You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@beam.apache.org by "Nir Gazit (Jira)" <ji...@apache.org> on 2021/04/28 17:16:00 UTC

[jira] [Created] (BEAM-12244) Failure to connect to expansion service with GRPC local secure channel

Nir Gazit created BEAM-12244:
--------------------------------

             Summary: Failure to connect to expansion service with GRPC local secure channel
                 Key: BEAM-12244
                 URL: https://issues.apache.org/jira/browse/BEAM-12244
             Project: Beam
          Issue Type: Bug
          Components: io-py-kafka
            Reporter: Nir Gazit
            Assignee: Nir Gazit


There seem to be a discrepancy between how Beam Python SDK connects to the job endpoint and the artifact endpoint (where it always uses insecure GRPC connection), and the expansion service (where it [tries to use|https://github.com/apache/beam/blob/master/sdks/python/apache_beam/transforms/external.py#L322] local secure channel if available in the GRPC module).

This has caused an error to be thrown when trying to connect to a remote Flink cluster: 
{code:none}
File "/Users/nir.gazit/.pyenv/versions/beam/lib/python3.7/site-packages/apache_beam/transforms/ptransform.py", line 573, in __ror__
    result = p.apply(self, pvalueish, label)
  File "/Users/nir.gazit/.pyenv/versions/beam/lib/python3.7/site-packages/apache_beam/pipeline.py", line 689, in apply
    pvalueish_result = self.runner.apply(transform, pvalueish, self._options)
  File "/Users/nir.gazit/.pyenv/versions/beam/lib/python3.7/site-packages/apache_beam/runners/runner.py", line 188, in apply
    return m(transform, input, options)
  File "/Users/nir.gazit/.pyenv/versions/beam/lib/python3.7/site-packages/apache_beam/runners/runner.py", line 218, in apply_PTransform
    return transform.expand(input)
  File "/Users/nir.gazit/.pyenv/versions/beam/lib/python3.7/site-packages/apache_beam/transforms/external.py", line 325, in expand
    response = service.Expand(request)
  File "/Users/nir.gazit/.pyenv/versions/beam/lib/python3.7/site-packages/grpc/_channel.py", line 923, in __call__
    return _end_unary_response_blocking(state, call, False, None)
  File "/Users/nir.gazit/.pyenv/versions/beam/lib/python3.7/site-packages/grpc/_channel.py", line 826, in _end_unary_response_blocking
    raise _InactiveRpcError(state)
grpc._channel._InactiveRpcError: <_InactiveRpcError of RPC that terminated with:
        status = StatusCode.UNAVAILABLE
        details = "failed to connect to all addresses"
        debug_error_string = "{"created":"@1619629304.651319000","description":"Failed to pick subchannel","file":"src/core/ext/filters/client_channel/client_channel.cc","file_line":5397,"referenced_errors":[{"created":"@1619629304.651278000","description":"failed to connect to all addresses","file":"src/core/ext/filters/client_channel/lb_policy/pick_first/pick_first.cc","file_line":398,"grpc_status":14}]}"
{code}

I think all GRPC services (job, artifact, expansion) should use the same way to connect (currently insecure_channel)



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