Infrastructure Concepts
About
Infrastructure Concepts in cloud computing refer to the foundational building blocks that cloud providers use to construct, deploy, and manage cloud services across the globe. These include how data centers are organized geographically, how services are isolated for high availability, and how resources are logically grouped and managed.
Understanding these concepts is essential for cloud architects, developers, administrators, and anyone who designs or manages workloads in the cloud. These concepts are also cloud-provider-neutral, but many map closely to implementations in providers like AWS, Azure, and GCP.
1. Global Infrastructure Organization
Geography
A Geography represents a broad geopolitical area where cloud services operate. It typically corresponds to a continent or large region, such as North America, Europe, or Asia Pacific.
Why it matters:
Data residency and compliance (e.g., GDPR)
Latency considerations
Regulatory boundaries
Example: Microsoft Azure defines separate geographies like "Asia Pacific" or "US Government".
Region
A Region is a specific geographic location that contains multiple data centers, often called Availability Zones (AZs). Regions are designed to be isolated from one another to provide fault tolerance.
Key points:
Each region is independent
Services may vary slightly between regions
Customers choose regions based on proximity and legal requirements
Example:
AWS:
us-east-1,ap-south-1Azure:
East US,North EuropeGCP:
asia-south1
Availability Zone (AZ)
An Availability Zone is an isolated data center (or a cluster of them) within a region. These zones are connected via high-speed, low-latency links and are designed to be independent in power, cooling, and networking.
Importance:
Deploying across AZs ensures high availability
Faults in one zone do not affect others
Example: AWS typically offers 3+ AZs per region, labeled as us-east-1a, us-east-1b, etc.
Region Pair (Azure-specific)
A Region Pair is a Microsoft Azure concept where two regions are linked within the same geography for disaster recovery and compliance. Data is replicated between these paired regions.
Benefits:
Geo-redundant storage
Prioritized updates in one region to avoid simultaneous failures
Example: West Europe is paired with North Europe.
2. Availability and Fault Isolation
Availability Set (Azure-specific)
An Availability Set is a logical group of VMs that Azure distributes across multiple fault domains and update domains to protect against localized failures and planned maintenance.
Why it matters:
Ensures that not all VMs are rebooted at the same time
Reduces the risk of downtime during hardware failure or patching
Fault Domain
A Fault Domain is a group of hardware components (servers, switches, power supply) that share a single point of failure. VMs deployed across fault domains are protected from simultaneous hardware failure.
Used in: Azure Availability Sets, on-premise hardware planning
Update Domain
An Update Domain is a logical grouping where Azure performs sequential updates and reboots. Only one update domain is updated at a time to avoid service disruption.
Result: Your VMs are not all restarted at once during scheduled maintenance.
3. Resource Management and Grouping
Resource Group (Azure concept, applicable elsewhere)
A Resource Group is a logical container that holds related resources like VMs, databases, networks, and storage accounts.
Benefits:
Unified billing and monitoring
Easier access control and role assignment
Enables automation of deployment (ARM templates, Bicep)
Example: A web app, its database, and a load balancer can all be grouped into one resource group.
Subscription (Common in Azure/GCP)
A Subscription acts as a billing boundary and access control unit for managing cloud usage.
In Azure, subscriptions belong to a tenant (organization).
Resources from multiple subscriptions can be managed together via management groups.
Tags (Cross-platform concept)
Tags are key-value metadata pairs attached to resources for better organization, automation, and cost tracking.
Example:
Environment: production, Owner: devops-team, CostCenter: IT
Last updated