This sounds to me like a perfect use case to be solved queueing jobs and processing them in the background.
Advantages:
- since retrieving and zipping the files requires a variable (and possibly significant) time, that should be decoupled from the HTTP request/response cycle;
- multiple jobs will be serialized for execution in the task queue.
The second advantage is particularly desirable since you’re prepared to receive multiple concurrent requests.
I would also consider using a “task” Django model with a FileField to be used as a container for the resulting zip file, so it will be statically and efficiently served by Nginx from the media folder. As an additional benefit, you will monitor what’s going on directly from he Django admin user interface.
I’ve used a similar approach in many Django project, and that has proven to be quite robust and manageable; you might want to take a quick look at the following django app I’m using for that: https://github.com/morlandi/django-task
To summarize:
- write a “task” Model with a FileField to be used as a container for the zipped result
- upon receiving a request, insert a new record in the “task” table, and a new job in the background queue
- the background job is responsible for collecting resources and zipping them; this is common Python stuff
- on completion, save the result in the FileField and send a notification to the user
- the user will follow the received url to download the zip file as a static file