About Document Merging
Although the Process Engine allows you to create a process model that will merge documents in the pipeline, this configuration can yield unpredictable behavior when you attempt to do this with different instances of an IS document with the same name. For example:
You pass different instances of an IS document with the same name to two branches within a process and later merge them with an AND join.
You implement transition loop in the process, where the process flow loops back to a join step where two or more documents of the same name are merged.
Important: The Process Engine does not support parallel data management and you are strongly advised to avoid allowing the same document in a process to be accessed concurrently on different branches within the process.
Tip: Best practices recommend that you manage the pipeline by implementing services within a process loop to ensure that old document references are removed from the pipeline when a looping condition occurs, as well as when they are no longer needed.