Description
This should be considered part 2 of #593.
A quick and least disruptive solution was implemented to get production out of the water (#617).
This new issue is so we can implement a better fix.
This better fix is required so we can switch back to the regular official Nginx docker image, instead of our custom build with cron and logrotate.
Concerned Organizations
@fmigneault @mishaschwartz