In some use cases it is important to be able to identify an entry after adding it to the JRJOBIMPORT table, e.g. to read the instance number or other data JobRouter writes back. The task_id column is used for this. You can enter any unique key you want to use to identify this entry again into this column.