CCNA Cloud Concepts Questions

This choice is correct because Azure SQL Database is a PaaS offering. Microsoft is responsible for managing the host operating system, including applying security updates, patching, and maintaining the infrastructure. The customer only interacts with the database instance at the logical level.

Vertical scaling (or scaling up/down) refers to increasing or decreasing the resources (CPU, memory, etc.) of a single virtual machine or resource. This change increases the CPU capacity of the existing VM, which is a classic example of scaling up.

In an IaaS deployment, Microsoft protects the physical datacenter, server hardware, the hypervisor layer, and the physical network, while the customer must secure the guest OS, including patching, hardening, and configuration, along with its applications and data. This distinction precisely defines the shared responsibility model: the cloud provider manages the infrastructure up to the virtualization layer, and the customer manages everything deployed on top of it.

This is the defining benefit of the public cloud: a pay-as-you-go operating expenditure model replaces upfront capital investment in data-center hardware. You pay only for consumed compute, storage, and bandwidth, enabling startups to launch with minimal CapEx. Azure's reserved instances and scaling allow cost alignment with actual usage, while hardware procurement, maintenance, and refresh costs remain with the provider.

Hybrid cloud joins on-premises infrastructure with Azure public cloud services, letting workloads burst into Azure during peak demand while retaining existing data centre resources. This directly matches the requirement to extend capacity without abandoning on-premises investment.

Elasticity is correct because the application's unpredictable workload demands automatic resource adjustments. Azure Elasticity, implemented through Autoscale and VM Scale Sets, adds or removes instances on the fly based on metrics like CPU or request count, without manual intervention. This ensures the application can absorb sudden spikes in load and consume fewer resources during lulls, optimizing cost and performance. It is the only option that directly addresses the need for dynamic, demand-driven scaling.

Adding five more virtual machines to the existing pool increases the instance count from three to eight. Horizontal scaling, or scaling out, adds more instances of the same resource to distribute load, rather than resizing an existing VM's CPU or memory.

This accurately describes the two primary scaling strategies in cloud computing. Vertical scaling (scaling up/down) increases or decreases the capacity of a single resource, such as moving a virtual machine to a larger SKU with more CPU and memory, which often requires a restart and has hardware limits. Horizontal scaling (scaling out/in) adds or removes additional instances of a resource, such as placing multiple VMs or containers behind a load balancer, and enables elasticity, redundancy, and near-unlimited growth via distributed architecture. Azure implements this with capabilities like VM scale sets and App Service autoscaling.

Vertical scaling changes the size or tier of a single resource, such as moving a VM to a larger SKU, which often requires stopping and restarting the instance to apply the change. Elasticity, by contrast, automatically adds or removes identical instances based on real-time demand, spreading load across them without interrupting existing operations. This distinction is fundamental: vertical scaling resizes one resource, while elasticity manages a fleet of instances dynamically and without downtime.

High availability is the ability of a system to remain operational and accessible even when individual components or entire datacenters fail. In this scenario, deploying the application across multiple Azure regions and using Traffic Manager to automatically route traffic away from a failed region achieves geographic redundancy and failover. This directly supports the requirement for uninterrupted operation, which is the essence of high availability. The configuration specifically targets minimizing downtime due to infrastructure failures, not improving performance or reducing costs.

Elasticity automatically scales resources out and back in to match fluctuating demand, so sudden traffic spikes from viral campaigns are absorbed without degradation. This satisfies the requirement to meet unexpected workload demand while avoiding over-provisioning during normal periods.

Agility is the cloud benefit that enables quick and flexible deployment of resources. The team can spin up environments in minutes and tear them down just as fast, which directly supports rapid development and testing cycles.

Private cloud is the correct choice because it provisions dedicated physical servers exclusively for one organization, delivering single-tenant infrastructure with no resource contention. This complete control over compute, storage, and networking allows consistent, predictable performance because workloads are isolated from outside traffic and 'noisy neighbor' effects. You can also tailor hardware configurations and capacity exactly to the workload's requirements, ensuring SLAs are met reliably.

Correct. Elasticity is the cloud characteristic that enables automatic provisioning and de-provisioning of resources to match workload demand. The autoscale behavior described—adding instances during high CPU usage and removing them when usage drops—is a textbook example of elasticity.

This is correct because moving to cloud shifts your workload to Azure's global infrastructure, eliminating the need to own and operate physical data centers. You no longer pay for server hardware, power, cooling, physical security, or the staff required to refresh and repair equipment, converting large capital expenses into manageable operating expenses. This lets your team focus on application development and business innovation rather than infrastructure maintenance.

Cloud providers operate multiple isolated regions, so replicating workloads across them delivers geographic redundancy and automated failover that on-premises DR sites cannot match without duplicating hardware and facilities. Consumption-based pricing removes the standing cost of idle secondary infrastructure, satisfying the lower-cost constraint.

This is the correct answer because the Azure SLA is a commitment to service availability—specifically uptime and connectivity—not to elasticity, throughput, or response times under load. Concurrent user count has no bearing on Microsoft's promised uptime percentage; even a completely idle service still receives the same SLA, and a heavily loaded service does not affect the contractual availability guarantee.

The shared responsibility model is the correct framework because it explicitly divides security duties between the cloud provider and the customer. Under this model, the provider maintains the security of the physical infrastructure, network, and hypervisor, while the customer is accountable for data, identities, access management, and—depending on the service model—configuration, patching, and encryption. For patient health records, this division is critical to understand for compliance (e.g., HIPAA) since the hospital, as the customer, retains responsibility for protecting the data itself regardless of where it is stored.

Reserved Instances provide a substantial discount (up to 72%) on virtual machine compute costs in exchange for a one- or three-year commitment. For a predictable, always-on workload, this is the most cost-effective option to minimize total ownership cost.

The customer is always responsible for their own data, including data classification, encryption, and access management. In SaaS, the provider does not have insight into which users should have access; the customer must manage identities and protect data.

Elasticity is the cloud capability that automatically provisions and releases compute, storage, and other resources to match the exact level of current demand, scaling out during traffic spikes and scaling in during lulls. This dynamic, reactive adjustment eliminates the need for IT teams to pre-purchase or over-provision capacity based on forecasted usage, because the platform handles scaling in real time. Elasticity directly supports pay-per-use cost models by ensuring you only consume what you actually need at any given moment.

Moving to Azure shifts spending from capital expenditure (CAPEX) to operational expenditure (OPEX). In an on-premises data center, you must purchase and depreciate physical servers, storage, and networking hardware. Azure's pay-as-you-go, consumption-based pricing means you pay only for resources you use, with no upfront infrastructure purchases; this is a fundamental shift from owning assets to renting services.

Correct. Migrating from on-premises hardware to Azure changes the cost model from purchasing and owning physical servers (capitalized and depreciated as CapEx) to paying recurring subscription fees for virtualized compute and storage (treated as OpEx). This removes large upfront capital investments, shifts IT costs to the operational budget, and aligns spending with actual usage.

Consumption-based pricing is the defining financial model of cloud computing: you are billed for metered usage such as VM-hours, storage GB-months, or outbound data transfer, typically on a recurring monthly cycle. Costs accrue only when resources are actively in use or data resides in the cloud, and you can deallocate resources to stop paying entirely. This shifts IT spend from capital expenditure (CapEx) to operating expenditure (OpEx), with no upfront hardware investment and the flexibility to align costs directly with actual demand.

Correct. On-demand self-service allows users to provision cloud resources (such as virtual machines) automatically without requiring human interaction with the IT team or cloud provider, matching the scenario where developers can provision resources in minutes via the Azure portal.

When two or more VMs are placed in the same Availability Set, Azure guarantees that you will have connectivity to at least one instance at least 99.95% of the time. The Availability Set spreads VMs across fault domains (racks with independent power and networking) and update domains (groups rebooted sequentially during maintenance), reducing the risk of concurrent downtime. This redundancy level specifically yields the 99.95% SLA, making it the correct answer for the stated scenario.

Disaster recovery (DR) is the structured process of restoring systems, applications, and data to an operational state after a catastrophic event such as a region-wide outage, cyberattack, or natural disaster. It is governed by recovery time objective (RTO) and recovery point objective (RPO), which specify the maximum acceptable downtime and the maximum acceptable data loss. This definition directly matches the option: restoring functionality with minimal downtime after a catastrophic loss.

Across every cloud service model—IaaS, PaaS, and SaaS—the cloud provider is solely responsible for physical security of the data center. This includes protecting the facilities with biometric access controls, multi-factor authentication for entry, continuous video monitoring, and physically securing server racks and hardware. Customers interact only with logical resources through the cloud interface, never with the physical hardware or facilities, so responsibility for physical security cannot be transferred or shared.

In IaaS, the cloud provider is responsible for the physical datacenter, including security, power, cooling, and the host hardware (servers, storage, and networking) that underpin the virtualized environment. The customer manages everything above the hypervisor, such as VMs, operating systems, and applications. This is the correct answer because the question specifically asks for a responsibility that belongs to the provider.

A private cloud is built for the exclusive use of one organization, with dedicated physical servers, storage, and network infrastructure. It can be hosted on-premises or at a third-party facility, but in all cases the infrastructure is not shared with any other tenant. This singular tenancy directly satisfies regulatory mandates that require absolute isolation and no shared hardware.

Cloud services are delivered over the internet, so connectivity to providers remains a hard dependency. Claiming cloud computing eliminates internet reliance is false; loss of connectivity disrupts access to workloads and data, making this the stated non-benefit.

Cloud providers measure resource usage (e.g., compute hours, storage) and charge only for what is consumed. By deallocating VMs during off-hours, the company avoids paying for compute time during those periods, directly leveraging the consumption-based cost model. This is a key advantage of the cloud over traditional on-premises infrastructure, where hardware costs are fixed regardless of usage.

On-demand self-service is the correct characteristic. It enables users to provision and manage computing resources automatically, without requiring human interaction with the service provider. The developer's ability to create a VM instantly via the Azure portal without IT approval exemplifies this concept.

Elasticity matches the requirement because it automatically scales compute resources out and in as real-time demand fluctuates, so capacity tracks the holiday spike and contracts afterwards. This satisfies the constraint of avoiding payment for idle resources outside peak periods, unlike scalability, which only describes the system's ability to grow to meet demand.

High availability means designing services to stay operational with minimal downtime, typically through redundancy and failover across zones or regions. This directly satisfies the stem's requirement by describing continuity of service rather than scalability, elasticity or durability, which address capacity, demand fluctuation and data retention respectively.

Durability is the property that assures data, once committed, will not be lost or corrupted even when hardware fails, accomplished through redundant copies, erasure coding, and cross-region replication. Azure achieves this with options like locally redundant storage (LRS), zone-redundant storage (ZRS), and geo-redundant storage (GRS), each providing a different level of persistence. Because the question asks about ensuring data is preserved over time, durability exactly matches the required guarantee.

High availability is a design principle that minimizes application downtime by eliminating single points of failure and enabling automatic failover when a component fails. This is achieved through redundancy (multiple instances, Azure Availability Zones, or Availability Sets), health probes, and load balancers that reroute traffic to healthy replicas. In the e-commerce scenario, HA ensures the platform remains reachable and functional during a component failure, directly addressing the stated requirement to keep services running.

Global reach (also called geographic distribution) is the cloud benefit that allows organizations to deploy applications and services in data centers around the world, reducing latency for users in different regions. The company is using this capability to serve customers in Europe, Asia, and North America with local instances.

Azure's consumption-based pricing charges you only for what you use, making costs variable in theory. However, with tools like the Azure Pricing Calculator, TCO Calculator, and Cost Management + Budgets, you can accurately forecast spending and set limits or alerts. For example, you can define budgets and configure alerts to notify you when spending approaches a threshold, and use Azure Advisor to optimize resource utilization. Thus, 'predictable costs' in Azure mean the ability to plan and govern spending, not a fixed bill.

Reliability is the Azure Well-Architected Framework pillar that enables systems to recover from failures and continue functioning while meeting defined availability targets. It includes concrete design techniques such as deploying across availability zones, using fault domains, implementing load balancing, and configuring backup/Disaster Recovery strategies to achieve specific RTO (Recovery Time Objective) and RPO (Recovery Point Objective) metrics. Because the scenario explicitly asks about recovering from failures and sustaining operations, this is the correct answer.

Multi-tenancy is the defining model of public cloud whereby a provider pools physical compute, storage, and networking hardware and serves many customers from that shared infrastructure. Logical isolation is enforced through hypervisors and software-defined networking, so each tenant's data and workloads remain private even though they reside on the same physical host. This sharing drives economies of scale, allowing the provider to offer pay-as-you-go pricing while maintaining security boundaries. Multi-tenancy is precisely the concept described in the question stem.

An Azure SLA defines specific uptime percentages (e.g., 99.9% for most services) and outlines the remedies—typically service credits—that customers receive when Microsoft fails to meet those commitments. This is its core purpose: to establish clear performance expectations and provide a contractual recourse when availability targets are missed.

Hybrid cloud is the correct model because it explicitly unites an organization's on-premises infrastructure with public cloud services through secure, dedicated connectivity such as VPN or ExpressRoute. This architecture allows the bank to keep sensitive customer data on premises to satisfy regulatory requirements while still leveraging cloud scalability for less-critical or transient workloads. The defining characteristic is the orchestrated data and application portability between the two environments, which neither private nor public cloud offered alone.

Correct. A hybrid cloud combines on-premises infrastructure with a public cloud like Azure. The institution keeps sensitive financial data on-premises (private cloud) and runs analytics on anonymized data in Azure (public cloud), meeting both regulatory compliance and scalability needs.

Broad network access is the characteristic that cloud resources are available over the network and can be accessed by standard protocols from a wide variety of client devices (e.g., smartphones, laptops). The scenario explicitly describes this capability.

Data sovereignty is the correct choice because it directly addresses legal and regulatory requirements concerning the physical and jurisdictional location of data. When evaluating cloud providers, organizations must ensure that their data is stored and processed in regions that comply with applicable laws, such as the GDPR in the EU or data localization laws in other countries. This goes beyond mere physical storage, encompassing who has legal authority to access the data and under what legal framework.

Cloud providers aggregate demand from millions of customers, enabling them to purchase hardware, bandwidth, and energy at volume discounts. These procurement advantages lower the per-unit cost of computing, which providers pass on to customers in the form of pay-as-you-go pricing. This is the definition of economies of scale in cloud computing.

Resource pooling is the correct answer. The provider's physical and virtual resources are pooled to serve many customers, with strong isolation between tenants. This allows Microsoft to achieve economies of scale while keeping each customer's data separate.

Correct. The customer always owns their data and identities, regardless of the service model. Data classification, access control, and identity management are perpetual customer responsibilities. Even in SaaS, the customer must manage who has access to the application and what data they can see.

Platform as a Service (PaaS) is the correct choice for building a custom mobile app backend because it abstracts the underlying infrastructure, including servers, storage, and networking, while also providing a fully managed platform for application hosting. The cloud provider handles critical operational concerns such as automatic scaling, patching, and high availability, allowing the development team to concentrate on writing application code and business logic. PaaS also offers built-in features like database management, authentication, and app monitoring, which are directly relevant to backend development. This makes it the most efficient model for deploying a complete, custom application without the burden of infrastructure maintenance.

Elasticity is the ability to dynamically allocate and deallocate cloud resources in response to changing workload demands. The company wants to automatically add VMs during sales traffic spikes and remove them afterward, paying only for what is used. This is a classic example of elasticity.

Pay-as-you-go is the correct pricing model because Azure meters resources actually consumed—compute hours, storage, data transfer—and bills monthly with no upfront cost, no termination fee, and no long-term commitment. This gives a CFO full flexibility to scale usage up or down dynamically, converting on-premises capital expenditure into operational expenditure. Unlike spot or reserved options, pay-as-you-go is the standard default for all Azure services and provides predictable, consumption-based billing for any workload type.

Elasticity is the correct answer because a startup expecting rapid growth needs a cloud model that automatically provisions and de-provisions compute resources to match real-time demand. Azure Virtual Machine Scale Sets and Autoscale capabilities add or remove instances based on metrics like CPU load or queue length, ensuring the application can handle spikes without manual intervention and avoid over-provisioning during lulls. This dynamic scaling directly addresses unpredictable growth patterns, making elasticity the foundational characteristic for this scenario.

Hybrid cloud is correct because it combines a private cloud or on-premises data center with the public Azure cloud, linked by secure connectivity (e.g., VPN, ExpressRoute). This architecture lets the company store PII on its own physical servers to meet data residency rules, while running computation-heavy analytics in Azure, scaling beyond local capacity. Tools like Azure Arc provide unified management and governance across both environments, ensuring compliance and operational consistency.

Consumption-based pricing charges per unit of resource used, so costs track actual usage rather than pre-provisioned capacity. This eliminates upfront capital expenditure and the waste of paying for idle resources, precisely matching the statement's description.

Pay-as-you-go charges only for actual usage with no upfront commitment, making it the lowest-cost option for a predictable but intermittent workload like a nightly 4-hour batch job.

An SLA is a formal contract that defines the guaranteed uptime percentage (like 99.95%) for a service, and specifies the service credits or compensation if the provider fails to meet that commitment. In Azure, SLAs are per-service and tier-specific, so the 99.95% figure for virtual machines reflects a contractual promise, not just a measurement or target. The guarantee is legally binding and comes with defined remedies for customers.

Consumption-based pricing (pay-as-you-go) is the core cloud model that lets the startup rent GPU instances only during model training or inference, with zero capital expenditure and no idle-time charges. Because ML workloads are often bursty, the startup can scale up compute for hours or days, then scale to zero, paying strictly for vCPU-hours, memory, and GPU-seconds consumed. This directly aligns with the scenario's need to run complex ML without buying expensive on-premises hardware.

Azure IaaS delivers virtualized compute, storage, and networking, with the cloud provider owning and maintaining the physical servers, power, cooling, and hardware lifecycle. The customer gains full control over the guest OS, middleware, and applications, allowing custom configurations and software installations. This means you avoid managing physical servers entirely—there is no rack mounting or hardware replacement—yet you still need to patch the OS and manage the application stack, which is what makes IaaS distinct for accommodating development workloads that require control. In this context, IaaS is correct because it completely abstracts the hardware layer while permitting the level of control needed to deploy a custom environment.

Azure's global footprint includes dozens of regions worldwide, and you can deploy the same application to multiple regions in minutes using tools like ARM templates, Azure DevOps, or Azure Front Door/Traffic Manager for routing. This 'go global' speed comes from having capacity already built and available on demand—no procurement, construction, or hardware installation is necessary. It directly answers the phrase 'go global' in the context of cloud agility.

For Azure VMs (IaaS), the customer maintains full administrative control over the guest operating system, including installing updates, patches, and security configurations. Microsoft's responsibility ends at the physical host, hypervisor, and Azure-managed infrastructure, so the customer must patch the guest OS to keep the VM secure and compliant. This is a core tenet of the shared responsibility model: the customer owns everything inside the VM, while Azure provides the underlying infrastructure.

A private cloud is built for exclusive use by one organization, either on-premises or in a dedicated hosted environment. All virtualization, orchestration, storage, and network components are dedicated to that single tenant, allowing the organization to select custom hypervisors, define isolation policies, and manage the full infrastructure stack. This dedicated, single-tenant basis directly provides the maximum infrastructure control specified in the scenario.

Azure's carbon footprint reduction stems from a combination of renewable energy procurement (such as wind and solar power purchase agreements), energy-efficient data center design (including advanced cooling like adiabatic evaporative cooling), and investments in carbon-free energy generation to match electricity consumption. These measures directly lower the carbon intensity of cloud workloads, which is the specific mechanism by which a company using Azure can reduce its own operational emissions.

Azure region pairs are two regions within the same geography that are at least 300 miles apart and are directly coupled by Azure for two critical purposes: disaster recovery and coordinated platform updates. When a region fails, the paired region is prioritized for recovery, and during planned maintenance, Azure sequences updates so that only one region in the pair is updated at a time, reducing the risk of simultaneous downtime. This ensures that customers using paired regions for geo-redundancy (e.g., with GRS) have a designed recovery path that is physically and logically isolated from the primary region.

PaaS delivers a managed runtime where the provider patches the OS, middleware and scaling, so the startup uploads code only. This satisfies the stem's constraint of no server purchasing or management, unlike IaaS, which still leaves patching and capacity to the customer.

The composite SLA for the ERP system is determined by multiplying the availability percentages of its dependent components. Since the ERP application (99.9% SLA) relies directly on the Azure SQL Database (99.99% SLA), their individual availabilities are multiplied: 0.999 * 0.9999 = 0.9989001. This calculation reflects the serial dependency where the failure of either component impacts the entire system's uptime, resulting in an overall availability of approximately 99.89%.

Pay-as-you-go bills only for consumed compute, with no upfront commitment, and scales elastically with demand. This directly satisfies the stem's constraints of zero upfront cost and usage-based charging, unlike reserved or spot models that require commitment or tolerate interruption.

An Azure virtual machine is an IaaS (Infrastructure-as-a-Service) workload. Under the shared responsibility model, Microsoft manages the physical hosts and hypervisor, but the customer is fully responsible for the guest operating system — including security patches, updates, and OS-level configuration. Because this is the only resource listed where the company must perform OS patching, this is the correct answer.

Moving to Azure eliminates the capital expenditure (CapEx) for purchasing and maintaining physical servers, storage, and datacenter infrastructure. Instead of buying hardware and managing refresh cycles, you subscribe to a pay-as-you-go operational expenditure (OpEx) model, paying only for the resources you use and scaling on demand. This shift removes the burden of hardware maintenance, power, cooling, and facility costs, making it a core financial benefit of cloud adoption.

This is correct. Under the shared responsibility model for SaaS, the customer manages user identities, data classification, and access control. The provider secures the platform and infrastructure, but the customer must ensure only authorized users access the data and that data is handled appropriately.

Elimination of internet connectivity requirements is incorrect because Azure's control plane and data plane are both reached over network connections; even hybrid deployments using Azure Stack Hub still synchronize with Azure over the internet or ExpressRoute. Therefore cloud computing does not remove connectivity needs—it makes them a prerequisite, so this is a dependency, not a benefit.

Agility is the cloud characteristic that enables rapid adaptation to changing business needs through fast provisioning of resources. The scenario directly demonstrates this by reducing deployment time from weeks to hours, which allows the company to respond quickly to new opportunities or internal demands. Unlike scalability, agility is specifically about the speed and flexibility of provisioning, not just the ability to adjust capacity.