I've created this repository to investigate how DB connections behave under the stack of SQLAlchemy 2, asyncio and Pubsub.

Given you have Python properly installed with poetry:

1. Create .env from env.sample

2. Start Pubsub

make pubsub_build
make pubsub_up
make pubsub_create_topic
make pubsub_create_subscription
make pubsub_up

3. Start poetry

poetry install
poetry shell

4. Start database

make db_up
make db_upgrade

task.py file connects to database, you have 4 options:

async_session as async_session,
async_session_null as async_session_null,
session_factory,
session_null_factory,

So you can use:

async for db_session in session_factory():

# OR

async for db_session in session_null_factory():

# OR

async with async_session() as db_session:

# OR

async with async_session_null() as db_session:

• Jaeger
• Grafana

In order to configure grafana you must add Jaeger data source (URL: http://jaeger:16686)

make pubsub_producer

make run_task

http://localhost:8080