1.1Cloud Computing and Its Benefits

Cloud computing means renting computing resources—servers, storage, databases, networking, and software—over the internet, on demand. Instead of building your own datacenter, you use up-to-date infrastructure and pay only for what you use. Microsoft Azure is one leading example of such a (public) cloud.

Why did the cloud take off? Traditionally you bought lots of servers up front to handle peak demand, then installed and ran them yourself. That meant heavy up-front investment, weeks of procurement, and wasted capacity outside of peaks. The cloud turned this "own and hold" model into a "rent when needed" model. You can start a server in minutes and stop it—and its billing—when you are done. This on-demand and pay-as-you-go nature is the source of nearly all of the cloud benefits.

1.1.1Core characteristics of the cloud

• On-demand self-service: provision resources yourself, instantly, from a portal or CLI—no waiting on an admin.
• Broad network access: usable from anywhere over the internet.
• Resource pooling: the provider shares a huge pool of resources across many customers, achieving economies of scale.
• Measured service (pay-as-you-go): usage is metered and you are billed for what you consume.

These characteristics stand out against traditional on-premises (owning and running your own gear). On-premises, every expansion meant buying hardware and spending weeks installing, cabling, and setting up the OS. In the cloud, the same is done self-service in minutes, and deleting resources stops their cost. The table contrasts the two.

1.1.2Key benefits of the cloud

AZ-900 asks about a family of terms describing cloud benefits. Four of them—high availability, reliability, scalability, and elasticity—are close in meaning and easily confused, so we first separate them carefully one by one.

• High availability (HA): keeping the service running even when failures occur—achieved with redundancy and availability zones, and promised via an SLA (covered later).
• Reliability: the ability to recover from and withstand failures or disasters—e.g., geo-replication and backups. If HA means "do not fall," reliability means "get back up if you do."
• Scalability: the ability to change capacity with demand, possibly manually. Scale up (vertical) = a bigger machine / scale out (horizontal) = more machines.
• Elasticity: scaling out/in automatically with load (autoscaling). Think of it as scalability made automatic.
• Agility: provisioning resources in minutes, shortening time to market.
• Predictability: predictability of both performance (capacity you can count on) and cost (you can estimate spend).
• Security, governance, manageability: strong security features, policy-based control (governance), and easier operations via templates and automation (manageability).

The remaining security, governance, and manageability matter too. Security: beyond hardened physical datacenters, Azure provides advanced features like encryption, firewalls, and threat detection (though, per the shared responsibility model in the next section, configuration duties remain with you). Governance: Azure Policy, tags, and locks let you enforce "use that follows organizational rules" by design. Manageability: templates (ARM/Bicep), automation, and the portal/CLI/PowerShell let you codify, reproduce, and automate your setup. These pay off most for larger organizations that need control.

1.1.2.1Distinguishing the four confusing terms

1.1.2.2Scale up vs scale out

1.1.2.3Why it works: economies of scale and how HA is achieved

Behind the cloud being cheaper and faster lie economies of scale. Providers procure and operate enormous fleets of servers globally and share (pool) them across many customers. Co-using a scale no single company could afford lowers per-unit cost and spreads the latest hardware and security investments across everyone. From your side, the essence is that you no longer buy for the peak yourself, eliminating the waste of over-provisioning.

How is high availability actually achieved? The key is redundancy—provisioning multiple components of the same role so the rest carry on if one fails. In Azure, a region contains physically separated availability zones (datacenter groups with independent power, cooling, and networking); distributing across them keeps the service running even if one zone goes down. Providers also commit to uptime (e.g., 99.9%) by contract via an SLA (service-level agreement). The relationship: "HA = make it hard to go down by design," "SLA = guarantee that level numerically" (regions and availability zones are covered in the next chapter).

1.1.3Peripheral knowledge: agility and global reach

The value of the cloud is not only "it is cheaper." Often the bigger gain is agility: from idea to experiment, you no longer wait weeks for procurement—you spin up an environment in minutes and throw it away if it fails. That speed directly drives business competitiveness. Azure also has regions (clusters of datacenters) worldwide, so you can deploy globally in a few clicks (global reach). These are advantages that are hard to get by "owning."

1.1.4Section summary

• The cloud = renting resources over the internet on demand, pay-as-you-go (e.g., Azure)
• HA = stay up / reliability = recover / scalability = change capacity / elasticity = auto change
• Scale up = vertical (bigger machine) / scale out = horizontal (more machines)
• Benefits go beyond cost: agility, global reach, predictability (performance/cost), security/governance/manageability