At the moment, it uses existing functions to find submissions already created for a student when picking an assignment in a Variants Collection. For n variants, this results in n checks (=n sqls) to find out whether there exists a submission for any of the variants for a student. This does not scale for larger n.

While this only happens once per student for a variants collection (the picked variant gets cached), it may become a problem when 500 students concurrently view a variant collection with larger n for the first time.

Submitted by Elmer van Chastelet on 1 October 2020 at 10:19