Skip to content
Tutorials Center
My learning

Topics

Lesson 2 of 57 ยท Cloud and Azure Basics

Azure at a Glance

A tour of Microsoft Azure: regions, region pairs, availability zones, geographies, the main service families and how a request travels to a datacenter.

Azure is Microsoft's public cloud: a global network of datacenters plus hundreds of services built on top of them, from virtual machines to AI models. Almost every decision you make in Azure starts with one question: where will this run? The answer affects speed, price, legal compliance and how well your system survives failures.

This lesson draws the map. You will learn the geographic building blocks (regions, region pairs, availability zones, geographies), the families of services you will meet in the rest of the tutorial, and how a user's request finds its way to your code. The commands and diagrams are read-only. The runnable example is a JavaScript stand-in that scores regions by latency, price and compliance.

Regions

An Azure region is a set of datacenters, deployed within a latency-defined perimeter and connected by a dedicated low-latency network. Examples are eastus, westeurope, uksouth, japaneast and brazilsouth. When you create most resources, you pick a region, and the resource lives there.

Regions differ in four ways that matter:

  • Distance to your users. Light in fibre needs roughly 5 ms per 1,000 km one way, so a user in Tokyo talking to a server in Virginia pays a real latency penalty on every request.
  • Price. The same virtual machine size can cost different amounts in different regions.
  • Service availability. New services and the newest hardware arrive in some regions first. Not every region offers every service or VM size.
  • Compliance and data residency. Laws and company policies may require data to stay in a country or economic area.

You can list the regions your subscription can use with the CLI (described in Portal, Azure CLI and PowerShell):

Terminal
az account list-locations --query "[].{name:name, display:displayName}" --output table

Availability zones

Inside many regions, Azure offers availability zones: physically separate locations, each with its own power, cooling and networking. A zone-enabled region has at least three of them. If one zone loses power, the others keep running.

You use zones by placing copies of a resource in different zones, or by choosing a zone-redundant service tier, where Azure spreads the replicas for you. Zones protect against a datacenter-level failure while keeping latency between copies very low, usually a couple of milliseconds. They do not protect against a whole-region disaster; that needs a second region.

Two kinds of zone-aware services are worth knowing now:

  • Zonal services let you pin a resource to a specific zone (for example, a VM in zone 2).
  • Zone-redundant services replicate across zones automatically (for example, zone-redundant storage).

Region pairs and geographies

Many regions are paired with another region in the same geography, usually hundreds of kilometres away. Examples are East US paired with West US, and North Europe paired with West Europe. Pairing matters for a few platform behaviours: Azure rolls out planned updates to one region of a pair at a time, prioritizes recovery of one region in each pair during a widespread outage, and some geo-redundant storage options replicate to the paired region.

A geography is a discrete market, typically one or more countries, that preserves data residency and compliance boundaries. Regions in the same geography share the same legal framework. Asia Pacific, Europe, United States and Brazil are examples. Some newer regions are not paired but still offer availability zones, so check each service's documentation rather than assuming a pair exists.

flowchart TB
    G["Geography: Europe"] --> R1["Region: North Europe"]
    G --> R2["Region: West Europe"]
    R1 --> Z1["Zone 1"]
    R1 --> Z2["Zone 2"]
    R1 --> Z3["Zone 3"]
    R1 <-->|"region pair"| R2

A geography contains regions; a region contains zones; some regions are paired with a neighbour.

Service families

Azure has several hundred products. They fall into families, and the tutorial chapters follow the same grouping:

Family What it gives you Examples Chapter
Compute Run code Virtual Machines, App Service, Functions Compute
Storage Keep files and disks Blob, Files, Queues, Disks Storage
Networking Connect and protect Virtual Network, Load Balancer, DNS Networking
Databases Structured data Azure SQL, Cosmos DB, PostgreSQL Databases
Messaging Decouple systems Service Bus, Event Grid, Event Hubs Messaging
Containers Run images Container Registry, Container Apps, AKS Containers
Identity Who can do what Microsoft Entra ID, RBAC Identity
Management Observe and govern Monitor, Policy, Cost Management Monitoring

There are also families this tutorial only touches: AI and machine learning, analytics, Internet of Things, developer tools and more. The rule of thumb for beginners is to learn compute, storage, networking and identity first; the other families reuse the same ideas.

How a request reaches a datacenter

When someone opens your site, the path usually looks like this:

  1. Their browser asks DNS for your domain's address. If you use a global entry point such as Azure Front Door or Traffic Manager, the answer points to a good entry location for that user.
  2. The request enters Microsoft's global network at a nearby point of presence (an edge location) and then travels over Microsoft's private backbone, not the open internet, to the region that hosts your app.
  3. Inside the region, a load balancer or gateway forwards the request to a healthy instance, in some zone, of your app.
  4. The app may call a database or storage account in the same region, which keeps the extra latency small.

Load Balancer, Application Gateway and Front Door and DNS and Traffic Manager cover steps 1 to 3 in depth.

Choosing a region: a stand-in

A beginner's checklist for the first region of a project: stay close to users, stay inside the legal area your data requires, check that every service and VM size you need is offered there, and compare price last. The stand-in below applies exactly that order. The regions, latencies and prices are invented teaching numbers, not Azure's real data.

JS Pick a region

Press Run (or Ctrl + Enter) to see the console output here.

No output.

The last request has no answer, and that is a useful lesson: constraints can contradict each other, and you find out at design time rather than after launch.

Zone-aware resilience: a stand-in

The second example shows why zones matter. A service has instances spread across zones; we fail zones and count survivors.

JS Losing a zone

Press Run (or Ctrl + Enter) to see the console output here.

No output.

Six instances in one zone give zero survivors when that zone fails, while the same six spread over three zones keep four running after one failure. Remember that this is a simplified model: real zone failures are rare, and your design must also have enough spare capacity in the remaining zones to carry the load.

How Azure works, and its strengths and weaknesses

Under the portal, Azure is a very large pool of servers in datacenters. A control layer called the fabric tracks every physical server, decides where each new VM or container will run, and restarts workloads on healthy hardware when a server fails. When you ask for a resource, your request passes through the Azure Resource Manager API, the fabric picks a host with free capacity in the region you chose, a virtualization layer isolates your workload from the other customers on the same host, and you pay for what you use. You never see the fabric; you see the resources it creates.

That design brings real strengths:

  • Scalability and elasticity. Capacity is added or removed in minutes, and many services scale automatically.
  • Global reach. Dozens of regions let you run close to users and satisfy data-residency rules.
  • Hybrid capabilities. Azure Arc, VPN and ExpressRoute connect it to your own datacenter (see Azure Arc, Lighthouse and Hybrid Management). Organizations that already use Windows Server, SQL Server or Microsoft 365 often get licensing benefits and easy sign-in integration.
  • Built-in security and compliance tooling, such as Defender for Cloud, Policy and dozens of compliance certifications.
  • Pay as you go, with discounts for commitments.

And real weaknesses:

  • A learning curve. Hundreds of services, overlapping options and frequent renames mean you need platform expertise to choose well (see Choosing a Compute Service).
  • Cost surprises. Idle resources, data leaving the cloud and forgotten disks all bill quietly (see Cost Management).
  • Lock-in. Managed services speed you up but tie designs to one provider; containers and open standards reduce the tie.
  • Shared responsibility. Azure secures the datacenter, but misconfigured storage or open ports are still your fault.

Azure and the other big clouds

Azure's main competitors are Amazon Web Services (AWS) and Google Cloud Platform (GCP), with smaller players such as Oracle Cloud and Alibaba Cloud. They offer the same building blocks under different names: a virtual machine is an EC2 instance on AWS and a Compute Engine instance on GCP; blob storage is S3 on AWS and Cloud Storage on GCP; identity is IAM on AWS and Microsoft Entra ID on Azure. Skills transfer well, because the concepts (regions, networks, identity, managed databases) are shared. If you know the vocabulary for one, learning another is mostly learning new names and new defaults.

Reaching Azure for the first time

Open the portal at portal.azure.com and sign in with your account. For a command line without installing anything, open Cloud Shell from the portal's top bar (see Portal, Azure CLI and PowerShell). The portal's Quickstart Center offers guided setup checklists, and a free account gives you starter credits so you can follow this tutorial's read-only commands against a real subscription at low risk.

Check your understanding

Quick check ยท 4 questions

of 4 answered You scored / 4 ยท perfect!

  1. 1. What is an Azure availability zone?

    Zones are separate datacenter locations within one region. Placing copies in different zones protects against a single-datacenter failure.
  2. 2. Which problem do availability zones NOT protect against?

    All zones are inside one region. Surviving a region-wide disaster needs resources in a second region.
  3. 3. A law requires customer data to stay in the EU. Which factor should you check first when choosing a region?

    Legal requirements are hard constraints. Price and performance are compared only among the regions that satisfy them.
  4. 4. What is a geography in Azure?

    A geography groups regions in one market and legal framework so data residency can be preserved.

Summary

  • Azure is a global set of datacenters in regions; most resources live in one region you choose.
  • Availability zones are separate locations inside a region and give protection against datacenter failures; region pairs and a second region cover regional disasters.
  • A geography is a compliance and data-residency boundary containing one or more regions.
  • Services come in families: compute, storage, networking, databases, messaging, containers, identity and management.
  • Choose a region by compliance first, then latency, service availability and price.

Next: Accounts, Tenants and Subscriptions.