Default mappings save you from having to map each column individually on the Table Properties, Mapping page and therefore can be very useful if you have many tables to distribute
Default column mappings are created implicitly, that is, without your explicit selection of the columns, after any of the following events:
-
You use the Distribute Tables dialog to select one or more of your available tables for distribution, and the Connect CDC Director uses a Table-mapping algorithm to find a table with the same name (or the same name except for capitalization) on the target server as the table you selected on the source server.
-
You select the Defaults button on the Mapping page of the Sending Table Properties or Receiving Table Properties dialog.
-
You specify in RSL a CREATE SHARED TABLE statement that includes the ROUTE keyword and a specified routing-set-name.
In the default mapping:
Each receiving table column is mapped to the sending table column that has the same name (or the same name except for capitalization).
System variables are mapped to receiving table columns that have the names designated in the initialization file installed with the Connect CDC Director. Using system variables can help track what the inserts in the target table represent, whether they are inserts, updates, or deletes. Refer to the section, “Using System Variables”, in Chapter 8 of the Advanced User Guide for details.
The Connect CDC Director chooses the columns that become Distribution Keys.
If a primary key is defined for a table, it is used as the Distribution Key. If the table has no primary key but at least one unique index, the first unique index defined for the table is used. If the table is a receiving table, and no primary key or unique index is defined, the source table distribution keys are used if possible.
The mappings are displayed on the Mapping page of the Sending Table Properties or Receiving Table Properties dialog.
You must make sure, prior to invoking the mapping, that your receiving tables contain the appropriately named columns.
No further table mapping is necessary if the default column mapping includes all the columns you intend to distribute and if the column datatypes are compatible and Distribution Key rules are satisfied.