Execute row SQL script

Description

The Execute Row SQL Script transform executes an SQL script for every input row the transform receives.

The SQL to execute can be passed as pipeline fields or read from a file.

Options

Option	Description
Transform name	Name of the transform; This name has to be unique in a single pipeline
Connection	Select a database connection to use
Commit	Number of rows to send before doing a database commit
SQL field name	The field that either contains the SQL to execute or optionally specifies the path to a file that contains SQL to execute
Read SQL from file	If checked, then the SQL field name option specifies a file that contains SQL to execute, otherwise the SQL field name option specifies the actual SQL to execute. (Available since 5.0)
Field to contain insert stats	Optional: If you want to get an additional field in our stream with the number or records that where inserted, please define the field name here.
Field to contain update stats	Same as insert stats, but for updated rows.
Field to contain delete stats	Same as insert stats, but for deleted rows.
Field to contain read stats	Same as insert stats, but for read rows.

Option

Description

Transform name

Name of the transform; This name has to be unique in a single pipeline

Connection

Select a database connection to use

Commit

Number of rows to send before doing a database commit

SQL field name

The field that either contains the SQL to execute or optionally specifies the path to a file that contains SQL to execute

Read SQL from file

If checked, then the SQL field name option specifies a file that contains SQL to execute, otherwise the SQL field name option specifies the actual SQL to execute. (Available since 5.0)

Field to contain insert stats

Optional: If you want to get an additional field in our stream with the number or records that where inserted, please define the field name here.

Field to contain update stats

Same as insert stats, but for updated rows.

Field to contain delete stats

Same as insert stats, but for deleted rows.

Field to contain read stats

Same as insert stats, but for read rows.

The counts come from the number of affected rows the JDBC driver reports for each statement, attributed by the statement’s verb (also behind a WITH clause). MERGE, UPSERT and REPLACE statements report a single count for the rows they inserted, updated and deleted together; that count is reported as updated rows. Rows returned by SELECT statements count as read rows, except when the SQL is sent as a single statement.

The same counts show up in the transform’s own metrics: read rows as input (I), inserted rows as output (O), updated and deleted rows as updated (U).

Notes

Because of the scripting/dynamic nature of the transform, it does not use prepared statements and thus is not intended to work fast or optimal. For good performance turn to dedicated transforms like Table Output (insert into), Table Input (Select), Update, Delete, etc.