Metadata Extraction Rules for Sources and Destinations
Metadata Extraction Rules for Sources and Destinations
Metadata Manager extracts metadata from sources and destinations in the data flow according to different metadata extraction rules.
Metadata Manager applies different metadata extraction rules for the following types of sources and destinations in a data flow:
Destination
Metadata Manager applies the following metadata extraction rules for flat file and OLE DB destinations:
Metadata Manager derives data type and length for target columns from the corresponding source column, not from the corresponding column in the destination data store.
For example, a data flow writes data from a source column with data type CHAR[50] to a column in a database with data type CHAR[100]. When you view the target column in Metadata Manager, Metadata Manager displays the data type for the target column as CHAR[50].
When multiple source columns load data to the same destination column and the source column lengths differ, Metadata Manager determines the target column length from either source column. Metadata Manager selects the source column that it encounters first in the metadata source.
For example, a data flow contains a Union All transformation that merges data in the NAME column from two source columns with data types CHAR[255] and CHAR[100]. Metadata Manager displays the NAME column length in the target object as either 255 or 100.
OLE DB Destination
If you use an SQL command to load data, Metadata Manager displays the SQL command in the
SQL
property in the package summary.
Destination for error output
Metadata Manager displays the destination for error output as a target object.
Source
If there is a mismatch in length between an external column in a flat file or OLE DB source and the corresponding output column, Metadata Manager derives the column length from the shorter of the two columns.
OLE DB Source
If you use an SQL command to extract data, Metadata Manager displays the SQL command in the