We are trying to load completion records for the LinkedIn Learning (LIL) courses that our users have taken (given that there is not way to en masse create > 7,000 LIL courses (using AICC packages) so our users could access them directly from our LMS). We have downloaded the .CSV file LIL and created the .CSV files necessary for uploading completions. We understand that all these Completion records will be created as Other Evidence since the actual courses don't exist in our LMS.
However, as we were testing the upload in our DEV environment, we needed to change the data in the .CSV file and rerun the uploads about 4 times until we got it right. When we went in to a users record to see what the completion looked like, however, we noticed that there were 4 records instead of one. Each had come from one of the uploads that we did. We had only changed completion dates (some were empty, some had text instead of a date) but had not changed the course shortnames, the course id's, or the user id's which have the information needed to overwrite a completion instead of adding new one. We did have the "Override current course completions" setting checked but that only appears to work for Courses and not Evidence.
While this is okay for our DEV environment, I am concerned that once we do this completion upload into Production and make it part of our regular maintenance, we may end up with duplicates if the file we request from LIL overlaps a previous one in any way and/or if the file we upload into production happens to have any errors so we need to rerun the completion upload.
What do we need to do in order to not create these duplicates. Do we need to create course shells for all these LIL courses so these completions upload as into those shells? If we do that and the broader integration between LIL and Totara happens in the future, how will these course shells be impacted.
Thank you.
Carol.