1.1 What Is Cloud Computing?

Quick Answer: Cloud computing is the delivery of computing services — servers, storage, databases, networking, software, analytics, and intelligence — over the internet ("the cloud"). You rent capacity on demand and pay only for what you consume, trading large upfront hardware purchases (CapEx) for variable operating costs (OpEx).

Exam Context

The AZ-900: Microsoft Azure Fundamentals exam contains roughly 40-60 questions, gives you 45 minutes of answering time (about 65 minutes of total seat time), and requires a scaled score of 700 out of 1000 to pass. The fee is USD $99 (regional pricing varies). Domain 1, Describe cloud concepts, is worth 25-30% of your score — so 10-18 questions here typically decide pass or fail. These are vocabulary and scenario questions, not calculations. Memorize the precise meaning of each benefit term, because Microsoft writes distractors using adjacent words (scalability vs elasticity, availability vs reliability).

The NIST Definition

The National Institute of Standards and Technology (NIST) provides the widely accepted definition of cloud computing through five essential characteristics:

Traditional IT vs Cloud Computing

Traditional (on-premises)

• Buy hardware upfront — servers, SAN storage, switches, firewalls.
• Build and staff data centers — power, cooling, physical security, fire suppression.
• Guess capacity months ahead: over-provision (wasted money) or under-provision (outages).
• Own all maintenance — patching, firmware, hardware replacement, refresh cycles.
• Capital expenditure (CapEx): large up-front spend, depreciated over years.
• Slow provisioning: weeks to months to procure and rack new gear.

Cloud (Azure)

• Rent on demand — no hardware purchases.
• Microsoft runs the data centers, hardware, cooling, power, and physical security.
• Scale dynamically in minutes to match real demand.
• Pay only for what you use with per-second or per-minute granularity.
• Operating expenditure (OpEx): variable, consumption-based, deductible in-year.
• Fast provisioning: seconds to minutes.

On the Exam: Microsoft's stock list of cloud advantages — trade CapEx for OpEx, stop guessing capacity, increase speed and agility, stop running data centers, benefit from economies of scale, go global in minutes — appears almost verbatim in exam stems. Recognize the phrasing.

The Benefits Vocabulary (heavily tested)

Microsoft groups cloud benefits into a specific set of terms. Map each scenario word to the right term:

• High availability — the service stays reachable. Azure publishes uptime in Service Level Agreements (SLAs) such as 99.9% or 99.99%.
• Reliability — the system recovers from failures and keeps working (resilience). Closely paired with disaster recovery and the design of resources across regions and availability zones.
• Scalability — add/remove capacity to meet demand. Vertical = scale up/down (a bigger VM, e.g., 4 vCPU to 8 vCPU). Horizontal = scale out/in (more instances, e.g., 2 VMs to 10).
• Elasticity — scaling that happens automatically in response to live metrics (CPU, memory, queue length).
• Agility — deploy and reconfigure in minutes instead of weeks, so teams experiment faster.
• Predictability — both cost predictability (Pricing Calculator, Cost Management) and performance predictability (autoscale, load balancing).
• Security — defense-in-depth, encryption, and tools like Microsoft Defender for Cloud.
• Governance — enforce standards with Azure Policy, blueprints, and tags.
• Manageability — manage of the cloud (autoscale, alerts) and in the cloud (portal, CLI, templates).

CapEx vs OpEx

Common Trap: A scenario that says "automatically add servers during a Black Friday spike" is elasticity, not plain scalability. A scenario that says "the website kept serving users even when one data center failed" is reliability/high availability, not scalability.

Worked Scenario: Reading the Vocabulary

Consider a startup that today owns ten servers in a leased rack. Three problems recur. First, every December its traffic triples and the servers cannot keep up, so checkout fails — this is a scalability and elasticity gap, because the company cannot add capacity fast enough, let alone automatically. Second, when one server's power supply dies, an entire service goes offline for hours — a reliability and high availability gap, because there is no redundancy or fast failover. Third, the finance team cannot predict next quarter's hardware bill — a cost predictability gap.

Moving to Azure addresses each one with a distinct benefit term, and AZ-900 expects you to map the symptom to the term:

• December spikes are absorbed by VM Scale Sets that autoscale (elasticity) and can scale out across more instances (horizontal scalability).
• A failed component no longer causes downtime because resources are spread across availability zones and backed by an SLA (high availability and reliability).
• The finance team uses the Azure Pricing Calculator and Cost Management to forecast and track spend (cost predictability and manageability).

Distinguishing the Easily-Confused Pairs

Keep these distinctions sharp: Microsoft routinely offers two correct-sounding benefit terms in the same question, and only the precise match earns the point.