AWS Integration

UniverSQL allows you to efficiently query Snowflake Iceberg tables stored in AWS S3 using your local compute resources, resulting in significant cost savings on Snowflake warehouse compute.

Setup AWS Credentials

1. First, install the AWS CLI
2. Configure your AWS credentials

# If using AWS client ID/secret
aws configure
 
# If using AWS SSO
aws configure sso

Using AWS Profile with UniverSQL

By default, UniverSQL uses your default AWS profile. You can specify a different profile using the --aws-profile option:

universql snowflake --account your_account --aws-profile custom_profile

Features

• Direct access to S3-stored Iceberg tables without Snowflake warehouse compute
• Smart caching of Parquet files for better performance
• Support for AWS IAM authentication
• Automatic handling of AWS credentials and regions
• Predicate pushdown support for partitioned tables

Cost Benefits

When querying Iceberg tables through UniverSQL:

• No Snowflake warehouse compute costs
• Only pay for S3 egress costs
• Local compute utilization for queries
• Smart caching reduces repeated data transfers

Best Practices

1. Use appropriate IAM roles with minimal required permissions
2. Configure AWS credentials securely
3. Utilize partitioned tables for better performance
4. Monitor S3 egress costs