Cloud Cert Notes

Original study notes for cloud certification exams — Google Cloud first, more clouds in build.

Home › Section 2: Planning and Configuring a Cloud Solution › Practice Quiz

Section 2 Practice Quiz: Planning and Configuring a Cloud Solution

By the Cloud Cert Notes Editorial Team
Last updated: November 2026

This quiz focuses on planning before you build: choosing the right compute, storage, database, and network options for a workload, estimating costs with the pricing calculator, planning data movement and growth, and designing for availability and region placement. Good planning questions on the exam test trade-offs between cost, performance, scalability, and operational effort, including the common distractors and trade-offs that appear in real exam scenarios.

Choose the best answer for each question, then press the button at the end to score yourself. Every question includes a full explanation — if JavaScript is unavailable, open the explanation under each question to check the answer yourself.

1. A stateless web application experiences unpredictable traffic spikes and the team wants to minimize operational management while paying only for requests handled. Which compute option is the best fit?

A. A single Compute Engine virtual machine with a fixed size B. A self-managed Kubernetes cluster the team patches themselves C. Cloud Run, a fully managed serverless container platform that scales with requests D. An App Engine standard environment application with manual scaling set to one instance

Show the answer and explanation

Correct answer: C. Cloud Run runs containers serverlessly, scales automatically including to zero when idle, and bills for resources used while handling requests, which matches unpredictable stateless traffic with minimal operations. A single fixed virtual machine cannot scale with spikes, a self-managed cluster adds patching and operations, and manual scaling to one instance prevents the elasticity the workload needs. This distinction is frequently tested in scenario-based Associate Cloud Engineer practice and exam questions.

2. A team needs a relational database with strong consistency, familiar SQL, and automatic backups, but does not need global multi-region writes. Which service is most appropriate?

A. Cloud SQL for MySQL or PostgreSQL B. Bigtable C. BigQuery D. Firestore in Datastore mode only

Show the answer and explanation

Correct answer: A. Cloud SQL is a fully managed relational service for MySQL, PostgreSQL, and SQL Server offering transactions, backups, replication, and high availability within a region. Bigtable is a wide-column NoSQL store for very large analytical and operational workloads, BigQuery is an analytical data warehouse rather than a transactional database, and Firestore is a document database, not a traditional relational SQL system. This distinction is frequently tested in scenario-based Associate Cloud Engineer practice and exam questions.

3. When estimating the cost of a proposed solution for stakeholders, which tool should you use to model compute, storage, and networking charges together?

A. Cloud Monitoring dashboards B. The gcloud alpha billing simulator command C. BigQuery cost tables alone without any estimate D. The Google Cloud Pricing Calculator

Show the answer and explanation

Correct answer: D. The Google Cloud Pricing Calculator lets you add products, configure machine types, storage amounts, regions, and usage patterns, then produce a shareable estimate. Monitoring shows observed usage after deployment, not a pre-purchase model. BigQuery billing exports help analyze actual spend later, but they are not a substitute for an upfront estimate built with the calculator. This distinction is frequently tested in scenario-based Associate Cloud Engineer practice and exam questions.

4. A workload requires data to remain available if an entire zone fails, and the team is choosing a storage strategy for virtual machine disks. What should they plan?

A. Use only local SSDs attached to instances in one zone B. Use regional Persistent Disks or regular snapshots replicated across zones, and deploy instances across zones C. Use zonal Persistent Disks in a single zone with no snapshots D. Store all data on the boot disk of a single instance

Show the answer and explanation

Correct answer: B. Regional Persistent Disks synchronously replicate data between two zones in a region, and snapshots provide another way to recover across zones. Deploying across multiple zones protects against a single-zone outage. Local SSDs are ephemeral and tied to the host, zonal disks in one zone have no cross-zone redundancy, and relying on one boot disk creates a single point of failure. This distinction is frequently tested in scenario-based Associate Cloud Engineer practice and exam questions.

5. A company in Europe needs low latency for European users and must keep data in Europe for compliance. How should region selection influence the plan?

A. Always choose the cheapest region regardless of location B. Choose a US region because all Google Cloud data is stored there anyway C. Choose a region in Europe close to users, and use multi-region resources only where the service and compliance needs allow D. Region choice has no effect on latency or compliance

Show the answer and explanation

Correct answer: C. Regions are geographic locations for resources; choosing one near users reduces latency, and choosing one within a required jurisdiction helps meet data residency needs. Price varies by region but should not override latency and compliance requirements. Data is stored in the regions and locations you select for regional and multi-region services, not automatically in the US, and region choice directly affects both factors.

6. Which database is best suited for a globally distributed application that needs relational semantics, strong consistency, and horizontal scaling across regions?

A. Spanner B. Cloud SQL with a single primary instance C. Cloud Storage with CSV files D. Memorystore for Redis as the primary system of record

Show the answer and explanation

Correct answer: A. Spanner is a fully managed, globally distributed relational database that combines SQL, strong consistency, and horizontal scaling. Cloud SQL is regional and scales primarily vertically with read replicas for reads, not global multi-region writes. Cloud Storage files are not a relational database, and Memorystore is an in-memory cache and store, not designed as a global relational system of record. This distinction is frequently tested in scenario-based Associate Cloud Engineer practice and exam questions.

7. A team is planning network design for multiple projects that need to share a common VPC network and centralized firewall rules. Which feature should they plan to use?

A. Separate VPC networks in every project with no peering or sharing B. Shared VPC, with a host project owning the network and service projects using it C. Cloud NAT in every project as the sharing mechanism D. External HTTP load balancers to connect the projects

Show the answer and explanation

Correct answer: B. Shared VPC lets an organization associate multiple projects with a single VPC network owned by a host project, enabling centralized network administration, subnets shared to service projects, and consistent firewall and routing control. Separate unconnected networks do not share resources centrally, Cloud NAT provides outbound internet access for private instances, and load balancers distribute traffic rather than share a network. This distinction is frequently tested in scenario-based Associate Cloud Engineer practice and exam questions.

8. An analytics team will run ad hoc SQL queries over petabytes of historical data with no need to manage servers or indexes. Which service should the plan center on?

A. Cloud SQL B. Bigtable with manual cluster sizing for every query C. Firestore with collection scans D. BigQuery, a serverless data warehouse

Show the answer and explanation

Correct answer: D. BigQuery is a serverless, highly scalable data warehouse designed for SQL analytics over very large datasets without managing infrastructure. Cloud SQL is sized for transactional workloads far smaller than petabyte analytics, Bigtable requires cluster management and is optimized for key-based operational access rather than ad hoc SQL, and Firestore is a transactional document database, not an analytical warehouse. This distinction is frequently tested in scenario-based Associate Cloud Engineer practice and exam questions.

9. A startup wants the lowest operational overhead for storing unstructured files that will be served to users worldwide. Which storage plan fits best?

A. Use Cloud Storage buckets with lifecycle and location settings matched to access patterns B. Attach Persistent Disks to a file server they manage and patch C. Store files inside a Cloud SQL database as binary columns D. Keep files on local SSDs for permanence

Show the answer and explanation

Correct answer: A. Cloud Storage is managed object storage for unstructured data, offering location types, storage classes, lifecycle rules, and global access options without running a file server. Managed disks and self-run file servers add patching and scaling work, databases are a poor fit for large file serving, and local SSDs are ephemeral and not durable for permanent file storage. This distinction is frequently tested in scenario-based Associate Cloud Engineer practice and exam questions.

10. During planning, a team discovers a workload needs GPUs for machine learning training that runs a few hours per week. How should this affect the compute plan?

A. Buy a permanently running largest GPU instance to be safe B. Use Cloud Run because it provides unlimited GPUs for training C. Use Compute Engine instances with attached GPUs that can be started for training and stopped when idle, considering cost controls D. Avoid GPUs and accept training that takes weeks on CPUs

Show the answer and explanation

Correct answer: C. Compute Engine supports attaching GPUs to virtual machines, so a plan can provision GPU instances for training jobs and stop them when not in use to control cost, using scheduling and budgets. Leaving the largest instance running permanently wastes money. Cloud Run GPU support is limited and aimed at inference-style serving, not a general replacement for planned batch training capacity. This distinction is frequently tested in scenario-based Associate Cloud Engineer practice and exam questions.

Study aid disclaimer: Cloud Cert Notes is an independent educational study aid. It is not affiliated with, endorsed by, or sponsored by Google, Microsoft, Amazon, Oracle, or Alibaba. Google Cloud, Associate Cloud Engineer, and Cloud Digital Leader are trademarks of Google LLC; all other certification names and product names are trademarks of their respective owners. Always confirm service details against the official provider documentation before sitting an exam.