`test_steal_twice` flaky

[gjoseph92]

On Windows, 3.8: https://github.com/dask/distributed/pull/4925/checks?check_run_id=2889177384#step:10:2191

The traceback is just an asyncio.TimeoutError

______________________________ test_steal_twice _______________________________

    def test_func():
        result = None
        workers = []
        with clean(timeout=active_rpc_timeout, **clean_kwargs) as loop:
    
            async def coro():
                with dask.config.set(config):
                    s = False
                    for _ in range(60):
                        try:
                            s, ws = await start_cluster(
                                nthreads,
                                scheduler,
                                loop,
                                security=security,
                                Worker=Worker,
                                scheduler_kwargs=scheduler_kwargs,
                                worker_kwargs=worker_kwargs,
                            )
                        except Exception as e:
                            logger.error(
                                "Failed to start gen_cluster: "
                                f"{e.__class__.__name__}: {e}; retrying",
                                exc_info=True,
                            )
                            await asyncio.sleep(1)
                        else:
                            workers[:] = ws
                            args = [s] + workers
                            break
                    if s is False:
                        raise Exception("Could not start cluster")
                    if client:
                        c = await Client(
                            s.address,
                            loop=loop,
                            security=security,
                            asynchronous=True,
                            **client_kwargs,
                        )
                        args = [c] + args
                    try:
                        future = func(*args)
                        if timeout:
                            future = asyncio.wait_for(future, timeout)
                        result = await future
                        if s.validate:
                            s.validate_state()
                    finally:
                        if client and c.status not in ("closing", "closed"):
                            await c._close(fast=s.status == Status.closed)
                        await end_cluster(s, workers)
                        await asyncio.wait_for(cleanup_global_workers(), 1)
    
                    try:
                        c = await default_client()
                    except ValueError:
                        pass
                    else:
                        await c._close(fast=True)
    
                    def get_unclosed():
                        return [c for c in Comm._instances if not c.closed()] + [
                            c
                            for c in _global_clients.values()
                            if c.status != "closed"
                        ]
    
                    try:
                        start = time()
                        while time() < start + 60:
                            gc.collect()
                            if not get_unclosed():
                                break
                            await asyncio.sleep(0.05)
                        else:
                            if allow_unclosed:
                                print(f"Unclosed Comms: {get_unclosed()}")
                            else:
                                raise RuntimeError("Unclosed Comms", get_unclosed())
                    finally:
                        Comm._instances.clear()
                        _global_clients.clear()
    
                    return result
    
>           result = loop.run_sync(
                coro, timeout=timeout * 2 if timeout else timeout
            )

distributed\utils_test.py:966: 
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
C:\Miniconda3\envs\dask-distributed\lib\site-packages\tornado\ioloop.py:530: in run_sync
    return future_cell[0].result()
distributed\utils_test.py:925: in coro
    result = await future
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

fut = <Task cancelled name='Task-157632' coro=<test_steal_twice() done, defined at D:\a\distributed\distributed\distributed\tests\test_steal.py:657>>
timeout = 30

    async def wait_for(fut, timeout, *, loop=None):
        """Wait for the single Future or coroutine to complete, with timeout.
    
        Coroutine will be wrapped in Task.
    
        Returns result of the Future or coroutine.  When a timeout occurs,
        it cancels the task and raises TimeoutError.  To avoid the task
        cancellation, wrap it in shield().
    
        If the wait is cancelled, the task is also cancelled.
    
        This function is a coroutine.
        """
        if loop is None:
            loop = events.get_running_loop()
        else:
            warnings.warn("The loop argument is deprecated since Python 3.8, "
                          "and scheduled for removal in Python 3.10.",
                          DeprecationWarning, stacklevel=2)
    
        if timeout is None:
            return await fut
    
        if timeout <= 0:
            fut = ensure_future(fut, loop=loop)
    
            if fut.done():
                return fut.result()
    
            await _cancel_and_wait(fut, loop=loop)
            try:
                fut.result()
            except exceptions.CancelledError as exc:
                raise exceptions.TimeoutError() from exc
            else:
                raise exceptions.TimeoutError()
    
        waiter = loop.create_future()
        timeout_handle = loop.call_later(timeout, _release_waiter, waiter)
        cb = functools.partial(_release_waiter, waiter)
    
        fut = ensure_future(fut, loop=loop)
        fut.add_done_callback(cb)
    
        try:
            # wait until the future completes or the timeout
            try:
                await waiter
            except exceptions.CancelledError:
                if fut.done():
                    return fut.result()
                else:
                    fut.remove_done_callback(cb)
                    # We must ensure that the task is not running
                    # after wait_for() returns.
                    # See https://bugs.python.org/issue32751
                    await _cancel_and_wait(fut, loop=loop)
                    raise
    
            if fut.done():
                return fut.result()
            else:
                fut.remove_done_callback(cb)
                # We must ensure that the task is not running
                # after wait_for() returns.
                # See https://bugs.python.org/issue32751
                await _cancel_and_wait(fut, loop=loop)
>               raise exceptions.TimeoutError()
E               asyncio.exceptions.TimeoutError

C:\Miniconda3\envs\dask-distributed\lib\asyncio\tasks.py:501: TimeoutError

[crusaderky]

I ran this test many times per environment between 6 and 17 aug over 3 mass stress test runs.

It failed 38 times out of 66 runs on Windows 3.7. It failed only once out of 512 runs on all other environments combined.

The error I'm getting is, invariably,

        assert a.in_flight_tasks == 0
>       assert b.in_flight_tasks == 0
E       assert 1 == 0
E         +1
E         -0

[hendrikmakait]

test_steal_twice has not appeared as a flaking test on the test report for the last 50 workflows, so this seems to be fixed.

[jacobtomlinson]

This test has been marked as problematic in #8437 so I'm going to reopen this.