Regions and Availability Zones

39 Regions, each with at least three isolated zones

ONE REGION, for example eu-west-1 (Ireland)a geographic areaAvailability Zone 1aone or more data centresown power, cooling, networkAvailability Zone 1bone or more data centresown power, cooling, networkAvailability Zone 1cone or more data centresown power, cooling, networkjoined by private low-latency fibre, kilometres apart39Regions124Availability Zones3+zones in every RegionAWS figures, November 2026A Region is a place on the map. An Availability Zone is an isolated group of data centres inside it.Regions are separate from each other: nothing is copied between them unless you ask. Inside a Region, each zonehas its own power, cooling and network, far enough from the others that one fire or flood cannot take out two,and close enough to replicate data quickly. The exam phrase: zones "do not share single points of failure".

A Region is a geographic area. An Availability Zone is one or more isolated data centres inside it.

Regions and Availability Zones · Going deeper

Names, distances and global services

Regiona separate geographic area with a code such as us-east-1; isolated from every other RegionAvailability Zoneone or more discrete data centres with redundant power, networking and connectivityBetween zonesphysically separate, within 100 km of each other, linked by redundant private fibreZone namesthe Region code plus a letter: us-east-1a, us-east-1b, us-east-1cMost servicesare regional: you pick a Region in the console and see only that Region's resourcesGlobal servicesIAM, Route 53, CloudFront and AWS Organizations are not tied to one RegionHierarchy to remember: Region, then Availability Zone, then data centre. Edge locations sit outside it.A common trap offers "an Availability Zone contains several Regions". It is the other way round.

Most services are regional. IAM, Route 53, CloudFront and Organizations are global.

High availability

Two zones and a load balancer survive a data centre failure

UsersLoad balancerspans both zonesAvailability Zone apower lostAvailability Zone ball traffic goes to the healthy zoneHigh availability on AWS means running in at least two Availability Zones.The load balancer checks health and stops sending requests to the failed zone. The database keeps a standbycopy in the other zone (RDS Multi-AZ). Nothing in one zone depends on the other. One zone is never the answerto an availability question; two or more always is.

High availability on AWS means at least two Availability Zones.

Multiple Regions

A second Region needs one of four reasons

ReasonThe situationWhat a second Region does1Disaster recoverya whole Region is disrupteda copy in a second Region takes over2Business continuitythe business must keep trading through anythingrun active in two Regions at once3Low latency for end usersusers are far from your Regionserve each continent from a Region near it4Data sovereigntylaw says data must stay in a countrystore it in a Region inside that borderMultiple zones are for availability. Multiple Regions are for these four reasons, and the exam lists them.More Regions cost more and add complexity, so they need a reason. Each of these four appears in the exam guideword for word. If a question says "disaster", "continuity", "users on another continent" or "data must remainin the country", the answer involves Regions, not zones.

Use several Regions for disaster recovery, business continuity, low latency or data sovereignty.

Multiple Regions · Going deeper

How to choose a Region

Choosing a Region, in order1ComplianceDoes law or contract say where the data must be? This can decide alone.2LatencyWhich Region is closest to the users?3Service availabilityIs every service you need offered there? New services launch in a few Regions first.4PriceThe same service costs different amounts in different Regions.Compliance first, then latency, then services, then price.Not every service is in every Region, and prices are not uniform. Both are common "true or false" style options.

Choose by compliance, latency, service availability, then price.

Check yourself

One RegionZone aZone bZone c

A company must keep serving customers even if an entire data centre fails. What is the minimum design?

Edge locations

750+ small sites exist only to be near users

Regions39Edge locations750+CloudFront points of presence, drawn to the same scaleA viewer in Lagosa few msEdge locationhas a cached copyonly on a missOriginin a distant RegionWhat runs at the edgeAmazon CloudFront: cachingAmazon Route 53: DNS answersAWS Global Accelerator: entry pointAWS WAF and Shield: filteringEdge locations far outnumber Regions, because their only job is to be near users.An edge location is a small site, not a Region: you cannot launch servers there. It caches content and answers DNSso that a user's first hop is short. The benefit the exam wants: lower latency for users everywhere withoutdeploying your application in every Region.

Edge locations cache content and answer DNS close to users. They are not Regions.

Edge locations · Going deeper

Local Zones, Wavelength, Outposts and clue words

When the question saysThe service isBecause it"cache content close to users"Edge location (CloudFront)hundreds of small sites worldwide"single-digit ms latency in a big city"AWS Local Zonescompute and storage near a metro area"ultra-low latency for 5G mobile apps"AWS WavelengthAWS inside a telecom operator's network"AWS services in our own data centre"AWS OutpostsAWS-run racks on your premises"static IPs and the AWS backbone"AWS Global Acceleratorroutes users to the nearest healthy endpoint"survive a data centre failure"Multiple Availability Zoneshigh availability inside a Region"survive a regional disaster"Multiple Regionsdisaster recoveryThe infrastructure, largest to smallest: Region, Availability Zone, Local Zone, edge location.Outposts is the only one of these that lives in a building you own.

Outposts is AWS in your building; Local Zones are AWS near a city.

Ways to access AWS

Console, CLI and SDK are three doors to the same API

Management Consolea website: click and seeAWS CLIcommands and scriptsAWS SDKsfrom your code: Python, JavaAWS APIsevery action is a callAWS SERVICESEC2S3RDSInfrastructure as code: a template makes the calls for youConsole, CLI and SDK are three doors to the same API. Everything in AWS is an API call.One-off or exploring: the console. Repeating or automating: the CLI, an SDK, or a template.

The console for people, the CLI for scripts, SDKs for code: all call the same APIs.

Infrastructure as code

A template builds the same environment every time

Resources:  WebServer:    Type: EC2 instance  Database:    Type: RDS database  Files:    Type: S3 bucketOne template, in version controlAWS CloudFormationreads it, builds a stackStack: DevelopmentStack: TestStack: Productionidentical every timeInfrastructure as code turns "how we built it" into a file that builds it the same way every time.A CloudFormation template lists the resources; CloudFormation creates, updates and deletes them together as astack. The benefits the exam names: repeatable, consistent, reviewable and fast. When a question says "repeatableprocess", "many environments" or "avoid manual errors", the answer is infrastructure as code.

AWS CloudFormation builds a stack from a template: repeatable, consistent and reviewable.

Infrastructure as code · Going deeper

Tools for provisioning and operating

When the question saysThe service isBecause it"define infrastructure in a template"AWS CloudFormationinfrastructure as code"just upload my web application"AWS Elastic Beanstalkprovisions and manages it for you"patch and run commands on many servers"AWS Systems Manageroperations at scale"approved products staff may launch"AWS Service Cataloga governed catalogue"set up many accounts with guardrails"AWS Control Towera landing zone"automate build, test and release"AWS CodePipeline with CodeBuildcontinuous delivery"check current limits and ask for more"Service Quotasper-account limits"track software licences"AWS License Managerlicence rules and usageOne-time operation: console. Repeatable process: a template, a script or a pipeline.Task 3.1 asks you to "evaluate requirements to determine whether to use one-time operations or repeatable processes".

One-time: the console. Repeatable: CloudFormation, scripts and pipelines.

Check yourself

Account 1Account 2Account 3Account 4Account 5

A team must create the same environment in five accounts, and prove each one matches. Which approach?

Deployment models

Cloud, on-premises, or hybrid with the two connected

CLOUDeverything runs on AWSno data centre of your ownHYBRIDVPN orDirect Connectsome here, some on AWS,connectedON-PREMISESyour own data centrealso called private cloudThree deployment models: all in the cloud, all on your premises, or hybrid, with the two connected.Hybrid is the usual state of a company part-way through a migration, or one that must keep certain systemslocal. The link is AWS Direct Connect or a VPN. AWS Outposts brings AWS hardware into the local side. Do notconfuse deployment models (where it runs) with service models (IaaS, PaaS, SaaS: who manages what).

The three deployment models are cloud, hybrid and on-premises.

Check yourself

RegionsEdgenumber of sites, to scale

Which statement about edge locations is correct?

Lesson complete

You finished “AWS global infrastructure and ways to deploy”

Answer the questions along the way to see your score here.

Open in a new tab