Modify the SDE and SIL mappings of all the related fact tables (fact tables that need to be linked to this dimension).
The custom fact SDE mappings must populate the foreign key column of the changed dimension (using a custom map table process to convert from Siebel's row IDs to the external data source row IDs). The custom SIL mapping should be modified to use the appropriate DATASRC_NUM_ID, because the standard SIL mappings assume DATASRC_NUM_ID for the dimensions are the same as the fact table's DATASRC_NUM_ID.
It is very important to decide when the data is going to be loaded. If it is going to be loaded along with the Siebel source data, you must be careful with how you handle failure recovery. The preconfigured workflows truncate the target staging table prior to loading. Upon failure, when the DAC server restarts the task, all the data is truncated and all the data is loaded again.
If the data from the external source gets loaded into the same staging table, be careful with how you handle this situation, since you cannot use the truncate table functionality. The data migrating into the staging table is not incrementally loaded, and, therefore, should be cleaned up prior to attempting to load this table again.
In such a case, it is recommended that you encapsulate the extract part from both the sources inside an Informatica workflow. Upon failure of either of the extracts, the whole workflow gets rerun. Note that the data from both the sources should be run at the same time, all the time.
If it is decided that the data is going to be loaded at different time frequencies, then the new SDE workflows need not depend on the preconfigured SDE workflows and can use the Truncate Table option for failure recovery. In this case, in the DAC Design view, define a new execution plan in the Execution Plans tab, and define the new data source in the Database Connections subtab. Make sure the shared SIL process depends on the SDE processes from both sources.
Figure 16 provides a graphical representation of this process.