Awaiting multiple times on a single task that failed with an exception results in an unbounded increase in memory usage. Enough repeated "await"s of the task can result in an OOM.

The same pattern on a task that didn't raise an exception behaves as expected.

The attached short script ends up using more than 1GB of memory in less than a minute.

This is a duplicate of bpo-45924.  The traceback accumulates another frame every time the exception is raised. To see that, change main in your script to 

async def main():
    task = asyncio.create_task(task_that_raise())
    while True:
        try:
            await task
        except Exception as e:
            print("<<<<<<<<<<<<<<<<<<<<<<<<<<<")
            traceback.print_exception(e)
            print(">>>>>>>>>>>>>>>>>>>>>>>>>>>")



and notice that in the output, X grows every time in the line like

    [Previous line repeated X more times]