In most cases, once each interval has been successfully processed there exist data results for that interval. The data from all intervals will need to be merged back into a single result. This is achieved by the use of a Collation process. The Collation refers to another custom process that is responsible for taking the data of each interval and combining them all into a single result (as if the program had been run sequentially without the concept of intervals). This single result is saved against the instance.
It is possible that a program does not require any collation, for example, a program that creates outbound IDocs.
Post your comment on this topic.