Serhii Charykov

OK, maybe there is more important problem here: https://github.com/redis/redis-om-python/blob/00ed537fcaf43d93ea26e0332d5cb2f1a4c1c4a1/aredis_om/checks.py#L13-L14 I think lru_cache caches coroutine creation, doesn't it? So it is always `not ` here https://github.com/redis/redis-om-python/blob/00ed537fcaf43d93ea26e0332d5cb2f1a4c1c4a1/aredis_om/model/model.py#L1468-L1474 and more over, this: ```python...

And what if user uses async API: Model would get async connection, then the check _that should be sync(because in `__init__`)_ gets async connection... Either we create async and sync...

Any input will be appreciated

Hi @chayim, I've already mentioned this in #234 and #239. Could you look at them and answer some questions there?

And related to `aioredis`, could you look at #233?

Any input will be appreciated

I've got similar behavior in GKE Standard Cluster and managed to trace the source of my "Reconfiguring" every 30 seconds: (not 100% sure) This is the place where admission server...

I think I found problem. I have encountered similar issue in the past. It's related to the way how aiohttp handle stopping(and asyncio.run() because aiohttp.run_app is analogous). [aiohttp web.run_app](https://github.com/aio-libs/aiohttp/blob/8a9dbf9e82dfe49c62c18176af89b9fe0f2fc501/aiohttp/web.py#L434-L440): ```python...

I'm not very familiar with either aio_pika or aiormq inner workings but after taking a peek in [aiormq.connection.Connection._\_\_reader()_](https://github.com/mosquito/aiormq/blob/63a8b0d6d50eb93f9ee8c851fcfda01e3d5f122b/aiormq/connection.py#L374) maybe something like this could work: ```python async def __reader(all_rpc_futures_are: asyncio.Future): cancellation_in_progress...

@multun Yes, I am quite happy that aiohttp now waits for cleanup tasks before cancelling all tasks. But some slight changes to the example above: ```python import asyncio import logging...