Hey @amirreza8002 — Thanks for posting.
I guess, at this point, what I’d like to see is an async (first) backend in a third-party package. That would allow pulling it in, configuring, using, … and then we can work out the remaining integration difficulties. I need to check out the django-valkey implementation. (Also, an overview of the existing options here is probably useful. Having not been paying attention, I can’t say off-hand what the state of play is, so likely others can’t other.)
Ref whether we need a separate sync backend (or just an async_to_sync adaptor, or even code generation for a sync version): I think we just have to wait and see. We already have the sync backends. It’s a long road to remove/update them. I don’t think we need to legislate in advance.
Rather, having async backends available would mean we could start adopting in earnest. I think usage will drive the direction.