Register source in SQL replication

First of all, you should know what you want to replicate to the target from the source database. So you should tell SQL replication about them by registering them. You can register sources that are DB2 tables and views, or tables on non-DB2 relational databases.

Note: When you register source tables on non-DB2 relational databases, you use SQL replication together with DB2 Relational Connect. You map your source database to a federated database and you create nicknames for each source table.

The replication control tables are playing as the core of replication, where the components of replication communicate by creating, reading, and updating the control tables' data. The control tables are viewable and manually able to be updated. The design divides the control tables into three sets based on the functionality they are related to: capture, apply, and monitoring. Each set of control tables is independent. The sets can be stored in separate databases, in separate instances, and on separate servers.

Replication sources are selected from the objects of the capture control server and registered to the capture control server, which implies that your capture control server must be the one where your source data for replication resides. Remote journaled tables on DB2 UDB for iSeries is the only exception to this rule. The capture control tables should already be created for that server, and the capture control server should already be added to the replication center.

xxxx