Insert Into

The INSERT INTO statement is used to send data from the stream into a sink. Each stream requires at least one insert into statement to mark which data should be output from the stream.

INSERT INTO table_name
select_statement

Example

INSERT INTO outputtable
SELECT column1, column2 FROM inputtable

Primary keys

A sink groups the rows it writes by a primary key. Most connectors find the primary key on their own, for instance by looking at the destination table, but the PRIMARY KEY clause lets the statement decide which columns to use instead.

INSERT INTO table_name PRIMARY KEY (column1, column2)
select_statement

This is useful when the destination has no primary key, when its primary key is not written by the stream, or when a stream writes to several tables that each need different keys.

Example:

INSERT INTO outputtable PRIMARY KEY (column1)
SELECT column1, column2 FROM inputtable

Every declared column must be written by the statement, a column that is not part of the select is an error. The clause is also placed after the column list when one is given:

INSERT INTO outputtable (column1, column2) PRIMARY KEY (column1)
SELECT a, b FROM inputtable

Note

Not all connectors can use declared primary keys. Building the stream fails if the statement declares primary keys for a sink that does not support them, rather than writing the data with other keys than the declared ones.