Four benefits the exam expects you to name and distinguish: availability & scalability, reliability & predictability, security & governance, and manageability. Each one is a short, sharp idea — the goal today is precision, not depth.
Your app is getting hammered by a traffic spike. You respond by adding three more VM instances behind a load balancer, rather than upgrading the CPU/RAM on the one VM you had. What's that called? Keep your answer in mind — the next section names it precisely.
High availability means a system stays up despite disruptions. Azure backs this with per-service SLAs (service-level agreements) — uptime guarantees like 99.9% or 99.99% that you design around.
Scalability means adjusting resources to meet demand — and because pricing is consumption-based, scaling down when demand drops also cuts your bill. There are two distinct kinds, and the exam expects you to tell them apart:
| Type | What changes | Example |
|---|---|---|
| Vertical scaling (scale up/down) | The capability of one resource | Adding more CPU/RAM to a single VM |
| Horizontal scaling (scale out/in) | The number of resource instances | Adding more VMs or containers behind a load balancer |
So the calibration scenario — adding more VM instances — is horizontal scaling, specifically scaling out.
Reliability is a system's ability to recover from failure and keep functioning. The cloud supports this through decentralization — your app can span multiple Azure regions, so a catastrophic event in one region doesn't take you down everywhere. (This is one pillar of the Azure Well-Architected Framework, which you'll see referenced again later.)
Predictability splits into two flavors:
Templates enforce technical/regulatory standards on every resource you deploy, and cloud-based auditing flags anything that drifts out of compliance — so governance scales with you instead of being a manual, one-off check. On the security side, your choice of service type sets the tradeoff: IaaS gives you maximum control (you patch the OS yourself), while PaaS/SaaS hand patching and maintenance to the provider automatically. Cloud providers also absorb large-scale threats like DDoS attacks that would be expensive to defend against on your own.
This one has a genuinely exam-tested split — two different things both called "manageability," distinguished by preposition:
| Category | What it means | Examples |
|---|---|---|
| Management of the cloud | Managing your resources — the things happening to your infrastructure | Auto-scaling, deployment templates, health monitoring with auto-replace, metric-based alerts |
| Management in the cloud | How you reach in and control your environment — the tools you use | Web portal, CLI, APIs, PowerShell |
The "of the cloud" vs. "in the cloud" wording is easy to blur under exam pressure — if question 4 below doesn't feel solid yet, that's worth flagging to your teacher now rather than on exam day.