Openflow Connector for Salesforce Bulk API: Iceberg table destinations¶

The Openflow Connector for Salesforce Bulk API supports writing to Snowflake-managed Apache Apache Iceberg™ tables as an opt-in destination format. The connector supports Iceberg v2 and v3. Existing connectors that write to standard Snowflake tables aren’t affected.

To use Iceberg table destinations, configure the Snowflake destination database with Iceberg defaults, then set the connector’s Table Storage Format parameter to ICEBERG and choose an Iceberg Version before the connector creates destination tables.

Prerequisites¶

Before you begin:

Complete the steps in Openflow Connector for Salesforce Bulk API: Set up Snowflake.
Choose either Snowflake storage or an external volume in your cloud storage for the Iceberg table files.

Configure the destination database¶

Configure the destination database with Iceberg defaults before starting the connector. The connector reads the database defaults for the external volume and storage serialization policy when it creates Iceberg tables.

Option A: Snowflake storage¶

To use Snowflake storage for Apache Apache Iceberg™ tables, configure the database with EXTERNAL_VOLUME = 'SNOWFLAKE_MANAGED':

ALTER DATABASE <my_salesforce_db> SET
  EXTERNAL_VOLUME = 'SNOWFLAKE_MANAGED'
  STORAGE_SERIALIZATION_POLICY = <COMPATIBLE|OPTIMIZED>;

With Snowflake storage, Snowflake stores and manages the Iceberg table files. You don’t need to configure external cloud storage or grant access to an external volume.

Option B: External volume in your cloud storage¶

To store the Iceberg table files in your cloud storage, configure the database with the external volume:

ALTER DATABASE <my_salesforce_db> SET
  EXTERNAL_VOLUME = '<external_volume>'
  STORAGE_SERIALIZATION_POLICY = <COMPATIBLE|OPTIMIZED>;

COMPATIBLE produces Parquet files readable by external engines. OPTIMIZED enables Snowflake-specific query optimizations. Choose based on your data query needs. For more information, see STORAGE_SERIALIZATION_POLICY.

Grant the connector role USAGE on the external volume:

GRANT USAGE ON EXTERNAL VOLUME <external_volume> TO ROLE <Salesforce_connector_role_name>;

This grant isn’t required when you use Snowflake storage (EXTERNAL_VOLUME = 'SNOWFLAKE_MANAGED').

Note

CATALOG = 'SNOWFLAKE' is set automatically by the connector when it creates Iceberg tables. Don’t set it at the database level.

Configure the connector¶

Set the following destination parameters before the connector creates destination tables:

Table Storage Format: Set to ICEBERG. The default is STANDARD.
Iceberg Version: Set to 2 or 3. The default is 3.

Don’t change Table Storage Format or Iceberg Version after ingestion begins. To switch between standard and Iceberg destinations, or between Iceberg v2 and v3, recreate the connector and perform a fresh load into new destination tables.

For the full connector configuration workflow, see Openflow Connector for Salesforce Bulk API: Configure the connector.

Verify Iceberg table creation¶

After the initial load completes, verify that the destination tables were created as Iceberg tables:

SELECT GET_DDL('TABLE', '<my_salesforce_db>.<my_salesforce_schema>.<object_table>');

The returned DDL should include ICEBERG TABLE.

You can also verify the database-level Iceberg defaults:

SHOW PARAMETERS LIKE 'EXTERNAL_VOLUME' IN DATABASE <my_salesforce_db>;
SHOW PARAMETERS LIKE 'STORAGE_SERIALIZATION_POLICY' IN DATABASE <my_salesforce_db>;

Limitations and behavior¶

Tri-Secret Secure accounts and Snowflake storage: Accounts with Tri-Secret Secure (TSS) enabled may be unable to create new Snowflake-managed Iceberg tables that use Snowflake storage for Apache Apache Iceberg™ tables. For details, see Encryption.
Collation isn’t supported on Iceberg tables: When Table Storage Format is set to ICEBERG, the connector doesn’t include DEFAULT_DDL_COLLATION in table creation parameters.
Salesforce time fields use microsecond precision: Salesforce fields of type time are created as TIME(6) on Iceberg tables.
Formula views can be created over Iceberg tables: When Enable Views Creation is set to true, the connector can create formula views over Iceberg base tables. The formula field limitations described in Openflow Connector for Salesforce Bulk API: Salesforce formula fields still apply.
Hard deletes aren’t supported: The connector still represents deleted Salesforce records with the isDeleted column rather than hard-deleting rows from the destination table.