apache/beam · error · RuntimeError

You cannot turn on runtime_type_check and performance_runtim

Error message

You cannot turn on runtime_type_check and performance_runtime_type_check simultaneously. Pick one or the other.

What it means

Pipeline._run_internal rejects enabling both runtime_type_check and performance_runtime_type_check in TypeOptions, since they are two mutually exclusive runtime type-hint checking implementations (the eager checker vs. the cheaper sampled one). Enabling both is ambiguous, so a RuntimeError is raised.

Source

Thrown at sdks/python/apache_beam/pipeline.py:609

        # runner API check should be skipped for such pipelines.

        # The InteractiveRunner relies on a constant pipeline reference, skip
        # it.
        test_runner_api = (
            not is_fnapi_compatible and
            not self.contains_external_transforms and
            self.runner.__class__.__name__ != 'InteractiveRunner')

      # When possible, invoke a round trip through the runner API.
      if test_runner_api and self._verify_runner_api_compatible():
        return Pipeline.from_runner_api(
            self.to_runner_api(use_fake_coders=True),
            self.runner,
            self._options).run(False)

      if (self._options.view_as(TypeOptions).runtime_type_check and
          self._options.view_as(TypeOptions).performance_runtime_type_check):
        raise RuntimeError(
            'You cannot turn on runtime_type_check '
            'and performance_runtime_type_check simultaneously. '
            'Pick one or the other.')

      if self._options.view_as(TypeOptions).runtime_type_check:
        from apache_beam.typehints import typecheck
        self.visit(typecheck.TypeCheckVisitor())

      if self._options.view_as(TypeOptions).performance_runtime_type_check:
        from apache_beam.typehints import typecheck
        self.visit(typecheck.PerformanceTypeCheckVisitor())

      if self._options.view_as(SetupOptions).save_main_session:
        # If this option is chosen, verify we can pickle the main session early.
        tmpdir = tempfile.mkdtemp()
        try:
          pickler.dump_session(os.path.join(tmpdir, 'main_session.pickle'))
        finally:

View on GitHub (pinned to 12126d8942)

Solutions

1. Remove one of the two flags from your PipelineOptions.
2. Prefer --performance_runtime_type_check for lower overhead in most jobs.
3. Check the environment/launcher scripts for a flag being injected automatically.

Example fix

// before
PipelineOptions(['--runtime_type_check', '--performance_runtime_type_check'])
// after
PipelineOptions(['--performance_runtime_type_check'])

from apache_beam.options.pipeline_options import TypeOptions
opts = PipelineOptions(sys.argv[1:])
t = opts.view_as(TypeOptions)
assert not (t.runtime_type_check and t.performance_runtime_type_check), 'pick one type check mode'

When it happens

Trigger: Setting both flags simultaneously, typically via PipelineOptions(['--runtime_type_check', '--performance_runtime_type_check']) or by assigning both on TypeOptions before pipeline.run().

Common situations: Merging CLI flag sets from different scripts/docs; a default config already setting one flag while the user adds the other on the command line; copy-pasted option lists.

Related errors

• Timing number 0b" + timingNumber.toString(2) + " has more th
• Unknown PaneInfo encoding: " + encodingNibble
• Both a BigQuery table and a query were specified. Please spe
• Transform node %r was not replaced as expected.
• A transform with label "%s" already exists in the pipeline.

AI-assisted analysis of apache/beam@12126d8942 (2026-09-13). Data as JSON: /api/errors/5aa84f1529b66dd5. Report an issue: GitHub.