Trino

This page describes how to use Trino or Starburst as a source in Census.

Getting Started

Census can use Trino (and any supported Trino catalog) as a source. Census has been tested with the open-source Trino as well as Starburst Galaxy and Starburst Enterprise.

• Open Census and navigate to the Sources page.

• Click New Source and select Trino from the list.

• Configure your Trino connection:

  • Enter your Trino host name

  • Enter your Trino user name

  • Enter your Trino password

  • (Optional) If your Trino instance does not run on port 443, enter the port here. Census requires a TLS connection to Trino.

• You’re all set! Head over to the Syncs page to activate your data.

How to find your Hostname in Starburst:

Census utilizes JDBC in order to connect to Trino/Starburst. In order to get the correct hostname to input into your Census connection you will want to follow the instructions linked here in Starburst's documentation to get your JDBC url for your desired cluster.

Once you have the JDBC url you'll only need to input the subdomain into the Census hostname field.

Example JDBC url: jdbc:trino://census-example-cluster.trino.galaxy.starburst.io:443?user=[e-mail]/accountadmin Hostname value to input into Census census-example-cluster.trino.galaxy.starburst.io

⚙️ Using the Advanced Sync Engine

Trino supports both of Census' Sync Engines: Basic and Advanced. In order to use the Advanced Sync Engine with Trino, all of the following must be true:

• Your Trino cluster must have a catalog named CENSUS containing a schema named CENSUS

• The connector you use for the CENSUS catalog must support:

  • CREATE TABLE and DROP TABLE

  • Table writes, including row-level INSERT, DELETE, and UPDATE operations. Tables that support these operations are sometimes called "transactional" tables in Trino connector documentation, although Census does not require true ACID transactions for the Advanced Sync Engine.

  • The CREATE OR REPLACE TABLE statement, which was added to Trino in October 2023 and is available in some Starburst releases (check the release notes for your Starburst Enterprise version or for Starburst Galaxy).

• The account (service account or user) that Census uses to connect to your Trino cluster must have full permissions for the CENSUS.CENSUS schema.

We have successfully tested Census' Advanced Sync Engine with the following configurations - it's possible that other configurations are supported, and we encourage you to use the Census connection tester to obtain diagnostics if needed:

• CENSUS catalog using the MySQL, Postgres, and Snowflake connectors in read-write mode

• CENSUS catalog using the Iceberg connector with S3 object storage and the AWS Glue catalog

• CENSUS catalog using the Starburst Delta Lake connector, on both the AWS Glue catalog and the Starburst Galaxy catalog

Census is unable to provide any additional Trino table options (such as location) to the WITH clause when creating or managing tables in the CENSUS catalog, so please ensure your catalog and schema are configured with any needed default table options.

Notes

As of December 2023, Warehouse Writeback is not yet supported but is coming soon - please reach out to your Census account executive for details.

Need help connecting to Trino?

Contact us via [e-mail] or start a conversation with us via the in-app chat.