| name | azure-patterns |
| description | Azure Patterns: Functions, Container Apps, Cosmos DB, Azure SQL, AKS, serverless. |
| triggers | {"extensions":[".py",".ts",".yaml"],"directories":["azure/","infrastructure/"],"keywords":["azure","functions","container apps","cosmos db","azure sql","aks","app service","bicep","arm"]} |
| auto_load_when | Building on Microsoft Azure or designing enterprise architectures |
| agent | cloud-architect |
| tools | ["Read","Write","Bash"] |
Azure Architecture Patterns
Focus: Enterprise integration, hybrid cloud, managed services
1. Compute Selection
When to use what:
├── Azure Functions
│ ├── Event-driven, HTTP, timers
│ ├── Premium: VNET, longer running
│ └── Consumption: pay per execution
│
├── Container Apps
│ ├── Microservices, event-driven
│ ├── KEDA-based autoscaling
│ └── Dapr integration for state
│
├── App Service
│ ├── Web apps (ASP.NET, Node, Python)
│ ├── PaaS for web apps
│ └── Managed, easy to use
│
├── AKS (Kubernetes)
│ ├── Full Kubernetes
│ ├── Enterprise features
│ └── Windows containers support
│
└── Azure VMs
│ ├── Legacy lift-and-shift
│ └── Specific requirements
2. Data Layer Patterns
Database selection:
├── Cosmos DB
│ ├── Global distributed NoSQL
│ ├── Multiple APIs (SQL, MongoDB, Cassandra, Gremlin)
│ ├── Multi-master, any region write
│ └── Serverless option
│
├── Azure SQL
│ ├── Managed SQL Server
│ ├── Intelligent database (auto-tuning)
│ └── Hyperscale for large DBs
│
├── PostgreSQL / MySQL
│ ├── Flexible Server (managed)
│ └── Serverless option
│
├── Azure Cache for Redis
│ ├── Caching, session store
│ └── Redis Enterprise for clustering
│
└── Blob Storage
├── Object storage
├── Hot/Cool/Archive tiers
└── Azure Data Lake integration
3. Integration Patterns
Azure Integration Services:
├── Service Bus
│ ├── Enterprise messaging
│ ├── Topics for pub/sub
│ └── Reliable delivery
│
├── Logic Apps
│ ├── No-code workflow automation
│ ├── 400+ connectors
│ └── Visual workflow designer
│
├── API Management
│ ├── API gateway, developer portal
│ ├── Rate limiting, caching
│ └── Mock APIs for development
│
├── Event Grid
│ ├── Event routing
│ └── Push-based, near real-time
│
└── Functions + Service Bus
└── Durable execution patterns
4. Enterprise Features
Azure Enterprise Patterns:
├── Identity
│ ├── Microsoft Entra ID (formerly AAD)
│ ├── SSO, MFA, Conditional Access
│ └── RBAC, managed identities
│
├── Hybrid Cloud
│ ├── Azure Arc (hybrid management)
│ ├── Azure Stack (on-prem Azure)
│ └── ExpressRoute (private connection)
│
├── Governance
│ ├── Azure Policy
│ ├── Management groups
│ └── Blueprints for compliance
│
└── Monitoring
│ ├── Application Insights
│ ├── Log Analytics
└── Azure Monitor
5. Infrastructure as Code
IaC in Azure:
├── ARM Templates
│ ├── JSON-based, verbose
│ └── Native Azure
│
├── Bicep (recommended)
│ ├── Simplified, transpiles to ARM
│ └── Better syntax, modular
│
├── Terraform
├── Multi-cloud, popular
└── Official Azure provider
│
└── Ansible
├── Configuration management
└── Azure modules available
Key Patterns
- Enterprise integration - Best for Microsoft ecosystem
- Cosmos DB for global - Multi-region, any API
- Logic Apps - No-code for integrations
- Entra ID - Identity and access management
- Bicep - Simplified IaC (or Terraform)
Anti-Patterns
❌ Using VMs for new workloads
✅ App Service / Container Apps / Functions
❌ Single region for production
✅ Multi-region with Traffic Manager
❌ Not using managed identities
✅ System/user assigned MI for RBAC
❌ No Azure Policy enforcement
✅ Governance from day one
❌ Ignoring Azure Cost Management
✅ Budget alerts, cost analysis
Quick Reference
| Service | Use Case | Key Feature |
|---|
| Functions | Serverless | Event-driven |
| Container Apps | Microservices | KEDA scaling |
| App Service | Web apps | Managed |
| Cosmos DB | Global NoSQL | Multi-API |
| Azure SQL | Managed SQL | Intelligent |
| Service Bus | Messaging | Enterprise |
| AKS | Kubernetes | Enterprise |
| Blob Storage | Object storage | Tiers |
🌍 Universal Language Support
- Turkish Native: This skill natively supports Turkish. If the user prompt is in Turkish, all analysis, formatting, and output MUST be entirely in Turkish. You do not need explicit "write in Turkish" instructions.
