This chapter introduces the foundational concept of cloud computing, which is the first topic you'll encounter in the AZ-900 exam (Domain: Cloud Concepts, Objective 1.1). Understanding cloud computing is critical because it underpins every Azure service you'll learn about later. This domain accounts for approximately 25-30% of the exam questions, so mastering this chapter is essential. We'll define cloud computing, explain its key characteristics, and explore the business problems it solves, setting the stage for deeper dives into Azure services.
Jump to a section
Imagine you run a small bakery. You need electricity to power your ovens, mixers, and lights. Instead of building your own power plant (which would cost millions and take years), you simply plug into the existing electrical grid. You pay only for the electricity you consume, and the grid is maintained by a utility company that handles all the complexity of generation, distribution, and reliability. Cloud computing works the same way: instead of building and maintaining your own data center (the power plant), you rent computing resources (servers, storage, databases) from a cloud provider like Microsoft Azure. You pay only for what you use (pay-as-you-go), and Azure handles the underlying infrastructure, security, and scaling. Just as the electrical grid can handle a sudden surge in demand during a holiday rush without you needing to build a new generator, Azure can automatically scale resources up or down to meet your business needs. The key mechanism is virtualization and multitenancy: Azure uses hypervisors to partition physical servers into multiple virtual machines, and it manages these resources across thousands of customers, achieving economies of scale that make it affordable for even the smallest bakery.
What is Cloud Computing and What Business Problem Does It Solve?
Cloud computing is the delivery of computing services—including servers, storage, databases, networking, software, analytics, and intelligence—over the internet ("the cloud") to offer faster innovation, flexible resources, and economies of scale. You typically pay only for cloud services you use, helping lower your operating costs, run your infrastructure more efficiently, and scale as your business needs change.
The core business problem cloud computing solves is the high cost and complexity of owning and operating physical data centers. Traditionally, companies had to purchase servers, software licenses, networking equipment, and backup power supplies, and hire staff to manage it all. This capital-intensive model required large upfront investments (CapEx) and long procurement cycles. Cloud computing shifts this to an operational expenditure (OpEx) model, where you pay as you go, reducing financial risk and enabling faster time to market.
How Cloud Computing Works: The Mechanism Step by Step
Cloud computing relies on two key technologies: virtualization and multitenancy. Virtualization allows a single physical server to host multiple virtual machines (VMs), each running its own operating system and applications. A hypervisor (like Microsoft Hyper-V) manages this abstraction, allocating CPU, memory, and storage to each VM as needed. Multitenancy means that the same physical infrastructure serves multiple customers (tenants) securely, with isolation between tenants.
When you provision a virtual machine in Azure, the following happens behind the scenes: 1. You specify the VM size, operating system, and region. 2. Azure's fabric controller (a management system) identifies a physical server with available capacity in that region. 3. The hypervisor creates a new VM with the requested resources, loading the OS image from Azure's blob storage. 4. The VM is connected to a virtual network, and a public IP (if requested) is assigned. 5. You can now connect to the VM via Remote Desktop or SSH.
This entire process takes minutes, compared to weeks for ordering and setting up physical hardware.
Key Components, Tiers, and Pricing Models
Cloud computing services fall into three main categories: - Infrastructure as a Service (IaaS): Provides virtualized computing resources over the internet. Examples: Azure Virtual Machines, Azure Virtual Networks. You manage the OS, middleware, and applications; Azure manages the physical hardware. - Platform as a Service (PaaS): Provides a managed platform for developing, running, and managing applications. Examples: Azure App Service, Azure SQL Database. You deploy your code; Azure handles the OS, runtime, and infrastructure. - Software as a Service (SaaS): Provides fully managed software applications over the internet. Examples: Microsoft 365, Dynamics 365. You just use the application; everything else is managed.
Pricing models include: - Pay-as-you-go: You pay for what you use, per hour or per minute, with no upfront commitment. - Reserved Instances: You commit to a 1- or 3-year term for a discount (up to 72% compared to pay-as-you-go). - Spot pricing: You bid on unused Azure capacity for significant discounts, but the instance can be evicted when Azure needs the capacity back.
Comparison to On-Premises Equivalent
In an on-premises data center, you own all hardware and software. You must plan for peak capacity, leading to overprovisioning and waste. Maintenance (power, cooling, hardware replacements) is your responsibility. Scaling requires purchasing and installing new hardware, which can take weeks. Security updates and compliance are also your burden.
With cloud computing, Azure handles all physical infrastructure, provides elastic scaling (you can add or remove resources in minutes), and offers built-in security and compliance certifications (e.g., ISO 27001, SOC 2). You only manage the services you choose (e.g., VMs, databases). The trade-off is less control over the physical layer and potential dependency on internet connectivity.
Azure Portal and CLI Touchpoints
You can interact with Azure cloud services through: - Azure Portal: A web-based GUI for managing all Azure resources. You can create VMs, configure networks, and monitor usage. - Azure CLI: A command-line tool for scripting and automation. For example, to create a VM:
az vm create --resource-group myResourceGroup --name myVM --image UbuntuLTS --admin-username azureuser --generate-ssh-keysAzure PowerShell: Similar to CLI but uses PowerShell cmdlets. Example:
New-AzVm -ResourceGroupName 'myResourceGroup' -Name 'myVM' -Location 'EastUS' -Image 'UbuntuLTS'Azure Cloud Shell: An interactive, browser-accessible shell in the Azure Portal that includes both CLI and PowerShell.
These tools allow you to manage your cloud resources programmatically, enabling infrastructure as code (IaC) practices.
Define Your Requirements
Before moving to the cloud, assess your workloads. Determine which applications are suitable for migration (e.g., stateless web apps are easier than legacy databases). Identify compliance requirements (e.g., data residency, regulatory standards). Estimate resource needs: CPU, memory, storage, and network bandwidth. This step ensures you choose the right Azure services and avoid overprovisioning.
Choose a Cloud Deployment Model
Decide between public cloud (Azure), private cloud (on-premises or hosted), or hybrid cloud (mix of both). Public cloud offers maximum scalability and lower upfront costs. Private cloud provides more control and isolation. Hybrid cloud allows you to keep sensitive data on-premises while leveraging cloud for burst capacity. For AZ-900, know the definitions and use cases of each.
Select a Service Model
Choose IaaS, PaaS, or SaaS based on your level of control required. IaaS gives you full control over the OS and applications but requires more management. PaaS abstracts the OS and runtime, letting you focus on code. SaaS is fully managed. The exam tests your ability to match business scenarios to the appropriate service model.
Provision Resources in Azure
Using the Azure Portal, CLI, or PowerShell, create a resource group (logical container) and then deploy services like VMs, databases, or web apps. For example, to create a storage account: `az storage account create --name mystorageaccount --resource-group myResourceGroup --location eastus --sku Standard_LRS`. Azure's fabric controller allocates physical resources and configures networking automatically.
Configure Scaling and Monitoring
Set up autoscaling rules (e.g., add VM instances when CPU > 75%) and monitoring with Azure Monitor to track performance and costs. Use Azure Cost Management to analyze spending. This step ensures your cloud environment remains cost-effective and responsive to demand. The exam may ask about scaling options (vertical vs. horizontal) and monitoring tools.
Scenario 1: Startup Launching a Mobile App A startup wants to launch a mobile app with unpredictable traffic. They use Azure App Service (PaaS) to host the backend API and Azure Cosmos DB for the database. They configure autoscaling to handle spikes during marketing campaigns. The team uses Azure DevOps for CI/CD. Cost is low initially (pay-as-you-go) and scales with success. A common mistake: not setting budget alerts, leading to unexpected bills when traffic surges.
Scenario 2: Enterprise Migrating Legacy Apps A large bank must migrate a legacy .NET application to the cloud but keep sensitive customer data on-premises due to regulations. They choose a hybrid cloud approach: the app runs on Azure VMs (IaaS) in a virtual network connected to the on-premises data center via Azure ExpressRoute (a private dedicated connection). The database remains on-premises but is replicated to Azure SQL Database for disaster recovery. This reduces capital expenditure on new hardware while meeting compliance. Pitfall: misconfiguring network security groups (NSGs) can expose the on-premises connection.
Scenario 3: Media Company Processing Video A media company needs to render high-definition videos on demand. They use Azure Batch (PaaS) to spin up hundreds of VM instances only when rendering jobs are submitted. Spot VMs are used to cut costs by 80%. After rendering, the VMs are deallocated, so they pay only for compute time. A common issue: not using low-priority VMs correctly, resulting in interruptions during rendering.
Objective 1.1: Describe cloud computing The AZ-900 exam tests your understanding of cloud computing definitions, the shared responsibility model, and the differences between CapEx and OpEx. Specifically, you must:
Define cloud computing.
Describe the shared responsibility model.
Define cloud models (public, private, hybrid).
Describe the consumption-based model.
Compare CapEx vs. OpEx.
Common Wrong Answers and Why Candidates Choose Them 1. "Cloud computing means you own the hardware." – Wrong. In the cloud, you rent resources; the provider owns the physical hardware. Candidates confuse cloud with traditional hosting. 2. "In a public cloud, resources are dedicated to one customer." – Wrong. Public cloud uses multitenancy; resources are shared but isolated. Candidates think "public" means shared like a public park. 3. "The cloud provider is responsible for everything." – Wrong. The shared responsibility model divides security: the provider secures the physical infrastructure, but you secure your data and configurations. Candidates assume the provider handles all security. 4. "CapEx is better because it's cheaper long-term." – Wrong. While CapEx may be cheaper over many years, it requires high upfront investment and lacks flexibility. The exam favors OpEx for agility and lower risk.
Specific Terms and Values - Consumption-based model: You pay for what you use (no upfront cost). - CapEx: Capital Expenditure (upfront, fixed cost). - OpEx: Operational Expenditure (ongoing, variable cost). - Shared responsibility model: Security of the cloud vs. security in the cloud. - Cloud models: Public, private, hybrid.
Edge Cases and Tricky Distinctions - The exam may describe a scenario where a company uses both on-premises and cloud resources—this is hybrid cloud, not public or private. - If a company uses only its own on-premises data center, it's a private cloud (if it uses cloud-like virtualization) or just traditional on-premises. The exam expects you to recognize that private cloud implies cloud characteristics (self-service, scalability) even if on-premises. - The consumption-based model is often contrasted with fixed-cost models; remember that reserved instances are still consumption-based but with a commitment discount.
Memory Trick Use the acronym "PCH" for cloud models: Public (shared infrastructure), Community (shared by several organizations), Hybrid (mix of public/private). But for AZ-900, focus on Public, Private, Hybrid.
Decision Tree for Eliminating Wrong Answers - If the question asks about who manages physical security? → Provider. - If it asks about who manages data encryption? → Customer (unless using a managed service like Azure SQL). - If it asks about cost model and the scenario has unpredictable usage? → OpEx/pay-as-you-go. - If the scenario mentions a company with its own data center and also uses Azure? → Hybrid cloud.
Cloud computing delivers computing services over the internet with pay-as-you-go pricing.
The shared responsibility model: Microsoft secures the cloud infrastructure; you secure your data and configurations.
Three cloud deployment models: public, private, hybrid.
Three service models: IaaS, PaaS, SaaS.
CapEx involves upfront costs; OpEx is variable and consumption-based.
Azure offers a 99.9% SLA for most services (uptime guarantee).
The consumption-based model aligns costs with actual usage, reducing waste.
Cloud computing enables rapid elasticity: resources scale up/down automatically.
Azure regions are pairs of data centers for disaster recovery (e.g., East US paired with West US).
Azure provides a free account with $200 credit for 30 days and 12 months of popular services free.
These come up on the exam all the time. Here's how to tell them apart.
Capital Expenditure (CapEx)
Upfront cost to purchase hardware/software
Fixed monthly cost (depreciation)
Higher financial risk if demand drops
Long procurement cycles (weeks/months)
Assets are owned and appear on balance sheet
Operational Expenditure (OpEx)
No upfront cost; pay as you go
Variable cost based on usage
Lower financial risk; scales with demand
Instant provisioning (minutes)
No ownership; costs are operational expenses
Public Cloud
Shared infrastructure across multiple tenants
Managed by third-party provider (e.g., Azure)
Accessible over public internet
Lower cost due to economies of scale
Less control over physical location
Private Cloud
Dedicated infrastructure for one organization
Managed by organization or third-party
Accessible via private network
Higher cost due to dedicated resources
Full control over security and compliance
Mistake
Cloud computing is just someone else's computer.
Correct
While true at a basic level, cloud computing involves complex orchestration, virtualization, and multitenancy that provide scalability, reliability, and security far beyond a single remote computer. Azure manages thousands of servers globally, with automatic failover and load balancing.
Mistake
The cloud is always cheaper than on-premises.
Correct
Cloud can be cheaper, but not always. For predictable, steady-state workloads, reserved instances can be cost-effective, but for 24/7 high-usage workloads, on-premises may be cheaper. The cloud's advantage is flexibility, not necessarily lower cost.
Mistake
Once you move to the cloud, you're locked in forever.
Correct
Vendor lock-in is a risk, but Azure supports open standards and provides tools like Azure Migrate to move workloads to other clouds or on-premises. However, migrating large datasets can be costly and time-consuming.
Mistake
Cloud computing is less secure than on-premises.
Correct
Azure has more security certifications than most on-premises data centers. Security is a shared responsibility; the provider secures the infrastructure, but you must configure your services correctly. Misconfigurations (e.g., open storage containers) are the most common cause of breaches.
Mistake
You need an internet connection to use cloud services.
Correct
While internet is typical, Azure offers offline data transfer via Azure Data Box and private connections via ExpressRoute, which bypass the internet. Some services (like Azure Stack) can run disconnected.
IaaS provides virtualized hardware (VMs, storage, networks) that you manage. PaaS provides a managed platform (OS, runtime) where you deploy your code. SaaS provides ready-to-use software (e.g., Microsoft 365). Think of it like pizza: IaaS is the dough (you add toppings), PaaS is the ready-made crust (you add sauce and cheese), SaaS is the delivered pizza (everything is done). For the exam, match scenarios: if the company manages the OS, it's IaaS; if they only manage the application, it's PaaS; if they just use the software, it's SaaS.
It defines who is responsible for security in the cloud. Microsoft is responsible for the security OF the cloud (physical data centers, network, hypervisor). You are responsible for security IN the cloud (your data, configurations, access policies). For example, Azure secures the VM host, but you must patch the guest OS and configure firewalls. The exam may ask: who is responsible for patching a VM's OS? Answer: you (the customer), unless using a PaaS service like Azure App Service.
CapEx (Capital Expenditure) is an upfront cost to purchase assets (e.g., servers) that are depreciated over time. OpEx (Operational Expenditure) is ongoing costs for services (e.g., cloud subscriptions). Cloud computing shifts from CapEx to OpEx because you pay as you go instead of buying hardware. The exam favors OpEx for its flexibility and lower risk. A common question: 'A startup wants to avoid large upfront costs. Which model should they use?' Answer: OpEx/consumption-based.
The six key benefits are: High availability (services remain available despite failures), Scalability (resources adjust to demand), Elasticity (automatic scaling), Agility (rapid deployment), Geo-distribution (deploy globally), and Disaster recovery (data backup and recovery). The exam may ask you to identify which benefit matches a scenario. For example, 'A company needs to handle sudden traffic spikes' → Elasticity.
A hybrid cloud combines a public cloud (like Azure) with a private cloud or on-premises infrastructure, allowing data and applications to be shared between them. For example, a bank might keep sensitive data on-premises but use Azure for compute-intensive analytics. The exam tests that hybrid cloud provides flexibility and leverages existing on-premises investments. A common wrong answer is to confuse hybrid with multi-cloud (using multiple public clouds).
Azure uses multiple data centers within a region, and region pairs for disaster recovery. Services like Azure Load Balancer distribute traffic across healthy instances. Availability Zones (physically separate data centers within a region) protect against local failures. The SLA (Service Level Agreement) guarantees uptime, e.g., 99.9% for a single VM. The exam may ask about availability zones vs. availability sets.
It means you pay only for the resources you actually use, with no upfront commitment. For example, you pay per hour for a VM or per GB for storage. This model reduces waste and allows you to scale costs with demand. The exam contrasts this with fixed pricing (like reserved instances, which also use consumption but with a discount for commitment). A common question: 'Which pricing model is best for unpredictable workloads?' Answer: pay-as-you-go.
You've just covered What is Cloud Computing? — now see how well it sticks with free AZ-900 practice questions. Full explanations included, no account needed.
Done with this chapter?