Identifying worker by ID || running code on single worker

[b3n4kh]

I want to write a custom APScheduler scheduler.

I want the scheduler to only have a single instance of the scheduler actually "started" to not execute tasks multiple times, if I have multiple workers.

My current approach would be to use a PostgreSQL advisory lock or something similar.

Is there a way to identify workers by some kind of unique id, from within python?
Any other suggestions as how I could achieve my goal easier / better with granian?

[b3n4kh]

I managed to get the IDs like this but I think it will not help me solve my underlying issue the way i thought it would and I am going in another direction and will start the apschedule "beside" granian during my granian startup script.

    g: MTServer | MPServer = Granian(
        "core", interface=Interfaces.WSGI, address=address, port=port, log_level=loglevel, workers=workers, respawn_failed_workers=True
    )
    spawn_target = g._spawn_wsgi_worker
    target_loader = partial(app_loader, g.target)
    g.startup(spawn_target, target_loader)
    worker_ids = [w._id() for w in g.wrks]
    print(f"Workers: {worker_ids}")
    g._serve_loop(spawn_target, target_loader)
    g.shutdown()

[gi0baro]

I am going in another direction and will start the apschedule "beside" granian during my granian startup script

This is the way.

[b3n4kh]

Just for documentation if someone finds this:

I spent a lot of time trying to get APScheduler to use the granian asynicio event loop with an AsyncIOScheduler and after taking a deep dive into the APScheduler codebase and trying I am pretty confident this doesnt work with APScheduler 3.

Therefore, I am abandoning this approach. I think the best chance to get this running the way I intended will be waiting for APScheduler 4