def test_pipeline_step_key_subset_execution_wrong_step_key_in_step_output_handles(
):
pipeline_def = define_addy_pipeline()
old_run_id = str(uuid.uuid4())
environment_dict = env_with_fs(
{'solids': {
'add_one': {
'inputs': {
'num': {
'value': 3
}
}
}
}})
result = execute_pipeline(pipeline_def,
environment_dict=environment_dict,
run_config=RunConfig(run_id=old_run_id))
assert result.success
assert result.run_id == old_run_id
new_run_id = str(uuid.uuid4())
with pytest.raises(DagsterExecutionStepNotFoundError):
execute_pipeline(
pipeline_def,
environment_dict=environment_dict,
run_config=RunConfig(
run_id=new_run_id,
reexecution_config=ReexecutionConfig(
previous_run_id=result.run_id,
step_output_handles=[
StepOutputHandle('invalid_in_step_output_handles')
],
),
step_keys_to_execute=['add_two.compute'],
),
)
def test_execution_plan_reexecution_with_in_memory():
pipeline_def = define_addy_pipeline()
instance = DagsterInstance.ephemeral()
old_run_id = str(uuid.uuid4())
environment_dict = {'solids': {'add_one': {'inputs': {'num': {'value': 3}}}}}
result = execute_pipeline(
pipeline_def,
environment_dict=environment_dict,
run_config=RunConfig(run_id=old_run_id),
instance=instance,
)
assert result.success
## re-execute add_two
new_run_id = str(uuid.uuid4())
in_memory_run_config = RunConfig(
run_id=new_run_id,
reexecution_config=ReexecutionConfig(
previous_run_id=result.run_id, step_output_handles=[StepOutputHandle('add_one.compute')]
),
)
execution_plan = create_execution_plan(
pipeline_def, environment_dict=environment_dict, run_config=in_memory_run_config
)
with pytest.raises(DagsterInvariantViolationError):
execute_plan(
execution_plan,
environment_dict=environment_dict,
run_config=in_memory_run_config,
step_keys_to_execute=['add_two.compute'],
instance=instance,
)
def _do_execute_plan(graphene_info, execution_params, dauphin_pipeline):
check.inst_param(graphene_info, 'graphene_info', ResolveInfo)
check.inst_param(execution_params, 'execution_params', ExecutionParams)
event_records = []
run_config = RunConfig(
run_id=execution_params.execution_metadata.run_id,
mode=execution_params.mode,
tags=execution_params.execution_metadata.tags,
event_callback=event_records.append,
)
execution_plan = create_execution_plan(
pipeline=dauphin_pipeline.get_dagster_pipeline(),
environment_dict=execution_params.environment_dict,
run_config=run_config,
)
if execution_params.step_keys:
for step_key in execution_params.step_keys:
if not execution_plan.has_step(step_key):
raise UserFacingGraphQLError(
graphene_info.schema.type_named('InvalidStepError')(invalid_step_key=step_key)
)
execute_plan(
execution_plan=execution_plan,
environment_dict=execution_params.environment_dict,
run_config=run_config,
step_keys_to_execute=execution_params.step_keys,
)
def to_graphql_event(event_record):
return from_dagster_event_record(
graphene_info, event_record, dauphin_pipeline, execution_plan
)
return graphene_info.schema.type_named('ExecutePlanSuccess')(
pipeline=dauphin_pipeline,
has_failures=any(
er
for er in event_records
if er.is_dagster_event and er.dagster_event.event_type == DagsterEventType.STEP_FAILURE
),
step_events=list(
map(to_graphql_event, filter(lambda er: er.is_dagster_event, event_records))
),
)
def test_using_gcs_for_subplan(gcs_bucket):
pipeline_def = define_inty_pipeline()
environment_dict = {'storage': {'gcs': {'config': {'gcs_bucket': gcs_bucket}}}}
run_id = str(uuid.uuid4())
execution_plan = create_execution_plan(
pipeline_def, environment_dict=environment_dict, run_config=RunConfig(run_id=run_id)
)
assert execution_plan.get_step_by_key('return_one.compute')
step_keys = ['return_one.compute']
instance = DagsterInstance.ephemeral()
pipeline_run = PipelineRun.create_empty_run(
pipeline_def.name, run_id=run_id, environment_dict=environment_dict
)
return_one_step_events = list(
execute_plan(
execution_plan.build_subset_plan(step_keys),
environment_dict=environment_dict,
pipeline_run=pipeline_run,
instance=instance,
)
)
assert get_step_output(return_one_step_events, 'return_one.compute')
with scoped_pipeline_context(pipeline_def, environment_dict, pipeline_run, instance) as context:
store = GCSIntermediateStore(
gcs_bucket, run_id, client=context.scoped_resources_builder.build().gcs.client
)
assert store.has_intermediate(context, 'return_one.compute')
assert store.get_intermediate(context, 'return_one.compute', Int).obj == 1
add_one_step_events = list(
execute_plan(
execution_plan.build_subset_plan(['add_one.compute']),
environment_dict=environment_dict,
pipeline_run=pipeline_run,
instance=instance,
)
)
assert get_step_output(add_one_step_events, 'add_one.compute')
with scoped_pipeline_context(pipeline_def, environment_dict, pipeline_run, instance) as context:
assert store.has_intermediate(context, 'add_one.compute')
assert store.get_intermediate(context, 'add_one.compute', Int).obj == 2
def test_multiproc_event_sink():
pipeline = ExecutionTargetHandle.for_pipeline_python_file(
__file__, 'define_diamond_pipeline'
).build_pipeline_definition()
sink = InMemoryEventSink()
result = execute_pipeline(
pipeline,
run_config=RunConfig(event_sink=sink),
environment_dict={'storage': {'filesystem': {}}, 'execution': {'multiprocess': {}}},
)
assert result.success
assert len(result.event_list) == len(sink.dagster_event_records)
def do_execute_command(pipeline, env_file_list, mode=None):
check.inst_param(pipeline, 'pipeline', PipelineDefinition)
env_file_list = check.opt_list_param(env_file_list,
'env_file_list',
of_type=str)
environment_dict = load_yaml_from_glob_list(
env_file_list) if env_file_list else {}
return execute_pipeline(
pipeline,
environment_dict=environment_dict,
run_config=RunConfig(mode=mode),
instance=DagsterInstance.get(),
)
def do_execute_command(pipeline, env_file_list, raise_on_error):
check.inst_param(pipeline, 'pipeline', PipelineDefinition)
env_file_list = check.opt_list_param(env_file_list,
'env_file_list',
of_type=str)
environment_dict = load_yaml_from_glob_list(
env_file_list) if env_file_list else {}
return execute_pipeline(
pipeline,
environment_dict=environment_dict,
run_config=RunConfig(executor_config=InProcessExecutorConfig(
raise_on_error=raise_on_error)),
)
def test_ingest_pipeline_fast_filesystem_storage(postgres):
ingest_config_dict = load_yaml_from_globs(
config_path('local_base.yaml'),
config_path('local_fast_ingest.yaml'),
config_path('filesystem_storage.yaml'),
)
ingest_config_dict = enviroment_overrides(ingest_config_dict)
result_ingest = execute_pipeline(
ingest_pipeline_def,
ingest_config_dict,
run_config=RunConfig(mode='local'),
instance=DagsterInstance.local_temp(),
)
assert result_ingest.success
def test_no_repo_registration_error():
with pytest.raises(
DagstermillError,
match=
'Error occurred during the execution of Dagstermill solid no_repo_reg',
) as exc:
execute_pipeline(define_no_repo_registration_error_pipeline())
assert (
'If Dagstermill solids have outputs that require serialization strategies'
in exc.value.original_exc_info[1].args[0])
with exec_for_test(
define_no_repo_registration_error_pipeline(),
run_config=RunConfig.nonthrowing_in_process()) as result:
assert not result.success
def test_error_notebook():
with pytest.raises(
DagstermillError, match='Error occurred during the execution of Dagstermill solid'
) as exc:
with exec_for_test('define_error_pipeline') as result:
pass
assert 'Someone set up us the bomb' in exc.value.original_exc_info[1].args[0]
with exec_for_test(
'define_error_pipeline', run_config=RunConfig.nonthrowing_in_process()
) as result:
assert not result.success
assert result.step_event_list[1].event_type.value == 'STEP_MATERIALIZATION'
assert result.step_event_list[2].event_type.value == 'STEP_FAILURE'
def test_execute_multi_mode_loggers_with_single_logger():
pipeline_def, foo_logger_captured_results, bar_logger_captured_results = (
define_multi_mode_with_loggers_pipeline()
)
execute_pipeline(
pipeline_def,
run_config=RunConfig(mode='foo_mode'),
environment_dict={'loggers': {'foo': {'config': {'log_level': 'DEBUG'}}}},
)
assert not bar_logger_captured_results
original_messages = parse_captured_results(foo_logger_captured_results)
assert len([x for x in original_messages if 'Here we are' in x]) == 1
def test_download_csv_locally_pipeline(mocker):
# Setup download mocks
mocker.patch('dagster_examples.bay_bikes.solids.requests')
mocker.patch('dagster_examples.bay_bikes.solids._write_chunks_to_fp')
mocker.patch('dagster_examples.bay_bikes.solids._unzip_file',
side_effect=mock_unzip_csv)
# execute tests
result = execute_pipeline_with_preset(monthly_bay_bike_etl_pipeline,
preset_name='dev',
run_config=RunConfig(mode='local'))
assert result.success
with open('/tmp/test_bucket/key_storage.json') as fp:
key_storage = json.load(fp)
assert len(key_storage.items()) == 1
def test_filesystem_run_storage_from_run_config():
@solid
def check_run_storage(context):
assert isinstance(context.get_system_context().run_storage,
FileSystemRunStorage)
pipeline = PipelineDefinition(name='filesystem_run_storage_test',
solids=[check_run_storage])
result = execute_pipeline(
pipeline, run_config=RunConfig(storage_mode=RunStorageMode.FILESYSTEM))
assert result.success
assert os.path.isdir(os.path.join(base_run_directory(), result.run_id))
def test_execution_plan_wrong_invalid_step_key():
pipeline_def = define_addy_pipeline()
old_run_id = str(uuid.uuid4())
environment_dict = {'solids': {'add_one': {'inputs': {'num': {'value': 3}}}}}
result = execute_pipeline(
pipeline_def,
environment_dict=environment_dict,
run_config=RunConfig(storage_mode=RunStorageMode.FILESYSTEM, run_id=old_run_id),
)
new_run_id = str(uuid.uuid4())
run_config = RunConfig(
run_id=new_run_id,
reexecution_config=ReexecutionConfig(
previous_run_id=result.run_id,
step_output_handles=[StepOutputHandle('not_valid.transform')],
),
storage_mode=RunStorageMode.FILESYSTEM,
)
execution_plan = create_execution_plan(pipeline_def, environment_dict=environment_dict)
with pytest.raises(DagsterExecutionStepNotFoundError) as exc_info:
execute_plan(
execution_plan,
environment_dict=environment_dict,
run_config=run_config,
step_keys_to_execute=['add_two.transform'],
)
assert str(exc_info.value) == (
'Step not_valid.transform was specified as a step from a previous run. '
'It does not exist.'
)
def test_injected_tags():
called = {}
@solid
def check_tags(context):
assert context.get_tag('foo') == 'bar'
called['yup'] = True
pipeline_def = PipelineDefinition(name='injected_run_id',
solid_defs=[check_tags])
result = execute_pipeline(pipeline_def,
run_config=RunConfig(tags={'foo': 'bar'}))
assert result.success
assert called['yup']
def test_pyspark_emr(mock_wait):
client = boto3.client('emr', region_name='us-west-1')
run_job_flow_args = dict(
Instances={
'InstanceCount': 1,
'KeepJobFlowAliveWhenNoSteps': True,
'MasterInstanceType': 'c3.medium',
'Placement': {
'AvailabilityZone': 'us-west-1a'
},
'SlaveInstanceType': 'c3.xlarge',
},
JobFlowRole='EMR_EC2_DefaultRole',
LogUri='s3://mybucket/log',
Name='cluster',
ServiceRole='EMR_DefaultRole',
VisibleToAllUsers=True,
)
job_flow_id = client.run_job_flow(**run_job_flow_args)['JobFlowId']
result = execute_pipeline(
example_pipe,
environment_dict={
'solids': {
'blah': {
'config': {
'foo': 'a string',
'bar': 123
}
}
},
'resources': {
'pyspark': {
'config': {
'pipeline_file': __file__,
'pipeline_fn_name': 'example_pipe',
'job_flow_id': job_flow_id,
'staging_bucket': 'dagster-scratch-80542c2',
'region_name': 'us-west-1',
}
}
},
},
run_config=RunConfig(mode='prod'),
)
assert result.success
assert mock_wait.called_once
def _launch_pipeline_execution(graphene_info, execution_params, is_reexecuted=False):
check.inst_param(graphene_info, 'graphene_info', ResolveInfo)
check.inst_param(execution_params, 'execution_params', ExecutionParams)
if is_reexecuted:
# required fields for re-execution
execution_metadata = check.inst_param(
execution_params.execution_metadata, 'execution_metadata', ExecutionMetadata
)
check.str_param(execution_metadata.root_run_id, 'root_run_id')
check.str_param(execution_metadata.parent_run_id, 'parent_run_id')
error_type = 'RunLauncherNotDefinedError'
success_type = (
'LaunchPipelineExecutionSuccess'
if not is_reexecuted
else 'LaunchPipelineReexecutionSuccess'
)
instance = graphene_info.context.instance
run_launcher = instance.run_launcher
if run_launcher is None:
return graphene_info.schema.type_named(error_type)()
pipeline_def = get_pipeline_def_from_selector(graphene_info, execution_params.selector)
get_validated_config(
graphene_info,
pipeline_def,
environment_dict=execution_params.environment_dict,
mode=execution_params.mode,
)
execution_plan = create_execution_plan(
pipeline_def,
execution_params.environment_dict,
run_config=RunConfig(
mode=execution_params.mode, previous_run_id=execution_params.previous_run_id
),
)
_check_start_pipeline_execution_errors(graphene_info, execution_params, execution_plan)
run = instance.launch_run(_create_pipeline_run(instance, pipeline_def, execution_params))
return graphene_info.schema.type_named(success_type)(
run=graphene_info.schema.type_named('PipelineRun')(run)
)
def start_pipeline_execution(graphene_info, execution_params,
reexecution_config):
check.inst_param(graphene_info, 'graphene_info', ResolveInfo)
check.inst_param(execution_params, 'execution_params', ExecutionParams)
check.opt_inst_param(reexecution_config, 'reexecution_config',
ReexecutionConfig)
pipeline_run_storage = graphene_info.context.pipeline_runs
dauphin_pipeline = get_dauphin_pipeline_from_selector(
graphene_info, execution_params.selector)
get_validated_config(
graphene_info,
dauphin_pipeline,
environment_dict=execution_params.environment_dict,
mode=execution_params.mode,
)
execution_plan = create_execution_plan(
dauphin_pipeline.get_dagster_pipeline(),
execution_params.environment_dict,
run_config=RunConfig(mode=execution_params.mode),
)
_check_start_pipeline_execution_errors(graphene_info, execution_params,
execution_plan, reexecution_config)
run = pipeline_run_storage.create_run(
run_id=execution_params.execution_metadata.run_id
if execution_params.execution_metadata.run_id else make_new_run_id(),
selector=execution_params.selector,
env_config=execution_params.environment_dict,
mode=execution_params.mode,
reexecution_config=reexecution_config,
step_keys_to_execute=execution_params.step_keys,
)
pipeline_run_storage.add_run(run)
graphene_info.context.execution_manager.execute_pipeline(
graphene_info.context.get_handle(),
dauphin_pipeline.get_dagster_pipeline(),
run,
raise_on_error=graphene_info.context.raise_on_error,
)
return graphene_info.schema.type_named('StartPipelineExecutionSuccess')(
run=graphene_info.schema.type_named('PipelineRun')(run))
def test_failure_midstream():
'''
A
\\
C (fails) = D (skipped)
//
B
'''
solid_a = create_root_success_solid('A')
solid_b = create_root_success_solid('B')
def fail_fn(_context, inputs):
check.failed('user error')
return [inputs['A'], inputs['B'], {'C': 'compute_called'}]
def success_fn(_context, inputs):
return [inputs['C'], {'D': 'compute_called'}]
solid_c = single_output_solid(
name='C',
input_defs=[InputDefinition(name='A'), InputDefinition(name='B')],
compute_fn=fail_fn,
output_def=OutputDefinition(),
)
solid_d = single_output_solid(
name='D',
input_defs=[InputDefinition(name='C')],
compute_fn=success_fn,
output_def=OutputDefinition(),
)
pipeline_def = PipelineDefinition(
solid_defs=[solid_a, solid_b, solid_c, solid_d],
dependencies={
'C': {'A': DependencyDefinition(solid_a.name), 'B': DependencyDefinition(solid_b.name)},
'D': {'C': DependencyDefinition(solid_c.name)},
},
)
pipeline_result = execute_pipeline(pipeline_def, run_config=RunConfig.nonthrowing_in_process())
assert pipeline_result.result_for_solid('A').success
assert pipeline_result.result_for_solid('B').success
assert not pipeline_result.result_for_solid('C').success
assert pipeline_result.result_for_solid('C').failure_data.error.cls_name == 'CheckError'
assert not pipeline_result.result_for_solid('D').success
assert pipeline_result.result_for_solid('D').skipped
def test_mem_storage_error_pipeline_multiprocess():
with pytest.raises(DagsterInvariantViolationError) as exc_info:
execute_pipeline(
define_diamond_pipeline(),
run_config=RunConfig(executor_config=MultiprocessExecutorConfig(
ExecutionTargetHandle.for_pipeline_fn(define_error_pipeline))),
)
assert ('While invoking '
'pipeline diamond_execution. You have attempted to use the '
'multiprocessing executor while using system storage in_memory '
'which does not persist intermediates. This means there would '
'be no way to move data between different processes. Please '
'configure your pipeline in the storage config section to use '
'persistent system storage such as the filesystem.') in str(
exc_info.value)
def test_execute_multi_mode_loggers_with_multiple_loggers_single_config():
pipeline_def, foo_logger_captured_results, bar_logger_captured_results = (
define_multi_mode_with_loggers_pipeline()
)
execute_pipeline(
pipeline_def,
run_config=RunConfig(mode='foo_bar_mode'),
environment_dict={'loggers': {'foo': {'config': {'log_level': 'DEBUG'}}}},
)
foo_original_messages = parse_captured_results(foo_logger_captured_results)
assert len(list(filter(lambda x: x == '"Here we are"', foo_original_messages))) == 1
assert not bar_logger_captured_results
def test_local():
result = execute_pipeline(
example_pipe,
environment_dict={
'solids': {
'blah': {
'config': {
'foo': 'a string',
'bar': 123
}
}
},
},
run_config=RunConfig(mode='local'),
)
assert result.success
def test_using_file_system_for_subplan_invalid_step():
pipeline = define_inty_pipeline()
environment_dict = {'storage': {'filesystem': {}}}
execution_plan = create_execution_plan(pipeline, environment_dict=environment_dict)
run_id = str(uuid.uuid4())
with pytest.raises(DagsterExecutionStepNotFoundError):
execute_plan(
execution_plan,
environment_dict=environment_dict,
run_config=RunConfig(run_id=run_id),
step_keys_to_execute=['nope'],
)
def test_airline_pipeline_1_warehouse(docker_compose_db):
warehouse_config_object = load_yaml_from_globs(
script_relative_path(
'../../dagster_examples/airline_demo/environments/local_base.yaml'
),
script_relative_path(
'../../dagster_examples/airline_demo/environments/local_warehouse.yaml'
),
)
warehouse_config_object = enviroment_overrides(warehouse_config_object)
result_warehouse = execute_pipeline(
define_airline_demo_warehouse_pipeline(),
warehouse_config_object,
run_config=RunConfig(mode='local'),
)
assert result_warehouse.success
def test_dask_cluster():
result = execute_on_dask(
ExecutionTargetHandle.for_pipeline_module(
'dagster_examples.toys.hammer', 'define_hammer_pipeline'),
env_config={
'storage': {
's3': {
's3_bucket': 'dagster-airflow-scratch'
}
}
},
run_config=RunConfig(storage_mode=RunStorageMode.S3),
dask_config=DaskConfig(address='%s:8786' % os.getenv('DASK_ADDRESS')),
)
assert result.success
assert result.result_for_solid('total').transformed_value() == 4
def test_error_resource(snapshot):
result = execute_pipeline(
resource_error_pipeline,
environment_dict={'storage': {
'filesystem': {}
}},
run_config=RunConfig(executor_config=InProcessExecutorConfig(
raise_on_error=False)),
)
assert not result.success
assert len(result.event_list) == 1
init_failure_event = result.event_list[0]
assert init_failure_event.event_type_value == 'PIPELINE_INIT_FAILURE'
snapshot.assert_match(init_failure_event.message)
def test_s3_object_store_with_composite_type_storage_plugin():
run_id = str(uuid.uuid4())
# FIXME need a dedicated test bucket
object_store = S3ObjectStore(
run_id=run_id,
s3_bucket='dagster-airflow-scratch',
types_to_register={String.inst(): FancyStringS3TypeStoragePlugin},
)
with yield_pipeline_execution_context(PipelineDefinition([]), {},
RunConfig(run_id=run_id)) as context:
with pytest.raises(check.NotImplementedCheckError):
object_store.set_value(['hello'], context,
resolve_to_runtime_type(List_(String_)),
['obj_name'])
def _create_pipeline_run(instance, pipeline, execution_params):
step_keys_to_execute = execution_params.step_keys
if not execution_params.step_keys and execution_params.previous_run_id:
execution_plan = create_execution_plan(
pipeline,
execution_params.environment_dict,
run_config=RunConfig(
mode=execution_params.mode,
previous_run_id=execution_params.previous_run_id,
tags=execution_params.execution_metadata.tags,
),
)
step_keys_to_execute = get_retry_steps_from_execution_plan(
instance, execution_plan)
return pipeline_run_from_execution_params(execution_params,
step_keys_to_execute)
def test_using_file_system_for_subplan_missing_input():
pipeline = define_inty_pipeline()
environment_dict = {'storage': {'filesystem': {}}}
execution_plan = create_execution_plan(pipeline, environment_dict=environment_dict)
run_id = str(uuid.uuid4())
with pytest.raises(DagsterStepOutputNotFoundError):
execute_plan(
execution_plan,
DagsterInstance.ephemeral(),
environment_dict=environment_dict,
run_config=RunConfig(run_id=run_id),
step_keys_to_execute=['add_one.compute'],
)
def get_pipeline_run_observable(graphene_info, run_id, after=None):
check.inst_param(graphene_info, 'graphene_info', ResolveInfo)
check.str_param(run_id, 'run_id')
check.opt_int_param(after, 'after')
instance = graphene_info.context.instance
run = instance.get_run(run_id)
if not run:
def _get_error_observable(observer):
observer.on_next(
graphene_info.schema.
type_named('PipelineRunLogsSubscriptionFailure')(
missingRunId=run_id,
message='Could not load run with id {}'.format(run_id)))
return Observable.create(_get_error_observable) # pylint: disable=E1101
if not instance.can_watch_events:
def _get_error_observable(observer):
observer.on_next(
graphene_info.schema.type_named(
'PipelineRunLogsSubscriptionFailure')
(message=
'Event log storage on current DagsterInstance is not watchable.'
))
return Observable.create(_get_error_observable) # pylint: disable=E1101
pipeline = get_dauphin_pipeline_from_selector(graphene_info, run.selector)
execution_plan = create_execution_plan(pipeline.get_dagster_pipeline(),
run.environment_dict,
RunConfig(mode=run.mode))
# pylint: disable=E1101
return Observable.create(
PipelineRunObservableSubscribe(
instance, run_id, after_cursor=after)).map(
lambda events: graphene_info.schema.
type_named('PipelineRunLogsSubscriptionSuccess')(
runId=run_id,
messages=[
from_event_record(graphene_info, event, pipeline,
execution_plan) for event in events
],
))
