Whether a failed processing operation can be restarted depends on the integration used and the way the process is set up.

In the event of temporary problems, such as a network fault, running the process again may be a suitable solution.

However, it must first be checked whether the original processing actually failed.

It may well be the case that a receiving system has already processed the data, even though the confirmation has not yet been received.

By using unique identifiers and checking against existing records, it is possible to prevent data from being duplicated when a new attempt is made.