Unexpected Memory Leak When Using Python 3.10 with Asyncio and Context Managers
I'm integrating two systems and I've been banging my head against this for hours... I'm encountering a memory leak in my Python application that utilizes asyncio and context managers, specifically with Python 3.10. The application manages multiple async tasks that perform I/O operations, and I'm using context managers to handle resources. Despite my efforts to ensure proper resource cleanup, I'm noticing an increase in memory usage that does not decrease, even after tasks are completed. Here's a simplified version of the code: ```python import asyncio import contextlib class Resource: def __init__(self): self.data = [i for i in range(100000)] # Simulated large resource def release(self): del self.data async def async_task(): async with contextlib.AsyncExitStack() as stack: resource = Resource() await stack.enter_async_context(resource) # Simulate some async I/O operation await asyncio.sleep(1) # Resource should be released automatically resource.release() async def main(): tasks = [async_task() for _ in range(10)] await asyncio.gather(*tasks) if __name__ == '__main__': asyncio.run(main()) ``` I expected the memory usage to stabilize after the tasks complete, but the memory footprint keeps growing as I run the application for an extended period. I've tried using `gc.collect()` after my tasks, but it doesn't seem to help. Additionally, I've verified that all references to the Resource instances are indeed removed, and I’ve checked for circular references using `gc` module. Could there be an issue with how context managers interact with asyncio? Is there anything I might be overlooking in managing the lifecycle of resources? Any insights or suggestions would be greatly appreciated! Any help would be greatly appreciated! Any ideas what could be causing this?