How to use Boto3 to remove tags from AWS Glue Resources

In this article, we will see how to remove tags from AWS Glue Resources using the boto3 library. AWS Glue resources can have tags for organization and cost tracking, and sometimes you need to remove specific tags programmatically.

Problem Statement

Use boto3 library in Python to remove tags from AWS Glue Resources like databases, tables, crawlers, and jobs.

Required Parameters

The untag_resource function requires two main parameters:

• resource_arn ? The Amazon Resource Name (ARN) of the Glue resource
• tags_list ? List of tag keys to remove (format: ["key1", "key2"])

AWS Glue Resource ARN Formats

Implementation Steps

1. Import boto3 and handle exceptions with botocore
2. Create an AWS session and Glue client
3. Use the untag_resource method with proper parameters
4. Handle potential errors during the operation

Example Code

Here's how to remove tags from an AWS Glue database −

import boto3
from botocore.exceptions import ClientError

def remove_tags_from_glue_resource(resource_arn, tags_to_remove):
    """
    Remove specified tags from AWS Glue resource
    
    Args:
        resource_arn (str): ARN of the Glue resource
        tags_to_remove (list): List of tag keys to remove
    
    Returns:
        dict: Response from AWS API
    """
    try:
        # Create AWS session and Glue client
        session = boto3.session.Session()
        glue_client = session.client('glue')
        
        # Remove tags from the resource
        response = glue_client.untag_resource(
            ResourceArn=resource_arn,
            TagsToRemove=tags_to_remove
        )
        
        return response
        
    except ClientError as e:
        raise Exception(f"AWS Client error in remove_tags_from_glue_resource: {str(e)}")
    except Exception as e:
        raise Exception(f"Unexpected error in remove_tags_from_glue_resource: {str(e)}")

# Example usage
if __name__ == "__main__":
    # Define the resource ARN and tags to remove
    database_arn = "arn:aws:glue:us-east-1:123456789012:database/sample-db"
    tags_to_remove = ["glue-db", "environment"]
    
    try:
        result = remove_tags_from_glue_resource(database_arn, tags_to_remove)
        print("Tags removed successfully!")
        print(f"Response: {result}")
    except Exception as e:
        print(f"Error: {e}")

Expected Output

When the tags are successfully removed, you'll see a response like this −

Tags removed successfully!
Response: {
    'ResponseMetadata': {
        'RequestId': 'c9f418b0-****-****-****-************fb96',
        'HTTPStatusCode': 200,
        'HTTPHeaders': {
            'date': 'Fri, 02 Apr 2021 08:04:54 GMT',
            'content-type': 'application/x-amz-json-1.1',
            'content-length': '27',
            'connection': 'keep-alive'
        },
        'RetryAttempts': 0
    }
}

Key Points

• Ensure your AWS credentials are properly configured
• The resource ARN must be valid and accessible
• Only specify tag keys in the TagsToRemove parameter, not key-value pairs
• Handle both ClientError and generic exceptions for robust error handling

Conclusion

Using boto3's untag_resource method makes it easy to remove tags from AWS Glue resources programmatically. Always handle exceptions properly and ensure your AWS credentials have the necessary permissions for tag management operations.