Django store aggregated data in the database?

Question

I have a use case where I have some complex functions (involving aggregations) that take a few seconds to run, and are affecting the UX of the app (even when the rows are just about 25-30k and the relevant fields are indexed). I am thinking of storing the aggregations in the database itself (and run them nightly) since real-time-liness of the data is not very important here. Is that a common practice with Django?

(I couldnt find much discussion on that on SO though)

Answer 1

Do you have several options:

• @ittus's and @Andrey Shipilov's approach: store result in cache system like Memcache or redis. You can check github.com/niwinz/django-redis

• @Anonymous's approach: use some sort of materialized view (depends on your db), and then create a unmanaged model in django to query it.

• Create new models/fields for aggregate data and figure up it nightly though a django custom comand or a celery task.

Conclusions:

• Your question should be close because is a primarily opinion-based question, like this one Django - what is best practice - Calculating field values.

Notice:

• If you need real time calculation then you can overwrite save method on models to keep data or just connect a post_save signal.