如何用Python AWS CDK创建带数据表的Aurora 3 MySQL数据库

核心实现步骤 #

1. 安装依赖 #

确保本地已安装所需的CDK库：

pip install aws-cdk-lib aws-cdk.aws-rds aws-cdk.aws-ec2

2. 完整CDK代码示例 #

创建堆栈类，包含Aurora集群搭建、表结构初始化逻辑：

from aws_cdk import (
    Stack,
    aws_rds as rds,
    aws_ec2 as ec2,
    CfnOutput,
)
from constructs import Construct

class AuroraMysqlStack(Stack):
    def __init__(self, scope: Construct, construct_id: str, **kwargs) -> None:
        super().__init__(scope, construct_id, **kwargs)

        # 创建VPC（若已有可用VPC，可替换为导入现有资源）
        vpc = ec2.Vpc(self, "AuroraVpc",
            max_azs=2,
            subnet_configuration=[
                ec2.SubnetConfiguration(
                    name="Public",
                    subnet_type=ec2.SubnetType.PUBLIC
                ),
                ec2.SubnetConfiguration(
                    name="Private",
                    subnet_type=ec2.SubnetType.PRIVATE_ISOLATED
                )
            ]
        )

        # 搭建Aurora 3 MySQL集群
        aurora_cluster = rds.DatabaseCluster(self, "AuroraMysqlCluster",
            engine=rds.DatabaseClusterEngine.aurora_mysql(
                version=rds.AuroraMysqlEngineVersion.VER_3_03_0
            ),
            credentials=rds.Credentials.from_generated_secret("admin"),
            instance_props=rds.InstanceProps(
                vpc=vpc,
                vpc_subnets=ec2.SubnetSelection(subnet_type=ec2.SubnetType.PRIVATE_ISOLATED),
                instance_type=ec2.InstanceType.of(ec2.InstanceClass.BURSTABLE3, ec2.InstanceSize.SMALL)
            ),
            instances=1,
            default_database_name="mydb"
        )

        # 定义表结构SQL脚本
        init_sql = """
        CREATE TABLE IF NOT EXISTS table1 (
            id INT NOT NULL AUTO_INCREMENT PRIMARY KEY,
            name VARCHAR(255) NOT NULL
        );

        CREATE TABLE IF NOT EXISTS table2 (
            id INT NOT NULL AUTO_INCREMENT PRIMARY KEY,
            table1_id INT NOT NULL,
            data TEXT NOT NULL,
            FOREIGN KEY (table1_id) REFERENCES table1(id) ON DELETE CASCADE
        );
        """

        # 绑定初始化脚本，集群启动后自动执行
        rds.DatabaseInstance.from_database_instance_attributes(
            self, "ClusterInstanceRef",
            instance_identifier=aurora_cluster.instance_identifiers[0],
            port=aurora_cluster.cluster_endpoint.port,
            security_groups=aurora_cluster.connections.security_groups,
            instance_endpoint_address=aurora_cluster.cluster_endpoint.hostname
        ).add_init_command(rds.InitCommand.sql(init_sql))

        # 输出关键资源信息
        CfnOutput(self, "ClusterEndpoint", value=aurora_cluster.cluster_endpoint.hostname)
        CfnOutput(self, "AdminSecretArn", value=aurora_cluster.secret.secret_arn)

3. 部署操作 #

• 若为新项目，先初始化CDK环境：

cdk init app --language python

• 将上述代码替换项目中app.py的堆栈定义
• 执行部署：

cdk deploy

关键细节说明 #

• VPC配置：示例新建了VPC，如需用现有VPC，可通过ec2.Vpc.from_lookup方法导入
• 表初始化：通过add_init_command绑定SQL脚本，确保集群就绪后自动创建表和外键约束
• 访问控制：CDK会自动配置基础安全组规则，若需外部客户端访问，需手动添加入站规则开放MySQL端口（3306）

内容的提问来源于stack exchange，提问作者TheMemeMachine