There is an issue with the ETO data export routines that occurs when you select an export to be run for a date range that is in the past. This issue is related to the way the export routine in ETO was scripted as it is only including the records that are active within that period but will not include exit, service, or assessment data after the export end date. While this may make sense at first, this can cause an issue if the exports are run out of chronological order.
The reason for this is the import routines in the warehouse read the new import file and process records according to the EnrollmentID in the Entry.csv table. If an EnrollmentID is found, the warehouse removes all records associated with that EnrollmentID and replaces the data with the data from the recently imported batch. If a client had an exit record in the warehouse that indicated that he or she exited in 2018, and a batch was uploaded for January 1, 2017 to December 31, 2017, then that exit record would no longer exist in the warehouse. This in turn would make it appear as if the client is still actively enrolled when in fact he or she had exited.
To address this, users should export a full batch of data from 10/1/2012 to the present day. Going forward, if data needs to be refreshed for a prior year, just make sure the export end date is today’s date.