Class ReplayExceptionClient

Declaration

Documentation

Adds the following methods (and their async variants)to the given client:

- ``recreate_error_locally``: main user method
- ``get_futures_error``: gets the task, its details and dependencies,
    responsible for failure of the given future.

Methods

• ▷ def __init__(self, client)

  source link
• ▶ def get_futures_error(self, future)

  Ask the scheduler details of the sub-task of the given failed future

  source link

  When a future evaluates to a status of "error", i.e., an exception
  was raised in a task within its graph, we an get information from
  the scheduler. This function gets the details of the specific task
  that raised the exception and led to the error, but does not fetch
  data from the cluster or execute the function.

  Parameters

  • future : future that failed, having ``status=="error"``, typically
    

    after an attempt to ``gather()`` shows a stack-stace.

  Returns

  See also

  ReplayExceptionClient.recreate_error_locally

• ▶ def recreate_error_locally(self, future)

  For a failed calculation, perform the blamed task locally for debugging.

  source link

  This operation should be performed after a future (result of ``gather``,
  ``compute``, etc) comes back with a status of "error", if the stack-
  trace is not informative enough to diagnose the problem. The specific
  task (part of the graph pointing to the future) responsible for the
  error will be fetched from the scheduler, together with the values of
  its inputs. The function will then be executed, so that ``pdb`` can
  be used for debugging.

  Parameters

  • future : future or collection that failed
    

    The same thing as was given to ``gather``, but came back with
    an exception/stack-trace. Can also be a (persisted) dask collection
    containing any errored futures.

  Returns

  • Nothing; the function runs and should raise an exception, allowing

  Examples

  >>> future = c.submit(div, 1, 0)         # doctest: +SKIP
  >>> future.status                        # doctest: +SKIP
  'error'
  >>> c.recreate_error_locally(future)     # doctest: +SKIP
  ZeroDivisionError: division by zero
  
  If you're in IPython you might take this opportunity to use pdb
  
  >>> %pdb                                 # doctest: +SKIP
  Automatic pdb calling has been turned ON
  
  >>> c.recreate_error_locally(future)     # doctest: +SKIP
  ZeroDivisionError: division by zero
        1 def div(x, y):
  ----> 2     return x / y
  ipdb>

• ▷ def scheduler(self) @property

  @property
  def scheduler(self)

  source link

Generated by nedoc v0.9 at 2020-12-29 14:05