Cloud region and availability zone
Also known as: cloud region, availability zone
A cloud region is a geographic area, such as Canada Central in Toronto or Canada East in Quebec City, where a cloud provider operates data centres.
Each region usually contains several availability zones, physically separate data centre sites with their own power and networking, close enough for fast connections but far enough apart that one failure won't take down the others. Choosing regions affects speed, cost, and where data is stored. While spreading systems across zones protects against most outages, problems in a provider's shared services can still affect an entire region.
Cloud region and availability zone, explained at three levels
At home
A cloud region is like a city where a company has warehouses, and availability zones are separate warehouses across town. If one floods, the others keep working.
At work
Choosing Canadian cloud regions can keep data in Canada and improve speed for Canadian users.
At home
Apps feel faster when their servers are in a nearby region. A game server in Montreal will feel snappier from Ottawa than one in California.
At work
Deploy across at least two availability zones for high availability. Some services and features launch in certain regions first, and prices vary by region.
At home
Some services store data outside Canada regardless of where you live. In my opinion, it's worth checking privacy policies if that matters to you.
At work
Design for zone failures by default, consider multi-region designs for critical services, test failover, account for data residency requirements, and remember that a provider's global control systems can still be single points of failure.
Real-world examples
At home
Carlos plays an online game that lets him choose a server region. Picking the Canadian region cut his lag noticeably.
At work
Lena designs systems for a Toronto hospital network. She chose a Canadian region and spread servers across three availability zones for resilience.
How it compares
| feature | Cloud region | Availability zone |
|---|---|---|
| Scope | Geographic area | Separate site within a region |
| Distance | Hundreds of km or more apart | Several km apart |
| Protects against | Regional disasters | Single data centre failures |
| Latency between | Higher | Very low |
Zones protect against local failures. Regions protect against wider ones.
For analysts
In cloud requirements, I'd specify data residency and resilience targets in plain terms. Region and zone design should follow from those, not the other way around.
History & origin
Amazon Web Services introduced multiple regions and the 'availability zone' concept in 2008, letting customers spread systems across separate facilities. Microsoft and Google adopted similar designs. Providers expanded rapidly worldwide, opening Canadian regions to meet customer demand for data stored in Canada: Microsoft in Toronto and Quebec City in 2016, AWS in Montreal in 2016 and Calgary in 2023, and Google in Montreal in 2018 and Toronto in 2021. Region-wide outages, such as the AWS US-East-1 disruption in October 2025, showed the limits of relying on a single region. In my view, region choice is one of the most important yet least discussed cloud decisions.