Setting the feed name

Each collection task requires a feed name to uniquely identify input data from a particular source. The feed name corresponds to a directory in a process folder where all collected data is retained for audit and recovery purposes. The assembly step pulls in collected data for all feeds of a process for a given select date. It is common practice when working out processing issues to disable the collector itself and re-run a job starting with the assembly step.

The feedName property is set on the collection tasklet itself along with the collector bean. Each collection for a process should have a unique feed name, as in the following syntax:

<property name="feedName" value="<nameOfFeed>" />

The feed name is typically an instance name for the external system from which the data originated. The collection tasklet creates a file of the collected records under a feed subdirectory of the appropriate process directory. It is important that it follow the naming rules defined for the host operating system (for example, no special characters).