The connection mode that you specify in the Google BigQuery source or target connection during runtime must match with the connection mode that you specify when you import a Google BigQuery source or target definition.

You cannot configure key range partitioning to read data from a Google BigQuery source.

You cannot read hierarchical data from a Google BigQuery source, convert the hierarchical input to relational output, and write data to relational targets.

You cannot read data from relational sources, convert the relational input to hierarchical output, and write data to a Google BigQuery target.

You cannot create a target at run time.

You must define a primary key in the target table for update and delete operations.

To perform an upsert operation on a column, you must ensure that the column in the target table is set to Required in Google BigQuery.

When you perform an upsert operation on a Google BigQuery target table, you cannot perform an update operation on multiple rows.

You cannot perform more than one operation with a pre SQL or post SQL command.

When you perform an upsert operation on a Google BigQuery target table, you must not select streaming mode to write data to a Google BigQuery target.

When you configure a Google BigQuery connection in simple mode, you can perform an upsert operation on columns of primitive data types only.

If you perform an UPDATE or DELETE operation with a pre SQL or post SQL command, you must specify the following parameter in the pre SQL configuration or post SQL configuration:

When you use staging mode to read data from Google BigQuery or bulk mode to write data to Google BigQuery, you must increase the maximum heap size in the

Java SDK Maximum Memory

property to 2048 MB or 4096 MB based on the amount data you want to process.

When you use the

pmrep

command to create a Google BigQuery connection, you must specify one of the following values for the

Connection mode

option:

Simple
Hybrid
Complex