Timing based RayTaskError(JSONDecodeError) error?

Ray Core
1

I ran into an issue with GitHub - DerwenAI/ray_tutorial: An introductory tutorial about leveraging Ray core features for distributed patterns. not sure if it’s Ray or not causing the issue.

RayTaskError(JSONDecodeError)             Traceback (most recent call last)
<ipython-input-13-2a75d4802248> in <module>
      1 wc_sum = defaultdict(int)
      2 
----> 3 for wc in ray.get(work):
      4     for token, count in wc.items():
      5         wc_sum[token] += count
/opt/conda/lib/python3.7/site-packages/ray/_private/client_mode_hook.py in wrapper(*args, **kwargs)
     45         if client_mode_enabled and _client_hook_enabled:
     46             return getattr(ray, func.__name__)(*args, **kwargs)
---> 47         return func(*args, **kwargs)
     48 
     49     return wrapper
/opt/conda/lib/python3.7/site-packages/ray/worker.py in get(object_refs, timeout)
   1454                     worker.core_worker.dump_object_store_memory_usage()
   1455                 if isinstance(value, RayTaskError):
-> 1456                     raise value.as_instanceof_cause()
   1457                 else:
   1458                     raise value

It’s in ex_para_iter.ipynb [link].

Is it expected to fail if execution of this cell:

nb_items = list(Path(".").glob("ex_*.ipynb"))

iter3 = ray.util.iter.from_items(nb_items, num_shards=3)

work = [nb_word_count.remote(shard) for shard in iter3.shards()]

is quickly succeeded by

wc_sum = defaultdict(int)for wc in ray.get(work):
   for token, count in wc.items():
       wc_sum[token] += count