为 Bash 创建 Azure CLI 脚本

已完成

在 Azure 门户中几乎可以完成的每项任务，您也可以使用 Azure CLI 参考命令来完成。 使用 Azure 门户 了解 Azure 是一个很好的起点。 但是，我们建议使用 Azure CLI 或 Azure PowerShell 大规模管理 Azure 资源 。

请考虑为全球公司管理 Azure 的方案。 每天收到针对新 资源组、 Azure 逻辑应用、 存储帐户、 Azure 数据工厂管道和 Azure SQL 数据库的多个请求。 所有团队都在开发、测试和生产环境中工作。 对于每个请求，都需要创建三个遵循公司命名标准和安全策略的类似 Azure 资源 。 是时候使用 Azure CLI 脚本了！

了解 Azure CLI 脚本

Azure CLI 脚本可实现大规模重复任务和基础结构部署的自动化。 脚本提供一致性、减少人为错误，并启用基础结构即代码（IaC）做法。 无论是管理开发、过渡环境还是生产环境， Azure CLI 脚本都简化了资源预配和管理。

脚本结构基础知识

结构良好的 Azure CLI 脚本通常包括：

• Shebang （#!/bin/bash 或 #!/usr/bin/env bash）： 指定执行脚本的翻译。
• 变量： 存储可重用的值，例如资源名称、位置和 SKU 配置。
• 评论： 记录脚本逻辑和参数以保持可维护性。
• 错误处理： 实现检查以正常处理故障。
• 循环和条件： 根据条件创建多个资源或应用逻辑。
• 输出格式： 以可读格式显示结果进行验证。

创建资源部署脚本

让我们来看看一个实际的 Bash 脚本，该脚本以一致、可重复的方式创建多个 Azure 存储帐户 。

#!/bin/bash

# Script: Create multiple Azure storage accounts
# Description: Automates storage account creation with consistent naming and configuration

# Define deployment parameters
resourceGroupName="rg-storage-prod-eastus"
saCount=3
saLocation="eastus"
saNamePrefix="stprod"
saSku="Standard_GRS"
saKind="StorageV2"

# Validate Azure CLI is installed and authenticated
if ! command -v az &> /dev/null; then
    echo "Error: Azure CLI is not installed. Please install Azure CLI first."
    exit 1
fi

if ! az account show &> /dev/null; then
    echo "Error: Not authenticated to Azure. Please run 'az login' first."
    exit 1
fi

echo "Starting storage account deployment..."
echo "Resource Group: $resourceGroupName"
echo "Location: $saLocation"
echo "Count: $saCount"
echo ""

# Create resource group if it doesn't exist
if ! az group show --name $resourceGroupName &> /dev/null; then
    echo "Creating resource group $resourceGroupName..."
    az group create --name $resourceGroupName --location $saLocation
else
    echo "Resource group $resourceGroupName already exists."
fi

# Loop to create multiple storage accounts
for i in $(seq 1 $saCount)
do
    # Generate unique identifier using timestamp and random number
    timestamp=$(date +%s)
    let "randomIdentifier=$RANDOM"
    saName="${saNamePrefix}${timestamp}${randomIdentifier}"

    # Ensure name is lowercase and within 24 character limit
    saName=$(echo $saName | tr '[:upper:]' '[:lower:]' | cut -c1-24)

    echo "Creating storage account $saName..."

    # Create storage account with error handling
    if az storage account create \
        --name $saName \
        --resource-group $resourceGroupName \
        --location $saLocation \
        --sku $saSku \
        --kind $saKind \
        --tags Environment=Production ManagedBy=AzureCLI \
        --output none; then
        echo "✓ Successfully created storage account $saName"
    else
        echo "✗ Failed to create storage account $saName"
    fi
done

echo ""
echo "Deployment complete. Verifying results..."
echo ""

# Verify results with formatted output
az storage account list \
    --resource-group $resourceGroupName \
    --query "[].{Name:name, Location:location, SKU:sku.name, Status:statusOfPrimary}" \
    --output table

脚本分解

让我们检查关键组件：

身份验证验证：

if ! az account show &> /dev/null; then
    echo "Error: Not authenticated to Azure. Please run 'az login' first."
    exit 1
fi

确保在执行资源作之前对用户进行身份验证，防止脚本失败。

使用存在检查创建资源组：

if ! az group show --name $resourceGroupName &> /dev/null; then
    az group create --name $resourceGroupName --location $saLocation
fi

仅当资源组尚不存在时，才创建资源组，使脚本具有幂等性。

独特名称生成：

timestamp=$(date +%s)
let "randomIdentifier=$RANDOM"
saName="${saNamePrefix}${timestamp}${randomIdentifier}"
saName=$(echo $saName | tr '[:upper:]' '[:lower:]' | cut -c1-24)

组合时间戳和随机数以生成唯一的存储帐户名称，转换为小写，并按 Azure 命名要求限制为 24 个字符。

错误处理：

if az storage account create ...; then
    echo "✓ Successfully created storage account $saName"
else
    echo "✗ Failed to create storage account $saName"
fi

验证每个存储帐户的创建，并提供明确的成功/失败反馈。

创建参数化脚本

生产脚本应接受用于跨环境灵活性的参数：

#!/bin/bash

# Script: create-storage-accounts.sh
# Usage: ./create-storage-accounts.sh <resource-group> <location> <count> <environment>

# Accept command-line parameters
resourceGroupName=$1
saLocation=$2
saCount=$3
environment=$4

# Validate parameters
if [ -z "$resourceGroupName" ] || [ -z "$saLocation" ] || [ -z "$saCount" ] || [ -z "$environment" ]; then
    echo "Usage: $0 <resource-group> <location> <count> <environment>"
    echo "Example: $0 rg-storage-dev eastus 3 Development"
    exit 1
fi

# Validate count is a number
if ! [[ "$saCount" =~ ^[0-9]+$ ]]; then
    echo "Error: Count must be a number"
    exit 1
fi

echo "Parameters received:"
echo "  Resource Group: $resourceGroupName"
echo "  Location: $saLocation"
echo "  Count: $saCount"
echo "  Environment: $environment"
echo ""

# Set environment-specific configurations
case $environment in
    Development)
        saSku="Standard_LRS"
        saKind="StorageV2"
        ;;
    Staging)
        saSku="Standard_GRS"
        saKind="StorageV2"
        ;;
    Production)
        saSku="Standard_RAGRS"
        saKind="StorageV2"
        ;;
    *)
        echo "Error: Environment must be Development, Staging, or Production"
        exit 1
        ;;
esac

echo "Using SKU: $saSku for $environment environment"
echo ""

# Continue with resource creation...

执行参数化脚本

# Make script executable
chmod +x create-storage-accounts.sh

# Execute with parameters
./create-storage-accounts.sh rg-storage-dev eastus 3 Development

创建配置驱动脚本

对于复杂的部署，请使用外部配置文件：

config.json：

{
  "resourceGroup": {
    "name": "rg-app-prod-eastus",
    "location": "eastus"
  },
  "storageAccounts": [
    {
      "namePrefix": "stappdata",
      "sku": "Standard_RAGRS",
      "kind": "StorageV2",
      "count": 2
    },
    {
      "namePrefix": "stappbackup",
      "sku": "Standard_GRS",
      "kind": "StorageV2",
      "count": 1
    }
  ],
  "tags": {
    "Environment": "Production",
    "CostCenter": "Engineering",
    "ManagedBy": "AzureCLI"
  }
}

deploy-from-config.sh：

#!/bin/bash

# Load configuration file
configFile="config.json"

if [ ! -f "$configFile" ]; then
    echo "Error: Configuration file $configFile not found"
    exit 1
fi

# Parse configuration using jq (JSON processor)
resourceGroupName=$(jq -r '.resourceGroup.name' $configFile)
resourceGroupLocation=$(jq -r '.resourceGroup.location' $configFile)
tagsJson=$(jq -r '.tags | to_entries | map("\(.key)=\(.value)") | join(" ")' $configFile)

echo "Deploying resources from configuration file..."
echo "Resource Group: $resourceGroupName"
echo "Location: $resourceGroupLocation"
echo ""

# Create resource group
az group create --name $resourceGroupName --location $resourceGroupLocation

# Read storage account configurations
storageConfigCount=$(jq '.storageAccounts | length' $configFile)

for i in $(seq 0 $(($storageConfigCount - 1)))
do
    namePrefix=$(jq -r ".storageAccounts[$i].namePrefix" $configFile)
    sku=$(jq -r ".storageAccounts[$i].sku" $configFile)
    kind=$(jq -r ".storageAccounts[$i].kind" $configFile)
    count=$(jq -r ".storageAccounts[$i].count" $configFile)

    echo "Creating $count storage account(s) with prefix: $namePrefix"

    for j in $(seq 1 $count)
    do
        timestamp=$(date +%s)
        saName="${namePrefix}${timestamp}${RANDOM}"
        saName=$(echo $saName | tr '[:upper:]' '[:lower:]' | cut -c1-24)

        az storage account create \
            --name $saName \
            --resource-group $resourceGroupName \
            --location $resourceGroupLocation \
            --sku $sku \
            --kind $kind \
            --tags $tagsJson \
            --output none

        echo "  ✓ Created: $saName"
    done
done

echo ""
echo "Deployment complete."

配置驱动的优势

• 关注点分离： 使配置与部署逻辑分离。
• 环境提升：针对开发、暂存、生产环境使用相同脚本但配置文件不同。
• 版本控制： 跟踪配置更改以及代码。
• 验证： 在部署之前验证配置结构。

使用脚本删除 Azure 资源

创建和测试脚本时，请记住删除测试资源以避免不必要的成本。 Azure CLI 脚本支持安全、一致的资源清理。

按创建日期删除存储帐户

删除在特定日期和时间或之后创建的所有 存储帐户 ：

#!/bin/bash

# Define cleanup parameters
cutoffDate="2025-10-08T00:00:00.000000+00:00"
resourceGroup="rg-storage-dev-eastus"

echo "Finding storage accounts created on or after $cutoffDate..."

# Get list of storage accounts matching criteria
saList=$(az storage account list \
    --resource-group $resourceGroup \
    --query "[?creationTime >='$cutoffDate'].{Name:name, Created:creationTime}" \
    --output table)

echo "$saList"
echo ""

# Confirm deletion
read -p "Delete these storage accounts? (yes/no): " confirm

if [ "$confirm" == "yes" ]; then
    for saId in $(az storage account list \
        --resource-group $resourceGroup \
        --query "[?creationTime >='$cutoffDate'].id" \
        --output tsv); do
        echo "Deleting storage account: $saId"
        az storage account delete --ids $saId --yes
    done
    echo "Cleanup complete."
else
    echo "Deletion cancelled."
fi

使用记录删除资源组

删除与具有全面日志记录的命名模式匹配的 资源组 ：

#!/bin/bash

# Define cleanup parameters
namePattern="rg-dev-*"
logFileLocation="cleanup-$(date +%Y%m%d-%H%M%S).log"

echo "Resource Group Cleanup" > $logFileLocation
echo "Started: $(date)" >> $logFileLocation
echo "Pattern: $namePattern" >> $logFileLocation
echo "----------------------------------------" >> $logFileLocation

# Find matching resource groups
echo "Finding resource groups matching pattern: $namePattern"
matchingGroups=$(az group list \
    --query "[?starts_with(name, 'rg-dev-')].name" \
    --output tsv)

if [ -z "$matchingGroups" ]; then
    echo "No resource groups found matching pattern: $namePattern"
    echo "No resource groups found" >> $logFileLocation
    exit 0
fi

# Display matches
echo "Found resource groups:"
echo "$matchingGroups"
echo ""

# Log matches
echo "Resource groups found:" >> $logFileLocation
echo "$matchingGroups" >> $logFileLocation
echo "" >> $logFileLocation

# Confirm deletion
read -p "Delete these resource groups? (yes/no): " confirm

if [ "$confirm" == "yes" ]; then
    echo "Starting deletion..." >> $logFileLocation

    for rgName in $matchingGroups
    do
        echo "Deleting resource group: $rgName"
        echo "$(date): Deleting $rgName" >> $logFileLocation

        # Delete with --no-wait for background execution
        if az group delete --name $rgName --yes --no-wait; then
            echo "  ✓ Deletion initiated for $rgName"
            echo "$(date): ✓ Deletion initiated for $rgName" >> $logFileLocation
        else
            echo "  ✗ Failed to initiate deletion for $rgName"
            echo "$(date): ✗ Failed to delete $rgName" >> $logFileLocation
        fi
    done

    echo ""
    echo "Deletion operations initiated. Resources will be removed in the background."
    echo "Check deletion status with: az group list --query \"[?starts_with(name, 'rg-dev-')].name\""

    echo "" >> $logFileLocation
    echo "Completed: $(date)" >> $logFileLocation
    echo "Log saved to: $logFileLocation"

    # Display log
    echo ""
    echo "=== Cleanup Log ==="
    cat $logFileLocation
else
    echo "Deletion cancelled by user" >> $logFileLocation
    echo "Deletion cancelled."
fi

安全删除做法

在删除之前实施安全检查：

#!/bin/bash

# Delete resources safely with multiple confirmations
resourceGroup="rg-storage-test-eastus"

# Check if resource group exists
if ! az group show --name $resourceGroup &> /dev/null; then
    echo "Error: Resource group '$resourceGroup' not found"
    exit 1
fi

# Display resources that will be deleted
echo "Resources in resource group '$resourceGroup':"
az resource list --resource-group $resourceGroup \
    --query "[].{Name:name, Type:type, Location:location}" \
    --output table

echo ""
echo "⚠️  WARNING: This will delete ALL resources in '$resourceGroup'"
echo ""

# First confirmation
read -p "Are you sure you want to delete '$resourceGroup'? (yes/no): " confirm1

if [ "$confirm1" != "yes" ]; then
    echo "Deletion cancelled."
    exit 0
fi

# Second confirmation with exact name
read -p "Type the resource group name to confirm: " confirm2

if [ "$confirm2" != "$resourceGroup" ]; then
    echo "Resource group name does not match. Deletion cancelled."
    exit 1
fi

# Final countdown
echo "Deleting in 5 seconds. Press Ctrl+C to cancel."
sleep 5

echo "Deleting resource group '$resourceGroup'..."
az group delete --name $resourceGroup --yes

echo "Resource group deleted successfully."

警告

在删除之前，请始终验证资源。 删除 资源组 会永久删除其中的所有资源。 此操作不可撤消。

生产脚本的最佳做法

生产就绪 的 Azure CLI 脚本需要仔细规划和实施。 本单元中提供的脚本演示了基本概念，但生产环境需要其他注意事项。

基本生产脚本组件

适用于生产环境的可靠 Azure CLI 脚本应包括：

1.身份验证和授权：

#!/bin/bash

# Authenticate with service principal (for automation)
if [ -n "$AZURE_CLIENT_ID" ] && [ -n "$AZURE_CLIENT_SECRET" ] && [ -n "$AZURE_TENANT_ID" ]; then
    echo "Authenticating with service principal..."
    az login --service-principal \
        --username $AZURE_CLIENT_ID \
        --password $AZURE_CLIENT_SECRET \
        --tenant $AZURE_TENANT_ID
else
    echo "Error: Service principal credentials not found"
    exit 1
fi

# Set subscription
az account set --subscription $AZURE_SUBSCRIPTION_ID

# Verify authentication
if ! az account show &> /dev/null; then
    echo "Error: Authentication failed"
    exit 1
fi

2. 全面的错误处理：

#!/bin/bash

# Exit immediately if command fails
set -e

# Trap errors and perform cleanup
trap 'cleanup_on_error' ERR

cleanup_on_error() {
    echo "Error occurred on line $1"
    echo "Performing cleanup..."
    # Add cleanup logic here
    exit 1
}

# Enable error trapping with line numbers
trap 'cleanup_on_error $LINENO' ERR

3. 幂等性检查：

#!/bin/bash

# Check if resource group exists before creation
create_resource_group() {
    local rgName=$1
    local location=$2

    if az group show --name $rgName &> /dev/null; then
        echo "Resource group $rgName already exists. Skipping creation."
    else
        echo "Creating resource group $rgName..."
        az group create --name $rgName --location $location
    fi
}

# Check if storage account exists
create_storage_account() {
    local saName=$1
    local rgName=$2

    if az storage account show --name $saName --resource-group $rgName &> /dev/null; then
        echo "Storage account $saName already exists. Skipping creation."
        return 0
    else
        echo "Creating storage account $saName..."
        az storage account create --name $saName --resource-group $rgName --location eastus
    fi
}

4. 结构化日志记录：

#!/bin/bash

# Define log file with timestamp
logFile="deployment-$(date +%Y%m%d-%H%M%S).log"
errorLog="errors-$(date +%Y%m%d-%H%M%S).log"

# Logging function
log() {
    local level=$1
    shift
    local message="$@"
    local timestamp=$(date "+%Y-%m-%d %H:%M:%S")

    echo "[$timestamp] [$level] $message" | tee -a $logFile

    if [ "$level" == "ERROR" ]; then
        echo "[$timestamp] $message" >> $errorLog
    fi
}

# Usage
log "INFO" "Starting deployment..."
log "ERROR" "Failed to create resource"
log "SUCCESS" "Deployment complete"

5.配置验证：

#!/bin/bash

# Validate required configuration
validate_config() {
    local errors=0

    if [ -z "$RESOURCE_GROUP" ]; then
        echo "Error: RESOURCE_GROUP not defined"
        ((errors++))
    fi

    if [ -z "$LOCATION" ]; then
        echo "Error: LOCATION not defined"
        ((errors++))
    fi

    # Validate location exists
    if ! az account list-locations --query "[?name=='$LOCATION']" | grep -q "$LOCATION"; then
        echo "Error: Invalid location: $LOCATION"
        ((errors++))
    fi

    if [ $errors -gt 0 ]; then
        echo "Configuration validation failed with $errors error(s)"
        exit 1
    fi

    echo "Configuration validation passed"
}

6. 标记策略：

#!/bin/bash

# Define standard tags
environment=$1  # Development, Staging, Production
costCenter=$2
owner=$3

# Create tags string
tags="Environment=$environment CostCenter=$costCenter Owner=$owner \
      CreatedBy=AzureCLI CreatedDate=$(date +%Y-%m-%d) \
      ManagedBy=Infrastructure-Team Project=WebApp"

# Apply tags to resources
az group create \
    --name $resourceGroup \
    --location $location \
    --tags $tags

az storage account create \
    --name $saName \
    --resource-group $resourceGroup \
    --location $location \
    --tags $tags

7. 回滚功能：

#!/bin/bash

# Track created resources for rollback
createdResources=()

create_with_tracking() {
    local resourceType=$1
    local resourceName=$2

    # Attempt resource creation
    if create_resource "$resourceType" "$resourceName"; then
        createdResources+=("$resourceType:$resourceName")
        echo "✓ Created: $resourceType - $resourceName"
        return 0
    else
        echo "✗ Failed: $resourceType - $resourceName"
        return 1
    fi
}

rollback() {
    echo "Rolling back created resources..."

    for resource in "${createdResources[@]}"; do
        IFS=':' read -r type name <<< "$resource"
        echo "Deleting $type: $name"
        delete_resource "$type" "$name"
    done

    echo "Rollback complete"
}

# Use trap to rollback on error
trap rollback ERR

脚本组织结构

将脚本组织到逻辑模块中：

azure-infrastructure/
├── config/
│   ├── dev.json
│   ├── staging.json
│   └── prod.json
├── scripts/
│   ├── deploy-infrastructure.sh
│   ├── deploy-storage.sh
│   └── cleanup.sh
├── lib/
│   ├── common-functions.sh
│   ├── logging.sh
│   └── validation.sh
├── logs/
└── README.md

可重用函数库

common-functions.sh：

#!/bin/bash

# Load this file in other scripts: source ./lib/common-functions.sh

# Check if Azure CLI is installed
check_azure_cli() {
    if ! command -v az &> /dev/null; then
        echo "Error: Azure CLI not installed"
        exit 1
    fi
}

# Wait for resource to be ready
wait_for_resource() {
    local resourceId=$1
    local maxAttempts=30
    local attempt=1

    echo "Waiting for resource to be ready..."

    while [ $attempt -le $maxAttempts ]; do
        if az resource show --ids $resourceId &> /dev/null; then
            echo "Resource is ready"
            return 0
        fi

        echo "Attempt $attempt/$maxAttempts - waiting..."
        sleep 10
        ((attempt++))
    done

    echo "Timeout waiting for resource"
    return 1
}

# Generate unique name
generate_unique_name() {
    local prefix=$1
    local timestamp=$(date +%s)
    local random=$RANDOM
    echo "${prefix}${timestamp}${random}" | tr '[:upper:]' '[:lower:]' | cut -c1-24
}

安全地测试脚本

在隔离环境中测试脚本：

#!/bin/bash

# Test mode flag
TEST_MODE=${TEST_MODE:-false}

execute_command() {
    local command=$1

    if [ "$TEST_MODE" == "true" ]; then
        echo "[TEST MODE] Would execute: $command"
    else
        echo "Executing: $command"
        eval $command
    fi
}

# Usage
execute_command "az group create --name rg-test --location eastus"

在测试模式下运行：

TEST_MODE=true ./deploy-infrastructure.sh

其他资源

• Azure CLI 示例存储库：github.com/Azure-Samples/azure-cli-samples。
• Azure CLI 脚本指南：learn.microsoft.com/cli/azure/azure-cli-vm-tutorial。
• Bash 脚本最佳做法：www.gnu.org/software/bash/manual/。

本单元中的脚本演示了将变量、循环、条件和错误处理与 Azure CLI 命令结合使用时的 Azure CLI 功能。 这些模式从管理少量资源扩展到协调复杂的企业基础设施部署。