The Union transformation is designed to merge multiple heterogeneous or homogeneous data streams into a single target stream. For the integration service to process the data correctly, the structure of the input data must be consistent. Therefore, all input groups must have the same number of ports, and the corresponding ports across these groups must have identical metadata attributes: datatype, precision, and scale. This ensures that the data from different sources can be combined into a single, structurally coherent output group without data type conflicts or loss of precision.

B. Mismatched ports or differing datatypes, precisions, or scales between input groups would result in a mapping validation error, as the transformation cannot create a unified output stream.

C. The Union transformation is analogous to the SQL UNION ALL operator. It combines all incoming rows from all input groups and does not identify or remove duplicate records.

D. All port names within a single transformation instance must be unique. Therefore, a port name used in one input group cannot be reused in another input group of the same Union transformation.

1. Informatica PowerCenter 10.5 - Transformation Guide

Chapter 25: Union Transformation

"Rules and Guidelines for Union Transformations"

Page 465.

The guide states: "All input groups and the output group must have matching ports. The precision

data type

and scale for the ports must be identical across all groups." This directly supports answer A and refutes B.

It also states: "The Union transformation does not remove duplicate rows." This refutes answer C.

Regarding port names

the general transformation rule applies: "Port names must be unique within the transformation." This refutes answer D.

2. University of Washington

Data Warehousing Course Material (CSE 590DB)

Lecture on ETL Transformations.

Course notes often describe the Union transformation's function as a UNION ALL operation

requiring schematic consistency across inputs. The material specifies that for a successful merge

the "number

order

and data types of the columns must be the same" for all input data sets

which aligns with the requirements stated in option A.