Add 12 hour cache to stats endpoint

[jjnesbitt]

Closes #1412

Notably, this doesn't actually change how the stats are computed, it just adds a cache. Using the asset paths models to compute this results in a slightly different value than we have now. Rather than try to determine what course of action we should take (a behavior change), I just wanted to get the difference making optimization in.

[waxlamp]

@AlmightyYakob, is this ready for review or are there still pending tasks before it can be moved out of draft mode?

[jjnesbitt]

@waxlamp It's now good to go

[danlamanna]

Is there already a configured cache backend?

[jjnesbitt]

Is there already a configured cache backend?

The default is local memory caching. Do you think we should use a DB cache instead?

[jjnesbitt]

Note: Since we're using a db cache, we'll need to run python manage.py createcachetable manually once, after this is deployed.

[dandibot]

🚀 PR was released in v0.3.13 🚀

[jjnesbitt]

Note: Since we're using a db cache, we'll need to run python manage.py createcachetable manually once, after this is deployed.

This has been successfully performed.

[waxlamp]

Note: Since we're using a db cache, we'll need to run python manage.py createcachetable manually once, after this is deployed.

This has been successfully performed.

@AlmightyYakob, can you update the documentation to explain how and when to carry out this step?