Skip to main content
When creating an Iceberg source, sink, or internal table, you must configure a catalog. The catalog is responsible for managing the table’s metadata, including its schema, partitioning strategy, and data file locations. RisingWave supports two main types of catalogs: hosted catalogs, which are managed by RisingWave, and external catalogs, which you manage.

​
Catalog parameters

You specify the catalog configuration in the WITH clause of a CREATE SOURCE, CREATE SINK, or CREATE CONNECTION statement.
ParameterDescription
catalog.typeThe type of catalog. Supported values: 'storage', 'rest', 'hive', 'jdbc', 'glue', and 'snowflake'. Defaults to 'storage' if not specified.
catalog.uriThe URI for the catalog service. Required for rest, hive, jdbc, and snowflake catalog types.
catalog.nameA user-defined name for the catalog. Required for all types except storage.
hosted_catalogSet to true when using a hosted catalog with an internal Iceberg table. See Hosted Iceberg catalogs.
catalog.jdbc.userUsername for the JDBC catalog.
catalog.jdbc.passwordPassword for the JDBC catalog.
catalog.credentialCredentials for the REST catalog’s OAuth2 flow.
catalog.tokenA bearer token for the REST catalog.
catalog.oauth2_server_uriThe token endpoint URI for the REST catalog’s OAuth2 flow.
catalog.scopeThe OAuth2 scope for the REST catalog.

​
Hosted catalogs

A hosted catalog is a built-in service managed by RisingWave that simplifies the setup for internal Iceberg tables by removing the need to manage an external catalog. This is the recommended approach for getting started. For a complete guide, see Hosted Iceberg catalogs.

​
External catalogs

External catalogs are required when you need to connect to Iceberg tables that are managed by other systems. RisingWave supports several types of external catalogs.

​
Glue catalog

Integrates with AWS Glue Data Catalog. 
CREATE SINK my_glue_sink FROM my_mv WITH (
    connector = 'iceberg',
    type = 'upsert',
    primary_key = 'id',
    catalog.type = 'glue',
    warehouse.path = 's3://my-bucket/warehouse/',
    s3.region = 'us-west-2',
    s3.access.key = '...',
    s3.secret.key = '...',
    database.name = 'my_db',
    table.name = 'my_table'
);

​
JDBC catalog

Stores metadata in a relational database like PostgreSQL. 
CREATE SINK my_jdbc_sink FROM my_mv WITH (
    connector = 'iceberg',
    type = 'upsert',
    primary_key = 'id',
    catalog.type = 'jdbc',
    catalog.uri = 'jdbc:postgresql://postgres:5432/iceberg_db',
    catalog.jdbc.user = 'user',
    catalog.jdbc.password = 'password',
    warehouse.path = 's3://my-bucket/warehouse/',
    s3.region = 'us-west-2',
    s3.access.key = '...',
    s3.secret.key = '...',
    database.name = 'my_db',
    table.name = 'my_table'
);

​
REST catalog

Uses a RESTful API to manage table metadata. 
CREATE SINK my_rest_sink FROM my_mv WITH (
    connector = 'iceberg',
    type = 'upsert',
    primary_key = 'id',
    catalog.type = 'rest',
    catalog.uri = 'http://rest-catalog:8181',
    warehouse.path = 's3://my-bucket/warehouse/',
    s3.region = 'us-west-2',
    s3.access.key = '...',
    s3.secret.key = '...',
    database.name = 'my_db',
    table.name = 'my_table'
);

​
Amazon S3 Tables

Amazon S3 Tables is an AWS-native Iceberg catalog service that uses the REST protocol with AWS SigV4 authentication. To connect, you use catalog.type = 'rest' and provide additional parameters for authentication. Required REST parameters for S3 Tables:
ParameterDescriptionValue for S3 Tables
catalog.uriThe S3 Tables REST endpoint for your region.https://s3tables.<your-region>.amazonaws.com/iceberg
catalog.rest.signing_regionThe AWS region for signing requests.e.g., us-east-1
catalog.rest.signing_nameThe service name for signing requests.s3tables
catalog.rest.sigv4_enabledEnables SigV4 signing. Must be true.true
Example CREATE SINK with S3 Tables: 
CREATE SINK my_s3_tables_sink FROM my_mv
WITH (
    connector = 'iceberg',
    type = 'upsert',
    primary_key = 'id',
    warehouse.path = 'arn:aws:s3tables:<your-region>:<account-id>:bucket/<bucket-name>',
    s3.access.key = '...',
    s3.secret.key = '...',
    s3.region = '<your-region>',
    catalog.type = 'rest',
    catalog.uri = 'https://s3tables.<your-region>.amazonaws.com/iceberg',
    catalog.rest.signing_region = '<your-region>',
    catalog.rest.sigv4_enabled = true,
    catalog.rest.signing_name = 's3tables',
    database.name = '<your-database-name>',
    table.name = '<your-table-name>'
);

​
Hive Metastore catalog

Stores metadata using a Hive Metastore backend. 
CREATE SINK my_hive_sink FROM my_mv WITH (
    connector = 'iceberg',
    type = 'upsert',
    primary_key = 'id',
    catalog.type = 'hive',
    catalog.uri = 'thrift://hive-metastore:9083',
    warehouse.path = 's3://my-bucket/warehouse/',
    s3.region = 'us-west-2',
    s3.access.key = '...',
    s3.secret.key = '...',
    database.name = 'my_db',
    table.name = 'my_table'
);

​
Snowflake catalog

Uses Snowflake’s built-in Iceberg catalog functionality. 
CREATE SINK my_snowflake_sink FROM my_mv WITH (
    connector = 'iceberg',
    type = 'upsert',
    primary_key = 'id',
    catalog.type = 'snowflake',
    catalog.uri = 'jdbc:snowflake://<account_identifier>.snowflakecomputing.com/',
    catalog.jdbc.user = '<your_username>',
    catalog.jdbc.password = '<your_password>',
    warehouse.path = '<warehouse_path>',
    database.name = '<your_database>',
    table.name = '<your_table>'
);
⌘I