AWS CDK S3 Tables Alpha

Common errors

• ModuleNotFoundError: No module named 'aws_cdk_aws_s3tables_alpha'

  cause The `aws-cdk-aws-s3tables-alpha` Python package has not been installed.
  fix Install the package: `pip install aws-cdk-aws-s3tables-alpha`.

• jsii.errors.JavaScriptError: Cannot read properties of undefined (reading 'PARQUET')

  cause This error often occurs when `DataFormat` or `Column` are not imported or referenced correctly within the `TableOptions` or `S3Table` definition.
  fix Ensure `from aws_cdk_aws_s3tables_alpha import DataFormat, Column` is present and that `DataFormat.PARQUET` (or other formats) and `Column(...)` objects are correctly passed to `TableOptions`.

• jsii.errors.JavaScriptError: The 'databaseName' property is required for S3Table.

  cause Essential properties like `database_name` or `table_name` were omitted when instantiating `S3Table`.
  fix Provide all required properties in the `S3Table` constructor. For example: `S3Table(self, 'Id', bucket=my_bucket, table_name='my-table', database_name='my-database', table_options=...)`.

Warnings

• breaking This is an ALPHA module. Stability is not guaranteed. Breaking changes can and do occur frequently, even between minor versions. Always consult the latest module documentation before updating.
  Fix: Pin your module versions (e.g., `aws-cdk-aws-s3tables-alpha==2.250.0a0`) and review changelogs or module-specific READMEs carefully before upgrading. Be prepared to refactor code when updating.
• gotcha CDK v1 vs. v2 import paths: This module is exclusively for AWS CDK v2. Trying to import `aws_cdk` components in the v1 style (e.g., `import aws_cdk.aws_s3`) will lead to `ModuleNotFoundError` or other incompatibilities if using CDK v2.
  Fix: Ensure all CDK imports follow the v2 pattern (e.g., `from aws_cdk import aws_s3 as s3`). If migrating from v1, consult the CDK v1 to v2 migration guide.
• gotcha The `S3Table` construct creates AWS Glue Data Catalog tables. Ensure your CDK deployment role has sufficient IAM permissions to create, update, and delete Glue databases and tables, as well as S3 bucket permissions for the data.
  Fix: Verify that the IAM role associated with your CDK deployment (e.g., `cdk deploy`) has policies like `AWSGlueConsoleFullAccess` (for development) or more granular `glue:CreateDatabase`, `glue:CreateTable`, `s3:GetObject`, `s3:PutObject` permissions for production environments.

Install

• pip install aws-cdk-aws-s3tables-alpha aws-cdk-lib constructs Install with pip

Imports

• S3Table

  from aws_cdk_aws_s3tables_alpha import S3Table

• TableOptions

  from aws_cdk_aws_s3tables_alpha import TableOptions

• DataFormat

  from aws_cdk_aws_s3tables_alpha import DataFormat

• Column

  from aws_cdk_aws_s3tables_alpha import Column

• aws_s3 as s3
  wrong:

  import aws_cdk.aws_s3 as s3

  correct:

  from aws_cdk import aws_s3 as s3

  CDK v2 imports AWS service modules as `aws_service` directly from `aws_cdk`.

Quickstart

This quickstart demonstrates how to create an S3 bucket and then define an S3-backed Glue Data Catalog table using `S3Table`. The table is configured with a PARQUET data format and several columns. Replace placeholder values like `product_sales` and `sales_data_db` with your specific requirements. This code synthesizes a CDK CloudFormation template.

import os
from aws_cdk import App, Stack, Environment, aws_s3 as s3
from aws_cdk.aws_s3 import Bucket, RemovalPolicy
from aws_cdk_aws_s3tables_alpha import S3Table, TableOptions, DataFormat, Column

class MyS3TableStack(Stack):
    def __init__(self, scope: App, id: str, **kwargs) -> None:
        super().__init__(scope, id, **kwargs)

        # Define an S3 Bucket to store the table data
        # In production, use RemovalPolicy.RETAIN and consider bucket versioning
        data_bucket = Bucket(self, "MyDataLakeBucket",
            versioned=True,
            removal_policy=RemovalPolicy.DESTROY # CAUTION: Deletes bucket and contents on stack delete
        )

        # Define the S3 Table construct, creating a Glue Table
        s3_glue_table = S3Table(self, "MyProductSalesS3Table",
            bucket=data_bucket,
            table_name="product_sales",
            database_name="sales_data_db", # The AWS Glue database name
            table_options=TableOptions(
                data_format=DataFormat.PARQUET, # Common for data lakes
                columns=[
                    Column(name="product_id", type="string"),
                    Column(name="sale_date", type="date"),
                    Column(name="amount", type="double"),
                    Column(name="region", type="string")
                ],
            ),
            # Example: Partition by year and month for efficient querying
            # partition_keys=[
            #     Column(name="year", type="string"),
            #     Column(name="month", type="string")
            # ]
        )

        # You can access properties like table_name or database_name
        # CfnOutput(self, "S3BucketName", value=data_bucket.bucket_name)
        # CfnOutput(self, "GlueTableName", value=s3_glue_table.table_name)

# Standard CDK application entry point
app = App()
MyS3TableStack(app, "MyS3TableStack",
    env=Environment(
        account=os.environ.get("CDK_DEFAULT_ACCOUNT"),
        region=os.environ.get("CDK_DEFAULT_REGION")
    )
)
app.synth()