使用AZD部署容器应用示例

使用 AZD 部署容器应用示例

此目录包含使用 Azure Developer CLI (AZD) 将容器化应用部署到 Azure Container Apps 的全面示例。这些示例展示了真实世界的模式、最佳实践以及生产级配置。

概述

Azure Container Apps 是一个完全托管的无服务器容器平台，可让您运行微服务和容器化应用，而无需管理基础设施。结合 AZD，您可以获得以下优势：

简化部署：通过单条命令即可部署容器和基础设施
自动扩展：根据 HTTP 流量或事件自动扩展至零或扩展至多个实例
集成网络：内置服务发现和流量分流功能
托管身份：安全认证 Azure 资源
成本优化：仅为使用的资源付费

先决条件

开始之前，请确保您已具备以下条件：


# Check AZD installation
azd version

# Check Azure CLI
az version

# Check Docker (for building custom images)
docker --version

# Login to Azure
azd auth login
az login

所需的 Azure 资源：

活跃的 Azure 订阅
创建资源组的权限
容器应用环境的访问权限

快速入门示例

1. 简单 Web API (Python Flask)

使用 Azure Container Apps 部署一个基础 REST API。

示例：Python Flask API


# azure.yaml
name: flask-api-demo
metadata:
  template: flask-api-demo@0.0.1-beta
services:
  api:
    project: ./src/api
    language: python
    host: containerapp

部署步骤：


# Initialize from template
azd init --template todo-python-mongo

# Provision infrastructure and deploy
azd up

# Test the deployment
azd show
curl $(azd show --output json | jq -r '.services.api.endpoint')/health

主要功能：

自动扩展至 0 到 10 个副本
健康探测和存活检查
环境变量注入
应用洞察集成

2. Node.js Express API

部署一个带有 MongoDB 集成的 Node.js 后端。


# Initialize Node.js API template
azd init --template todo-nodejs-mongo

# Configure environment variables
azd env set DATABASE_NAME todosdb
azd env set COLLECTION_NAME todos

# Deploy
azd up

# View logs
azd logs api

基础设施亮点：


// Bicep snippet from infra/main.bicep
resource containerApp 'Microsoft.App/containerApps@2023-05-01' = {
  name: 'api-${resourceToken}'
  location: location
  properties: {
    managedEnvironmentId: containerEnv.id
    configuration: {
      ingress: {
        external: true
        targetPort: 3000
        transport: 'auto'
      }
      secrets: [
        {
          name: 'mongodb-connection'
          value: mongoConnection
        }
      ]
    }
    template: {
      containers: [
        {
          name: 'api'
          image: containerImage
          env: [
            {
              name: 'DATABASE_URL'
              secretRef: 'mongodb-connection'
            }
          ]
        }
      ]
      scale: {
        minReplicas: 0
        maxReplicas: 10
      }
    }
  }
}

3. 静态前端 + API 后端

部署一个完整的全栈应用，包括 React 前端和 API 后端。


# Initialize full-stack template
azd init --template todo-csharp-sql-swa-func

# Review configuration
cat azure.yaml

# Deploy both services
azd up

# Open the application
azd show --output json | jq -r '.services.web.endpoint' | xargs start

生产级示例

示例 1: 微服务架构

场景：具有多个微服务的电商应用

目录结构：


microservices-demo/
├── azure.yaml
├── infra/
│   ├── main.bicep
│   ├── app/
│   │   ├── container-env.bicep
│   │   ├── product-service.bicep
│   │   ├── order-service.bicep
│   │   └── payment-service.bicep
│   └── core/
│       ├── storage.bicep
│       └── database.bicep
└── src/
    ├── product-service/
    ├── order-service/
    └── payment-service/

azure.yaml 配置：


name: microservices-ecommerce
services:
  product-service:
    project: ./src/product-service
    language: python
    host: containerapp
    
  order-service:
    project: ./src/order-service
    language: csharp
    host: containerapp
    
  payment-service:
    project: ./src/payment-service
    language: nodejs
    host: containerapp

部署：


# Initialize project
azd init

# Set production environment
azd env new production

# Configure production settings
azd env set ENVIRONMENT production
azd env set MIN_REPLICAS 2
azd env set MAX_REPLICAS 50

# Deploy all services
azd up

# Monitor deployment
azd monitor --overview

示例 2: AI 驱动的容器应用

场景：集成 Azure OpenAI 的 AI 聊天应用

文件：src/ai-chat/app.py


from flask import Flask, request, jsonify
from azure.identity import DefaultAzureCredential
from azure.keyvault.secrets import SecretClient
import openai

app = Flask(__name__)

# Use Managed Identity for secure access
credential = DefaultAzureCredential()
vault_url = "https://{vault-name}.vault.azure.net"
client = SecretClient(vault_url=vault_url, credential=credential)

@app.route('/api/chat', methods=['POST'])
def chat():
    user_message = request.json.get('message')
    
    # Get OpenAI key from Key Vault
    openai_key = client.get_secret("openai-api-key").value
    openai.api_key = openai_key
    
    response = openai.ChatCompletion.create(
        model="gpt-4",
        messages=[{"role": "user", "content": user_message}]
    )
    
    return jsonify({"response": response.choices[0].message.content})

if __name__ == '__main__':
    app.run(host='0.0.0.0', port=8000)

文件：azure.yaml


name: ai-chat-app
services:
  api:
    project: ./src/ai-chat
    language: python
    host: containerapp

文件：infra/main.bicep


param location string = resourceGroup().location
param environmentName string

var resourceToken = uniqueString(subscription().id, environmentName, location)

// Container Apps Environment
module containerEnv './app/container-env.bicep' = {
  name: 'container-env-${resourceToken}'
  params: {
    location: location
    environmentName: environmentName
  }
}

// Key Vault for secrets
resource keyVault 'Microsoft.KeyVault/vaults@2023-02-01' = {
  name: 'kv-${resourceToken}'
  location: location
  properties: {
    sku: {
      family: 'A'
      name: 'standard'
    }
    tenantId: subscription().tenantId
    enableRbacAuthorization: true
  }
}

// Container App with Managed Identity
module aiChatApp './app/container-app.bicep' = {
  name: 'ai-chat-app-${resourceToken}'
  params: {
    location: location
    environmentId: containerEnv.outputs.environmentId
    containerImage: 'your-registry.azurecr.io/ai-chat:latest'
    keyVaultName: keyVault.name
  }
}

部署命令：


# Set up environment
azd init --template ai-chat-app
azd env new dev

# Configure OpenAI
azd env set AZURE_OPENAI_ENDPOINT "https://your-openai.openai.azure.com/"
azd env set AZURE_OPENAI_DEPLOYMENT "gpt-4"

# Deploy
azd up

# Test the API
curl -X POST $(azd show --output json | jq -r '.services.api.endpoint')/api/chat \
  -H "Content-Type: application/json" \
  -d '{"message": "Hello, how are you?"}'

示例 3: 队列处理的后台工作器

场景：带消息队列的订单处理系统

目录结构：


queue-worker/
├── azure.yaml
├── infra/
│   ├── main.bicep
│   ├── app/
│   │   ├── api.bicep
│   │   └── worker.bicep
│   └── core/
│       ├── storage-queue.bicep
│       └── servicebus.bicep
└── src/
    ├── api/
    └── worker/

文件：src/worker/processor.py


import os
from azure.storage.queue import QueueClient
from azure.identity import DefaultAzureCredential

def process_orders():
    credential = DefaultAzureCredential()
    queue_url = os.getenv('AZURE_QUEUE_URL')
    
    queue_client = QueueClient.from_queue_url(
        queue_url=queue_url,
        credential=credential
    )
    
    while True:
        messages = queue_client.receive_messages(max_messages=10)
        for message in messages:
            # Process order
            print(f"Processing order: {message.content}")
            
            # Complete message
            queue_client.delete_message(message)

if __name__ == '__main__':
    process_orders()

文件：azure.yaml


name: order-processing
services:
  api:
    project: ./src/api
    language: python
    host: containerapp
    
  worker:
    project: ./src/worker
    language: python
    host: containerapp

部署：


# Initialize
azd init

# Deploy with queue configuration
azd up

# Scale worker based on queue length
az containerapp update \
  --name worker \
  --resource-group rg-order-processing \
  --scale-rule-name queue-scaling \
  --scale-rule-type azure-queue \
  --scale-rule-metadata queueName=orders accountName=storageaccount

高级模式

模式 1: 蓝绿部署


# Create new revision without traffic
azd deploy api --revision-suffix blue --no-traffic

# Test the new revision
curl https://api--blue.nicegrass-12345.eastus.azurecontainerapps.io/health

# Split traffic (20% to blue, 80% to current)
az containerapp ingress traffic set \
  --name api \
  --resource-group rg-myapp \
  --revision-weight latest=80 blue=20

# Full cutover to blue
az containerapp ingress traffic set \
  --name api \
  --resource-group rg-myapp \
  --revision-weight blue=100

模式 2: 使用 AZD 的金丝雀部署

文件：.azure/dev/config.json


{
  "deploymentStrategy": "canary",
  "canary": {
    "initialTrafficPercentage": 10,
    "incrementPercentage": 10,
    "intervalMinutes": 5
  }
}

部署脚本：


#!/bin/bash
# deploy-canary.sh

# Deploy new revision with 10% traffic
azd deploy api --revision-mode multiple

# Monitor metrics
azd monitor --service api --duration 5m

# Increase traffic gradually
for i in {20..100..10}; do
  echo "Increasing traffic to $i%"
  az containerapp revision set-traffic \
    --name api \
    --resource-group rg-myapp \
    --revision-weight latest=$i
  
  sleep 300  # Wait 5 minutes
done

模式 3: 多区域部署

文件：azure.yaml


name: global-app
services:
  api:
    project: ./src/api
    language: python
    host: containerapp
    regions:
      - eastus
      - westeurope
      - southeastasia

文件：infra/multi-region.bicep


param regions array = ['eastus', 'westeurope', 'southeastasia']

module containerApps './app/container-app.bicep' = [for region in regions: {
  name: 'app-${region}'
  params: {
    location: region
    environmentName: environmentName
  }
}]

// Traffic Manager for global load balancing
resource trafficManager 'Microsoft.Network/trafficManagerProfiles@2022-04-01' = {
  name: 'tm-global-app'
  location: 'global'
  properties: {
    trafficRoutingMethod: 'Performance'
    endpoints: [for i in range(0, length(regions)): {
      name: 'endpoint-${regions[i]}'
      type: 'Microsoft.Network/trafficManagerProfiles/externalEndpoints'
      properties: {
        target: containerApps[i].outputs.fqdn
        endpointStatus: 'Enabled'
      }
    }]
  }
}

部署：


# Deploy to all regions
azd up

# Verify endpoints
azd show --output json | jq '.services.api.endpoints'

模式 4: Dapr 集成

文件：infra/app/dapr-enabled.bicep


resource containerApp 'Microsoft.App/containerApps@2023-05-01' = {
  name: 'dapr-app'
  properties: {
    configuration: {
      dapr: {
        enabled: true
        appId: 'order-service'
        appPort: 8000
        appProtocol: 'http'
      }
    }
    template: {
      containers: [
        {
          name: 'app'
          image: containerImage
        }
      ]
    }
  }
}

带 Dapr 的应用代码：


from flask import Flask
from dapr.clients import DaprClient

app = Flask(__name__)

@app.route('/orders', methods=['POST'])
def create_order():
    with DaprClient() as client:
        # Save state
        client.save_state(
            store_name='statestore',
            key='order-123',
            value={'status': 'pending'}
        )
        
        # Publish event
        client.publish_event(
            pubsub_name='pubsub',
            topic_name='orders',
            data={'orderId': '123'}
        )
    
    return {'status': 'created'}

最佳实践

1. 资源组织


# Use consistent naming conventions
azd env set AZURE_ENV_NAME "myapp-prod"
azd env set AZURE_LOCATION "eastus"

# Tag resources for cost tracking
azd env set AZURE_TAGS "Environment=Production,CostCenter=Engineering"

2. 安全最佳实践


// Always use managed identity
resource containerApp 'Microsoft.App/containerApps@2023-05-01' = {
  identity: {
    type: 'SystemAssigned'
  }
}

// Store secrets in Key Vault
resource keyVault 'Microsoft.KeyVault/vaults@2023-02-01' = {
  properties: {
    enableRbacAuthorization: true
    networkAcls: {
      defaultAction: 'Deny'
      bypass: 'AzureServices'
    }
  }
}

// Use private endpoints
resource privateEndpoint 'Microsoft.Network/privateEndpoints@2023-04-01' = {
  properties: {
    subnet: {
      id: subnetId
    }
    privateLinkServiceConnections: [
      {
        name: 'containerapp-connection'
        properties: {
          privateLinkServiceId: containerApp.id
        }
      }
    ]
  }
}

3. 性能优化


# azure.yaml with performance settings
services:
  api:
    project: ./src/api
    host: containerapp
    resources:
      cpu: 1.0
      memory: 2Gi
    scale:
      minReplicas: 2
      maxReplicas: 20
      rules:
        - name: http-rule
          http:
            concurrent: 100

4. 监控和可观察性


# Enable Application Insights
azd env set APPLICATIONINSIGHTS_CONNECTION_STRING "InstrumentationKey=..."

# View logs in real-time
azd logs api --follow

# Monitor metrics
azd monitor --service api

# Create alerts
az monitor metrics alert create \
  --name high-cpu-alert \
  --resource-group rg-myapp \
  --scopes $(azd show --output json | jq -r '.services.api.resourceId') \
  --condition "avg CPU > 80" \
  --description "Alert when CPU exceeds 80%"

5. 成本优化


# Scale to zero when not in use
az containerapp update \
  --name api \
  --resource-group rg-myapp \
  --min-replicas 0

# Use spot instances for dev environments
azd env set CONTAINER_APP_REPLICA_TYPE "Spot"

# Set up budget alerts
az consumption budget create \
  --budget-name myapp-budget \
  --amount 100 \
  --time-grain Monthly \
  --threshold 80

6. CI/CD 集成

GitHub Actions 示例：


name: Deploy to Azure Container Apps

on:
  push:
    branches: [main]

jobs:
  deploy:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v3
      
      - name: Setup AZD
        uses: Azure/setup-azd@v1
      
      - name: Login to Azure
        run: |
          azd auth login --client-id ${{ secrets.AZURE_CLIENT_ID }} \
            --client-secret ${{ secrets.AZURE_CLIENT_SECRET }} \
            --tenant-id ${{ secrets.AZURE_TENANT_ID }}
      
      - name: Deploy
        run: azd up --no-prompt
        env:
          AZURE_ENV_NAME: ${{ secrets.AZURE_ENV_NAME }}
          AZURE_LOCATION: ${{ secrets.AZURE_LOCATION }}

常用命令参考


# Initialize new container app project
azd init --template <template-name>

# Deploy infrastructure and application
azd up

# Deploy only application code (skip infrastructure)
azd deploy

# Provision only infrastructure
azd provision

# View deployed resources
azd show

# Stream logs
azd logs <service-name> --follow

# Monitor application
azd monitor --overview

# Clean up resources
azd down --force --purge

故障排除

问题：容器无法启动


# Check logs
azd logs api --tail 100

# View container events
az containerapp revision show \
  --name api \
  --resource-group rg-myapp \
  --revision latest

# Test locally
docker build -t api:local ./src/api
docker run -p 8000:8000 api:local

问题：无法访问容器应用端点


# Verify ingress configuration
az containerapp show \
  --name api \
  --resource-group rg-myapp \
  --query properties.configuration.ingress

# Check if internal ingress is enabled
az containerapp ingress update \
  --name api \
  --resource-group rg-myapp \
  --external true

问题：性能问题


# Check resource utilization
az monitor metrics list \
  --resource $(azd show --output json | jq -r '.services.api.resourceId') \
  --metric "CPUPercentage,MemoryPercentage"

# Scale up resources
az containerapp update \
  --name api \
  --resource-group rg-myapp \
  --cpu 2.0 \
  --memory 4Gi

其他资源和示例

贡献

要贡献新的容器应用示例：

创建一个包含示例的新子目录
包含完整的 azure.yaml、infra/ 和 src/ 文件
添加详细的 README 文件，包含部署说明
使用 azd up 测试部署
提交一个拉取请求

需要帮助？ 加入 Microsoft Foundry Discord 社区获取支持和解答。

免责声明：
本文档使用AI翻译服务Co-op Translator进行翻译。尽管我们努力确保翻译的准确性，但请注意，自动翻译可能包含错误或不准确之处。原始语言的文档应被视为权威来源。对于关键信息，建议使用专业人工翻译。我们不对因使用此翻译而产生的任何误解或误读承担责任。

作者与出处

原作者：microsoft

来源：microsoft

原文： https://github.com/microsoft/AZD-for-beginners

许可证：MIT

整理人：灏天文库整理

由灏天文库结构化整理，提供目录导航、全文检索与在线阅读，便于系统化学习