Cloud Concepts — Azure AZ-900 Fundamentals practice questions

OpEx (operational expenditure) is associated with:

1. On-premises servers
2. Cloud pay-as-you-go billing
3. Hardware purchases
4. Building a new datacenter

Answer: B — Cloud pay-as-you-go billing

A/C/D) All are CapEx — upfront capital purchases. B) Correct — cloud billing is OpEx (operational), expensed monthly as used. This shifts IT spend from capital to operating budget.

Which BEST distinguishes scalability from elasticity?

1. Scalability always happens automatically while elasticity is manual
2. Elasticity is a concept that applies only to on-premises kit
3. The two terms describe exactly the same platform capability
4. Scalability is the ability to grow; elasticity does it for you

Answer: D — Scalability is the ability to grow; elasticity does it for you

D) Scalability is the capability to add capacity; elasticity adds automation in both directions as demand changes. A) The description is reversed. B) Elasticity is a defining cloud characteristic. C) The terms overlap but are not identical.

A retail site must keep serving customers even if one datacenter becomes unavailable. Which cloud benefit is being pursued?

1. Elasticity, adjusting the amount of capacity to match demand
2. Manageability, administering resources from a single console
3. Governance, enforcing organisational standards on how resources are deployed
4. High availability, keeping the service reachable despite failures

Answer: D — High availability, keeping the service reachable despite failures

D) Correct — designing so a failure in one location does not take the service down is high availability. A) Elasticity is about matching capacity to load, which is a different goal. B) Manageability concerns administration, not fault tolerance. C) Governance controls how resources are configured but does not itself keep a site online.

An architect describes a workload as consuming a 'shared pool of configurable resources that can be rapidly provisioned and released'. Which cloud property is being described MOST directly?

1. Defense in depth, layering independent controls around the workload data
2. Governance, applying organisational standards to deployed resources
3. Elasticity, adding and removing capacity quickly as demand moves
4. Manageability, administering resources through a single control plane

Answer: C — Elasticity, adding and removing capacity quickly as demand moves

C) Correct — rapid provisioning and release of pooled capacity in response to demand is elasticity. A) Defense in depth is a security layering concept and is true of Azure but unrelated to provisioning speed. B) Governance is about enforcing standards, not about how fast capacity moves. D) Manageability describes how you administer resources, which is a different benefit from how quickly capacity scales.

A team runs a database on Azure SQL Database, a platform as a service offering. Which responsibility does the team retain that it would NOT retain for a software as a service application?

1. Securing the physical media on which the database files reside
2. Applying periodic firmware updates to the underlying storage controllers
3. Configuring the application logic and schema that use the database
4. Patching the database engine binaries used by the managed service

Answer: C — Configuring the application logic and schema that use the database

C) Correct — with platform as a service the customer still builds and owns the application and its data model on top of the managed platform, which a finished software as a service product would already provide. A) Physical media security is Microsoft's in every service model. B) Firmware on hardware is never a customer task in Azure. D) The managed service patches the engine itself; assuming the customer must is the common misconception about platform as a service.

An organization wants to handle 10x traffic spikes WITHOUT human intervention. They need:

1. Scalability only
2. High availability
3. Elasticity (auto-scale up and down)
4. Disaster recovery

Answer: C — Elasticity (auto-scale up and down)

A) Scalability allows growth but not automatic. C) Correct — elasticity = automatic scale up + scale DOWN with demand. B) HA = uptime, not auto-scale. D) DR = recovery from failure, not load handling.

Which statement about public cloud is accurate?

1. Resources are dedicated to a single customer on isolated hardware
2. Resources are owned by the provider and shared across many customers
3. Resources require the customer to purchase and rack physical servers
4. Resources can never be isolated logically from other tenants in any way at all

Answer: B — Resources are owned by the provider and shared across many customers

B) Correct — in public cloud the provider owns the infrastructure and offers it to many customers over the internet. A) That describes private cloud or specialised dedicated host options, not the public model generally. C) Purchasing and racking servers is exactly what public cloud removes. D) A misconception: logical isolation through virtual networks, subscriptions and role assignments is standard.

Multi-cloud means:

1. Running workloads in two Azure regions for redundancy
2. Running workloads in Azure and an on-premises datacenter
3. Running workloads across two or more cloud providers
4. Reaching cloud services over the public internet only

Answer: C — Running workloads across two or more cloud providers

C) Multi-cloud means using more than one public cloud provider, for example Azure alongside another hyperscaler. A) Two regions of one provider is still a single cloud. B) Cloud combined with on-premises is hybrid. D) Internet connectivity is a network detail, not a deployment model.

Fault tolerance is BEST described as:

1. Preventing every hardware component from ever failing
2. Delivering lower network latency to nearby end users
3. Continuing to operate while components are failing
4. Adding redundancy without any additional cost at all

Answer: C — Continuing to operate while components are failing

C) Fault tolerance means redundant components absorb a failure so the service keeps running. A) Failures cannot be eliminated, only tolerated. B) Latency reduction is a performance goal, not fault tolerance. D) Redundant capacity increases cost rather than removing it.

Which statement about capital expenditure and operational expenditure is accurate for an organisation moving a workload to Azure?

1. Cloud consumption is capital expenditure because the capacity is reserved in advance
2. Cloud consumption is operational expenditure billed against ongoing usage of services
3. Cloud consumption is capital expenditure that must be depreciated over several years
4. Cloud consumption removes the concept of expenditure classes from the workload entirely

Answer: B — Cloud consumption is operational expenditure billed against ongoing usage of services

B) Correct — cloud spending is an ongoing operating cost tied to consumption rather than an up-front asset purchase. A) Reservations do commit spend in advance, but they are still billed as an operating expense, so the tempting link to capital expenditure fails. C) Depreciation applies to owned assets such as purchased servers, not to consumed services. D) Expenditure classification still applies; only its category changes from capital to operating.

What does 'serverless' mean, and what are its two defining characteristics?

Serverless means the provider fully manages the underlying servers so you deploy only code or configuration. Its two hallmarks are automatic scaling driven by events or demand (including scaling to zero) and billing tied to execution rather than to reserved capacity.

SaaS — what's managed by customer?

Just data/config. Provider manages everything else. Example: Microsoft 365, Salesforce.

Shared responsibility — who owns identity?

Customer. Always — across all service models.

PaaS — what's managed by customer?

Applications + data only. Provider manages OS, runtime, middleware.