Flat File Management Guidelines

 The standards and guidelines in this post apply to situations where you must move flat files into the BW environment and load them through the BW flat file interface. You may only load from flat files when you absolutely cannot use the BW Service API, the DB Connect interface, or the Universal Data Connect interface to extract data into BW.

File Locations on the BW System:

As part of your solution implementation, you need to ask the Basis team to create at least one file system directory for your application’s flat files. You should request a folder under the following path;

/usr/sap/interfaces/…..

NOTE: Refer your client's SAP BW Naming Convention document for the details to follow naming convention.

Managing Flat Files in BW Process Chains

The BI Technology group has developed four custom process types to help you manage flat files in BW. You can find these process types under “General Services” in the Process Chain Maintenance tool. Their icons and descriptions appear below:

Each of these process types can perform its process on either the local BW system or on a remote system accessible from the BW server via ftp. In addition, each process type allows branching to one of two subsequent processes based on success or failure.

Standards

1. Use the standard process types for managing flat files within BW.
2. When creating process chains that depend on the presence of a flat file, always include the standard process for checking the existence of the file before processes that use the file.
3. If the absence of a flat file does not represent a normal, expected condition, then stop the process chain and include the name of the file in the notification that you send to Production Support.
4. Do not create your own methods of managing flat files in SAP BW unless you have requirements these tools cannot meet. In that event, please consult with the BI Architecture group to determine whether client should develop a new, reusable component to meet your requirements.

Guidelines

You can use the Rename process type to help ensure that the sender of a file does not overwrite it while the process chain runs. Assign one name for the sender of the file to use, and then assign another name for internal processes that use the file. At the top of the process chain, check for existence of the external file name and then immediately rename it to the internal one.