data needed to identify functions with the highest execution duration. Unlike logs or error reports, metrics are designed for numerical aggregation and can be used to create dashboards or alerts that rank functions by their p50, p95, or p99 latency values.

Exam trap

Google Cloud often tests the distinction between logs (Cloud Logging) and metrics (Cloud Monitoring), trapping candidates who think that because latency data appears in logs, querying logs is the correct method, when in fact metrics are the proper tool for numerical aggregation and ranking.

How to eliminate wrong answers

Option A is wrong because Cloud Audit Logs record administrative actions and access to resources, not the execution duration of individual function invocations. Option B is wrong because Error Reporting is designed to capture and analyze exceptions and errors, not to measure performance metrics like latency. Option D is wrong because Cloud Logging queries can retrieve individual log entries that may contain execution times, but they are not optimized for aggregating and ranking latency across many functions; Cloud Monitoring metrics are purpose-built for this numerical analysis.

168
MCQmedium

You are trying to create a Compute Engine VM instance and receive the error: "The resource 'projects/my-project' was not found." What is the most likely cause of the error?

A.The machine type is invalid
B.The zone is incorrect
C.The user lacks compute.instances.create permission
D.The project does not exist
AnswerD

When creating an instance, the Compute Engine API receives the project ID as part of the request path. If the project ID is mistyped, deleted, or never existed, the API returns a 404 error stating the project was not found. This error typically appears as 'The project X was not found' or 'Project not found', which matches the scenario described in the question.

Why this answer

The error message "The resource 'projects/my-project' was not found" indicates that the specified project does not exist or the project ID is incorrect. The most likely cause is that the project does not exist.

Exam trap

Google Cloud often tests the order of API validation steps — candidates mistakenly think permission errors come first, but Google Cloud validates project existence before checking IAM permissions.

How to eliminate wrong answers

Option A is wrong because an invalid machine type would produce a different error, such as 'Invalid machine type: [type]', not a project-level error. Option B is wrong because an incorrect zone would result in a 'zone does not exist' or 'zone not found' error, which is distinct from a project error. Option C is wrong because lacking compute.instances.create permission would return a 'Permission denied' or 'Forbidden' error, not a project not found error.

169
MCQmedium

A media company currently licenses proprietary software for video editing that costs $50,000 per seat annually. They are considering a cloud-based SaaS alternative at $5,000 per seat annually. Beyond the licensing cost, which additional financial benefits should they consider when calculating total cost of ownership (TCO)?

A.Only the licensing cost difference ($45,000 per seat) matters for the financial comparison.
B.Eliminated hardware costs, reduced IT maintenance staff, no upgrade cycles, and freed facilities costs — all lowering the true on-premises TCO that should be compared against the SaaS subscription.
C.The SaaS option has an internet dependency risk that may cost more than the savings.
D.The vendor's market capitalization, since larger companies are more financially stable.
AnswerB

The correct TCO comparison must account for the full set of costs eliminated by moving to SaaS: no upfront hardware procurement or refresh cycles, reduced IT maintenance staff for patching and monitoring, no annual software upgrade projects, and freed data center space for power, cooling, and cabling. When these hidden on-premises costs are quantified and subtracted, the SaaS subscription fee often becomes comparable or even lower over a 3–5 year horizon. This holistic view is exactly what financial decision-makers miss when they only compare license fees, and it is the reason the 'SaaS vs. on-prem' TCO analysis is so favorable to cloud.

Why this answer

The total cost of ownership (TCO) for on-premises software includes not just the licensing fee but also hardware acquisition, IT staff for maintenance, periodic upgrade costs, and physical facility expenses. By moving to a SaaS model, the company eliminates these variable costs, making the $5,000 per seat subscription a more accurate comparison against the fully-loaded on-premises TCO, which often exceeds the $50,000 license alone.

Exam trap

The GCDL exam often tests the misconception that only direct licensing costs matter, ignoring the broader TCO components like hardware, staff, and facilities that make on-premises solutions more expensive than they appear.

How to eliminate wrong answers

Option A is wrong because it ignores the hidden costs of on-premises infrastructure (hardware, maintenance, upgrades, facilities) that are part of the true TCO, leading to an incomplete financial comparison. Option C is wrong because while internet dependency is a risk, it is not a direct financial benefit of the SaaS option; the question asks for additional financial benefits, not risks or drawbacks. Option D is wrong because vendor market capitalization is a measure of financial stability, not a direct cost or benefit in TCO calculation; it does not affect the per-seat cost comparison.

170
MCQmedium

A retail company wants to build a recommendation engine that suggests products to customers based on their browsing history. The team has ML expertise but wants to use Google's pre-built ML infrastructure to train and deploy models at scale without managing compute resources. Which Google Cloud service should they use?

A.BigQuery ML
B.Vertex AI
C.Cloud AI Platform Notebooks (now Vertex AI Workbench)
D.Cloud Dataflow
AnswerB

Vertex AI is Google Cloud's unified machine learning platform that delivers exactly what the team needs: managed training on GPU/TPU clusters, AutoML for tabular/image/text, hyperparameter tuning, a centralized Model Registry, and Vertex AI Endpoints for autoscaling predictions. It also integrates with Vertex AI Pipelines for orchestration and Vertex AI Feature Store for consistent features. The team can bring their existing Python/TensorFlow expertise, and Vertex AI abstracts away the infrastructure management, letting them focus on building and iterating on models.

Why this answer

Vertex AI is the correct choice because it provides a fully managed, unified ML platform that handles the entire ML workflow—from data preparation and training to deployment and monitoring—without requiring the team to manage underlying compute infrastructure. It integrates with Google Cloud's pre-built ML infrastructure, including distributed training, AutoML, and custom model serving, making it ideal for building and scaling a recommendation engine.

Exam trap

The GCDL exam often tests the distinction between a managed ML platform (Vertex AI) and individual tools like BigQuery ML or Dataflow, trapping candidates who confuse data processing or SQL-based ML with end-to-end model deployment and infrastructure management.

How to eliminate wrong answers

Option A is wrong because BigQuery ML is designed for creating and executing machine learning models directly in BigQuery using SQL, which is suitable for simple, in-database ML but lacks the flexibility and infrastructure for custom model training, deployment, and scaling needed for a recommendation engine. Option C is wrong because Cloud AI Platform Notebooks (now Vertex AI Workbench) is a tool for creating and managing Jupyter notebooks for exploratory data analysis and model development, not a managed service for training and deploying models at scale without managing compute resources. Option D is wrong because Cloud Dataflow is a fully managed service for stream and batch data processing, not an ML platform; it can be used for data preprocessing but does not provide model training or deployment capabilities.

171
MCQmedium

A DevOps team wants to automatically build a Docker image from a GitHub repository and store it in a private registry whenever a new tag is pushed. Which Google Cloud services should they combine?

A.Cloud Build and Container Registry
B.Cloud Source Repositories and Container Registry
C.Cloud Functions and Artifact Registry
D.Cloud Build and Artifact Registry
AnswerD

Cloud Build natively supports build triggers tied to GitHub events, including tag push events, making it ideal for automatically building a Docker image when a tag is pushed. The built image can then be securely pushed to Artifact Registry, which is the fully managed, regional container registry with IAM integration, vulnerability scanning, and support for Docker and OCI artifacts. This combination is the recommended, maintainable CI/CD approach for tag-driven image builds.

Why this answer

Cloud Build is Google Cloud's fully managed CI/CD service that can be triggered by GitHub tag pushes via Cloud Build triggers. Artifact Registry is the modern, recommended private registry for storing Docker images (and other artifacts), replacing the older Container Registry. Together, they provide an automated build-and-store pipeline: Cloud Build builds the image from the GitHub repo, and Artifact Registry stores it privately.

Exam trap

GCDL often tests the distinction between Container Registry (legacy) and Artifact Registry (current), and candidates may pick Container Registry out of familiarity, missing that Artifact Registry is the recommended service for new deployments.

How to eliminate wrong answers

Option A is wrong because while Cloud Build and Container Registry can work together, Container Registry is deprecated and lacks the advanced features (e.g., multi-format support, fine-grained IAM) of Artifact Registry; the question asks for the best combination, and Artifact Registry is the current standard. Option B is wrong because Cloud Source Repositories is a private Git repository service, not a build service; it cannot automatically build Docker images from GitHub. Option C is wrong because Cloud Functions is a serverless compute service for event-driven functions, not for building Docker images; it cannot perform Docker builds.

172
Drag & Dropmedium

Drag and drop the steps to set up a Cloud NAT for private Compute Engine instances to access the internet in the correct order.

Drag or tap steps into the slots.

Steps
Order
1Step 1
2Step 2
3Step 3
4Step 4

Why this order

The setup requires a VPC, Cloud Router, NAT gateway, appropriate firewall rules, and verification.

173
MCQmedium

A software team deploys microservices using containers and wants Google Cloud to automatically manage the scaling, self-healing, and rollout of their containerized applications. They don't want to provision or manage underlying virtual machines. Which Google Cloud service best meets this need?

A.Compute Engine, Google Cloud's virtual machine service
B.Google Kubernetes Engine (GKE), which manages containerized applications with automatic scaling, self-healing, and rolling updates
C.Cloud Storage, for storing container images and serving them to instances
D.Cloud CDN, which distributes containerized applications globally
AnswerB

Google Kubernetes Engine (GKE) is a managed Kubernetes service that automates container orchestration, including deployment, scaling, and operations. It provides managed node pools with cluster autoscaling, health checks that automatically reschedule unhealthy pods, and rolling updates with configurable surge and unavailable budgets. Teams only define the desired application state via Deployments and HorizontalPodAutoscaler resources, and GKE continually reconciles the actual cluster state to match, fulfilling all requirements for automatic scaling, self-healing, and rolling updates.

Why this answer

Google Kubernetes Engine (GKE) is the correct choice because it provides a managed Kubernetes environment that automates scaling, self-healing (e.g., automatic pod restarts and node repair), and rollout management (e.g., rolling updates and rollbacks) for containerized applications. This meets the requirement of not provisioning or managing underlying virtual machines, as GKE abstracts the node infrastructure.

Exam trap

The trap here is that candidates may confuse Compute Engine (IaaS) with a managed container service, or assume Cloud Storage or Cloud CDN can orchestrate containers, when only GKE provides the automated lifecycle management described.

How to eliminate wrong answers

Option A is wrong because Compute Engine is an Infrastructure-as-a-Service (IaaS) offering that requires manual provisioning and management of virtual machines, not automatic scaling, self-healing, or rollout of containers. Option C is wrong because Cloud Storage is an object storage service for storing container images (e.g., via Artifact Registry integration), but it does not manage scaling, self-healing, or rollout of applications. Option D is wrong because Cloud CDN is a content delivery network that caches and distributes content globally, not a compute or orchestration service for managing containerized workloads.

174
MCQmedium

A company wants to protect its web application running on Google Cloud from DDoS attacks and SQL injection. Which service should they use?

A.reCAPTCHA Enterprise
B.Cloud Armor
C.Cloud CDN
D.VPC firewall rules
AnswerB

Cloud Armor is Google Cloud's DDoS protection and web application firewall (WAF) service. It enforces L7 filtering rules—such as the OWASP ModSecurity Core Rule Set for SQLi and XSS, rate limiting, and custom expressions—at the edge of Google's global load balancing and Cloud CDN infrastructure. In addition, it provides always-on network-layer DDoS mitigation, making it the correct answer for protecting a GKE-hosted web app. It can be configured with preconfigured rules or tailored to block specific attack signatures.

Why this answer

Google Cloud Armor is the edge security service that provides DDoS protection and WAF capabilities, including preconfigured rules for SQL injection (SQLi) and cross-site scripting (XSS). It attaches to HTTP(S) load balancers to filter malicious traffic before it reaches the application.

Exam trap

GCDL often tests the misconception that Cloud CDN or VPC firewall rules provide WAF/DDoS protection; only Cloud Armor delivers both DDoS mitigation and SQLi/XSS filtering at the edge.

How to eliminate wrong answers

Option A is wrong because reCAPTCHA Enterprise protects against bots and automated abuse, not network-layer DDoS or SQL injection at the WAF level. Option C is wrong because Cloud CDN caches and accelerates content; it does not provide WAF or DDoS mitigation by itself. Option D is wrong because VPC firewall rules operate at L3/L4 and cannot inspect HTTP payloads to block SQL injection.

175
MCQeasy

A business leader asks: 'What is the difference between a data center region and an availability zone in Google Cloud?' Which explanation is most accurate?

A.A region is a single data center building; an availability zone is a floor within that building
B.A region is a geographic area containing multiple independent zones; each zone is a distinct facility with independent power, cooling, and networking — failures in one zone don't affect other zones in the region
C.A region is equivalent to an availability zone; both terms refer to a single data center
D.An availability zone is larger than a region and spans multiple geographic areas for global redundancy
AnswerB

This is the correct definition. GCP regions (e.g., us-central1) contain multiple zones (a, b, c, d). Each zone is independent infrastructure. A zonal failure (power outage, cooling failure) doesn't propagate to other zones. Customers deploy across multiple zones within a region for high availability against zone-level failures.

Why this answer

In Google Cloud, a region is a specific geographic location composed of multiple zones, where each zone is an independent data center with its own power, cooling, and networking infrastructure. This isolation ensures that failures within one zone do not impact other zones in the same region, providing high availability and fault tolerance for applications.

Exam trap

The trap here is confusing the hierarchical relationship between regions and zones, leading candidates to incorrectly think a region is a single data center or that zones are larger than regions, which Google Cloud often tests by presenting false equivalencies or reversed sizes.

How to eliminate wrong answers

Option A is wrong because a region is not a single data center building; it is a collection of multiple independent zones, each of which is a separate facility. Option C is wrong because a region and an availability zone are not equivalent; a region contains multiple zones, and each zone is a distinct data center within that region. Option D is wrong because an availability zone is smaller than a region and does not span multiple geographic areas; it is confined to a single region, while regions themselves are separate geographic areas.

176
MCQmedium

A business intelligence team wants to create interactive dashboards and reports from their BigQuery data without writing code. They need to share reports with stakeholders who don't have GCP accounts. Which Google Cloud tool is most appropriate?

A.Vertex AI Workbench
B.Looker Studio (formerly Data Studio)
C.Cloud Dataprep by Trifacta
D.BigQuery Studio
AnswerB

Looker Studio (formerly Data Studio) is Google's free business intelligence and data visualization platform with first-class native BigQuery connectors, enabling analysts to build interactive, shareable dashboards without writing web application code. Reports are stored in Google Cloud and can be shared via a simple link with view-only access, so stakeholders do not need a GCP account, a BigQuery project, or any technical skills. The tool supports scheduled data refresh, interactive filters, and drag-and-drop charting, making it the ideal choice for non-technical consumers. This directly matches the requirement for a BI dashboarding tool with integrated BigQuery connectivity.

Why this answer

Looker Studio (formerly Data Studio) is the correct choice because it is a no-code, drag-and-drop business intelligence tool that connects directly to BigQuery, enabling the creation of interactive dashboards and reports. It also supports sharing reports via public links or embedded views, allowing stakeholders without GCP accounts to access them without needing IAM permissions.

Exam trap

The GCDL exam often tests the distinction between data preparation tools (Cloud Dataprep), ML development environments (Vertex AI Workbench), and BI/reporting tools (Looker Studio), trapping candidates who confuse 'no-code dashboards' with SQL-focused interfaces like BigQuery Studio.

How to eliminate wrong answers

Option A is wrong because Vertex AI Workbench is a Jupyter-based environment for building and deploying machine learning models, not for creating no-code dashboards or sharing reports with external stakeholders. Option C is wrong because Cloud Dataprep by Trifacta is a data preparation and cleaning tool, not a dashboarding or reporting solution; it focuses on transforming raw data rather than visualizing it interactively. Option D is wrong because BigQuery Studio is a unified interface for writing SQL queries and exploring data within BigQuery, but it does not provide native no-code dashboard creation or the ability to share reports with users who lack GCP accounts.

177
MCQhard

A company needs to run containerized workloads but wants to avoid managing Kubernetes cluster infrastructure (node pools, cluster upgrades, node autoscaling) entirely. They want to simply deploy container images and have Google Cloud manage all underlying infrastructure automatically. Which Google Cloud product best fits this fully managed container execution requirement?

A.Google Kubernetes Engine (GKE) Standard mode, where the team manages node pools and cluster configuration
B.Cloud Run, which executes container images on fully managed serverless infrastructure with automatic scaling and no cluster or node management
C.Compute Engine with a startup script that pulls and runs the container image from Artifact Registry
D.App Engine Flexible, which runs containers on managed virtual machines
AnswerB

Cloud Run is the answer. Developers provide a container image; Cloud Run handles all infrastructure: no node pools, no cluster upgrades, no capacity planning. It scales automatically from zero to thousands of instances based on request traffic. It's the most infrastructure-free container execution option on Google Cloud.

Why this answer

Cloud Run is the correct choice because it provides a fully managed serverless platform that executes container images without any cluster or node management. It automatically handles scaling, infrastructure provisioning, and maintenance, aligning perfectly with the requirement to avoid managing Kubernetes cluster infrastructure entirely.

Exam trap

The trap here is that candidates may confuse App Engine Flexible with a fully managed container service, but it still requires managing VM instances and does not provide the same serverless abstraction as Cloud Run.

How to eliminate wrong answers

Option A is wrong because GKE Standard mode requires the team to manage node pools, cluster upgrades, and node autoscaling, which contradicts the requirement to avoid managing Kubernetes cluster infrastructure. Option C is wrong because Compute Engine with a startup script still requires managing virtual machine instances, including patching, scaling, and lifecycle, which is not fully managed container execution. Option D is wrong because App Engine Flexible runs containers on managed virtual machines but still requires managing the underlying VM instances and does not provide the same level of automatic scaling and infrastructure abstraction as Cloud Run.

178
MCQhard

A company is migrating a legacy monolithic application to Google Cloud. They want to reduce operational overhead while maintaining the ability to run the application without rewriting code. Which migration strategy should they use?

A.Retiring
B.Replatforming (lift, tinker, and shift)
C.Refactoring (rearchitecting)
D.Lift and shift (rehosting)
AnswerD

Lift and shift (rehosting) involves copying the application and its data to cloud virtual machines, such as Google Compute Engine, with minimal or no modifications, preserving the current architecture and behavior. This approach reduces capital expenses and operational overhead associated with on-premises infrastructure while avoiding the risks and costs of code changes, making it the correct choice for this migration scenario.

Why this answer

Lift and shift (rehosting) moves the application to Google Cloud with minimal or no code changes, which directly satisfies the requirement to avoid rewriting code while reducing the operational burden of running on-premises hardware. It is the fastest migration strategy and preserves the existing monolithic architecture.

Exam trap

The trap is reading 'reduce operational overhead' and jumping to replatforming or refactoring — but the binding constraint is 'without rewriting code,' which forces the answer to lift and shift (rehosting).

How to eliminate wrong answers

Option A is wrong because retiring means decommissioning the application entirely — the company wants to keep running it, not remove it. Option B is wrong because replatforming (lift, tinker, and shift) involves making some modifications to the application or its dependencies (e.g., swapping a self-managed database for a managed service), which conflicts with the 'without rewriting code' constraint. Option C is wrong because refactoring/rearchitecting requires significant code changes to adopt cloud-native patterns like microservices, which is explicitly what the company wants to avoid.

179
MCQmedium

What is the key difference between a virtual machine (VM) and a container in terms of how they package and run applications?

A.VMs run on physical hardware; containers run in the cloud.
B.VMs include a full guest OS; containers share the host OS kernel and contain only the application and its dependencies.
C.Containers are less secure than VMs because they share hardware.
D.VMs are only for Linux applications; containers support all operating systems.
AnswerB

A VM bundles a complete guest OS—kernel, system libraries, and user-space binaries—under a hypervisor, making it gigabytes in size and relatively slow to boot. A container, by contrast, is an isolated user-space process that reuses the host kernel and carries only the application plus its dependencies, so it starts in seconds and remains lightweight. This fundamental packaging difference drives both resource efficiency and workload portability.

Why this answer

The key difference is that a virtual machine (VM) includes a full guest operating system (OS) running on top of a hypervisor, which virtualizes the underlying hardware. In contrast, a container packages only the application and its dependencies (libraries, binaries, configuration files) and shares the host OS kernel via the container runtime (e.g., Docker, containerd). This makes containers lightweight and faster to start, as they avoid the overhead of a separate OS instance.

Exam trap

Google Cloud often tests the misconception that containers are simply 'lightweight VMs' or that the difference is about location (cloud vs. on-prem), when the actual distinction is the presence or absence of a guest OS and kernel sharing.

How to eliminate wrong answers

Option A is wrong because VMs do not inherently run on physical hardware; they run on a hypervisor that abstracts the physical hardware, and containers can run on physical hardware, in the cloud, or on-premises — location is not the defining difference. Option C is wrong because containers share the host OS kernel, not the hardware directly, and security depends on isolation mechanisms (e.g., cgroups, namespaces, seccomp); a misconfigured container can be less secure, but VMs provide stronger isolation via a separate kernel, so the blanket statement 'containers are less secure because they share hardware' is technically inaccurate. Option D is wrong because VMs support multiple operating systems (Windows, Linux, etc.) and containers primarily run on Linux or Windows with kernel support; containers do not support all operating systems — for example, a Linux container cannot run natively on a Windows host without a Linux VM.

180
MCQhard

A CISO is evaluating Google Cloud's security posture and asks about independent third-party validation of Google's security practices. Which types of certifications and audit reports most directly provide this independent assurance?

A.Google's internal security policies and self-assessment reports published on its website
B.Third-party audit certifications such as ISO 27001, SOC 2 Type II, PCI DSS, and FedRAMP, which independently verify that Google's security controls meet defined international and industry standards
C.Google's Bug Bounty program, which shows that the public can report security vulnerabilities
D.Customer testimonials from large enterprises that use Google Cloud for sensitive workloads
AnswerB

These certifications are the gold standard for independent assurance. ISO 27001 and SOC 2 Type II involve rigorous independent audits. PCI DSS is required for payment data handling. FedRAMP provides US government-validated cloud security. A CISO can review these certifications as credible evidence that Google's security controls have been independently verified.

Why this answer

Independent third-party validation of Google Cloud's security posture is most directly provided by certifications and audit reports such as ISO 27001, SOC 2 Type II, PCI DSS, and FedRAMP. These are issued by accredited external auditors who verify that Google's security controls, processes, and infrastructure meet rigorous, internationally recognized standards. This gives customers objective assurance beyond Google's own claims.

Exam trap

Google Cloud often tests the distinction between internal self-assessments or informal programs (like bug bounties or testimonials) and formal, independent third-party audit certifications that provide legally defensible assurance of security controls.

How to eliminate wrong answers

Option A is wrong because internal security policies and self-assessment reports are not independent; they lack external verification and are considered first-party attestations, not third-party validation. Option C is wrong because the Bug Bounty program is a vulnerability disclosure mechanism that encourages ethical hacking, but it does not provide a systematic, audited certification of overall security controls or compliance with standards like ISO 27001 or SOC 2. Option D is wrong because customer testimonials, while valuable for reputation, are anecdotal and not a formal, audited certification; they do not constitute independent third-party validation of security practices.

181
MCQeasy

What is the primary purpose of a folder in the Google Cloud resource hierarchy?

A.To manage user identities
B.To define network topologies
C.To store billing information
D.To group projects and apply policies
AnswerD

Folders are the primary mechanism for grouping projects under an organization, enabling you to apply IAM policies, organization policies, and resource hierarchy rules at a sub-organization level. This design supports delegated management and policy inheritance, where permissions assigned to a folder flow down to all projects and resources inside it. A folder acts as a container that gives you granular control without needing to manage each project individually.

Why this answer

Folders in the Google Cloud resource hierarchy sit between the organization node and projects, letting you group related projects and attach IAM policies and organization policies that are inherited by everything beneath them. This makes folders the primary tool for delegated administration and policy scoping across project boundaries.

Exam trap

The trap is that all four options describe things that exist in Google Cloud (identities, networks, billing), so candidates pick based on keyword familiarity rather than understanding that only folders provide hierarchical grouping and policy inheritance.

How to eliminate wrong answers

Option A is wrong because user identities are managed by Cloud Identity or Google Workspace, not by folders — folders only reference identities in IAM bindings. Option B is wrong because network topologies are defined by VPC networks, subnets, and firewall rules, which are project-level resources, not hierarchy constructs. Option C is wrong because billing information is associated with the Cloud Billing account linked to projects, not stored in folders.

182
Multi-Selectmedium

Which TWO of the following are key benefits of cloud technology that are transforming business operations?

Select 2 answers
A.Physical security of data centers
B.Elasticity to automatically scale resources up and down
C.Access to managed services that reduce operational overhead
D.Data locality to keep data within national borders
E.Ability to purchase reserved instances for predictable workloads
AnswersB, C

Elasticity is a defining benefit of cloud computing because it enables resources to automatically scale up or down in response to real-time demand. This dynamic provisioning allows businesses to handle traffic spikes without manual intervention and avoid the cost of over-provisioning for peak loads, while also ensuring consistent performance during surges. The pay-per-use model ties directly to elasticity, meaning you only pay for what you consume, which is a fundamental shift from capacity planning with fixed on-premises infrastructure.

Why this answer

Cloud elasticity allows resources to automatically scale up or down based on demand, eliminating over-provisioning and underutilization. This dynamic adjustment is a key transformation driver, enabling businesses to handle variable workloads efficiently without manual intervention.

Exam trap

Google Cloud often tests the distinction between 'elasticity' (dynamic, automatic scaling) and 'reserved instances' (a pricing model for predictable capacity), leading candidates to mistakenly select reserved instances as a transformative benefit when it is merely a cost optimization tactic.

183
MCQeasy

An organization needs to apply security policies to protect their web application from DDoS attacks and SQL injection. Which Google Cloud service should they use?

A.Cloud NAT
B.Cloud Load Balancing
C.Cloud Armor
D.Cloud CDN
AnswerC

Cloud Armor is Google Cloud's web application firewall (WAF) and DDoS protection service, which integrates with Cloud Load Balancing to protect services at the edge. It enables fine-grained security policies based on IP addresses, geographic location, and preconfigured or custom rules that filter OWASP Top 10 threats like SQL injection and cross-site scripting. Cloud Armor also mitigates volumetric DDoS attacks with adaptive protection and scale, making it the correct service for applying security policies.

Why this answer

Cloud Armor is a web application firewall (WAF) that provides DDoS protection and security rules to block threats like SQL injection. Cloud CDN is for content caching. Load Balancing distributes traffic.

Cloud NAT is for outbound connectivity.

184
MCQeasy

A developer needs to estimate the monthly cost of running a set of Compute Engine instances with specific machine types, persistent disks, and network egress before deploying. Which tool should they use?

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

The Google Cloud Pricing Calculator is the authoritative pre-deployment estimation tool, letting users select specific services (e.g., Compute Engine machine types, Cloud Storage classes) and input usage parameters like region, sustained use, and committed use discounts to generate a monthly cost projection. It pulls current, publicly listed rates and supports SKU-level customization, making it the correct choice for forecasting costs before any resource is created or consumed.

Why this answer

The Google Cloud Pricing Calculator is the tool designed to estimate costs for specific resources like Compute Engine instances, persistent disks, and network egress before deployment. It lets you configure machine types, disk sizes, and egress volumes to produce a monthly cost estimate. This is exactly the pre-deployment estimation use case described.

Exam trap

GCDL often tests the difference between forward-looking estimation (Pricing Calculator) and retrospective cost analysis (Billing export, Cost Management dashboard) — candidates may pick a reporting tool when the question asks for pre-deployment estimation.

How to eliminate wrong answers

Option B is wrong because Billing export to BigQuery is for analyzing historical actual costs, not for pre-deployment estimation. Option C is wrong because Active Assist Recommendations provides optimization suggestions based on existing usage, not forward-looking cost estimates. Option D is wrong because the Cost Management dashboard shows actual incurred costs and trends, not estimates for planned resources.

185
MCQmedium

A company wants to replace its VPN-based remote access with a solution that grants access to internal web applications based on user identity and device context, without requiring a VPN. Which Google Cloud service should they use?

A.Access Transparency
B.Cloud Identity
C.Identity-Aware Proxy (IAP)
D.Cloud VPN
AnswerC

Identity-Aware Proxy (IAP) sits in front of HTTPS applications and evaluates each connection using Google credentials and additional context like IP address, geolocation, and device security posture before granting access. It enforces fine-grained role-based policies per resource, eliminating the need for a full VPN for remote users. IAP also supports TCP forwarding for SSH/RDP, so it can directly replace VPN for administrative access to VMs.

Why this answer

Identity-Aware Proxy (IAP) is the Google Cloud service that provides identity and context-aware access to internal web applications without a VPN. It verifies user identity and device context before granting access, making it the correct choice. Other options do not provide this functionality.

Exam trap

GCDL often tests the difference between IAP and other access management services, and candidates may confuse IAP with Cloud Identity or think a VPN is still required.

How to eliminate wrong answers

Option A is wrong because Access Transparency provides logs about Google personnel access, not user access control. Option B is wrong because Cloud Identity is for identity management, not application access. Option D is wrong because Cloud VPN still requires a VPN connection, which the company wants to eliminate.

186
Multi-Selecthard

A company runs a containerized microservices application on Google Kubernetes Engine (GKE). They want to expose a set of services externally with a single IP address, implement SSL termination, and protect against DDoS attacks. Which THREE Google Cloud services should they use together?

Select 3 answers
A.Cloud Armor
B.Cloud NAT
C.Cloud Load Balancing
D.Cloud VPN
E.Cloud CDN
AnswersA, C, E

Cloud Armor is Google Cloud's distributed denial-of-service (DDoS) protection and web application firewall (WAF) service. It attaches to an HTTP(S) Load Balancer and filters incoming traffic based on preconfigured or custom rules, mitigating OWASP Top 10 threats, IP-based blocklists, and rate-limiting before requests hit backend services. This makes it the decisive security layer for a containerized microservices application exposed to the internet, blocking malicious traffic at the edge.

Why this answer

Cloud Load Balancing (C) is correct because an external Application Load Balancer (HTTP(S)) provides a single global anycast IP address, routes traffic to GKE services via Ingress/BackendConfig, and performs SSL/TLS termination using Google-managed or self-managed certificates. Cloud Armor (A) is correct because it attaches security policies to the load balancer's backend service to filter and mitigate DDoS and Layer 7 attacks (e.g., via preconfigured WAF rules and rate limiting). Cloud CDN (E) is correct because enabling it on the load balancer's backend service caches content at Google's edge points of presence, absorbing traffic and adding another layer of DDoS resilience while reducing origin load.

Cloud NAT (B) is not appropriate because it provides outbound internet access for private instances, not inbound external exposure or SSL termination. Cloud VPN (D) is not appropriate because it only establishes encrypted tunnels between networks and does not provide public load balancing, SSL termination, or DDoS protection.

Exam trap

GCDL often tests the misconception that Cloud NAT or Cloud VPN provides inbound protection or external exposure; candidates must recognize that only Cloud Load Balancing + Cloud Armor + Cloud CDN form the ingress/edge security stack.

187
MCQeasy

A telecommunications company wants to launch new 5G services faster than its competitors. Which cloud characteristic most directly accelerates its ability to bring new services to market quickly?

A.On-demand provisioning that allows infrastructure to be deployed in minutes rather than waiting months for hardware procurement and installation
B.The ability to store large amounts of customer call records in the cloud at lower cost
C.Cloud providers' global data center network ensures low latency for all customer calls
D.Managed cloud databases that eliminate the need for database administrators
AnswerA

On-demand self-service is the cloud characteristic that most directly removes the hardware procurement bottleneck. By provisioning infrastructure in minutes, telecoms can test, iterate, and launch new services at a pace impossible with traditional infrastructure cycles.

Why this answer

On-demand provisioning allows the company to spin up virtual servers, networks, and storage in minutes via APIs, eliminating the months-long lead time required for traditional hardware procurement and installation. This directly reduces the time-to-market for new 5G services, as infrastructure can be scaled and configured on the fly to support rapid deployment and testing.

Exam trap

Google Cloud often tests the distinction between operational benefits (cost, latency, management) and the specific agility benefit of rapid provisioning, leading candidates to confuse a general cloud advantage with the one that directly accelerates time-to-market.

How to eliminate wrong answers

Option B is wrong because storing call records at lower cost addresses data retention and compliance, not the speed of launching new services. Option C is wrong because low latency for customer calls is a performance benefit for existing services, not a factor that accelerates the initial deployment of new services. Option D is wrong because managed databases reduce administrative overhead but do not directly impact the speed of provisioning the core infrastructure needed to launch a new service.

188
MCQmedium

A company's security team wants to detect and remediate public exposure of Cloud SQL instances. Which service should they use?

A.Cloud Armor
B.Security Command Center
C.Cloud Data Loss Prevention (DLP)
D.VPC Service Controls
AnswerB

Security Command Center is Google Cloud's security and risk management platform that continuously monitors resources for vulnerabilities, threats, and misconfigurations. It includes built-in detectors that flag publicly exposed Cloud SQL instances by checking assigned IP addresses and IAM policies against best practices. This makes it the correct choice for identifying and remediating the public exposure described in the scenario.

Why this answer

Security Command Center (SCC) is the correct service because it provides centralized visibility and monitoring of Google Cloud resources, including the ability to detect and alert on public exposure of Cloud SQL instances. SCC's built-in vulnerability and threat detection findings, such as 'Public SQL instance,' directly identify misconfigured Cloud SQL instances that are accessible from the internet, enabling the security team to remediate the exposure.

Exam trap

The trap here is that candidates often confuse services that enforce security (like VPC Service Controls or Cloud Armor) with services that detect and alert on misconfigurations, leading them to pick a tool that blocks or filters traffic rather than one that provides visibility and detection.

How to eliminate wrong answers

Option A is wrong because Cloud Armor is a web application firewall (WAF) and DDoS protection service that operates at the edge of Google Cloud, protecting HTTP(S) load-balanced applications, not detecting or remediating public exposure of Cloud SQL instances. Option C is wrong because Cloud Data Loss Prevention (DLP) is a service for inspecting, classifying, and de-identifying sensitive data within content, not for detecting network-level exposure of Cloud SQL instances. Option D is wrong because VPC Service Controls is a security perimeter service that prevents data exfiltration from managed services by defining perimeters around VPC networks, but it does not actively detect or alert on public exposure of Cloud SQL instances; it enforces access boundaries but does not provide visibility into existing public configurations.

189
MCQeasy

A company needs to estimate the monthly cost of running a set of Compute Engine instances, including network egress and Cloud Storage usage, before deploying the architecture. Which tool should they use?

A.Active Assist recommendations
B.Google Cloud Pricing Calculator
C.Budget alerts
D.Cost Management dashboard
AnswerB

The Google Cloud Pricing Calculator is the correct tool for estimating monthly costs before deployment because it accepts user-defined resource specifications, including machine type, region, persistent disk size, estimated usage hours, and committed use discount terms. It references Google Cloud's current public SKU pricing to generate a detailed, line-item cost estimate for a proposed workload, enabling architects to model multi-component architectures and assess the financial impact of configuration choices before committing to them.

Why this answer

The Google Cloud Pricing Calculator is a web-based estimation tool that lets you model Compute Engine instances, Cloud Storage, and network egress charges before any resources are provisioned. It produces a shareable, itemized monthly cost estimate based on SKU-level pricing, which is exactly what pre-deployment planning requires. It is the only option designed for forward-looking cost estimation rather than monitoring actual spend.

Exam trap

GCDL often tests the distinction between pre-deployment cost estimation (Pricing Calculator) and post-deployment cost visibility (Cost Management dashboard, budgets, and alerts), so candidates who see 'cost' and reflexively pick a monitoring tool get it wrong.

How to eliminate wrong answers

Option A is wrong because Active Assist recommendations analyze already-deployed resources to surface optimization opportunities (idle VMs, unattached disks), not to estimate costs for a future architecture. Option C is wrong because Budget alerts only notify you when actual or forecasted spend crosses a threshold on an existing billing account — they cannot model a hypothetical deployment. Option D is wrong because the Cost Management dashboard (Cloud Billing reports) visualizes historical and current incurred costs, not pre-deployment estimates.

190
MCQmedium

An organisation is migrating its on-premises Oracle database to Google Cloud. They need a fully managed, PostgreSQL-compatible database with high performance for transaction processing and built-in AI capabilities for predictive analytics. Which database service should they choose?

A.Bare Metal Solution for Oracle
B.Cloud SQL for PostgreSQL
C.Cloud Spanner
D.AlloyDB
AnswerD

AlloyDB is a fully managed, PostgreSQL-compatible database service architected for high performance, delivering up to 4x faster transactional throughput and up to 100x faster analytical queries compared to standard PostgreSQL. It includes AlloyDB AI with vector similarity search, predictive autoscaling, and columnar engine, making it ideal for migrating Oracle workloads while adding artificial intelligence features. This aligns exactly with the requirement for a managed, compatible database with built-in AI.

Why this answer

AlloyDB is a fully managed PostgreSQL-compatible database that offers high performance (4x faster than standard PostgreSQL) and integrated AI capabilities for vector search and predictive analytics. Cloud SQL for PostgreSQL is also managed but lacks the AI optimisations. Cloud Spanner is globally distributed but not PostgreSQL-compatible.

Bare Metal Solution runs Oracle on dedicated hardware, not managed.

191
Multi-Selectmedium

A company wants to implement a zero-trust security model for accessing internal applications. Which TWO Google Cloud services should they use together? (Choose 2)

Select 2 answers
A.Cloud Identity
B.Cloud Armor
C.Identity-Aware Proxy (IAP)
D.BeyondCorp Enterprise
E.Cloud VPN
AnswersC, D

Identity-Aware Proxy (IAP) enforces zero-trust access by intercepting requests to applications and verifying the user's identity and contextual attributes, such as device security and network origin, against Cloud IAM policies. It applies least-privilege principles at the application layer, allowing only authenticated and authorized users to reach resources. IAP is a key building block for zero trust because it turns access decisions from network-based to identity-based.

Why this answer

Identity-Aware Proxy (IAP) is correct because it enforces zero-trust access to internal applications by verifying a user's identity and context (via Google identity and IAM) before allowing requests to reach the app, rather than relying on network location. BeyondCorp Enterprise is correct because it provides the zero-trust access framework—context-aware access policies, device trust, and continuous authorization—that works with IAP to protect internal apps without a VPN. Together, BeyondCorp Enterprise supplies the policy/context engine and IAP enforces those policies at the application layer, which is exactly the zero-trust pattern for internal app access.

Cloud Identity is not the right pairing here because it is primarily an identity and device management service (IdP/endpoint management), not the access-enforcement or context-aware policy layer for internal apps. Cloud Armor is a WAF/DDoS protection service for external HTTP(S) load balancers, and Cloud VPN provides network-level connectivity, both of which rely on perimeter/network trust rather than zero-trust application access.

Exam trap

The trap here is confusing identity management (Cloud Identity) or network security (Cloud Armor, Cloud VPN) with zero-trust access services; candidates might pick Cloud Identity because it sounds related to identity, but it lacks the access enforcement capabilities of IAP and BeyondCorp Enterprise.

192
MCQmedium

A company uses Cloud Load Balancing to distribute traffic to Compute Engine VMs. They want to protect against SQL injection and cross-site scripting attacks. Which service should they enable?

A.Identity-Aware Proxy (IAP)
B.Cloud Armor
C.Cloud CDN
D.VPC Service Controls
AnswerB

Cloud Armor is a web application firewall (WAF) and DDoS protection service that works natively with Cloud Load Balancing to filter inbound traffic. It provides preconfigured rules to block common attacks like SQLi, XSS, and OWASP Top 10 threats, as well as custom rules for IP and geolocation-based blocking, making it the correct tool for defending against web attacks.

Why this answer

Cloud Armor is the correct service because it provides web application firewall (WAF) capabilities that can inspect HTTP/S traffic for malicious patterns, including SQL injection and cross-site scripting (XSS) signatures. It integrates directly with Cloud Load Balancing to filter requests before they reach backend Compute Engine VMs, using pre-configured rules from the ModSecurity Core Rule Set (CRS) to block these common OWASP Top 10 threats.

Exam trap

The trap here is that candidates confuse Identity-Aware Proxy (IAP) with a security filter for application-layer attacks, but IAP only authenticates and authorizes users, not inspects traffic for malicious payloads like SQL injection or XSS.

How to eliminate wrong answers

Option A is wrong because Identity-Aware Proxy (IAP) controls access based on user identity and context (e.g., OAuth, device state), but it does not inspect HTTP request payloads for attack patterns like SQL injection or XSS. Option C is wrong because Cloud CDN caches content at edge locations to improve latency and reduce load, but it does not provide a WAF or inspect traffic for malicious payloads. Option D is wrong because VPC Service Controls create a security perimeter around Google Cloud APIs and services (e.g., preventing data exfiltration via VPC peering), but they do not filter application-layer attacks like SQL injection or XSS.

193
MCQmedium

A healthcare company must store PHI in Cloud Storage. They require encryption at rest and in transit, and need to comply with HIPAA. Which combination of Google Cloud features should they implement?

A.Cloud Storage with SSE-C, HTTP for in-transit, and enable HIPAA compliance flag.
B.Use Cloud Storage with CSEK and disable public access.
C.Cloud Storage with SSE-GCP and use HTTPS, sign BAA with Google.
D.Use Cloud Storage with CMEK and use VPN for transit.
AnswerC

This is the correct combination because it addresses all three HIPAA requirements. SSE-GCP is Google's default encryption for data at rest, automatically applied without additional key management overhead. HTTPS (TLS) encrypts all data in transit between clients and Cloud Storage, preventing eavesdropping. Signing a Business Associate Agreement (BAA) with Google establishes the necessary legal framework for handling Protected Health Information, making this solution fully HIPAA compliant.

Why this answer

Cloud Storage with server-side encryption (SSE-GCP) provides encryption at rest by default, HTTPS ensures encryption in transit, and signing a Business Associate Agreement (BAA) with Google is a mandatory contractual requirement for HIPAA compliance. This combination satisfies all stated requirements: encryption at rest, encryption in transit, and HIPAA compliance.

Exam trap

The trap here is that candidates often confuse encryption mechanisms (SSE-C, CSEK, CMEK, SSE-GCP) with HIPAA compliance requirements, mistakenly thinking that any encryption method plus disabling public access or using a VPN is sufficient, when in fact a signed BAA is the non-negotiable contractual requirement for HIPAA compliance with Google Cloud.

How to eliminate wrong answers

Option A is wrong because SSE-C (Server-Side Encryption with Customer-Provided Keys) is a valid encryption-at-rest option, but HTTP does not encrypt data in transit, and there is no 'HIPAA compliance flag' to enable in Cloud Storage—HIPAA compliance requires a signed BAA. Option B is wrong because CSEK (Customer-Supplied Encryption Keys) is a valid encryption-at-rest method, but disabling public access alone does not ensure encryption in transit (HTTPS is required) and does not address HIPAA compliance (a signed BAA is needed). Option D is wrong because CMEK (Customer-Managed Encryption Keys) provides encryption at rest, but using a VPN for transit does not guarantee HTTPS for Cloud Storage access; Cloud Storage requires HTTPS for encryption in transit, and a VPN alone does not satisfy the HIPAA requirement for a signed BAA.

194
MCQmedium

A company wants to receive notifications when its monthly spending exceeds 80% of a $10,000 budget. Which set of steps will achieve this?

A.Create a budget alert in Cloud Monitoring with a metric threshold of 80%
B.Set up a budget in the Billing section of Cloud Console, define the amount, and add alert thresholds at 80%
C.Use the Google Cloud Pricing Calculator to set a spending limit
D.Configure a Cloud Scheduler job to check billing data in BigQuery and send a notification
AnswerB

Cloud Billing budgets, configured under the Billing section of Cloud Console, are the official service for monitoring spend against a defined amount. You set a budget amount and then add alert threshold rules, such as 80% of the budget, which trigger notifications when actual or forecasted spend crosses that threshold. These alerts are event-driven and can deliver email or Pub/Sub messages, enabling immediate reaction when spending approaches the limit. This directly satisfies the company's requirement to be notified when monthly spend reaches 80% of the planned amount.

Why this answer

To receive notifications when spending exceeds 80% of a $10,000 budget, the correct approach is to create a budget in the Billing section of Google Cloud Console, define the amount, and set alert thresholds at 80%. This directly triggers alerts based on actual or forecasted spend. Other options do not provide budget-based alerts.

Exam trap

GCDL often tests the distinction between billing budgets and other monitoring tools, and candidates may confuse Cloud Monitoring with Billing budgets or think the Pricing Calculator can set alerts.

How to eliminate wrong answers

Option A is wrong because Cloud Monitoring is for infrastructure metrics, not billing budgets. Option C is wrong because the Pricing Calculator is for estimating costs, not setting alerts. Option D is wrong because it requires custom scripting and does not natively integrate with budget alerts.

195
Multi-Selectmedium

A data analytics team wants to analyze large datasets using SQL and create dashboards with minimal latency. They need a serverless data warehouse and a BI tool. Which two services should they use? (Choose exactly 2.)

Select 2 answers
A.Dataflow
B.Looker Studio
C.Cloud Storage
D.Looker
E.BigQuery
AnswersD, E

Looker is an enterprise business intelligence and data analytics platform that provides a semantic modeling layer (LookML) to define business logic, enabling consistent and reusable metrics across the organization. It allows analysts to explore large datasets through a governed interface and create interactive dashboards, and it ties into cloud data warehouses like BigQuery for query execution. Looker is designed specifically for large-scale business analytics and is a correct choice for this use case.

Why this answer

BigQuery is a serverless data warehouse for SQL analytics. Looker is a BI platform integrated with BigQuery for dashboards. Dataflow is for data processing, not storage.

Cloud Storage is for object storage, not SQL analytics. Looker Studio is free but less feature-rich for enterprise needs.

196
MCQeasy

A startup wants to deploy a web application globally and expects traffic to be unpredictable — sometimes very low, sometimes very high. Which cloud characteristic ensures the startup only pays for the compute resources it actually uses?

A.Resource pooling, which allows the startup to share physical hardware with other tenants
B.Measured service / pay-per-use pricing, where the startup is billed only for actual resource consumption with no payment for idle capacity
C.Broad network access, which allows the application to be reached from any device globally
D.On-demand self-service, which allows the startup to provision resources through a web interface without calling a sales team
AnswerB

Measured service is the NIST cloud characteristic that directly answers this. The startup is metered — billed for actual CPU, memory, network, and storage consumed. During low-traffic periods, bills are low. During spikes, costs scale up. No capacity is pre-purchased and wasted.

Why this answer

The measured service / pay-per-use pricing model of cloud computing ensures that the startup is billed only for the compute resources it actually consumes, such as CPU hours, memory, or storage, with no charges for idle capacity. This directly addresses the need to handle unpredictable traffic spikes without incurring costs for unused resources during low-traffic periods. In Google Cloud, this is implemented through per-second billing for compute instances and autoscaling, which dynamically adjusts resources based on demand.

Exam trap

Google Cloud often tests the distinction between the 'what' (e.g., resource pooling, broad network access) and the 'why it matters for cost' (measured service), leading candidates to confuse a general cloud characteristic with the specific billing model that addresses pay-for-use.

How to eliminate wrong answers

Option A is wrong because resource pooling refers to the cloud provider's ability to serve multiple tenants from the same physical hardware using virtualization, but it does not directly determine billing based on actual usage; it is about multi-tenancy and efficiency, not cost proportionality. Option C is wrong because broad network access describes the ability to access cloud services over the network via standard protocols (e.g., HTTP/HTTPS) from any device, but it has no relation to billing models or paying only for resources used. Option D is wrong because on-demand self-service allows users to provision resources automatically without human interaction, but it does not define the pricing model; it enables rapid provisioning but does not guarantee pay-per-use billing.

197
MCQeasy

Which Google Cloud service provides a managed Redis or Memcached in-memory data store for caching and low-latency data access?

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

Memorystore delivers fully managed Redis and Memcached engines on Google Cloud, removing patching, failover and scaling overhead. It satisfies the stem's requirement for a managed in-memory data store serving caching and low-latency access, unlike persistent disk or relational alternatives.

Why this answer

Memorystore is Google Cloud's fully managed in-memory data store service that supports Redis and Memcached. It provides low-latency caching and data access, making it ideal for applications requiring high-speed data retrieval.

Exam trap

GCDL often tests the confusion between Memorystore and other data services like Firestore or Bigtable; candidates may incorrectly choose a database service when the question specifically asks for an in-memory data store.

How to eliminate wrong answers

Option A is wrong because Firestore is a NoSQL document database, not an in-memory cache; it provides persistent storage with higher latency than in-memory solutions. Option B is wrong because Bigtable is a wide-column NoSQL database for large analytical workloads, not an in-memory data store. Option C is wrong because Cloud SQL is a managed relational database service (MySQL, PostgreSQL, SQL Server) that stores data on disk, not in memory.

198
Multi-Selecteasy

Which TWO features are part of Cloud Data Loss Prevention (Cloud DLP)?

Select 2 answers
A.Classification of sensitive data such as credit card numbers
B.De-identification of data through masking, tokenization, and encryption
C.Network vulnerability scanning
D.Removal of malware from uploaded files
E.Creation of IAM policies
AnswersA, B

Cloud DLP identifies sensitive data by applying built-in infoType detectors that match patterns like credit card numbers, which are further validated with Luhn checksum and contextual evidence. This classification step is the foundation for subsequent data protection actions, allowing organizations to discover where sensitive data resides across stored and streaming content, and even supports custom regex detectors for proprietary formats.

Why this answer

Cloud DLP is Google Cloud's service for discovering, classifying, and protecting sensitive data, so option A is correct: it inspects content and metadata to detect and classify sensitive data types such as credit card numbers, using built-in infoType detectors (for example CREDIT_CARD_NUMBER). Option B is also correct because Cloud DLP provides de-identification transformations including masking, tokenization (crypto-based tokenization via CryptoKey), and encryption (format-preserving encryption, deterministic encryption, and pseudonymization) to protect detected sensitive values. Option C is not part of Cloud DLP; network vulnerability scanning is handled by Security Command Center/Web Security Scanner, not DLP.

Option D is not part of Cloud DLP; malware scanning of uploaded files is performed by services such as Cloud Storage malware scanning or third-party tools, not DLP. Option E is not part of Cloud DLP; IAM policies are created and managed through IAM, not through Cloud DLP.

Exam trap

Google Cloud often tests the distinction between data-level security (Cloud DLP) and infrastructure-level security (vulnerability scanning, malware removal, IAM) to see if candidates confuse Cloud DLP's content inspection capabilities with broader security services.

199
MCQeasy

Which term best describes when an organization uses cloud-based tools (video conferencing, cloud document collaboration, project management platforms) to enable employees to work productively from any location?

A.Outsourcing — moving work to contractors in lower-cost locations.
B.Cloud-enabled distributed work / remote work — employees collaborate productively from any location using cloud tools.
C.Digital transformation — the company is changing its business model.
D.Automation — replacing human work with AI and robotics.
AnswerB

Cloud-enabled distributed work refers to a work model where employees use cloud services—such as IaaS for virtual desktops, PaaS for application development, and SaaS tools like Google Workspace or Microsoft Teams—to access the same enterprise resources and collaborate in real time regardless of geographic location. The cloud abstracts physical infrastructure, providing secure, scalable, and low-latency access to data and applications, which makes remote collaboration as effective as co-located work. This capability directly aligns with the question's scenario: employees leveraging cloud tools to be productive from any location, representing a fundamental shift in how work is organized.

Why this answer

The scenario explicitly describes employees using cloud-based tools (video conferencing, cloud document collaboration, project management platforms) to work productively from any location. This is the definition of cloud-enabled distributed work or remote work, where cloud infrastructure provides the connectivity, storage, and collaboration capabilities that decouple work from a fixed physical office. The key enabler is the cloud's ability to deliver real-time synchronization and access to shared resources over the internet, which is the core technical mechanism here.

Exam trap

Google Cloud often tests the distinction between 'cloud-enabled distributed work' and 'digital transformation' by making candidates think any use of cloud technology automatically qualifies as a business model change, when in fact remote work is a specific operational model, not a transformation of the core business.

How to eliminate wrong answers

Option A is wrong because outsourcing involves contracting work to external parties in lower-cost locations, not enabling existing employees to work from any location using cloud tools. Option C is wrong because digital transformation refers to a fundamental change in a company's business model or processes using digital technology, which is broader and not specifically about location-independent work. Option D is wrong because automation involves replacing human tasks with AI or robotic processes, not enabling human collaboration and productivity from remote locations via cloud tools.

200
MCQeasy

A startup wants to deploy a containerised web application that scales automatically from zero to handle traffic spikes and charges only for the resources used during request processing. They want to avoid managing servers or Kubernetes clusters. Which compute service should they choose?

A.Compute Engine
B.Cloud Run
C.Google Kubernetes Engine (GKE)
D.App Engine Standard
AnswerB

Cloud Run is a fully managed, Knative-based serverless platform that executes stateless containers in a scale-to-zero model: when there are no incoming requests, it shuts down all instances and you pay nothing. Each request is billed in 100-millisecond increments of compute time plus a minimal per-request charge, and it automatically scales to handle variable traffic, including bursting to thousands of concurrent requests. It accepts any container image that uses an HTTP server, which directly matches the need to deploy a containerised web application without managing any underlying nodes or clusters.

Why this answer

Cloud Run is a serverless compute platform that runs containers, scales to zero when not in use, and charges for resources used during request processing. It fits the description perfectly.

201
MCQmedium

A company is migrating a legacy monolithic application to Google Cloud. They want to reduce operational overhead by eliminating server management while keeping the ability to run containers. The application has unpredictable traffic patterns and needs to scale to zero when idle. Which compute option is the best fit?

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

Cloud Run is a fully managed serverless compute platform that executes stateless containers delivered via a requests URL. It automatically scales your container from zero to whatever number of instances are needed to handle incoming traffic, and you are billed only for the resources consumed during request processing—when idle, you pay nothing. This makes it exceptionally well-suited for unpredictable traffic patterns without requiring any cluster or infrastructure management.

Why this answer

Cloud Run is a serverless container platform that automatically scales to zero when idle, eliminating server management and cost during idle periods. GKE requires cluster management, Compute Engine is not serverless, and App Engine Flex requires VMs always running.

202
MCQeasy

Which Google Cloud service provides a centralized view of an application's performance metrics, logs, and traces — enabling teams to monitor system health, set up alerts, and diagnose issues from a single platform?

A.Cloud Security Command Center
B.Cloud Monitoring (part of Google Cloud's operations suite)
C.BigQuery
D.Cloud Asset Inventory
AnswerB

Cloud Monitoring is the central operational observability component of Google Cloud's operations suite, ingesting metric time series from GCP services, Prometheus, and custom app instrumentation. It offers flexible dashboards, alerting policies based on thresholds or MQL, uptime checks, and native integration with Cloud Logging and Cloud Trace for a unified troubleshooting workflow. For example, you can correlate a spike in HTTP 500 responses with error log entries and trace samples to diagnose root cause — capabilities no other listed service provides.

Why this answer

Cloud Monitoring (part of Google Cloud's operations suite) is the correct answer because it provides a unified platform for collecting and visualizing metrics, logs, and traces from applications and infrastructure. It enables teams to set up alerting policies, create dashboards, and diagnose performance issues using a single interface, integrating with services like Cloud Logging and Cloud Trace for end-to-end observability.

Exam trap

Google Cloud often tests the distinction between security-focused services and operations-focused services, so the trap here is that candidates might confuse Cloud Security Command Center (a security tool) with a monitoring solution because both provide 'visibility' into cloud resources.

How to eliminate wrong answers

Option A is wrong because Cloud Security Command Center is a security and risk management service that provides visibility into threats and vulnerabilities, not application performance metrics, logs, or traces. Option C is wrong because BigQuery is a serverless data warehouse for analytics over large datasets, not a monitoring or observability tool for real-time application performance. Option D is wrong because Cloud Asset Inventory is used to track and manage cloud resources and their metadata, not to monitor application performance or collect logs and traces.

203
MCQmedium

A SRE team wants to alert when their service is consuming error budget faster than expected, rather than alerting only when the SLO threshold is crossed. Which Cloud Monitoring alerting strategy supports this approach?

A.Threshold alerting — alert when error rate exceeds 0.1%.
B.SLO burn rate alerting — alert when error budget is being consumed faster than the measurement window allows.
C.Uptime check alerting — alert when health checks fail.
D.Log-based alerting — alert when specific error messages appear in logs.
AnswerB

SLO burn rate alerting continuously calculates the rate at which errors occur relative to the SLO's error budget and the remaining time in the measurement window. When the current burn rate projects that the budget will be exhausted before the window ends, an alert triggers promptly, enabling teams to respond before the SLO is actually violated. This approach is predictive rather than reactive, and can be tuned with fast and slow burn rates to detect both acute and chronic budget consumption, making it the correct mechanism for SLO compliance monitoring.

Why this answer

B is correct because SLO burn rate alerting is specifically designed to detect when error budget is being consumed faster than the measurement window allows, enabling proactive alerts before the SLO threshold is breached. This approach uses a burn rate (e.g., 2x, 10x) to trigger alerts when the error budget depletion rate exceeds a predefined multiple of the expected rate, allowing the team to respond early. It directly addresses the requirement of alerting on error budget consumption speed rather than waiting for a hard SLO violation.

Exam trap

The trap here is that candidates confuse threshold alerting on a static error rate with SLO burn rate alerting, mistakenly thinking a fixed percentage threshold (like 0.1%) is sufficient to catch fast error budget consumption, when in fact burn rate alerting is the only method that measures consumption velocity relative to the SLO window.

How to eliminate wrong answers

Option A is wrong because threshold alerting on a static error rate (e.g., 0.1%) does not account for the error budget consumption rate over time; it only triggers when a fixed percentage is exceeded, which may be too late or too early depending on traffic volume. Option C is wrong because uptime check alerting only monitors synthetic health checks (e.g., HTTP 200 responses) and does not measure error budget consumption or SLO compliance, making it irrelevant to the scenario. Option D is wrong because log-based alerting reacts to specific error messages in logs, which is a reactive, pattern-matching approach that does not track error budget burn rate or SLO adherence.

204
MCQmedium

A media streaming company wants to reduce latency for users worldwide by serving content from edge locations. They also want to offload traffic from their origin servers. Which Google Cloud service should they implement?

A.Cloud Interconnect
B.Cloud Storage
C.Cloud Load Balancing
D.Cloud CDN
AnswerD

Cloud CDN is the correct choice because it leverages Google's globally distributed edge infrastructure to cache media content (such as video segments, images, and other static assets) at points of presence closer to end users. When a user requests a media file, the CDN serves it from the nearest edge cache, dramatically reducing round-trip time and network hops compared to fetching from the origin. It also offloads traffic from origin servers, reducing bandwidth costs and origin load, and it supports flexible cache modes and TTLs for both static and dynamic content—making it the ideal solution for low-latency media streaming.

Why this answer

Cloud CDN uses Google's global edge network to cache content close to users, reducing latency and origin load.

205
MCQeasy

A non-technical manager asks what a 'virtual machine' is and how it differs from the physical servers the company used to run in its own data center. Which explanation is most accurate and accessible?

A.A virtual machine is a web-based application that runs in a browser and replaces traditional desktop software
B.A virtual machine is a complete software-defined computer that runs on shared physical hardware, providing the same capabilities as a dedicated server but created and managed through software in minutes
C.A virtual machine is a physical server located in the cloud provider's data center that is reserved exclusively for one customer
D.A virtual machine is a type of database that stores data virtually rather than on physical disk
AnswerB

This accurately describes a VM: it behaves like a physical server (has CPU, memory, OS, storage) but exists as software running on shared hardware. The key management advantage is that it can be provisioned, modified, and terminated through software, unlike physical servers which require manual hardware work.

Why this answer

A virtual machine (VM) is a software-based emulation of a physical computer that runs on shared physical hardware via a hypervisor. It provides the same capabilities as a dedicated server (CPU, memory, storage, networking) but can be provisioned, cloned, and managed in minutes through software, which is the core difference from traditional on-premises physical servers that require manual setup and are tied to specific hardware.

Exam trap

The trap here is that The GCDL exam often tests the misconception that 'virtual' means 'web-based' or 'in the cloud as a service,' leading candidates to confuse VMs with SaaS applications or dedicated physical servers, when the key differentiator is the hypervisor-based abstraction of hardware.

How to eliminate wrong answers

Option A is wrong because a virtual machine is not a web-based application running in a browser; that describes a web app or SaaS, not a VM. Option C is wrong because a virtual machine is not a physical server reserved exclusively for one customer; that describes a dedicated physical host or bare-metal server, while VMs share underlying physical hardware with other tenants via a hypervisor. Option D is wrong because a virtual machine is not a type of database; databases store data, while VMs are complete computing environments that can host databases or any other software.

206
Multi-Selecteasy

Which THREE of the following are best practices for managing operations in Google Cloud? (Choose THREE.)

Select 3 answers
A.Set up budget alerts to monitor costs
B.Implement infrastructure as code using Deployment Manager or Terraform
C.Enable Cloud Audit Logs for security and compliance
D.Use Cloud Logging to store all logs indefinitely to ensure compliance
E.Use a single project for all workloads to simplify management
AnswersA, B, C

Budget alerts in Google Cloud are configured at the billing account or project level to send notifications when actual or forecasted spend exceeds defined thresholds. This proactive monitoring prevents unexpected charges and enables timely cost governance, ensuring that teams can adjust resource usage before overspending occurs. Alerts are not resource limits but essential visibility tools for financial accountability.

Why this answer

Setting up budget alerts in Google Cloud allows you to monitor costs proactively by triggering notifications when spending exceeds defined thresholds. This is a fundamental operational best practice to avoid unexpected bills and maintain financial control over your cloud resources.

Exam trap

The trap here is that candidates often confuse 'storing logs indefinitely' with a compliance requirement, but Google Cloud best practices emphasize cost-effective log retention policies and using log exports for long-term storage rather than keeping logs in Cloud Logging forever.

207
MCQeasy

What is virtualization in the context of cloud computing, and why is it fundamental to how cloud providers deliver services?

A.Virtualization is the process of converting physical servers into digital images for backup purposes.
B.Virtualization abstracts physical hardware into multiple isolated virtual machines, enabling many customers to share physical infrastructure efficiently and securely.
C.Virtualization is a networking technique that routes internet traffic more efficiently.
D.Virtualization is a backup strategy where data is stored in multiple geographic locations.
AnswerB

Virtualization abstracts physical hardware resources such as CPU, memory, and storage through a hypervisor, which partitions them into multiple isolated virtual machines. Each VM contains its own operating system and applications, sharing the same underlying server while maintaining strong isolation. This abstraction enables cloud providers to achieve high server utilization and multi-tenancy, allowing many customers to run workloads securely on the same physical infrastructure — the foundation of cloud economics.

Why this answer

Virtualization is fundamental to cloud computing because it decouples the operating system and applications from the underlying physical hardware through a hypervisor (e.g., VMware ESXi, KVM, Hyper-V). This abstraction allows a single physical server to host multiple isolated virtual machines (VMs), each with its own guest OS, enabling cloud providers to achieve high resource utilization, multi-tenancy, and rapid provisioning. Without virtualization, providers would be limited to one OS per physical server, drastically reducing efficiency and scalability.

Exam trap

The GCDL exam often tests the misconception that virtualization is only about backup or networking, so candidates mistakenly pick options that describe tangential technologies (disk imaging, traffic routing, or geo-replication) instead of the core abstraction of physical hardware into multiple isolated environments.

How to eliminate wrong answers

Option A is wrong because virtualization is not merely converting physical servers into digital images for backup; that describes disk imaging or backup processes, not the core abstraction of hardware into multiple VMs. Option C is wrong because virtualization is not a networking technique for routing traffic; that describes technologies like SDN or routing protocols (e.g., BGP, OSPF), not the hypervisor-based abstraction of compute resources. Option D is wrong because virtualization is not a backup strategy for geographic data replication; that describes disaster recovery or geo-redundancy, not the foundational multi-tenant resource sharing enabled by hypervisors.

208
MCQmedium

A company runs a mission-critical application that must be available 24/7. They want to ensure that if a Google Cloud region becomes unavailable (e.g., due to a natural disaster), the application automatically continues to serve users from another region. Which architecture pattern achieves this?

A.Deploy in a single region with a Managed Instance Group using 3 availability zones.
B.Deploy the application in multiple regions with a Global Load Balancer for automated failover.
C.Enable Cloud Armor on the load balancer to protect against regional failures.
D.Use Cloud Storage multi-region buckets for application data.
AnswerB

Deploying the application in multiple regions and placing the backends behind a global load balancer (e.g., Global HTTP(S) LB) provides active-active geographic redundancy. The GLB uses a single anycast IP and routes each request to the closest healthy backend based on latency and health check status; when the health checks for an entire region fail, the load balancer automatically shifts traffic to the remaining healthy regions. This protects against regional outages and is a core pattern for mission-critical applications requiring high availability. For full recovery, the application must also replicate state (e.g., Cloud Spanner or Firestore) across regions.

Why this answer

Deploying the application in multiple regions behind a Global Load Balancer (GLB) enables automated failover. The GLB uses health checks to detect regional failures and routes traffic only to healthy backends, ensuring continuous availability even if an entire region goes down. This aligns with the requirement for a multi-region active-passive or active-active architecture for disaster recovery.

Exam trap

The trap here is that candidates confuse zonal redundancy (Option A) with regional redundancy, mistakenly believing that three zones in one region provide the same disaster recovery protection as multiple regions, but a regional failure (e.g., earthquake, power grid collapse) can take down all zones simultaneously.

How to eliminate wrong answers

Option A is wrong because deploying in a single region with three availability zones protects against zonal failures (e.g., a single datacenter outage) but does not protect against a full regional failure, such as a natural disaster affecting the entire region. Option C is wrong because Cloud Armor is a web application firewall (WAF) and DDoS protection service; it does not provide failover or regional redundancy. Option D is wrong because Cloud Storage multi-region buckets provide geo-redundant object storage but do not automatically failover compute or application logic; the application itself must be deployed in multiple regions with a load balancer to serve traffic.

209
Multi-Selectmedium

A company is planning their resource hierarchy in Google Cloud. They need to separate environments (prod, non-prod) and teams (engineering, data science). They also need to apply common policies to all prod projects. Which TWO resources should they use?

Select 2 answers
A.Billing accounts
B.Folders
C.Organization policies
D.Labels
E.Projects
AnswersB, C

Folders are the correct hierarchical grouping node within Google Cloud's resource hierarchy, sitting between the organization node and projects. They allow you to organize projects into environments (e.g., dev, prod) or business units, and they propagate policies (IAM, Organization Policies) downward to all projects and resources inside them. This inheritance enables centralized governance and access control across multiple projects.

Why this answer

Folders (B) are the correct resource for modeling the hierarchy: you can nest folders under the organization to represent environments (prod, non-prod) and teams (engineering, data science), and IAM policies and Organization Policies inherit down through the folder tree to all contained projects. Organization policies (C) are the mechanism to apply common constraints (e.g., restrict VM external IPs, allowed locations) that inherit to every project under the prod folder, satisfying the requirement to enforce common policies across all prod projects. Billing accounts (A) only link projects to a payment method and do not model environment/team separation or policy inheritance.

Labels (D) are metadata for filtering, reporting, and billing attribution, not a policy or hierarchy boundary. Projects (E) are the leaf resources that hold workloads; they cannot themselves group other projects or enforce inherited common policies across a set.

Exam trap

GCDL often tests the distinction between resource hierarchy components and policy tools, confusing candidates into selecting labels or billing accounts for organizational purposes when folders and organization policies are the correct structural and policy enforcement mechanisms.

210
MCQeasy

Which statement best describes the shared responsibility model in Google Cloud?

A.Google is responsible for security of the infrastructure; customers are responsible for their data and configurations.
B.Google is responsible for all security aspects, including customer data.
C.Customers are responsible for physical security of data centers.
D.Security is entirely the customer's responsibility in IaaS.
AnswerA

In the shared responsibility model, Google Cloud operates and secures the physical infrastructure layer, including data centers, servers, storage, networking hardware, and the hypervisor. The customer retains responsibility for what they control: their data, identity and access management (IAM) policies, and configuration of services such as encryption, firewalls, and OS-level hardening. This split ensures that each party focuses on the components it is best positioned to protect.

Why this answer

The Google Cloud shared responsibility model divides security duties: Google secures the underlying infrastructure (physical data centers, hardware, network, hypervisor), while customers are responsible for securing what they put in the cloud — their data, configurations, access controls, and application-level security. Option A accurately captures this division.

Exam trap

GCDL often tests the misconception that Google secures everything in the cloud — candidates must remember that data, configurations, and access controls are always the customer's responsibility.

How to eliminate wrong answers

Option B is wrong because Google does not secure customer data or configurations; that responsibility always remains with the customer. Option C is wrong because physical security of data centers is Google's responsibility, not the customer's. Option D is wrong because even in IaaS, Google still secures the physical and foundational layers — security is never entirely the customer's responsibility.

211
MCQmedium

A company's cloud spending suddenly spikes by 300% for one week before returning to normal. The cloud team investigates and finds a developer accidentally left a large cluster of VMs running over the weekend. Which cloud financial management practice most effectively prevents this type of unexpected cost spike?

A.Requiring all cloud resource creation to go through a central IT approval process to prevent accidental VM creation
B.Setting Cloud Billing budget alerts that notify stakeholders when spending approaches defined thresholds, enabling early detection and response to abnormal spending patterns
C.Reviewing cloud bills at the end of each month to identify cost anomalies and address them retroactively
D.Using Reserved Instances for all VM workloads to reduce per-hour costs, making accidental long-running VMs less expensive
AnswerB

Budget alerts are the direct preventive control. A budget alert at 150% of normal daily spending would have triggered early Saturday morning, prompting investigation. This gives the team time to act before a full weekend of over-spending accumulates. Budget alerts with escalating thresholds (50%, 80%, 100%, 150%) are a best practice.

Why this answer

Cloud Billing budget alerts provide real-time notifications when spending exceeds defined thresholds, enabling the cloud team to detect and respond to the 300% spike immediately rather than after the fact. This proactive monitoring directly addresses the root cause—unexpected resource usage—by alerting stakeholders while the VMs are still running, allowing them to shut down the cluster and prevent further cost accumulation.

Exam trap

Google Cloud often tests the distinction between proactive prevention (alerts) and reactive cost optimization (Reserved Instances or monthly reviews), leading candidates to mistakenly choose D because they focus on reducing cost per unit rather than preventing the unexpected usage itself.

How to eliminate wrong answers

Option A is wrong because requiring central IT approval for resource creation introduces a bottleneck that slows down development agility and does not prevent a developer from leaving VMs running; it only controls creation, not runtime duration. Option C is wrong because reviewing bills at the end of each month is a retroactive practice that cannot prevent the cost spike—by the time the bill is reviewed, the 300% increase has already been incurred, making it a detection method rather than a prevention method. Option D is wrong because Reserved Instances reduce per-hour costs but do not prevent the unexpected spike; even at a lower rate, leaving a large cluster running for a full weekend would still result in a significant cost increase, and the practice does not address the behavioral or monitoring gap.

212
Multi-Selecteasy

An e-commerce platform uses Compute Engine instances in a managed instance group behind a Cloud Load Balancer. During a flash sale, the load balancer reports increased error rates. The operations team suspects the instances are overwhelmed. Which two steps should they take to troubleshoot the issue? (Choose TWO.)

Select 2 answers
A.Switch to a Network Load Balancer for higher throughput.
B.Increase the size of the instance group without investigation.
C.Enable HTTP health checks on the load balancer.
D.Check the CPU utilization of the instance group in Cloud Monitoring.
E.Review the load balancer logs in Cloud Logging for error messages.
AnswersD, E

Inspect the Cloud Monitoring metrics for the managed instance group, specifically the CPU utilization time series (compute.googleapis.com/instance/cpu/utilization). Sustained high CPU percentages indicate that the existing instances are computing at their capacity, which directly supports the hypothesis that backend overload is producing errors or latency. This observation helps determine whether autoscaling or code optimization is the correct next step, rather than blindly changing infrastructure.

Why this answer

Option D is correct because checking CPU utilization of the instance group in Cloud Monitoring directly tests the hypothesis that the instances are overwhelmed; sustained high CPU across the MIG is the key signal of saturation and helps confirm whether the error rate spike is caused by insufficient compute capacity. Option E is correct because reviewing the load balancer logs in Cloud Logging exposes the actual HTTP error codes and messages (for example 502/503 responses or backend timeouts), which distinguishes backend overload from other causes such as misconfigured health checks or client errors. Option A is not appropriate because switching to a Network Load Balancer is a design change, not a troubleshooting step, and it would not diagnose why the current load balancer is reporting errors.

Option B is not appropriate because blindly increasing the size of the instance group without investigation may mask the symptom, waste resources, and leave the root cause unidentified. Option C is not appropriate because enabling HTTP health checks is a configuration change rather than a diagnostic action, and health checks alone would not reveal the cause of the elevated error rates.

Exam trap

The trap here is that candidates may confuse switching load balancer types (Option A) with a performance fix, when in fact the issue is backend capacity, not frontend protocol handling.

213
MCQhard

An architect explains that her cloud application uses a 'loosely coupled architecture.' She contrasts it with a tightly coupled on-premises system where all components run in a single process. What is the primary operational benefit of loose coupling in a cloud environment?

A.Loosely coupled applications are always faster because messages are passed in memory rather than over the network
B.Loose coupling allows individual components to fail, scale, or be updated independently without cascading failures to the entire system
C.Loosely coupled architectures require less developer expertise and are easier to build than monolithic applications
D.Loose coupling reduces cloud costs because fewer network calls are made between services
AnswerB

This is the primary operational benefit. When components communicate through queues and APIs rather than direct coupling, a failure in one component doesn't automatically bring down others. Each component can also scale independently based on its own load, and teams can deploy updates without coordinating a system-wide release.

Why this answer

In a loosely coupled architecture, components communicate via well-defined interfaces (e.g., REST APIs, message queues) and are deployed as independent services. This means a failure in one component does not propagate to others, and each component can be scaled or updated without affecting the rest of the system. This isolation is the primary operational benefit in a cloud environment, enabling high availability and continuous delivery.

Exam trap

The GCDL exam often tests the misconception that 'loose coupling' implies faster performance or lower cost, when in fact the primary benefit is operational independence and fault isolation, not raw speed or expense.

How to eliminate wrong answers

Option A is wrong because loose coupling typically involves network calls (e.g., HTTP, AMQP) between services, which are slower than in-memory calls; the statement incorrectly claims messages are passed in memory. Option C is wrong because loosely coupled architectures (e.g., microservices) require significant developer expertise in areas like service discovery, distributed tracing, and eventual consistency, making them harder to build than monolithic applications. Option D is wrong because loose coupling often increases the number of network calls between services, which can increase latency and data transfer costs, not reduce them.

214
MCQhard

A company's risk management team wants to understand Google Cloud's approach to supply chain security — specifically, how Google ensures that the hardware and firmware running in its data centers have not been tampered with. Which Google security initiative addresses hardware supply chain integrity?

A.Google uses third-party antivirus software to scan all hardware components for tampering before installation
B.Google's Titan security chip, embedded in Google's servers, cryptographically attests boot firmware integrity and machine identity — providing hardware-level supply chain security assurance
C.Google relies on hardware manufacturers' security certifications to ensure supply chain integrity
D.Google encrypts all hardware components with AES-256 to prevent tampering
AnswerB

Titan is Google's hardware root of trust for supply chain security. It generates a cryptographic identity for the machine, verifies boot firmware hasn't been tampered with (preventing firmware attacks), and provides attestation that can be verified throughout the machine's lifecycle. This is a core component of Google's defense-in-depth security architecture.

Why this answer

Google's Titan security chip is a dedicated hardware root of trust that cryptographically verifies the boot firmware integrity and machine identity at every startup. This ensures that only Google-signed firmware runs on servers, preventing tampering during manufacturing, shipping, or deployment. Titan provides a hardware-anchored attestation chain that validates the entire supply chain from chip fabrication to rack installation.

Exam trap

The trap here is that candidates often confuse supply chain security with data protection mechanisms (like encryption) or rely on third-party certifications, missing that Google's proprietary hardware root of trust (Titan) is the specific initiative for hardware integrity.

How to eliminate wrong answers

Option A is wrong because Google does not use third-party antivirus software to scan hardware components; antivirus software operates at the OS level and cannot verify hardware or firmware integrity at the supply chain level. Option C is wrong because Google does not rely solely on hardware manufacturers' security certifications; instead, Google implements its own hardware security controls like Titan to independently verify integrity, as manufacturer certifications can be compromised or insufficient. Option D is wrong because AES-256 encryption protects data at rest or in transit, not hardware components themselves; encrypting hardware components would not prevent tampering with firmware or the physical device.

215
MCQhard

An engineer is deploying a global application that requires low latency for users in North America, Europe, and Asia. They plan to use Compute Engine instances in multiple regions. Which Google Cloud resource is essential to distribute traffic and provide a single anycast IP address?

A.Global external HTTP(S) load balancer
B.Cloud CDN
C.Traffic Director
D.Cloud NAT
AnswerA

The Global external HTTP(S) load balancer is correct because it provides a single global anycast IP address, leveraging Google's global network and advanced routing to direct each request to the closest region with available capacity. It supports HTTP(S) traffic, can be integrated with Cloud CDN, and routes to backend services across multiple regions, making it ideal for a global application requiring low latency and one entry point.

Why this answer

Global external HTTP(S) load balancer provides a single anycast IP and distributes traffic across regions, enabling low-latency access worldwide.

216
Multi-Selectmedium

A company is considering migrating its on-premises workloads to Google Cloud. Which TWO are common business drivers for cloud adoption? (Choose TWO.)

Select 2 answers
A.Cost optimization by converting CapEx to OpEx
B.Guaranteed 100% uptime for all services
C.Faster time to market through on-demand resource provisioning
D.Ability to avoid any vendor lock-in
E.Elimination of all security risks
AnswersA, C

Cloud migration converts capital expenditure (CapEx) on physical data centers and hardware into operational expenditure (OpEx) through pay-as-you-go pricing. This eliminates large upfront investments and reserves, replacing them with predictable, variable costs directly tied to consumption, which improves cash flow and financial flexibility.

Why this answer

Option A is correct because migrating to Google Cloud shifts IT spending from large upfront capital expenditures (CapEx) on hardware and data centers to pay-as-you-go operational expenditures (OpEx), which is a well-established financial driver for cloud adoption. Option C is correct because cloud providers like Google Cloud offer on-demand, self-service resource provisioning through services such as Compute Engine and Cloud Storage, enabling organizations to deploy and iterate faster and thereby accelerate time to market. Option B is incorrect because no cloud provider guarantees 100% uptime for all services; SLAs typically offer credits for failing to meet targets like 99.9% or 99.99%, not absolute uptime.

Option D is incorrect because avoiding vendor lock-in is not a guaranteed outcome of cloud adoption and often requires deliberate multi-cloud or portable architectures; it is a concern to manage rather than a common driver. Option E is incorrect because cloud adoption does not eliminate security risks; it changes the shared responsibility model, and customers remain responsible for securing their workloads, identities, and data.

Exam trap

GCDL often tests the misconception that cloud guarantees 100% uptime or eliminates security risks, which are exaggerated claims not supported by SLAs or shared responsibility models.

217
MCQhard

A company runs a globally distributed application with users in North America, Europe, and Asia. They need to serve static content (images, videos) with low latency from edge locations. They also need to protect against DDoS attacks. Which combination of services should they use?

A.Cloud CDN and Cloud NAT
B.Cloud Load Balancing and Cloud CDN
C.Cloud CDN and Cloud Armor
D.Cloud Armor and Cloud Interconnect
AnswerC

Cloud CDN offloads static content delivery to Google's global edge cache, absorbing sudden spikes in traffic and reducing origin load. Cloud Armor enforces security policies at the edge, offering HTTP(S) L3-7 DDoS protection, WAF rules, and per-user rate limits. Together, they provide both acceleration and protection: Cloud CDN handles legitimate cacheable content at scale, while Cloud Armor filters out attack traffic before it reaches Cloud CDN or the backend.

Why this answer

Cloud CDN caches static content at edge locations for low latency, and Cloud Armor provides DDoS protection and WAF capabilities.

218
Multi-Selectmedium

A company wants to implement a data pipeline that ingests streaming events from a global user base, processes them in real-time to detect anomalies, and stores the results in a database for low-latency querying. The solution must be fully managed. Which THREE services should they use? (Choose 3)

Select 3 answers
A.Dataflow
B.Cloud Storage
C.Cloud Bigtable
D.Pub/Sub
E.Cloud Functions
AnswersA, C, D

Dataflow is a fully managed, unified stream and batch data processing service based on Apache Beam. It performs real-time stream processing, including event-time windowing, aggregation, and anomaly detection with exactly-once semantics, auto-scaling, and low latency. It is the correct choice for the processing stage because it can consume from Pub/Sub, apply transformations (including anomaly detection), and write results to Bigtable.

Why this answer

Option D (Pub/Sub) is correct because it is a fully managed, globally scalable messaging service designed to ingest high-volume streaming events from a worldwide user base and reliably deliver them to downstream processing systems. Option A (Dataflow) is correct because it is a fully managed, serverless stream and batch processing service that can consume Pub/Sub messages and perform real-time anomaly detection using Apache Beam pipelines. Option C (Cloud Bigtable) is correct because it is a fully managed, low-latency NoSQL database that scales horizontally and is well suited for storing and querying the pipeline's results with millisecond latency.

Option B (Cloud Storage) is not appropriate as the primary sink for low-latency querying, since it is object storage optimized for large blobs rather than fast row-level reads. Option E (Cloud Functions) is not suitable as the core stream-processing engine here, because it is an event-driven serverless function service with execution time and concurrency limits, not a managed pipeline for continuous, high-throughput stream analytics.

Exam trap

GCDL often tests the difference between batch and streaming services, tricking candidates into selecting Cloud Storage or Cloud Functions for real-time processing when Dataflow is the correct managed stream-processing service.

219
MCQmedium

A company is evaluating Google Cloud and wants to know: what is Access Transparency, and how does it benefit customers with stringent governance requirements?

A.Access Transparency shows customers which Google Cloud services are available in their region.
B.Access Transparency logs when Google Cloud personnel access customer content, providing an audit trail for governance.
C.Access Transparency is a feature that makes all customer data visible to Google for quality improvement.
D.Access Transparency provides customers with real-time dashboards of their application's security vulnerabilities.
AnswerB

Access Transparency generates near-real-time audit logs whenever Google Cloud employees or support engineers access customer content, capturing the specific action, the data involved, the business justification (e.g., support ticket or legal request), and the timestamp. These logs are delivered to the customer's own Cloud Audit Logs bucket, so they can be stored, queried, and retained under the customer's governance policies. This provides independent, cryptographic evidence that Google respects data residency and access controls, which is critical for regulated industries like finance, healthcare, and public sector.

Why this answer

Access Transparency logs are a Google Cloud feature that provides customers with near real-time logs whenever Google personnel access their data. This creates a detailed audit trail, which is essential for customers with stringent governance or compliance requirements, as it allows them to monitor and verify that access is only for authorized purposes.

Exam trap

The GCDL exam often tests the distinction between 'logging access' and 'providing visibility into data' — the trap here is confusing Access Transparency (an audit log of Google personnel actions) with a feature that exposes or shares customer data with Google.

How to eliminate wrong answers

Option A is wrong because Access Transparency does not show which Google Cloud services are available in a region; that is the function of the Google Cloud region and service listing pages. Option C is wrong because Access Transparency does not make customer data visible to Google for quality improvement; it logs when Google personnel access data, and customers must opt in to share data for quality improvement through separate programs. Option D is wrong because Access Transparency does not provide real-time dashboards of security vulnerabilities; that is the role of services like Security Command Center or Web Security Scanner.

220
Multi-Selectmedium

A company is designing a disaster recovery plan for its critical application running on Google Kubernetes Engine (GKE) in us-central1. The application uses Cloud SQL (MySQL) as its database. Which TWO actions should the company take to achieve a Recovery Point Objective (RPO) of less than 5 minutes and a Recovery Time Objective (RTO) of less than 1 hour in the event of a regional outage? (Choose two.)

Select 2 answers
A.Use Cloud SQL's automated backups with a backup schedule every 4 hours.
B.Store container images in a regional registry in the same region.
C.Configure Cloud SQL for cross-region replication using a read replica in a different region.
D.Deploy a GKE cluster in a different region and use Multi-cluster Ingress to route traffic.
E.Use Cloud CDN to cache static content.
AnswersC, D

A cross-region read replica in Cloud SQL uses asynchronous replication to continuously apply changes from the primary instance to a replica in a different region. This near real-time stream typically has an RPO in the range of seconds, staying well under a 5-minute requirement. During a regional failure, you can promote the replica to become the new primary and redirect applications to it, providing a concrete failover path without restoring from backups. The read replica must be large enough to handle production traffic after promotion.

Why this answer

Option C is correct because Cloud SQL cross-region replication maintains a continuously updated read replica in a different region, so in a regional outage the replica can be promoted with only seconds to minutes of data loss, satisfying the RPO of under 5 minutes. Option D is correct because deploying a GKE cluster in a second region and using Multi-cluster Ingress provides a pre-provisioned, load-balanced failover target so the application can be served from the surviving region within the 1-hour RTO. Option A is incorrect because 4-hour automated backups cannot meet an RPO of less than 5 minutes, since up to 4 hours of data could be lost.

Option B is incorrect because a regional registry in the same region would itself be unavailable during a regional outage and does not address database recovery or RPO/RTO. Option E is incorrect because Cloud CDN only caches static content and does nothing to recover the GKE workload or Cloud SQL database during a regional failure.

Exam trap

Google Cloud often tests the distinction between backup-based recovery (which has higher RPO due to fixed intervals) and continuous replication (which can achieve near-zero RPO), leading candidates to mistakenly choose automated backups for low RPO requirements.

221
MCQmedium

A company is moving its financial reporting application to Google Cloud. The CFO asks: 'If Google Cloud experiences a data breach and our financial data is exposed, who is financially liable?' How should the cloud architect answer this question?

A.Google Cloud bears full financial liability for all data breaches involving customer data on its platform
B.Liability depends on where the breach originated: Google is responsible for failures in its infrastructure security; the customer is responsible for breaches resulting from misconfiguration, application vulnerabilities, or inadequate access controls in areas under their responsibility
C.The customer bears all liability for any breach because they chose to use cloud services
D.No party is liable because data breaches in cloud are force majeure events similar to natural disasters
AnswerB

This accurately describes the shared responsibility reality. If Google's physical security or hypervisor is breached, Google bears responsibility. If a misconfigured IAM policy exposes data (customer responsibility), the customer bears the consequences. The customer should also have cyber insurance to manage residual risk.

Why this answer

The Google Cloud Shared Responsibility Model explicitly delineates liability: Google is responsible for the security of the cloud (e.g., physical infrastructure, hypervisor, network controls), while the customer is responsible for security in the cloud (e.g., IAM policies, application code, data encryption). In a breach, liability is determined by where the failure occurred—if Google’s infrastructure (e.g., GKE node isolation) fails, Google bears liability; if the customer misconfigures a Cloud Storage bucket or leaves a Compute Engine firewall open, the customer bears liability. This aligns with the CFO’s question about financial liability, which is not absolute but contingent on the breach’s origin.

Exam trap

The trap here is that candidates assume Google Cloud automatically assumes all liability for any data breach, ignoring the Shared Responsibility Model’s clear division of accountability based on the breach’s origin (infrastructure vs. customer-managed layers).

How to eliminate wrong answers

Option A is wrong because Google Cloud does not bear full financial liability; the Shared Responsibility Model assigns liability based on the breach’s origin, and customers retain responsibility for their own configurations, applications, and access controls. Option C is wrong because the customer does not bear all liability; Google is liable for breaches caused by failures in its infrastructure security (e.g., hypervisor escapes, physical data center breaches). Option D is wrong because data breaches are not force majeure events; they are foreseeable risks addressed in Google Cloud’s SLA and contractual terms, and liability is governed by the Cloud Terms of Service, not natural disaster clauses.

222
MCQmedium

An organization runs its entire infrastructure on a single public cloud provider (Google Cloud). All applications, data, and services live in Google Cloud's infrastructure. Which deployment model describes this?

A.Private cloud
B.Public cloud
C.Hybrid cloud
D.Community cloud
AnswerB

Public cloud deployment means all infrastructure resources are provided and managed by a third-party provider like Google, hosted in Google's facilities, and shared with other customers under logical isolation. When an organization uses only Google Cloud for all workloads, with no dedicated hardware or self-managed servers, it is a textbook public cloud implementation. This model offers scalability, pay-as-you-go pricing, and reduced operational overhead, all characteristics of public cloud services.

Why this answer

The organization is using a single public cloud provider, Google Cloud, which delivers computing resources over the public internet on a pay-as-you-go basis. In a public cloud deployment, the infrastructure is owned and operated by the cloud provider and shared across multiple tenants, which matches the scenario where all applications, data, and services reside in Google Cloud's infrastructure.

Exam trap

The GCDL exam often tests the misconception that using a single public cloud provider is a 'private cloud' because the organization has exclusive use of that provider's resources, but the key distinction is that the provider's infrastructure is still shared among multiple customers, making it a public cloud deployment.

How to eliminate wrong answers

Option A is wrong because a private cloud is dedicated to a single organization and is typically hosted on-premises or in a single-tenant environment, not on a shared public cloud provider like Google Cloud. Option C is wrong because a hybrid cloud requires a combination of at least two distinct deployment models (e.g., public and private) that are interconnected, whereas the scenario describes all infrastructure on a single public cloud. Option D is wrong because a community cloud is shared by several organizations with common concerns (e.g., compliance or security requirements) and is not a single public cloud provider used exclusively by one organization.

223
MCQhard

A company's monolithic application is difficult to update because any change requires testing and redeploying the entire application, causing multi-hour downtime during updates. The team is considering a microservices architecture. What is the primary benefit of microservices in this context?

A.Microservices always cost less than monolithic applications to run.
B.Each service can be updated and deployed independently, enabling teams to release changes faster with lower risk and without full-application downtime.
C.Microservices eliminate the need for testing because each service is small enough to be bug-free.
D.Microservices allow applications to run on any hardware without modification.
AnswerB

Independent deployability is the defining operational advantage of microservices: because services communicate via well-defined APIs rather than sharing code and runtime state, changing service A requires only rebuilding and redeploying that service, leaving B, C, and D untouched. This shrinks the deployment blast radius, lets teams ship on their own cadence, and supports safer strategies like canary releases or blue-green deployments that would be far riskier and slower on a single monolithic unit. The result is faster feature delivery with less coordination overhead and no need to drain and redeploy the entire application.

Why this answer

The core benefit of microservices is independent deployability: each service has its own lifecycle, so a change to one service can be built, tested, and deployed without redeploying the entire application. This reduces blast radius, enables faster release cadence, and eliminates the full-application downtime described in the scenario.

Exam trap

GCDL often tests whether candidates confuse microservices' deployability benefit with cost or portability claims, or wrongly assume smaller services eliminate testing needs.

How to eliminate wrong answers

Option A is wrong because microservices do not inherently cost less — they often increase operational complexity and infrastructure overhead due to more services, networking, and observability tooling. Option C is wrong because microservices absolutely require testing; smaller services are not automatically bug-free, and distributed systems introduce new failure modes. Option D is wrong because portability across hardware is not a defining benefit of microservices — containerization (not microservices per se) provides that, and monolithic apps can also run on varied hardware.

224
Multi-Selectmedium

A company wants to connect its on-premises network to Google Cloud for consistent low-latency access to resources. They also need to use the same network policies across multiple projects. Which two Google Cloud products should they use?

Select 2 answers
A.Cloud Interconnect
B.Shared VPC
C.Cloud NAT
D.Cloud VPN
E.VPC Network Peering
AnswersA, B

Cloud Interconnect provides a dedicated, private connection between the on-premises network and Google Cloud, delivering the consistent low-latency access the company requires. It satisfies the connectivity constraint, while Shared VPC separately handles the cross-project policy requirement.

Why this answer

Cloud Interconnect (A) is correct because it provides a dedicated, private physical connection between the on-premises network and Google Cloud, delivering consistent low-latency and high-bandwidth access that a public internet-based VPN cannot guarantee. Shared VPC (B) is correct because it lets a host project share its VPC network with multiple service projects, so the same firewall rules, subnets, and network policies are applied uniformly across all those projects. Cloud NAT (C) is not appropriate here because it only provides outbound internet access for resources without external IPs and does not connect on-premises networks or unify policies across projects.

Cloud VPN (D) is not the best fit because it tunnels traffic over the public internet, which cannot ensure the consistent low latency the company requires. VPC Network Peering (E) connects separate VPC networks but does not centralize network policies across projects the way Shared VPC does.

Exam trap

Google Cloud often tests the distinction between connectivity products (Cloud Interconnect vs. Cloud VPN) and policy-sharing mechanisms (Shared VPC vs. VPC Network Peering), where candidates mistakenly choose VPC Network Peering for cross-project policy management, not realizing it only connects VPCs without centralizing policies.

225
MCQeasy

A company's security team wants to control which resources on the internet can communicate with their virtual machines in Google Cloud. Which fundamental cloud networking concept provides this control?

A.VPC firewall rules that control inbound and outbound network traffic to VMs based on IP ranges, protocols, and ports
B.Cloud IAM policies that grant or deny permissions for external systems to access VM resources
C.Cloud Storage bucket policies that restrict access to VM storage volumes
D.DNS records that determine which internet addresses can resolve the VM's hostname
AnswerA

VPC firewall rules are the correct answer. They are stateful, software-defined firewalls applied to VM network interfaces in Google Cloud VPCs. Rules specify what traffic is allowed or denied based on source/destination IPs, protocols, and ports — providing precise control over which internet resources can communicate with the VMs.

Why this answer

VPC firewall rules are the fundamental cloud networking construct that controls network traffic to and from virtual machine instances in Google Cloud. These rules operate at the network layer, filtering traffic based on source/destination IP ranges, protocols (e.g., TCP, UDP, ICMP), and port numbers, thereby governing which internet resources can communicate with the VMs.

Exam trap

The GCDL exam often tests the distinction between network-layer controls (firewall rules) and identity/access management (IAM), tempting candidates to choose IAM policies because they sound like 'control' over access, but IAM does not filter network traffic at the packet level.

How to eliminate wrong answers

Option B is wrong because Cloud IAM policies control identity-based permissions for who can perform actions on cloud resources (e.g., who can create VMs), not network-level traffic filtering between internet resources and VMs. Option C is wrong because Cloud Storage bucket policies govern access to objects in Cloud Storage, not to VM storage volumes; VM storage is handled by persistent disks or local SSDs, which are not controlled by bucket policies. Option D is wrong because DNS records resolve hostnames to IP addresses but do not enforce any access control or filtering of network traffic; they merely provide name resolution.

Page 2

Page 3 of 12

Page 4