Feature: simplified closing of old db connections about django-apscheduler HOT 4 OPEN

Would this achieve the same result and perhaps be easier to understand (untested)?

import functools
from apscheduler.schedulers.blocking import BlockingScheduler
from django import db

class BlockingDBScheduler(BlockingScheduler):
    @util.close_old_connections
    def add_job(self, func, *args, **kwargs):
        # Edited to swap `wrapper` with `func` as per the correction pointed out by @HeyHugo in the comments below.
        return super().add_job(func, *args, **kwargs)

from django-apscheduler.

@jcass77 Aha nice yes that works (just swap wrapper arg to func)

from django-apscheduler.

Ok. I think this approach could be handy for folks who only intend to schedule jobs that always read or mutate the database.

The drawback is that if you use something like BlockingDBScheduler for anything else then those extra calls to db.close_old_connections() are essentially superfluous and just unnecessary overhead. That feels like an unreasonable tradeoff for the sake of convenience and syntactical sugar, and maybe not something that we want as the default behaviour for new users.

I'm happy to leave this here for others to find and use in their projects if their use case is narrowly defined along the lines discussed above though.

Thanks for pointing this out and providing a coding example!

from django-apscheduler.

Ok that's fair and I agree.

I have not yet had a case with a scheduled task without db interaction, but I understand there could be such use cases where state lives elsewhere.

from django-apscheduler.