Redshift

On this page:

Overview

To query Tabular tables from Redshift, you must set up AWS Glue mirroring to sync the databases and tables in your Tabular warehouse to the Glue catalog. Redshift makes all Iceberg tables registered in Glue available for read-only queries using the awsdatacatalog catalog. See the AWS Redshift documentation for additional information.

Terminology: A Tabular database is semantically similar to a Redshift schema.

Current Limitations

  • Glue mirroring is a one-way replication; mirrored tables should be considered read-only.
  • Redshift does not leverage Tabular’s permission system to access tables. Rather, Redshift’s access to tables is governed by the IAM role used to execute queries in Redshift.

Configuration

Enable AWS Glue mirror [Tabular]

Enabling the AWS Glue mirror is available on the Connections -> Compute page.

Explore & Query Tables

Once you have successfully mirrored your Tabular warehouse to AWS Glue, all tables in that warehouse are immediately available for querying in the auto-provisioned awsdatacatalog catalog. Redshift is directly querying the Tabular managed Iceberg tables stored in S3, so any updates made to the tables from other engines such as AWS Athena or Apache Spark are immediately reflected.

Example:

-- Assuming we have previously created the taxi_tip_rate_per_passenger table
-- in the default database in our Tabular warehouse.
SELECT
    passenger_count,
    trip_count,
    format('%,.2f', tip_rate * 100) as tip_percentage
FROM awsdatacatalog.default.taxi_tip_rate_per_passenger
ORDER BY passenger_count;