I'm not sure why that would happen - the code that chooses what application
instance to use is here:
https://github.com/django/asgiref/blob/master/asgiref/server.py (under
get_or_create_application_instance).
If you're always using a single channel name, which gets passed as scope_id
(see here: ht
I am only using one channel name and my expectation is that only one
instance of the background worker class should be initiated/used. Instead,
I am seeing a new class being initiated for each call routed to the
background worker class. This is unexpected...
On Thursday, June 21, 2018 at 3:55:
The worker server tries to keep one application instance open per incoming
channel name, but that's no guarantee. It will always make a new instance
for different channel names. What kind of behaviour are you expecting?
Andrew
On Tue, Jun 19, 2018 at 6:45 AM itsnate_b wrote:
> I am using Django
I am using Django 2.0.5, Python 3.6.2, Django Channels 2.1.1, aiohttp 3.1.3.
I have created a background worker class which is initialized on a route to
the appropriate event/function. Problem is that it initializes a *new*
instance of the class on every triggered call, meaning that I cannot acc
4 matches
Mail list logo