Access AWS Resource In Another Account without STS

[Blasti]

The EC2 instance profile I setup in the master AWS account can assume an S3/Dynamo access role in another S3 account. How do i setup in Databricks/AWS so that when I use Python Boto3 to access S3 and Dynamo without using STS to assume the role. 

[Kaniz]

Hi @Blasti, To access Amazon S3 and DynamoDB from Databricks using Python Boto3 without explicitly assuming an STS role, you can follow these steps:

1. Amazon S3 Access:

   • Create a session using your AWS access key ID and secret access key. You can do this as follows:

     import boto3
     session = boto3.Session(
         aws_access_key_id='YOUR_ACCESS_KEY',
         aws_secret_access_key='YOUR_SECRET_KEY'
     )
     

   • Use the session to get an S3 resource:

     s3 = session.resource('s3')
     

   • Now you can perform S3 operations using the s3 resource.

2. DynamoDB Access:

   • Similarly, for DynamoDB, you can pass the access key and secret key directly in your code:

     import boto3
     session = boto3.Session(
         aws_access_key_id='YOUR_ACCESS_KEY',
         aws_secret_access_key='YOUR_SECRET_KEY'
     )
     client = session.client('dynamodb')
     

   • Now you can use the client to interact with DynamoDB.

Remember to replace 'YOUR_ACCESS_KEY' and 'YOUR_SECRET_KEY' with your actual AWS credentials. Keep in mind that embedding keys directly in the code is not the recommended way to access AWS resources with Boto3. It’s better to use IAM roles or environment variables for security reasons¹². 

[Blasti]

Hey Kaniz, i am sorry about the confusion. I should have made my question more clear. I mean to access without using IAM assume role or access key as if the i am access resource within the same aws account.