All expectations fail if one fails with 'unrecognized condition_parser None for Spark execution engine'

[iamamutt]

Describe the bug
Failure to specify condition_parser for a single expectation results in all expectations failing.

To Reproduce

import tempfile

from pprint import pprint

import great_expectations as gx
import pandas as pd

data = {
    'col1': [1, 2, 3, 4, 5],
    'col2': ['A', 'B', 'C', 'D', None],
    'col3': [1.1, None, 3.3, 4.4, 5.5],
}


def validate(dir_path: str, file_name: str):
    context = gx.get_context(mode='ephemeral')
    suite = context.suites.add(
        gx.ExpectationSuite(
            name='test-suite',
            expectations=[
                gx.expectations.ExpectColumnValuesToNotBeNull(
                    column='col1', result_format='COMPLETE'
                ),
                gx.expectations.ExpectColumnValuesToBeInSet(
                    column='col2',
                    value_set=['A', 'B', 'C'],
                    row_condition='col3 IS NOT NULL',
                    mostly=0.665,
                    # condition_parser='spark',
                    result_format='COMPLETE',
                ),
            ],
        )
    )

    return gx.ValidationDefinition(
        name='test-validation',
        data=(
            context.data_sources.add_spark_filesystem(
                name='test-spark-fs',
                base_directory=dir_path,
            )
            .add_csv_asset(
                name='csv-asset',
                sep=',',
                header=True,
                infer_schema=True,
            )
            .add_batch_definition_path(
                name='test-data',
                path=file_name,
            )
        ),
        suite=suite,
    ).run()


with tempfile.TemporaryDirectory() as dir_path:
    file_name = 'data.csv'
    pd.DataFrame(data).to_csv(f'{dir_path}/{file_name}', index=False)
    result = validate(dir_path, file_name)
    pprint(result.to_json_dict(), sort_dicts=False, width=100)

Expected behavior
Only expectations with errors should fail and show exception info

Environment (please complete the following information):

• Operating System: Linux
• Great Expectations Version: 1.2.4
• Data Source: Spark file

Additional context

Details

{
  "success": false,
  "results": [
    {
      "success": false,
      "expectation_config": {
        "type": "expect_column_values_to_not_be_null",
        "kwargs": {
          "result_format": "COMPLETE",
          "column": "col1",
          "batch_id": "test-spark-fs-csv-asset"
        },
        "meta": {},
        "id": "0e734de0-e872-43e7-adc7-07c868c689d7"
      },
      "result": {},
      "meta": {},
      "exception_info": {
        "('table.row_count', '0dfa72ce94f9f181a7dc04305a6c30f7', ())": {
          "exception_traceback": "Traceback (most recent call last):\n  File \"/home/jburling/.local/conda/envs/stonevalley/lib/python3.12/site-packages/great_expectations/execution_engine/execution_engine.py\", line 545, in _process_direct_and_bundled_metric_computation_configurations\n    self.resolve_metric_bundle(metric_fn_bundle=metric_fn_bundle_configurations)\n  File \"/home/jburling/.local/conda/envs/stonevalley/lib/python3.12/site-packages/great_expectations/execution_engine/sparkdf_execution_engine.py\", line 900, in resolve_metric_bundle\n    df: pyspark.DataFrame = self.get_domain_records(domain_kwargs=domain_kwargs)\n                            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n  File \"/home/jburling/.local/conda/envs/stonevalley/lib/python3.12/site-packages/great_expectations/execution_engine/sparkdf_execution_engine.py\", line 681, in get_domain_records\n    raise GreatExpectationsError(  # noqa: TRY003\ngreat_expectations.exceptions.exceptions.GreatExpectationsError: unrecognized condition_parser None for Spark execution engine\n\nThe above exception was the direct cause of the following exception:\n\nTraceback (most recent call last):\n  File \"/home/jburling/.local/conda/envs/stonevalley/lib/python3.12/site-packages/great_expectations/validator/validation_graph.py\", line 276, in _resolve\n    self._execution_engine.resolve_metrics(\n  File \"/home/jburling/.local/conda/envs/stonevalley/lib/python3.12/site-packages/great_expectations/execution_engine/execution_engine.py\", line 279, in resolve_metrics\n    return self._process_direct_and_bundled_metric_computation_configurations(\n           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n  File \"/home/jburling/.local/conda/envs/stonevalley/lib/python3.12/site-packages/great_expectations/execution_engine/execution_engine.py\", line 549, in _process_direct_and_bundled_metric_computation_configurations\n    raise gx_exceptions.MetricResolutionError(\ngreat_expectations.exceptions.exceptions.MetricResolutionError: unrecognized condition_parser None for Spark execution engine\n",
          "exception_message": "unrecognized condition_parser None for Spark execution engine",
          "raised_exception": true
        }
      }
    },
    {
      "success": false,
      "expectation_config": {
        "type": "expect_column_values_to_be_in_set",
        "kwargs": {
          "result_format": "COMPLETE",
          "column": "col2",
          "mostly": 0.665,
          "row_condition": "col3 IS NOT NULL",
          "value_set": [
            "A",
            "B",
            "C"
          ],
          "batch_id": "test-spark-fs-csv-asset"
        },
        "meta": {},
        "id": "c648e872-154c-4374-9cf1-cb8751e1c6d2"
      },
      "result": {},
      "meta": {},
      "exception_info": {
        "('table.column_types', 'e48bc318d7e9c92e270e3f7ab807c1b8', 'include_nested=True')": {
          "exception_traceback": "Traceback (most recent call last):\n  File \"/home/jburling/.local/conda/envs/stonevalley/lib/python3.12/site-packages/great_expectations/execution_engine/execution_engine.py\", line 532, in _process_direct_and_bundled_metric_computation_configurations\n    metric_computation_configuration.metric_fn(  # type: ignore[misc] # F not callable\n  File \"/home/jburling/.local/conda/envs/stonevalley/lib/python3.12/site-packages/great_expectations/expectations/metrics/metric_provider.py\", line 60, in inner_func\n    return metric_fn(*args, **kwargs)\n           ^^^^^^^^^^^^^^^^^^^^^^^^^^\n  File \"/home/jburling/.local/conda/envs/stonevalley/lib/python3.12/site-packages/great_expectations/expectations/metrics/table_metrics/table_column_types.py\", line 81, in _spark\n    df, _, _ = execution_engine.get_compute_domain(\n               ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n  File \"/home/jburling/.local/conda/envs/stonevalley/lib/python3.12/site-packages/great_expectations/execution_engine/sparkdf_execution_engine.py\", line 800, in get_compute_domain\n    data: pyspark.DataFrame = self.get_domain_records(domain_kwargs=domain_kwargs)\n                              ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n  File \"/home/jburling/.local/conda/envs/stonevalley/lib/python3.12/site-packages/great_expectations/execution_engine/sparkdf_execution_engine.py\", line 681, in get_domain_records\n    raise GreatExpectationsError(  # noqa: TRY003\ngreat_expectations.exceptions.exceptions.GreatExpectationsError: unrecognized condition_parser None for Spark execution engine\n\nThe above exception was the direct cause of the following exception:\n\nTraceback (most recent call last):\n  File \"/home/jburling/.local/conda/envs/stonevalley/lib/python3.12/site-packages/great_expectations/validator/validation_graph.py\", line 276, in _resolve\n    self._execution_engine.resolve_metrics(\n  File \"/home/jburling/.local/conda/envs/stonevalley/lib/python3.12/site-packages/great_expectations/execution_engine/execution_engine.py\", line 279, in resolve_metrics\n    return self._process_direct_and_bundled_metric_computation_configurations(\n           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n  File \"/home/jburling/.local/conda/envs/stonevalley/lib/python3.12/site-packages/great_expectations/execution_engine/execution_engine.py\", line 537, in _process_direct_and_bundled_metric_computation_configurations\n    raise gx_exceptions.MetricResolutionError(\ngreat_expectations.exceptions.exceptions.MetricResolutionError: unrecognized condition_parser None for Spark execution engine\n",
          "exception_message": "unrecognized condition_parser None for Spark execution engine",
          "raised_exception": true
        },
        "('table.row_count', 'e48bc318d7e9c92e270e3f7ab807c1b8', ())": {
          "exception_traceback": "Traceback (most recent call last):\n  File \"/home/jburling/.local/conda/envs/stonevalley/lib/python3.12/site-packages/great_expectations/execution_engine/execution_engine.py\", line 545, in _process_direct_and_bundled_metric_computation_configurations\n    self.resolve_metric_bundle(metric_fn_bundle=metric_fn_bundle_configurations)\n  File \"/home/jburling/.local/conda/envs/stonevalley/lib/python3.12/site-packages/great_expectations/execution_engine/sparkdf_execution_engine.py\", line 900, in resolve_metric_bundle\n    df: pyspark.DataFrame = self.get_domain_records(domain_kwargs=domain_kwargs)\n                            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n  File \"/home/jburling/.local/conda/envs/stonevalley/lib/python3.12/site-packages/great_expectations/execution_engine/sparkdf_execution_engine.py\", line 681, in get_domain_records\n    raise GreatExpectationsError(  # noqa: TRY003\ngreat_expectations.exceptions.exceptions.GreatExpectationsError: unrecognized condition_parser None for Spark execution engine\n\nThe above exception was the direct cause of the following exception:\n\nTraceback (most recent call last):\n  File \"/home/jburling/.local/conda/envs/stonevalley/lib/python3.12/site-packages/great_expectations/validator/validation_graph.py\", line 276, in _resolve\n    self._execution_engine.resolve_metrics(\n  File \"/home/jburling/.local/conda/envs/stonevalley/lib/python3.12/site-packages/great_expectations/execution_engine/execution_engine.py\", line 279, in resolve_metrics\n    return self._process_direct_and_bundled_metric_computation_configurations(\n           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n  File \"/home/jburling/.local/conda/envs/stonevalley/lib/python3.12/site-packages/great_expectations/execution_engine/execution_engine.py\", line 549, in _process_direct_and_bundled_metric_computation_configurations\n    raise gx_exceptions.MetricResolutionError(\ngreat_expectations.exceptions.exceptions.MetricResolutionError: unrecognized condition_parser None for Spark execution engine\n",
          "exception_message": "unrecognized condition_parser None for Spark execution engine",
          "raised_exception": true
        }
      }
    }
  ],
  "suite_name": "test-suite",
  "suite_parameters": {},
  "statistics": {
    "evaluated_expectations": 2,
    "successful_expectations": 0,
    "unsuccessful_expectations": 2,
    "success_percent": 0.0
  },
  "meta": {
    "great_expectations_version": "1.2.4",
    "batch_spec": {
      "path": "/tmp/tmpgf8032g5/data.csv",
      "reader_method": "csv",
      "reader_options": {
        "sep": ",",
        "header": true,
        "inferSchema": true
      }
    },
    "batch_markers": {
      "ge_load_time": "20241126T222652.157081Z"
    },
    "active_batch_definition": {
      "datasource_name": "test-spark-fs",
      "data_connector_name": "fluent",
      "data_asset_name": "csv-asset",
      "batch_identifiers": {
        "path": "data.csv"
      },
      "batching_regex": "(?P<path>data.csv)"
    },
    "validation_id": "cd188ea7-bedf-4f8a-9898-1cf823b69b5f",
    "checkpoint_id": null,
    "batch_parameters": null
  },
  "id": null
}

[adeola-ak]

Hi there, thank you for bringing this to our attention. I've shared it with the team and I will follow up with you once I have any updates

[adeola-ak]

Thank you for bringing this to our attention. We've identified the root cause, but unfortunately, the fix won't be prioritized at this time. I'll keep the issue open, as it's on our backlog to address. Thanks again for highlighting this.

[Mr-Neutr0n]

Opened a fix for this: #11658

The root cause is in ExecutionEngine._process_direct_and_bundled_metric_computation_configurations — all bundled metrics are passed to resolve_metric_bundle() in a single call, so if any one domain fails (e.g. missing condition_parser on Spark), the resulting MetricResolutionError lists every bundled metric as failed, even those from unrelated domains.

The fix groups bundled metrics by their compute domain before resolving, so each group is processed independently. A failure in one group no longer cascades to others.