The initial data transfer is used to transfer master data and transaction data that is relevant to SAP APO from an ERP system to SAP APO.
You must have made the basic settings for setting up the system infrastructure and for the data transfer in both the ERP system and in SAP APO.
● For SAP APO, see CIF Customizing in SAP APO.
For SAP R/3 3.1I,
for Setting Up the System Infrastructure.
As of SAP R/3 4.0B, an Implementation Guide is available. For more information, see CIF Customizing in the ERP System. In the IMG, choose Basic Settings for Creating the System Landscape and Basic Settings for the Data Transfer ® Initial Transfer.
In the integration model, the master data and transaction data is selected that is required for the SAP APO-relevant planning processes. These tasks include:
1. Defining the target system
2. Selecting the filter objects to be transferred, that is, compiling a data pool for the initial data supply
3. Triggering the initial data upload
As standard (that is, if you have not set parallelized initial data transfer (as of SAP R/3 Plug-In 2002.1)), the initial data transfer to SAP APO always runs using the same queue, irrespective of the object type transferred. In other words, it runs in serial. If you use serialization, you cannot trigger a second initial data transfer to the same SAP APO system, even if none of the selected object types overlap with the active integration model. During an initial data supply run, all transaction data postings are synchronized with the initial supply that is currently running. That means that the method and time of the transaction data transfer are dependent upon the processing status of the initial data transfer. This ensures that the system always transfers the most recent transaction data to SAP APO and helps avoid data inconsistencies. However, this can reduce the performance of the transaction data postings. For this reason, we recommend that you carry out the initial data transfer during a time when the number of postings is at a minimum (for example, at night).
You can improve system performance by parallelizing the initial data transfer.
It might also be useful to change to inbound queues if there is a large system load in the receiving system. For more information, see Queue Type Selection for the Initial Data Transfer.
● Create an integration model. See Generating Integration Models.
● Execute a consistency check for the integration model, if necessary. See Consistency Check
● Activate the integration model. See Activation and Deactivation of Integration Models