Connection

Catalog type

Glue id

Warehouse

Location of the warehouse, like: s3://some-bucket/some/path/

Region

AWS region used to connect to Glue

Warehouse

Location of the warehouse, like: s3a://some-bucket/some/path/

Use impersonated user

Access the files on HDFS as the end user, rather than as the DSS service user

Hadoop properties

Metastore uri

URI of the Hive metastore

Warehouse

Location of the Hive warehouse in HDFS

Use impersonated user

Access the files on HDFS as the end user, rather than as the DSS service user

Hadoop properties

Host

Role

URI

Warehouse

Name of a warehouse. Some iceberg catalogs use the name "catalog" for a warehouse.

URI

Warehouse

URI

Driver class

Warehouse

JDBC properties

Driver jars directory

Send driver jars to Spark

Experimental: Add driver jars to spark-submit subprocesses

Authentication type

User

Password

The AWS signature V4 authentication will use the default credentials provider chain or the credentials from the S3/Glue credentials set below.

Signing name

Signing region

User

Private key file

Use the key provided in the given full path on the DSS server

Private key

Or use the key provided as a string

Private key password

Optional, password to decode the key

Client id

Client secret

Scope

Authorization endpoint

Token endpoint

Rotate refresh tokens

Whether DSS should expect refresh tokens to be renewed when an access token is retrieved

Properties

Catalog properties

Credentials mode

Cloud storage credentials

These settings apply when the Iceberg catalog doesn't or can't provide vended credentials. DSS will then access the data files on the relevant cloud storage using credentials fetched from the connections selected below.

S3/Glue credentials

Name of a S3 connection to fetch AWS credentials from. If empty, use the environment's credentials. Used to connect to Glue as well.

GCS credentials

Name of a GCS connection to fetch GCS credentials from. If empty, use the environment's credentials.

Azure credentials

Name of a Azure connection to fetch Azure credentials from. If empty, use the environment's credentials.

Storage account container

Name of the container containing the files in the storage account. Needed when the Azure connection uses OAuth2.

Add endpoint properties

Add properties pertaining to cloud storage endpoints, on top of properties containing credentials.

Namespaces

Default namespace

Default namespace. Unqualified table lookups will be resolved in this namespace.

Auto-create namespaces

When possible, create namespaces as needed in the catalog when creating a table

Rules for new datasets

These settings define how managed Iceberg datasets are mapped to database tables. These settings are applied when creating a new managed Iceberg dataset. You can always modify these afterwards in the Iceberg dataset settings. See for more information.

Namespace

Managed datasets will be created in this namespace. You may reference variables here, they will be expanded dynamically

Allow Namespace override

If this is enabled, an option to choose a namespace will be presented to users each time they create a managed dataset on this connection

Table prefix

Table name for new managed datasets will be prefixed by this. You may reference variables here, they will be expanded dynamically

Table suffix

Table name for new managed datasets will be suffixed by this. You may reference variables here, they will be expanded dynamically

Custom properties

Advanced connection properties

For specific use cases

Catalog name

Name to use to register the Iceberg catalog internally and in Spark. Should only contain alphanumeric letters and underscores to be used in Spark. If not specified, the connection name is used.

Purge data on drop

Purge actual data when dropping a table from the Iceberg catalog.

Associated Trino conn.

Name of a Trino connection that has access to this Iceberg catalog.

Is Athena connection

Associated Trino catalog

Name of the Trino catalog corresponding to this Iceberg catalog.