Performance issues when requesting a vast amount of occurrences
Closed this issue · 4 comments
via the "table" API. This was never noticed because of our pagination, but it's been used by @sronveaux that consumes our API for other purposes.
I removed tons of duplicated queries already.
In case the user is authenticated, it is not enough because https://github.com/riparias/early-warning-webapp/blob/58ca7750bb6a796bf6a4a224ef49857f952f2edd/dashboard/models.py#L224-L228 is not using the result of prefect_related().
To be improved, but not problematic anymore since @sronveaux access the data as non-authenticated user.
This is now deployed on dev, hopefully for production release next week.
Hi @niconoe, I just tested and can confirm you the same request has a gain of 400% in dev compared to prod, nice job!
Thanks for that!
I tried to fix the remaining (authenticated) use-case, but since it's not that easy and there's no concrete use case for it now, I'm gonna close this issue (YAGNI).