Task was destroyed but is is pending

Hi, I’m receiving multiple messages stating the following:

Task was destroyed but it is pending!
task: <Task pending coro=<RedisPool._do_close() running at /Users/manuelmiranda/.virtualenvs/redis-cache/lib/python3.5/site-packages/aioredis/pool.py:102> wait_for=<Future pending cb=[Task._wakeup()]>>
Task was destroyed but it is pending!
task: <Task pending coro=<RedisConnection._read_data() running at /Users/manuelmiranda/.virtualenvs/redis-cache/lib/python3.5/site-packages/aioredis/connection.py:131> wait_for=<Future pending cb=[Task._wakeup()]> cb=[Future.set_result()]>
Task was destroyed but it is pending!

The code I’m using:

    async def _connect(self):
        if self._pool is None:
            self._pool = await aioredis.create_pool(
                (self.endpoint, self.port))

        return await self._pool

    async def get(self, key):
        with await self._connect() as client:
                await client.get(key)

Am I missing something? Isn’t the context_manager supposed to close the pool when exits?

Thanks!

Issue Analytics

State:
Created 7 years ago
Comments:14 (10 by maintainers)

Top GitHub Comments

3reactions

jp3049commented, Jan 21, 2021

@popravich Can you help me, maybe I’m doing something wrong in my Pool, I tried a lot of variants of this code - the result is the same

Task was destroyed but it is pending! task: <Task pending name='Task-57663' coro=<RedisConnection._read_data() done, defined at /layers/google.python.pip/pip/lib/python3.8/site-packages/aioredis/connection.py:180> wait_for=<Future pending cb=[<TaskWakeupMethWrapper object at 0x3ea2c3b858b0>()]> cb=[RedisConnection.__init__.<locals>.<lambda>() at /layers/google.python.pip/pip/lib/python3.8/site-packages/aioredis/connection.py:168]>

just solved this by downgrade python to 3.6

2reactions

argaencommented, Oct 5, 2016

test_redis.py: should not redis_cache fixture be a yield_fixture? maybe that is a problem… You can check aioredis test fixtures.

Nope, in pytest 3.0 if a fixture has a yield it implicitly becomes an “old” yield_fixture. It works out of the box now.

D’oh! I thought close() and wait_closed() were just the sync and async ways to close the redis connection respectively, my bad. After transforming the fixture to:

@pytest.fixture
def redis_cache(event_loop, mocker):
    cache = RedisCache(namespace="test", loop=event_loop)
    yield cache
    event_loop.run_until_complete(cache.delete(KEY))
    event_loop.run_until_complete(cache.delete("random"))
    cache._pool.close()
    event_loop.run_until_complete(cache._pool.wait_closed())

works perfectly. Thanks a lot for the help!