Regions

The cloud is in real places: a region is a location you choose

N. VirginiaOregonSão PauloIrelandFrankfurtCape TownMumbaiSingaporeTokyoSydneyus-east-1, the first (2006)Ten of the 38 AWS regions, at their real longitude and latitude. Continents are sketched, not to scale.The cloud is in real places. A region is a named location where a provider has built data centres.You choose the region for everything you create, and it stays there. Late 2025: AWS has 38 regions, Google Cloud 42,and Azure more than 60. Each is a separate copy of the cloud, with its own machines, its own price list and its own failures.

A region is a geographic area where a provider runs data centres. Everything you create lives in the region you pick.

Availability zones

A region is three or more zones; a zone is one or more data centres

Regionone geographic area, such as FrankfurtZone aZone bZone cprivate fibre1 to 2 mszones sit kilometres apart, and at most about 100 km: far enough for separate floods and fires, near enough to stay in stepA region is three or more zones. A zone is one or more data centres with their own power and cooling.Each zone has its own electricity supply, cooling and network links, so a fault in one building stays in one zone.The zones of a region are joined by the provider’s own fibre, fast enough to copy every write to a second zone as it happens.AWS and Azure call them availability zones, Google calls them zones. The idea and the numbers are the same.

An availability zone is an isolated group of data centres. A region has several, a few milliseconds apart.

Availability zones · Going deeper

Reading region and zone names on the three providers

AWSAzureGoogle CloudA regionus-east-1East USus-east1Its plain nameUS East (N. Virginia)East US (Virginia)Moncks Corner, S. CarolinaA zone in itus-east-1aEast US, zone 1us-east1-bZones per region3 or more3 in zone-enabled regions3 or moreRegions, late 20253860+42Zones, late 2025120not published as one total127A region code names the area; a letter or number after it names the zone.You type these codes in every command and every bill lists them, so they are worth recognising on sight.One trap on AWS: the letters are shuffled per account. Your us-east-1a may be another account’s us-east-1c, so zone IDssuch as use1-az1 are used when two accounts must agree on the same physical zone.

us-east-1a on AWS, East US zone 1 on Azure and us-east1-b on Google Cloud all name one zone of a region in the eastern United States.

Why zones exist

Zones exist so that one failure cannot take everything

usersthe load balancer sends each request to a healthy zoneZone apower lost: both servers downZone bZone ccarry on: two thirds of the capacity, no outageSpread across zones, a whole data centre can fail and your users see nothing.This is the main reason zones exist, and it is your choice to use them: a system placed in one zone is down when that zone is.The usual pattern is a copy in at least two zones behind a load balancer, with enough spare capacity that the survivors cancarry the full load. Providers call a design like this highly available, and it is the subject of lesson 10.

Run copies in two or more zones behind a load balancer and a failed data centre goes unnoticed.

Why zones exist · Going deeper

What the provider promises: one zone or two

Downtime the provider’s promise allows in a year, to scaleOne virtual machine, one zoneAWS; Azure and Google promise 99.9% for a single machine99.5%: 43.8 hoursMachines in two or more zonesAWS, Azure and Google Cloud99.99%: 53 minutesTwo zones turn a promise of 99.5% into 99.99%: from almost two days a year to under an hour.The figure is called an SLA, a service level agreement. Miss it and the provider refunds part of the bill as credit; it doesnot pay for your lost sales. 99.99% is said “four nines”. The compute SLAs of all three providers require two or morezones for it, which makes the exam answer to “how do I make this highly available?” almost always: more than one zone.

An SLA states the uptime a provider commits to. Across zones it is 99.99%; for one machine in one zone on AWS, 99.5%.

Latency

Distance is delay: about 1 ms for every 100 km, there and back

A user in LondonFastest possible round trip to a region in…Londonunder 1 mssame cityFrankfurt6 ms640 kmN. Virginia59 ms5,900 kmMumbai72 ms7,200 kmSão Paulo95 ms9,500 kmSydney170 ms17,000 kmDistance is delay: light in fibre covers about 200 km per millisecond, so 100 km costs 1 ms there and back.These are floors set by physics, computed from straight-line distance. Real routes are longer and add equipment, so expect1.5 to 2 times as much. A web page makes many round trips; at 170 ms each, a site hosted in Sydney feels slow in Londonhowever fast its servers are. People notice delay above about 100 ms.

Latency is the time a request takes to travel. Physics sets a floor, so serve users from a region near them.

Choosing a region

Four questions pick a region: law, distance, services, price

1. LawWhere must the data live?Some data may not leave acountry. This rules regionsout before anything else.2. DistanceWhere are the users?Closer is faster. Pick theregion nearest most of them.3. ServicesIs what I need offered there?New services reach a fewregions first, others lateror never.4. PriceWhat does it cost there?The same machine costs upto 60% more in some regions.Four questions pick a region, in this order: law, distance, services, price.AWS lists exactly these four for its exam: compliance, proximity to customers, available services and pricing.Law comes first because it removes choices outright. Price comes last because it is the only one you can trade againstthe others: a cheaper region further away is a real option for work no user waits on, such as nightly reports.

Choose a region by data residency rules first, then closeness to users, then service availability, then price.

Choosing a region · Going deeper · 1 of 2

The same machine, five prices

The same virtual machine (2 CPUs, 8 GB), on demand, per hourN. Virginia$0.096Mumbai$0.101 (+5%)Frankfurt$0.115 (+20%)Sydney$0.120 (+25%)São Paulo$0.153 (+59%)Prices differ by region because land, electricity, taxes and demand differ.The oldest and largest regions are usually the cheapest. Example list prices: AWS m5.large, 2025; bars are to scale.The same holds on Azure and Google Cloud. Data sent between regions is also billed, so splitting one system acrosstwo regions to save on machines can cost more in transfer than it saves.

On-demand prices vary by region. In 2025 an identical AWS machine costs about 60% more in São Paulo than in N. Virginia.

Choosing a region · Going deeper · 2 of 2

Data residency, sovereignty and the special regions

Where data sits decides whose law reaches it. Providers offer ordinary regions, and a few special ones for the strictest rules.

Three words that are easy to mix up

  • Data residency Where the data is physically stored. You control it by choosing the region.
  • Data sovereignty Which country’s law applies to the data. It usually follows where the data is stored, and sometimes also where the provider is based.
  • Compliance The rules you must prove you follow, such as the EU’s GDPR for personal data or national rules for health and government data.
The provider’s promise: data you put in a region stays in that region unless you copy or move it. Replication to a second region, a backup elsewhere, or a CDN are all things you switch on.

Regions built for particular rules

GovernmentAWS GovCloud (US) and Azure Government: separate regions run by screened US staff
Chinaoperated by local companies under Chinese law, with separate accounts
Sovereignthe AWS European Sovereign Cloud and similar offers: run and staffed entirely inside the EU

Microsoft also groups regions into geographies, such as Europe or the United Kingdom, and promises that data stays within the geography even when it replicates between a pair of regions.

Resource scope

Every resource has a scope: a zone, a region, or the world

Global: one for the whole accountIdentityusers, rolesDNS, CDNnames, edge cacheRegional: lives in one region, spans its zonesObject storagecopies in 3 zonesLoadbalancerZonal: zone aVirtual machineIts diskZonal: zone bVirtual machineIts diska machine and its disk share a zoneEvery resource has a scope: one zone, one region, or the whole world.Zonal things fail with their zone, so you make copies. Regional services spread over zones for you. Global services haveno region to pick. Knowing a service’s scope tells you what a zone failure will do to it, which is a common exam question.

A resource is zonal, regional or global, and its scope tells you what a failure will do to it.

Resource scope · Going deeper

Scopes on the three providers

ResourceAWSAzureGoogle CloudVirtual machinezonalzonalzonalIts diskzonalzonal, or zone-redundantzonal, or regionalObject storageregionalregional, or geo-redundantregional, dual or multi-regionPrivate networkregional (VPC)regional (VNet)global (VPC)Subnetone zoneone regionone regionIdentityglobal (IAM)global (Entra ID)global (IAM)The three providers agree on machines and identity, and differ on networks and storage.The difference that surprises people most: a Google Cloud network is global, with subnets per region, while on AWS a networkbelongs to one region and each subnet to one zone. Lesson 6 builds on this. For a provider exam, learn that provider’s column;for real work, check the scope of anything you depend on before you design around it.

Machines are zonal and identity is global everywhere. Networks differ: regional on AWS and Azure, global on Google Cloud.

Edge locations

Hundreds of small sites bring content close to users

Region: N. Virginiathe original filesSydneyedge site5 msLondonedge site5 msSão Pauloedge site5 msTokyoedge site5 msdashed: fetched once from the region,then kept at the edgeThe edge is hundreds of small sites near users that keep copies of your content.A content delivery network (CDN) answers from the nearest edge site, so a user in Sydney waits about 5 ms for a filewhose original is 170 ms away. AWS runs more than 700 edge sites against 38 regions. Edge sites also answer DNS lookupsand absorb attack traffic before it reaches you. Your servers and databases stay in the region.

A CDN serves copies of your content from the edge site nearest each user, cutting the round trip to a few milliseconds.

Edge locations · Going deeper

Four ways to get closer, and what each provider calls them

Getting closer to the userAWSAzureGoogle CloudContent delivery (CDN)CloudFrontAzure Front DoorCloud CDNDNSRoute 53Azure DNSCloud DNSA mini-zone in a big cityLocal ZonesExtended Zonesnot offeredIn your own buildingOutpostsAzure LocalDistributed CloudRegions are few and large. Everything else on this page exists to get closer to a user or a machine.The top two rows are the edge network: cached content and name lookups, plus small pieces of code that run at the edge(on AWS, CloudFront Functions and Lambda@Edge). The bottom two move real compute closer: a mini-zone in a city with noregion, or racks in your own building. AWS also places compute inside 5G mobile networks, as Wavelength.

CDNs and DNS form the edge network. Local zones and on-premises racks move real compute closer.

Data residency

Regions are isolated: nothing leaves one unless you send it

Region 1: Frankfurtthe systemits dataRegion 2: Irelanda standbya copyyou switch replication onabout 1,000 km apart: no storm, flood or power grid failure reaches bothRegions are isolated on purpose. Nothing leaves a region unless you send it.That isolation is what makes data residency possible: data stored in Frankfurt stays in Frankfurt. It also means a secondregion is something you build: for disaster recovery, when a whole region fails, or to serve users on another continent.Zones protect you from a building failing. Regions protect you from a city failing, at roughly double the cost.

A second region is for disaster recovery or for distant users, and replicating to it is your decision.

Data residency · Going deeper

Paired regions and multi-regions: the provider’s help with two regions

Building across regions is your job, and each provider offers ready-made pieces that make it easier.

Azure region pairs

Most Azure regions have a fixed partner in the same geography, at least about 300 miles (480 km) away where possible: East US with West US, North Europe (Ireland) with West Europe (Netherlands). Platform updates reach one region of a pair at a time, and if both fail, one is restored first. Geo-redundant storage copies your data to the paired region automatically.

Google Cloud multi-regions

Some Google services accept a location wider than a region. A storage bucket created in the multi-region EU or US, or in a named dual-region, keeps copies in more than one region and stays readable if one is lost.

AWS: you choose both ends

AWS has no fixed pairs. You pick any two regions and switch on a feature per service: cross-region replication for S3 storage, read replicas in another region for databases, global tables for DynamoDB.

Zonesprotect against a building or data centre failing; near-instant copying; small extra cost
Regionsprotect against a whole area failing; copying lags by seconds; roughly double the cost
Bothwhat systems that must never stop use: several zones in each of two regions
Exam shorthand: high availability means multiple zones. Disaster recovery means another region.

Lesson complete

You finished “Regions, zones and the edge”

Nicely done.

Open in a new tab