IBM Db2 for z/OS Models - Server Options

On the Server tab, you specify the Db2® for z/OS® system where the model is to be built.

Use upstream connection. (default) Uses the connection details specified in an upstream node, for example the Database source node. Note: This option works only if all upstream nodes are able to use SQL pushback. In this case, there is no need to move the data out of the database, as the SQL fully implements all of the upstream nodes.
Move data to connection. Moves the data to the database you specify here. Doing so allows modeling to work if the data is in another IBM® database, or a database from another vendor, or even if the data is in a flat file. In addition, data is moved back to the database specified here if the data has been extracted because a node did not perform SQL pushback. Click the Edit button to browse for and select a connection.

Note: The ODBC data source name is effectively embedded in each SPSS® Modeler stream. If a stream that is created on one host is executed on a different host, the name of the data source must be the same on each host. Alternatively, a different data source can be selected on the Server tab in each source or modeling node.