dask.distributed SLURM кластер Няня Тайм-аут - PullRequest
Новая
       94

dask.distributed SLURM кластер Няня Тайм-аут

1 голос
/ 05 марта 2020

Я пытаюсь использовать dask.distributed.SLURMCluster для отправки пакетных заданий в планировщик заданий SLURM в суперкомпьютерном кластере. Все задания отправляются как ожидается, но выдают ошибку через 1 минуту работы: asyncio.exceptions.TimeoutError: Nanny failed to start in 60 seconds. Как мне подключить няню?

Полная трассировка: 

distributed.nanny - INFO -         Start Nanny at: 'tcp://206.76.203.125:38324'
distributed.dashboard.proxy - INFO - To route to workers diagnostics web server please install jupyter-server-proxy: python -m pip install jupyter-server-proxy
distributed.worker - INFO -       Start worker at: tcp://206.76.203.125:37609
distributed.worker - INFO -          Listening to: tcp://206.76.203.125:37609
distributed.worker - INFO -          dashboard at:       206.76.203.125:35505
distributed.worker - INFO - Waiting to connect to:  tcp://129.114.63.43:35489
distributed.worker - INFO - -------------------------------------------------
distributed.worker - INFO -               Threads:                          8
distributed.worker - INFO -                Memory:                    2.00 GB
distributed.worker - INFO -       Local Directory: /home1/06729/tg860286/tests/dask-rsmas-presentation/dask-worker-space/worker-pu937jui
distributed.worker - INFO - -------------------------------------------------
distributed.worker - INFO - Waiting to connect to:  tcp://129.114.63.43:35489
distributed.worker - INFO - Waiting to connect to:  tcp://129.114.63.43:35489
distributed.worker - INFO - Waiting to connect to:  tcp://129.114.63.43:35489
distributed.worker - INFO - Waiting to connect to:  tcp://129.114.63.43:35489
distributed.nanny - INFO - Closing Nanny at 'tcp://206.76.203.125:38324'
distributed.worker - INFO - Stopping worker at tcp://206.76.203.125:37609
distributed.worker - INFO - Closed worker has not yet started: None
distributed.dask_worker - INFO - End worker
Traceback (most recent call last):
  File "/home1/06729/tg860286/miniconda3/envs/daskbase/lib/python3.8/site-packages/distributed/node.py", line 173, in wait_for
    await asyncio.wait_for(future, timeout=timeout)
  File "/home1/06729/tg860286/miniconda3/envs/daskbase/lib/python3.8/asyncio/tasks.py", line 490, in wait_for
    raise exceptions.TimeoutError()
asyncio.exceptions.TimeoutError

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "/home1/06729/tg860286/miniconda3/envs/daskbase/lib/python3.8/runpy.py", line 193, in _run_module_as_main
    return _run_code(code, main_globals, None,
  File "/home1/06729/tg860286/miniconda3/envs/daskbase/lib/python3.8/runpy.py", line 86, in _run_code
    exec(code, run_globals)
  File "/home1/06729/tg860286/miniconda3/envs/daskbase/lib/python3.8/site-packages/distributed/cli/dask_worker.py", line 440, in <module>
    go()
  File "/home1/06729/tg860286/miniconda3/envs/daskbase/lib/python3.8/site-packages/distributed/cli/dask_worker.py", line 436, in go
    main()
  File "/home1/06729/tg860286/miniconda3/envs/daskbase/lib/python3.8/site-packages/click/core.py", line 764, in __call__
    return self.main(*args, **kwargs)
  File "/home1/06729/tg860286/miniconda3/envs/daskbase/lib/python3.8/site-packages/click/core.py", line 717, in main
    rv = self.invoke(ctx)
  File "/home1/06729/tg860286/miniconda3/envs/daskbase/lib/python3.8/site-packages/click/core.py", line 956, in invoke
    return ctx.invoke(self.callback, **ctx.params)
  File "/home1/06729/tg860286/miniconda3/envs/daskbase/lib/python3.8/site-packages/click/core.py", line 555, in invoke
    return callback(*args, **kwargs)
  File "/home1/06729/tg860286/miniconda3/envs/daskbase/lib/python3.8/site-packages/distributed/cli/dask_worker.py", line 422, in main
    loop.run_sync(run)
  File "/home1/06729/tg860286/miniconda3/envs/daskbase/lib/python3.8/site-packages/tornado/ioloop.py", line 532, in run_sync
    return future_cell[0].result()
  File "/home1/06729/tg860286/miniconda3/envs/daskbase/lib/python3.8/site-packages/distributed/cli/dask_worker.py", line 416, in run
    await asyncio.gather(*nannies)
  File "/home1/06729/tg860286/miniconda3/envs/daskbase/lib/python3.8/asyncio/tasks.py", line 684, in _wrap_awaitable
    return (yield from awaitable.__await__())
  File "/home1/06729/tg860286/miniconda3/envs/daskbase/lib/python3.8/site-packages/distributed/node.py", line 176, in wait_for
    raise TimeoutError(
asyncio.exceptions.TimeoutError: Nanny failed to start in 60 seconds```

1 Ответ

0 голосов
/ 13 марта 2020

Похоже, ваши работники не смогли подключиться к планировщику. Я предполагаю, что вам нужно указать сетевой интерфейс. Вы должны спросить своего системного администратора, какой сетевой интерфейс следует использовать, а затем указать это с ключевым словом interface=.

Возможно, вы также захотите прочитать https://blog.dask.org/2019/08/28/dask-on-summit, что дает тематическое исследование общих проблем, которые возникают.

...