[ 
https://issues.apache.org/jira/browse/BEAM-9127?focusedWorklogId=375893&page=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-375893
 ]

ASF GitHub Bot logged work on BEAM-9127:
----------------------------------------

                Author: ASF GitHub Bot
            Created on: 22/Jan/20 21:16
            Start Date: 22/Jan/20 21:16
    Worklog Time Spent: 10m 
      Work Description: boyuanzz commented on issue #10665: [BEAM-9127] Fix 
output type declaration in xlang wordcount. (#10605)
URL: https://github.com/apache/beam/pull/10665#issuecomment-577389161
 
 
   R: @robertwb 
 
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
 
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org


Issue Time Tracking
-------------------

    Worklog Id:     (was: 375893)
    Time Spent: 2h 20m  (was: 2h 10m)

> postcommit: suites:portable:py2:crossLanguagePortableWordCount failing
> ----------------------------------------------------------------------
>
>                 Key: BEAM-9127
>                 URL: https://issues.apache.org/jira/browse/BEAM-9127
>             Project: Beam
>          Issue Type: Bug
>          Components: test-failures
>            Reporter: Udi Meiri
>            Assignee: Robert Bradshaw
>            Priority: Major
>          Time Spent: 2h 20m
>  Remaining Estimate: 0h
>
> Log: 
> https://builds.apache.org/job/beam_PostCommit_Python2/1445/timestamps/?time=HH:mm:ss&timeZone=GMT-8&appendLog&locale=en_US
> {code}
> 11:32:07  RuntimeError: Pipeline 
> BeamApp-root-0114193141-d68e793b_21a90e10-d9a1-41e8-bdba-154dddc532f3 failed 
> in state FAILED: java.lang.RuntimeException: Error received from SDK harness 
> for instruction 16: Traceback (most recent call last):
> 11:32:07    File 
> "/usr/local/lib/python2.7/site-packages/apache_beam/runners/worker/sdk_worker.py",
>  line 165, in _execute
> 11:32:07      response = task()
> 11:32:07    File 
> "/usr/local/lib/python2.7/site-packages/apache_beam/runners/worker/sdk_worker.py",
>  line 221, in <lambda>
> 11:32:07      lambda: self.create_worker().do_instruction(request), request)
> 11:32:07    File 
> "/usr/local/lib/python2.7/site-packages/apache_beam/runners/worker/sdk_worker.py",
>  line 350, in do_instruction
> 11:32:07      request.instruction_id)
> 11:32:07    File 
> "/usr/local/lib/python2.7/site-packages/apache_beam/runners/worker/sdk_worker.py",
>  line 384, in process_bundle
> 11:32:07      bundle_processor.process_bundle(instruction_id))
> 11:32:07    File 
> "/usr/local/lib/python2.7/site-packages/apache_beam/runners/worker/bundle_processor.py",
>  line 811, in process_bundle
> 11:32:07      data.transform_id].process_encoded(data.data)
> 11:32:07    File 
> "/usr/local/lib/python2.7/site-packages/apache_beam/runners/worker/bundle_processor.py",
>  line 206, in process_encoded
> 11:32:07      self.output(decoded_value)
> 11:32:07    File "apache_beam/runners/worker/operations.py", line 302, in 
> apache_beam.runners.worker.operations.Operation.output
> 11:32:07      def output(self, windowed_value, output_index=0):
> 11:32:07    File "apache_beam/runners/worker/operations.py", line 304, in 
> apache_beam.runners.worker.operations.Operation.output
> 11:32:07      cython.cast(Receiver, 
> self.receivers[output_index]).receive(windowed_value)
> 11:32:07    File "apache_beam/runners/worker/operations.py", line 178, in 
> apache_beam.runners.worker.operations.SingletonConsumerSet.receive
> 11:32:07      self.consumer.process(windowed_value)
> 11:32:07    File "apache_beam/runners/worker/operations.py", line 751, in 
> apache_beam.runners.worker.operations.SdfProcessSizedElements.process
> 11:32:07      with self.scoped_process_state:
> 11:32:07    File "apache_beam/runners/worker/operations.py", line 759, in 
> apache_beam.runners.worker.operations.SdfProcessSizedElements.process
> 11:32:07      delayed_application = 
> self.dofn_runner.process_with_sized_restriction(o)
> 11:32:07    File "apache_beam/runners/common.py", line 888, in 
> apache_beam.runners.common.DoFnRunner.process_with_sized_restriction
> 11:32:07      return self.do_fn_invoker.invoke_process(
> 11:32:07    File "apache_beam/runners/common.py", line 657, in 
> apache_beam.runners.common.PerWindowInvoker.invoke_process
> 11:32:07      return self._invoke_process_per_window(
> 11:32:07    File "apache_beam/runners/common.py", line 744, in 
> apache_beam.runners.common.PerWindowInvoker._invoke_process_per_window
> 11:32:07      output_processor.process_outputs(
> 11:32:07    File "apache_beam/runners/common.py", line 984, in 
> apache_beam.runners.common._OutputProcessor.process_outputs
> 11:32:07      def process_outputs(self, windowed_input_element, results):
> 11:32:07    File "apache_beam/runners/common.py", line 1024, in 
> apache_beam.runners.common._OutputProcessor.process_outputs
> 11:32:07      self.main_receivers.receive(windowed_value)
> 11:32:07    File "apache_beam/runners/worker/operations.py", line 178, in 
> apache_beam.runners.worker.operations.SingletonConsumerSet.receive
> 11:32:07      self.consumer.process(windowed_value)
> 11:32:07    File "apache_beam/runners/worker/operations.py", line 657, in 
> apache_beam.runners.worker.operations.DoOperation.process
> 11:32:07      with self.scoped_process_state:
> 11:32:07    File "apache_beam/runners/worker/operations.py", line 658, in 
> apache_beam.runners.worker.operations.DoOperation.process
> 11:32:07      delayed_application = self.dofn_receiver.receive(o)
> 11:32:07    File "apache_beam/runners/common.py", line 876, in 
> apache_beam.runners.common.DoFnRunner.receive
> 11:32:07      self.process(windowed_value)
> 11:32:07    File "apache_beam/runners/common.py", line 883, in 
> apache_beam.runners.common.DoFnRunner.process
> 11:32:07      self._reraise_augmented(exn)
> 11:32:07    File "apache_beam/runners/common.py", line 952, in 
> apache_beam.runners.common.DoFnRunner._reraise_augmented
> 11:32:07      raise_with_traceback(new_exn)
> 11:32:07    File "apache_beam/runners/common.py", line 881, in 
> apache_beam.runners.common.DoFnRunner.process
> 11:32:07      return self.do_fn_invoker.invoke_process(windowed_value)
> 11:32:07    File "apache_beam/runners/common.py", line 496, in 
> apache_beam.runners.common.SimpleInvoker.invoke_process
> 11:32:07      output_processor.process_outputs(
> 11:32:07    File "apache_beam/runners/common.py", line 1024, in 
> apache_beam.runners.common._OutputProcessor.process_outputs
> 11:32:07      self.main_receivers.receive(windowed_value)
> 11:32:07    File "apache_beam/runners/worker/operations.py", line 177, in 
> apache_beam.runners.worker.operations.SingletonConsumerSet.receive
> 11:32:07      self.update_counters_start(windowed_value)
> 11:32:07    File "apache_beam/runners/worker/operations.py", line 155, in 
> apache_beam.runners.worker.operations.ConsumerSet.update_counters_start
> 11:32:07      self.opcounter.update_from(windowed_value)
> 11:32:07    File "apache_beam/runners/worker/opcounters.py", line 214, in 
> apache_beam.runners.worker.opcounters.OperationCounters.update_from
> 11:32:07      self.do_sample(windowed_value)
> 11:32:07    File "apache_beam/runners/worker/opcounters.py", line 233, in 
> apache_beam.runners.worker.opcounters.OperationCounters.do_sample
> 11:32:07      
> self.coder_impl.get_estimated_size_and_observables(windowed_value))
> 11:32:07    File "apache_beam/coders/coder_impl.py", line 1161, in 
> apache_beam.coders.coder_impl.WindowedValueCoderImpl.get_estimated_size_and_observables
> 11:32:07      def get_estimated_size_and_observables(self, value, 
> nested=False):
> 11:32:07    File "apache_beam/coders/coder_impl.py", line 1171, in 
> apache_beam.coders.coder_impl.WindowedValueCoderImpl.get_estimated_size_and_observables
> 11:32:07      self._value_coder.get_estimated_size_and_observables(
> 11:32:07    File "apache_beam/coders/coder_impl.py", line 187, in 
> apache_beam.coders.coder_impl.CoderImpl.get_estimated_size_and_observables
> 11:32:07      return self.estimate_size(value, nested), []
> 11:32:07    File "apache_beam/coders/coder_impl.py", line 161, in 
> apache_beam.coders.coder_impl.CoderImpl.estimate_size
> 11:32:07      self.encode_to_stream(value, out, nested)
> 11:32:07    File "apache_beam/coders/coder_impl.py", line 489, in 
> apache_beam.coders.coder_impl.BytesCoderImpl.encode_to_stream
> 11:32:07      out.write(value, nested)
> 11:32:07  TypeError: Expected bytes, got unicode [while running 'split']
> {code}
> Culprit seems to be: https://github.com/apache/beam/pull/10558



--
This message was sent by Atlassian Jira
(v8.3.4#803005)

Reply via email to