INFRASTRUCTURE LEARNINGINFRASTRUCTURE LEARNING
インフラ大全インフラ大全IT INFRASTRUCTURE / LEARNING
Infrastructure learning guide
Find a lesson or lab/

Esc to close · Search terms stay on your device

LEARNING日本語

Core lesson 4 of 18Topic: AWS foundations

AWS Regions vs Availability Zones: placement and failure boundaries

A Region is a separate geographic area; an Availability Zone (AZ) is an isolated location within a Region. Two instances in one Region are not necessarily spread across AZs. Separate resource placement, the failure boundary and the application design needed to continue serving requests.

What you will learn

Explain the Region/AZ hierarchy, distinguish single-AZ, multi-AZ and multi-Region placement, and identify dependencies that placement alone does not fix.

Prerequisite

Know the roles of a VM and stored data. Diagrams are synthetic placement examples, not observations of real outages or service health.

AZs belong to Regions; instance count is not placement

Two illustrative Regions, each containing separate AZs

Region R

Region R — Placement example: each inner box is one AZAZ 1 and AZ 2 are in the same Region. Cross-AZ placement still needs application and data design.AZ 1EC2AZ 2EC2

AZ 1 and AZ 2 are in the same Region. Cross-AZ placement still needs application and data design.

Region S

Region S — Placement example: each inner box is one AZSeparate from Region R. EC2 resources are not automatically replicated here.AZ 1EC2AZ 2EC2

Separate from Region R. EC2 resources are not automatically replicated here.

If all instances occupy one location affected by a failure, all may be unavailable. Spreading instances across AZs addresses that placement risk, but placement alone does not implement traffic failover or data synchronization.

Choose a Region from requirements, not distance alone

Questions to resolve before choosing a Region
AreaWhat to check
CapabilitiesAre the required services and features available?
Users and connectivityWill latency and connection requirements be met? Distance alone is not a measurement.
Data and organizational requirementsReview location, legal/contractual needs and organizational policy.
AccountCheck account type, Region enablement and accessible AZs.
Costs and dependenciesReview regional terms, transfer and dependency placement.

When inspecting EC2 resources, establish both account and selected Region. An empty list in one Region does not establish that another Region has no resources. Keep that scope when investigating costs or cleanup.

Worked example: two app instances and one storage dependency

Identify the separated boundary and remaining dependency
LayoutBoundary separatedStill to establish
Two instances in AZ 1More than one instanceBoth still share the AZ’s failure boundary
One instance in AZ 1 and one in AZ 2Instance placement spans AZsTraffic distribution, data, capacity and shared dependencies
A deployment in another RegionPlacement spans RegionsReplication, failover, destinations, operations and cost

Choose an option to see the result and explanation.

Two EC2 instances run in the same AZ. Have you separated the AZ failure boundary?
Read the answer and reason

Answer: No: they still share that AZ boundary · Instance count is not placement diversity. Review AZ distribution and also the traffic, storage and other dependencies needed for continuity.

Use AZ IDs when matching physical placement across accounts

Do not infer cross-account physical placement from the AZ code’s final letter alone. AWS documents independently mapped codes in certain older Regions for accounts created before November 2025; accounts created starting that month receive the same mappings.

An AZ ID identifies the same physical AZ across accounts. Check the account and Region conditions and compare AZ IDs when coordinating placement. Avoid both blanket claims: that letters always differ and that they always match.

Review placement together with dependencies and recovery

  1. List the account, Region, AZ and relevant resources.
  2. Identify resources that share a failure boundary.
  3. For multiple AZs, design traffic handling and data access.
  4. For multiple Regions, do not assume automatic resource or data replication.
  5. Define recovery requirements and evaluate them in an authorized environment.

Check, then explain in your own words

  1. Draw AZs within their Region.
  2. Compare two instances in one AZ with instances in separate AZs.
  3. Include data, traffic failover and AZ IDs in a placement review.
Check your understanding: Does creating an EC2 instance in another Region automatically copy the original data?

No. Required resources, data replication and failover need their own design.

EC2 resources are not automatically replicated across Regions. More locations and a working continuity design are distinct outcomes.

Sources and conditions checked

The official references used in this lesson are collected here with their scope, conditions and corresponding sections. Independent expert review and live AWS comparison have not been performed.