AZ-900 Describe cloud concepts: 336 practice questions

12 of the 336 Describe cloud concepts questions in the Certsqill AZ-900 bank, shown in full below. Each one carries an explanation for every option, not just the correct one — the wrong answers are where the marks go.

Preparing for AZ-900? Take the free 5-min readiness check →

1. Obtain resources on demand over a network: Which cloud concept lets it obtain resources when needed without bu

1. Purchase and install servers before demand occurs.

   This requires upfront hardware ownership and does not provide on-demand cloud resource acquisition.
2. Build a private environment with scheduled project capacity.

   This describes dedicated infrastructure planning rather than obtaining provider resources on demand.
3. Reserve capacity for the semester.

   Reserved fixed capacity does not address changing weekly demand and may leave resources underused.
4. Obtain resources on demand over a network, without buying servers. ✓

   Cloud computing provides network-accessible resources that can be obtained as needed without purchasing equivalent physical infrastructure.

Cloud computing provides network-accessible resources that can be provisioned as demand changes.

2. Maintain the guest operating system inside the virtual: Which responsibility remains with the company?

1. Maintain the guest operating system inside the virtual machine. ✓

   In IaaS, the customer manages the guest operating system, applications, data, identities, and access configuration.
2. Maintain the provider's physical servers.

   The provider manages the physical infrastructure that hosts the virtual machine.
3. Manage the hypervisor on the physical host.

   The provider manages the hypervisor that supports the virtual machine.
4. Operate the provider's physical datacenter facilities and network infrastructure.

   The provider operates the physical datacenter facilities and infrastructure supporting the service.

IaaS leaves guest operating-system maintenance with the customer.

3. Deploy instances across separate failure locations: Select TWO actions that address these different requiremen

Select two. More than one option is correct — every correct one is ticked below.

1. Create a backup copy only after the service stops.

   A post-failure backup supports recovery but does not maintain service availability during the outage.
2. Move all processing to one larger instance.

   A larger instance can provide vertical capacity but leaves one primary failure point.
3. Deploy instances across separate failure locations. ✓

   Distribution across failure locations reduces dependence on one host or localized failure.
4. Configure capacity to increase as incoming workload grows. ✓

   Scaling capacity helps the service handle increased demand without relying on fixed resources.
5. Purchase additional storage without changing compute capacity.

   Extra storage may hold more data but does not address host failure or compute demand.

High availability reduces outage impact; scaling increases capacity for changing workload demand.

4. Public cloud infrastructure is provider-owned: Which statement is accurate?

1. Private cloud requires fewer security controls because it is dedicated to one organization.

   Dedicated use does not remove the need for identities, access controls, configuration, and other security practices.
2. Public cloud infrastructure is provider-owned, while customer data access remains controlled through responsibilities shared with the customer. ✓

   Public cloud infrastructure serves multiple customers, but customer data is not automatically publicly accessible. Access depends on provider and customer responsibilities, including customer identities and configuration.
3. The provider alone controls identities and application access for public cloud data.

   Customers retain responsibility for their identities, access, data, and customer-controlled settings.
4. Public cloud resources are shared, so customer data is accessible to other tenants by default.

   Shared provider infrastructure does not mean one customer's data is automatically accessible to other customers or the public.

Public cloud describes shared provider infrastructure, not public visibility of customer data.

5. The provider manages the underlying operating system: Select TWO accurate responsibility statements.

Select two. More than one option is correct — every correct one is ticked below.

1. PaaS removes the shop's responsibility for application configuration.

   Customers still configure applications and customer-controlled settings even when platform maintenance is managed.
2. The provider manages the underlying operating system and runtime. ✓

   PaaS shifts underlying operating-system and runtime maintenance to the cloud provider.
3. The shop must maintain physical servers supporting the platform.

   The provider operates physical infrastructure used to deliver the managed platform service.
4. The shop remains responsible for application code and customer data. ✓

   PaaS does not transfer ownership of customer-created code or business data to the provider.
5. The provider automatically defines which employees may access customer information.

   The customer must configure identities and access according to its business and security requirements.

PaaS manages platform components while customers retain application, data, identity, and access responsibilities.

6. Resize the instance: What scaling action is appropriate?

1. Resize the instance. ✓

   Vertical scaling changes one instance's capacity by increasing or decreasing its available resources.
2. Distribute requests across multiple instances.

   Load distribution supports horizontal scaling, which conflicts with retaining one machine.
3. Automatically reduce capacity whenever processing becomes expensive.

   Reducing capacity may control cost but does not provide the requested additional CPU and memory.
4. Add another instance beside it.

   Adding instances is horizontal scaling and changes the number of running service instances.

Vertical scaling changes the capacity of one instance rather than increasing the instance count.

7. The cloud is dedicated to one organization: What does that dedication primarily indicate?

1. The organization receives unlimited capacity automatically.

   Dedicated infrastructure does not create unlimited capacity or remove the need for capacity planning.
2. The provider cannot manage any infrastructure components.

   A provider or internal team may manage infrastructure; private describes dedication, not management absence.
3. The cloud is dedicated to one organization. ✓

   A private cloud is dedicated to a single organization, unlike shared public cloud infrastructure.
4. Every workload becomes secure without configuration.

   Security still depends on identities, access, data protection, and correctly configured customer controls.

Private cloud means cloud resources are dedicated to one organization.

8. SaaS: Which cloud service model fits?

1. PaaS for custom application development.

   PaaS supplies a managed development platform, requiring the customer to provide application code.
2. SaaS. ✓

   SaaS delivers a finished application while the customer manages users, data, access, and settings.
3. Private cloud hosting for dedicated infrastructure.

   Private cloud describes dedication and deployment, not specifically consumption of a finished application.
4. IaaS virtual machines for operating-system control.

   IaaS provides infrastructure control but leaves guest operating-system and application management with customers.

SaaS provides a finished application while minimizing infrastructure and platform management.

9. Add more service instances: Which action represents horizontal scaling?

1. Add more service instances. ✓

   Horizontal scaling increases capacity by adding instances that can process requests together.
2. Replace application code with a newer version.

   Code replacement changes application behavior or features but does not inherently increase running capacity.
3. Increase one server's size.

   Increasing one server's resources is vertical scaling, not an increase in service instance count.
4. Move the service to a different subscription.

   Changing subscription organization does not inherently add processing instances or increase application capacity.

Horizontal scaling increases capacity by adding additional service instances.

10. Hybrid cloud connecting the private datacenter with public: Which cloud model combines these environments?

1. Hybrid cloud connecting the private datacenter with public cloud resources ✓

   Hybrid cloud combines private or on-premises resources with public cloud resources in a connected approach.
2. Public cloud

   Public cloud does not describe retaining the regulated workload in the startup's own datacenter.
3. Multicloud using multiple public providers

   Multicloud uses multiple public cloud providers; it does not specifically combine on-premises resources with public cloud resources.
4. Private cloud

   A private cloud is dedicated to one organization and does not describe the required public cloud component.

Hybrid cloud combines private or on-premises resources with public cloud resources.

11. IaaS virtual machine with customer-managed guest OS: Which service model best matches this requirement?

1. IaaS virtual machine with customer-managed guest OS patching ✓

   IaaS virtual machines provide customer control of the guest operating system, including applying patches.
2. PaaS with a provider-managed runtime

   In PaaS, the provider manages the underlying operating system and runtime, so the customer does not patch the guest operating system.
3. Serverless computing while the provider manages server provisioning and platform infrastructure

   Serverless abstracts server provisioning and platform infrastructure; it does not provide customer control of a guest operating system.
4. SaaS for a finished training application

   SaaS provides a finished application and does not give the customer control of its guest operating system.

IaaS provides guest operating-system control, including responsibility for patching.

12. Elasticity through configured automatic scaling: What does this configuration primarily demonstrate?

1. Vertical scaling only

   Vertical scaling changes resource size, whereas this rule changes the number of running instances.
2. High availability without configuration

   High availability reduces downtime through design; demand-based instance changes address capacity instead.
3. Elasticity through configured automatic scaling ✓

   Elasticity adapts capacity to changing demand, provided scaling rules and automation are configured.
4. A guarantee that every application component automatically remains available during failures.

   Scaling responds to demand but does not guarantee component availability or failure recovery.

Elasticity adjusts capacity as demand changes, but automation requires explicit configuration and does not guarantee availability.

Other AZ-900 domains

• Describe Azure architecture and services — 456 questions →
• Describe Azure management and governance — 408 questions →
• All 1200 AZ-900 questions →
• AZ-900 certification: requirements, cost and exam format →