def test_coverage_transform(self):
ph = ParameterHelper(self.dataset_management, self.addCleanup)
pdict_id = ph.create_parsed()
stream_def_id = self.pubsub_management.create_stream_definition(
'ctd parsed', parameter_dictionary_id=pdict_id)
self.addCleanup(self.pubsub_management.delete_stream_definition,
stream_def_id)
stream_id, route = self.pubsub_management.create_stream(
'example',
exchange_point=self.exchange_point_name,
stream_definition_id=stream_def_id)
self.addCleanup(self.pubsub_management.delete_stream, stream_id)
ingestion_config_id = self.get_ingestion_config()
dataset_id = self.create_dataset(pdict_id)
self.ingestion_management.persist_data_stream(
stream_id=stream_id,
ingestion_configuration_id=ingestion_config_id,
dataset_id=dataset_id)
self.addCleanup(self.ingestion_management.unpersist_data_stream,
stream_id, ingestion_config_id)
publisher = StandaloneStreamPublisher(stream_id, route)
rdt = ph.get_rdt(stream_def_id)
ph.fill_parsed_rdt(rdt)
dataset_monitor = DatasetMonitor(dataset_id)
self.addCleanup(dataset_monitor.stop)
publisher.publish(rdt.to_granule())
self.assertTrue(dataset_monitor.event.wait(30))
replay_granule = self.data_retriever.retrieve(dataset_id)
rdt_out = RecordDictionaryTool.load_from_granule(replay_granule)
np.testing.assert_array_almost_equal(rdt_out['time'], rdt['time'])
np.testing.assert_array_almost_equal(rdt_out['temp'], rdt['temp'])
np.testing.assert_array_almost_equal(rdt_out['conductivity_L1'],
np.array([42.914]))
np.testing.assert_array_almost_equal(rdt_out['temp_L1'],
np.array([20.]))
np.testing.assert_array_almost_equal(rdt_out['pressure_L1'],
np.array([3.068]))
np.testing.assert_array_almost_equal(
rdt_out['density'], np.array([1021.7144739593881],
dtype='float32'))
np.testing.assert_array_almost_equal(
rdt_out['salinity'], np.array([30.935132729668283],
dtype='float32'))
def test_coverage_transform(self):
ph = ParameterHelper(self.dataset_management, self.addCleanup)
pdict_id = ph.create_parsed()
stream_def_id = self.pubsub_management.create_stream_definition('ctd parsed', parameter_dictionary_id=pdict_id)
self.addCleanup(self.pubsub_management.delete_stream_definition, stream_def_id)
stream_id, route = self.pubsub_management.create_stream('example', exchange_point=self.exchange_point_name, stream_definition_id=stream_def_id)
self.addCleanup(self.pubsub_management.delete_stream, stream_id)
ingestion_config_id = self.get_ingestion_config()
dataset_id = self.create_dataset(pdict_id)
self.ingestion_management.persist_data_stream(stream_id=stream_id, ingestion_configuration_id=ingestion_config_id, dataset_id=dataset_id)
self.addCleanup(self.ingestion_management.unpersist_data_stream, stream_id, ingestion_config_id)
publisher = StandaloneStreamPublisher(stream_id, route)
rdt = ph.get_rdt(stream_def_id)
ph.fill_parsed_rdt(rdt)
dataset_monitor = DatasetMonitor(dataset_id)
self.addCleanup(dataset_monitor.stop)
publisher.publish(rdt.to_granule())
self.assertTrue(dataset_monitor.wait())
replay_granule = self.data_retriever.retrieve(dataset_id)
rdt_out = RecordDictionaryTool.load_from_granule(replay_granule)
np.testing.assert_array_almost_equal(rdt_out['time'], rdt['time'])
np.testing.assert_array_almost_equal(rdt_out['temp'], rdt['temp'])
np.testing.assert_allclose(rdt_out['conductivity_L1'], np.array([42.914]))
np.testing.assert_allclose(rdt_out['temp_L1'], np.array([20.]))
np.testing.assert_allclose(rdt_out['pressure_L1'], np.array([3.068]))
np.testing.assert_allclose(rdt_out['density'], np.array([1021.7144739593881], dtype='float32'))
np.testing.assert_allclose(rdt_out['salinity'], np.array([30.935132729668283], dtype='float32'))
class TestTransformWorker(IonIntegrationTestCase):
def setUp(self):
self._start_container()
self.container.start_rel_from_url('res/deploy/r2deploy.yml')
# Instantiate a process to represent the test
process=TransformWorkerTestProcess()
self.dataset_management_client = DatasetManagementServiceClient(node=self.container.node)
self.pubsub_client = PubsubManagementServiceClient(node=self.container.node)
self.dataproductclient = DataProductManagementServiceClient(node=self.container.node)
self.dataprocessclient = DataProcessManagementServiceClient(node=self.container.node)
self.processdispatchclient = ProcessDispatcherServiceClient(node=self.container.node)
self.damsclient = DataAcquisitionManagementServiceClient(node=self.container.node)
self.rrclient = ResourceRegistryServiceClient(node=self.container.node)
self.imsclient = InstrumentManagementServiceProcessClient(node=self.container.node, process = process)
self.time_dom, self.spatial_dom = time_series_domain()
self.ph = ParameterHelper(self.dataset_management_client, self.addCleanup)
self.wait_time = CFG.get_safe('endpoint.receive.timeout', 10)
def push_granule(self, data_product_id):
'''
Publishes and monitors that the granule arrived
'''
datasets, _ = self.rrclient.find_objects(data_product_id, PRED.hasDataset, id_only=True)
dataset_monitor = DatasetMonitor(datasets[0])
rdt = self.ph.rdt_for_data_product(data_product_id)
self.ph.fill_parsed_rdt(rdt)
self.ph.publish_rdt_to_data_product(data_product_id, rdt)
assert dataset_monitor.wait()
dataset_monitor.stop()
@attr('LOCOINT')
@unittest.skipIf(os.getenv('CEI_LAUNCH_TEST', False), 'Skip test while in CEI LAUNCH mode')
def test_transform_worker(self):
# test that a data process (type: data-product-in / data-product-out) can be defined and launched.
# verify that the output granule fields are correctly populated
# test that the input and output data products are linked to facilitate provenance
self.dp_list = []
self.data_process_objs = []
self._output_stream_ids = []
self.granule_verified = Event()
self.worker_assigned_event_verified = Event()
self.dp_created_event_verified = Event()
self.heartbeat_event_verified = Event()
self.parameter_dict_id = self.dataset_management_client.read_parameter_dictionary_by_name(name='ctd_parsed_param_dict', id_only=True)
# create the StreamDefinition
self.stream_def_id = self.pubsub_client.create_stream_definition(name='stream_def', parameter_dictionary_id=self.parameter_dict_id)
self.addCleanup(self.pubsub_client.delete_stream_definition, self.stream_def_id)
# create the DataProduct that is the input to the data processes
input_dp_obj = IonObject( RT.DataProduct, name='input_data_product', description='input test stream',
temporal_domain = self.time_dom.dump(), spatial_domain = self.spatial_dom.dump())
self.input_dp_id = self.dataproductclient.create_data_product(data_product=input_dp_obj, stream_definition_id=self.stream_def_id)
# retrieve the Stream for this data product
stream_ids, assoc_ids = self.rrclient.find_objects(self.input_dp_id, PRED.hasStream, RT.Stream, True)
self.stream_id = stream_ids[0]
self.start_event_listener()
# create the DPD, DataProcess and output DataProduct
dataprocessdef_id, dataprocess_id, dataproduct_id = self.create_data_process()
self.dp_list.append(dataprocess_id)
# validate the repository for data product algorithms persists the new resources NEW SA-1
# create_data_process call created one of each
dpd_ids, _ = self.rrclient.find_resources(restype=OT.DataProcessDefinition, id_only=False)
# there will be more than one becuase of the DPDs that reperesent the PFs in the data product above
self.assertTrue(dpd_ids is not None)
dp_ids, _ = self.rrclient.find_resources(restype=OT.DataProcess, id_only=False)
# only one DP becuase the PFs that are in the code dataproduct above are not activated yet.
self.assertEquals(len(dp_ids), 1)
# validate the name and version label NEW SA - 2
dataprocessdef_obj = self.dataprocessclient.read_data_process_definition(dataprocessdef_id)
self.assertEqual(dataprocessdef_obj.version_label, '1.0a')
self.assertEqual(dataprocessdef_obj.name, 'add_arrays')
# validate that the DPD has an attachment NEW SA - 21
attachment_ids, assoc_ids = self.rrclient.find_objects(dataprocessdef_id, PRED.hasAttachment, RT.Attachment, True)
self.assertEqual(len(attachment_ids), 1)
attachment_obj = self.rrclient.read_attachment(attachment_ids[0])
log.debug('attachment: %s', attachment_obj)
# validate that the data process resource has input and output data products associated
# L4-CI-SA-RQ-364 and NEW SA-3
outproduct_ids, assoc_ids = self.rrclient.find_objects(dataprocess_id, PRED.hasOutputProduct, RT.DataProduct, True)
self.assertEqual(len(outproduct_ids), 1)
inproduct_ids, assoc_ids = self.rrclient.find_objects(dataprocess_id, PRED.hasInputProduct, RT.DataProduct, True)
self.assertEqual(len(inproduct_ids), 1)
# Test for provenance. Get Data product produced by the data processes
output_data_product_id,_ = self.rrclient.find_objects(subject=dataprocess_id,
object_type=RT.DataProduct,
predicate=PRED.hasOutputProduct,
id_only=True)
output_data_product_provenance = self.dataproductclient.get_data_product_provenance(output_data_product_id[0])
# Do a basic check to see if there were 3 entries in the provenance graph. Parent and Child and the
# DataProcessDefinition creating the child from the parent.
self.assertTrue(len(output_data_product_provenance) == 2)
self.assertTrue(self.input_dp_id in output_data_product_provenance[output_data_product_id[0]]['parents'])
self.assertTrue(output_data_product_provenance[output_data_product_id[0]]['parents'][self.input_dp_id]['data_process_definition_id'] == dataprocessdef_id)
# NEW SA - 4 | Data processing shall include the appropriate data product algorithm name and version number in
# the metadata of each output data product created by the data product algorithm.
output_data_product_obj,_ = self.rrclient.find_objects(subject=dataprocess_id,
object_type=RT.DataProduct,
predicate=PRED.hasOutputProduct,
id_only=False)
self.assertTrue(output_data_product_obj[0].name != None)
self.assertTrue(output_data_product_obj[0]._rev != None)
# retrieve subscription from data process
subscription_objs, _ = self.rrclient.find_objects(subject=dataprocess_id, predicate=PRED.hasSubscription, object_type=RT.Subscription, id_only=False)
log.debug('test_transform_worker subscription_obj: %s', subscription_objs[0])
# create a queue to catch the published granules
self.subscription_id = self.pubsub_client.create_subscription(name='parsed_subscription', stream_ids=[self.stream_id], exchange_name=subscription_objs[0].exchange_name)
self.addCleanup(self.pubsub_client.delete_subscription, self.subscription_id)
self.pubsub_client.activate_subscription(self.subscription_id)
self.addCleanup(self.pubsub_client.deactivate_subscription, self.subscription_id)
stream_route = self.pubsub_client.read_stream_route(self.stream_id)
self.publisher = StandaloneStreamPublisher(stream_id=self.stream_id, stream_route=stream_route )
for n in range(1, 101):
rdt = RecordDictionaryTool(stream_definition_id=self.stream_def_id)
rdt['time'] = [0] # time should always come first
rdt['conductivity'] = [1]
rdt['pressure'] = [2]
rdt['salinity'] = [8]
self.publisher.publish(rdt.to_granule())
# validate that the output granule is received and the updated value is correct
self.assertTrue(self.granule_verified.wait(self.wait_time))
# validate that the data process loaded into worker event is received (L4-CI-SA-RQ-182)
self.assertTrue(self.worker_assigned_event_verified.wait(self.wait_time))
# validate that the data process create (with data product ids) event is received (NEW SA -42)
self.assertTrue(self.dp_created_event_verified.wait(self.wait_time))
# validate that the data process heartbeat event is received (for every hundred granules processed) (L4-CI-SA-RQ-182)
#this takes a while so set wait limit to large value
self.assertTrue(self.heartbeat_event_verified.wait(200))
# validate that the code from the transform function can be retrieve via inspect_data_process_definition
src = self.dataprocessclient.inspect_data_process_definition(dataprocessdef_id)
self.assertIn( 'def add_arrays(a, b)', src)
# now delete the DPD and DP then verify that the resources are retired so that information required for provenance are still available
self.dataprocessclient.delete_data_process(dataprocess_id)
self.dataprocessclient.delete_data_process_definition(dataprocessdef_id)
in_dp_objs, _ = self.rrclient.find_objects(subject=dataprocess_id, predicate=PRED.hasInputProduct, object_type=RT.DataProduct, id_only=True)
self.assertTrue(in_dp_objs is not None)
dpd_objs, _ = self.rrclient.find_subjects(subject_type=RT.DataProcessDefinition, predicate=PRED.hasDataProcess, object=dataprocess_id, id_only=True)
self.assertTrue(dpd_objs is not None)
@attr('LOCOINT')
@unittest.skipIf(os.getenv('CEI_LAUNCH_TEST', False), 'Skip test while in CEI LAUNCH mode')
def test_transform_worker_with_instrumentdevice(self):
# test that a data process (type: data-product-in / data-product-out) can be defined and launched.
# verify that the output granule fields are correctly populated
# test that the input and output data products are linked to facilitate provenance
self.data_process_objs = []
self._output_stream_ids = []
self.event_verified = Event()
# Create CTD Parsed as the initial data product
# create a stream definition for the data from the ctd simulator
self.parameter_dict_id = self.dataset_management_client.read_parameter_dictionary_by_name('ctd_parsed_param_dict', id_only=True)
self.stream_def_id = self.pubsub_client.create_stream_definition(name='stream_def', parameter_dictionary_id=self.parameter_dict_id)
# create the DataProduct that is the input to the data processes
input_dp_obj = IonObject( RT.DataProduct, name='input_data_product', description='input test stream',
temporal_domain = self.time_dom.dump(), spatial_domain = self.spatial_dom.dump())
self.input_dp_id = self.dataproductclient.create_data_product(data_product=input_dp_obj, stream_definition_id=self.stream_def_id)
# retrieve the Stream for this data product
stream_ids, assoc_ids = self.rrclient.find_objects(self.input_dp_id, PRED.hasStream, RT.Stream, True)
self.stream_id = stream_ids[0]
log.debug('new ctd_parsed_data_product_id = %s' % self.input_dp_id)
# only ever need one device for testing purposes.
instDevice_obj,_ = self.rrclient.find_resources(restype=RT.InstrumentDevice, name='test_ctd_device')
if instDevice_obj:
instDevice_id = instDevice_obj[0]._id
else:
instDevice_obj = IonObject(RT.InstrumentDevice, name='test_ctd_device', description="test_ctd_device", serial_number="12345" )
instDevice_id = self.imsclient.create_instrument_device(instrument_device=instDevice_obj)
self.damsclient.assign_data_product(input_resource_id=instDevice_id, data_product_id=self.input_dp_id)
# create the DPD, DataProcess and output DataProduct
dataprocessdef_id, dataprocess_id, dataproduct_id = self.create_data_process()
self.addCleanup(self.dataprocessclient.delete_data_process, dataprocess_id)
self.addCleanup(self.dataprocessclient.delete_data_process_definition, dataprocessdef_id)
# Test for provenance. Get Data product produced by the data processes
output_data_product_id,_ = self.rrclient.find_objects(subject=dataprocess_id,
object_type=RT.DataProduct,
predicate=PRED.hasOutputProduct,
id_only=True)
output_data_product_provenance = self.dataproductclient.get_data_product_provenance(output_data_product_id[0])
# Do a basic check to see if there were 3 entries in the provenance graph. Parent and Child and the
# DataProcessDefinition creating the child from the parent.
self.assertTrue(len(output_data_product_provenance) == 3)
self.assertTrue(self.input_dp_id in output_data_product_provenance[output_data_product_id[0]]['parents'])
self.assertTrue(instDevice_id in output_data_product_provenance[self.input_dp_id]['parents'])
self.assertTrue(output_data_product_provenance[instDevice_id]['type'] == 'InstrumentDevice')
@attr('LOCOINT')
@unittest.skipIf(os.getenv('CEI_LAUNCH_TEST', False), 'Skip test while in CEI LAUNCH mode')
def test_transform_worker_with_platformdevice(self):
# test that a data process (type: data-product-in / data-product-out) can be defined and launched.
# verify that the output granule fields are correctly populated
# test that the input and output data products are linked to facilitate provenance
self.data_process_objs = []
self._output_stream_ids = []
self.event_verified = Event()
# Create CTD Parsed as the initial data product
# create a stream definition for the data from the ctd simulator
self.parameter_dict_id = self.dataset_management_client.read_parameter_dictionary_by_name('ctd_parsed_param_dict', id_only=True)
self.stream_def_id = self.pubsub_client.create_stream_definition(name='stream_def', parameter_dictionary_id=self.parameter_dict_id)
# create the DataProduct that is the input to the data processes
input_dp_obj = IonObject( RT.DataProduct, name='input_data_product', description='input test stream',
temporal_domain = self.time_dom.dump(), spatial_domain = self.spatial_dom.dump())
self.input_dp_id = self.dataproductclient.create_data_product(data_product=input_dp_obj, stream_definition_id=self.stream_def_id)
# retrieve the Stream for this data product
stream_ids, assoc_ids = self.rrclient.find_objects(self.input_dp_id, PRED.hasStream, RT.Stream, True)
self.stream_id = stream_ids[0]
log.debug('new ctd_parsed_data_product_id = %s' % self.input_dp_id)
# only ever need one device for testing purposes.
platform_device_obj,_ = self.rrclient.find_resources(restype=RT.PlatformDevice, name='TestPlatform')
if platform_device_obj:
platform_device_id = platform_device_obj[0]._id
else:
platform_device_obj = IonObject(RT.PlatformDevice, name='TestPlatform', description="TestPlatform", serial_number="12345" )
platform_device_id = self.imsclient.create_platform_device(platform_device=platform_device_obj)
self.damsclient.assign_data_product(input_resource_id=platform_device_id, data_product_id=self.input_dp_id)
# create the DPD, DataProcess and output DataProduct
dataprocessdef_id, dataprocess_id, dataproduct_id = self.create_data_process()
self.addCleanup(self.dataprocessclient.delete_data_process, dataprocess_id)
self.addCleanup(self.dataprocessclient.delete_data_process_definition, dataprocessdef_id)
# Test for provenance. Get Data product produced by the data processes
output_data_product_id,_ = self.rrclient.find_objects(subject=dataprocess_id,
object_type=RT.DataProduct,
predicate=PRED.hasOutputProduct,
id_only=True)
output_data_product_provenance = self.dataproductclient.get_data_product_provenance(output_data_product_id[0])
# Do a basic check to see if there were 3 entries in the provenance graph. Parent and Child and the
# DataProcessDefinition creating the child from the parent.
self.assertTrue(len(output_data_product_provenance) == 3)
self.assertTrue(self.input_dp_id in output_data_product_provenance[output_data_product_id[0]]['parents'])
self.assertTrue(platform_device_id in output_data_product_provenance[self.input_dp_id]['parents'])
self.assertTrue(output_data_product_provenance[platform_device_id]['type'] == 'PlatformDevice')
@attr('LOCOINT')
@unittest.skipIf(os.getenv('CEI_LAUNCH_TEST', False), 'Skip test while in CEI LAUNCH mode')
def test_event_transform_worker(self):
self.data_process_objs = []
self._output_stream_ids = []
self.event_verified = Event()
# test that a data process (type: data-product-in / event-out) can be defined and launched.
# verify that event fields are correctly populated
self.parameter_dict_id = self.dataset_management_client.read_parameter_dictionary_by_name(name='ctd_parsed_param_dict', id_only=True)
# create the StreamDefinition
self.stream_def_id = self.pubsub_client.create_stream_definition(name='stream_def', parameter_dictionary_id=self.parameter_dict_id)
self.addCleanup(self.pubsub_client.delete_stream_definition, self.stream_def_id)
# create the DataProduct
input_dp_obj = IonObject( RT.DataProduct, name='input_data_product', description='input test stream',
temporal_domain = self.time_dom.dump(), spatial_domain = self.spatial_dom.dump())
self.input_dp_id = self.dataproductclient.create_data_product(data_product=input_dp_obj, stream_definition_id=self.stream_def_id)
# retrieve the Stream for this data product
stream_ids, assoc_ids = self.rrclient.find_objects(self.input_dp_id, PRED.hasStream, RT.Stream, True)
self.stream_id = stream_ids[0]
# create the DPD and two DPs
self.event_data_process_id = self.create_event_data_processes()
# retrieve subscription from data process
subscription_objs, _ = self.rrclient.find_objects(subject=self.event_data_process_id, predicate=PRED.hasSubscription, object_type=RT.Subscription, id_only=False)
log.debug('test_event_transform_worker subscription_obj: %s', subscription_objs[0])
# create a queue to catch the published granules
self.subscription_id = self.pubsub_client.create_subscription(name='parsed_subscription', stream_ids=[self.stream_id], exchange_name=subscription_objs[0].exchange_name)
self.addCleanup(self.pubsub_client.delete_subscription, self.subscription_id)
self.pubsub_client.activate_subscription(self.subscription_id)
self.addCleanup(self.pubsub_client.deactivate_subscription, self.subscription_id)
stream_route = self.pubsub_client.read_stream_route(self.stream_id)
self.publisher = StandaloneStreamPublisher(stream_id=self.stream_id, stream_route=stream_route )
self.start_event_transform_listener()
self.data_modified = Event()
rdt = RecordDictionaryTool(stream_definition_id=self.stream_def_id)
rdt['time'] = [0] # time should always come first
rdt['conductivity'] = [1]
rdt['pressure'] = [2]
rdt['salinity'] = [8]
self.publisher.publish(rdt.to_granule())
self.assertTrue(self.event_verified.wait(self.wait_time))
@attr('LOCOINT')
@unittest.skipIf(os.getenv('CEI_LAUNCH_TEST', False), 'Skip test while in CEI LAUNCH mode')
def test_bad_argument_map(self):
self._output_stream_ids = []
# test that a data process (type: data-product-in / data-product-out) parameter mapping it validated during
# data process creation and that the correct exception is raised for both input and output.
self.parameter_dict_id = self.dataset_management_client.read_parameter_dictionary_by_name(name='ctd_parsed_param_dict', id_only=True)
# create the StreamDefinition
self.stream_def_id = self.pubsub_client.create_stream_definition(name='stream_def', parameter_dictionary_id=self.parameter_dict_id)
self.addCleanup(self.pubsub_client.delete_stream_definition, self.stream_def_id)
# create the DataProduct that is the input to the data processes
input_dp_obj = IonObject( RT.DataProduct, name='input_data_product', description='input test stream',
temporal_domain = self.time_dom.dump(), spatial_domain = self.spatial_dom.dump())
self.input_dp_id = self.dataproductclient.create_data_product(data_product=input_dp_obj, stream_definition_id=self.stream_def_id)
# two data processes using one transform and one DPD
dp1_func_output_dp_id = self.create_output_data_product()
# Set up DPD and DP #2 - array add function
tf_obj = IonObject(RT.TransformFunction,
name='add_array_func',
description='adds values in an array',
function='add_arrays',
module="ion_example.add_arrays",
arguments=['arr1', 'arr2'],
function_type=TransformFunctionType.TRANSFORM,
uri='http://sddevrepo.oceanobservatories.org/releases/ion_example-0.1-py2.7.egg'
)
add_array_func_id, rev = self.rrclient.create(tf_obj)
dpd_obj = IonObject(RT.DataProcessDefinition,
name='add_arrays',
description='adds the values of two arrays',
data_process_type=DataProcessTypeEnum.TRANSFORM_PROCESS
)
add_array_dpd_id = self.dataprocessclient.create_data_process_definition(data_process_definition=dpd_obj, function_id=add_array_func_id)
self.dataprocessclient.assign_stream_definition_to_data_process_definition(self.stream_def_id, add_array_dpd_id, binding='add_array_func' )
# create the data process with invalid argument map
argument_map = {"arr1": "foo", "arr2": "bar"}
output_param = "salinity"
with self.assertRaises(BadRequest) as cm:
dp1_data_process_id = self.dataprocessclient.create_data_process(data_process_definition_id=add_array_dpd_id, inputs=[self.input_dp_id],
outputs=[dp1_func_output_dp_id], argument_map=argument_map, out_param_name=output_param)
ex = cm.exception
log.debug(' exception raised: %s', cm)
self.assertEqual(ex.message, "Input data product does not contain the parameters defined in argument map")
# create the data process with invalid output parameter name
argument_map = {"arr1": "conductivity", "arr2": "pressure"}
output_param = "foo"
with self.assertRaises(BadRequest) as cm:
dp1_data_process_id = self.dataprocessclient.create_data_process(data_process_definition_id=add_array_dpd_id, inputs=[self.input_dp_id],
outputs=[dp1_func_output_dp_id], argument_map=argument_map, out_param_name=output_param)
ex = cm.exception
log.debug(' exception raised: %s', cm)
self.assertEqual(ex.message, "Output data product does not contain the output parameter name provided")
def create_event_data_processes(self):
# two data processes using one transform and one DPD
argument_map= {"a": "salinity"}
# set up DPD and DP #2 - array add function
tf_obj = IonObject(RT.TransformFunction,
name='validate_salinity_array',
description='validate_salinity_array',
function='validate_salinity_array',
module="ion.processes.data.transforms.test.test_transform_worker",
arguments=['a'],
function_type=TransformFunctionType.TRANSFORM
)
add_array_func_id, rev = self.rrclient.create(tf_obj)
dpd_obj = IonObject(RT.DataProcessDefinition,
name='validate_salinity_array',
description='validate_salinity_array',
data_process_type=DataProcessTypeEnum.TRANSFORM_PROCESS,
)
add_array_dpd_id = self.dataprocessclient.create_data_process_definition(data_process_definition=dpd_obj, function_id=add_array_func_id)
self.dataprocessclient.assign_stream_definition_to_data_process_definition(self.stream_def_id, add_array_dpd_id, binding='validate_salinity_array' )
# create the data process
dp1_data_process_id = self.dataprocessclient.create_data_process(data_process_definition_id=add_array_dpd_id, inputs=[self.input_dp_id],
outputs=None, argument_map=argument_map)
self.damsclient.register_process(dp1_data_process_id)
self.addCleanup(self.dataprocessclient.delete_data_process, dp1_data_process_id)
return dp1_data_process_id
def create_data_process(self):
# two data processes using one transform and one DPD
dp1_func_output_dp_id = self.create_output_data_product()
argument_map = {"arr1": "conductivity", "arr2": "pressure"}
output_param = "salinity"
# set up DPD and DP #2 - array add function
tf_obj = IonObject(RT.TransformFunction,
name='add_array_func',
description='adds values in an array',
function='add_arrays',
module="ion_example.add_arrays",
arguments=['arr1', 'arr2'],
function_type=TransformFunctionType.TRANSFORM,
uri='http://sddevrepo.oceanobservatories.org/releases/ion_example-0.1-py2.7.egg'
)
add_array_func_id, rev = self.rrclient.create(tf_obj)
dpd_obj = IonObject(RT.DataProcessDefinition,
name='add_arrays',
description='adds the values of two arrays',
data_process_type=DataProcessTypeEnum.TRANSFORM_PROCESS,
version_label='1.0a'
)
add_array_dpd_id = self.dataprocessclient.create_data_process_definition(data_process_definition=dpd_obj, function_id=add_array_func_id)
self.dataprocessclient.assign_stream_definition_to_data_process_definition(self.stream_def_id, add_array_dpd_id, binding='add_array_func' )
# create the data process
dp1_data_process_id = self.dataprocessclient.create_data_process(data_process_definition_id=add_array_dpd_id, inputs=[self.input_dp_id],
outputs=[dp1_func_output_dp_id], argument_map=argument_map, out_param_name=output_param)
self.damsclient.register_process(dp1_data_process_id)
#self.addCleanup(self.dataprocessclient.delete_data_process, dp1_data_process_id)
# add an attachment object to this DPD to test new SA-21
import msgpack
attachment_content = 'foo bar'
attachment_obj = IonObject( RT.Attachment,
name='test_attachment',
attachment_type=AttachmentType.ASCII,
content_type='text/plain',
content=msgpack.packb(attachment_content))
att_id = self.rrclient.create_attachment(add_array_dpd_id, attachment_obj)
self.addCleanup(self.rrclient.delete_attachment, att_id)
return add_array_dpd_id, dp1_data_process_id, dp1_func_output_dp_id
def create_output_data_product(self):
dp1_outgoing_stream_id = self.pubsub_client.create_stream_definition(name='dp1_stream', parameter_dictionary_id=self.parameter_dict_id)
dp1_output_dp_obj = IonObject( RT.DataProduct,
name='data_process1_data_product',
description='output of add array func',
temporal_domain = self.time_dom.dump(),
spatial_domain = self.spatial_dom.dump())
dp1_func_output_dp_id = self.dataproductclient.create_data_product(dp1_output_dp_obj, dp1_outgoing_stream_id)
self.addCleanup(self.dataproductclient.delete_data_product, dp1_func_output_dp_id)
# retrieve the id of the OUTPUT stream from the out Data Product and add to granule logger
stream_ids, _ = self.rrclient.find_objects(dp1_func_output_dp_id, PRED.hasStream, None, True)
self._output_stream_ids.append(stream_ids[0])
subscription_id = self.pubsub_client.create_subscription('validator', data_product_ids=[dp1_func_output_dp_id])
self.addCleanup(self.pubsub_client.delete_subscription, subscription_id)
def on_granule(msg, route, stream_id):
log.debug('recv_packet stream_id: %s route: %s msg: %s', stream_id, route, msg)
self.validate_output_granule(msg, route, stream_id)
self.granule_verified.set()
validator = StandaloneStreamSubscriber('validator', callback=on_granule)
validator.start()
self.addCleanup(validator.stop)
self.pubsub_client.activate_subscription(subscription_id)
self.addCleanup(self.pubsub_client.deactivate_subscription, subscription_id)
return dp1_func_output_dp_id
def validate_event(self, *args, **kwargs):
"""
This method is a callback function for receiving DataProcessStatusEvent.
"""
data_process_event = args[0]
log.debug("DataProcessStatusEvent: %s" , str(data_process_event.__dict__))
# if data process already created, check origin
if self.dp_list:
self.assertIn( data_process_event.origin, self.dp_list)
# if this is a heartbeat event then 100 granules have been processed
if 'data process status update.' in data_process_event.description:
self.heartbeat_event_verified.set()
else:
# else check that this is the assign event
if 'Data process assigned to transform worker' in data_process_event.description:
self.worker_assigned_event_verified.set()
elif 'Data process created for data product' in data_process_event.description:
self.dp_created_event_verified.set()
def validate_output_granule(self, msg, route, stream_id):
self.assertIn( stream_id, self._output_stream_ids)
rdt = RecordDictionaryTool.load_from_granule(msg)
log.debug('validate_output_granule rdt: %s', rdt)
sal_val = rdt['salinity']
np.testing.assert_array_equal(sal_val, np.array([3]))
def start_event_listener(self):
es = EventSubscriber(event_type=OT.DataProcessStatusEvent, callback=self.validate_event)
es.start()
self.addCleanup(es.stop)
def validate_transform_event(self, *args, **kwargs):
"""
This method is a callback function for receiving DataProcessStatusEvent.
"""
status_alert_event = args[0]
np.testing.assert_array_equal(status_alert_event.origin, self.stream_id )
np.testing.assert_array_equal(status_alert_event.values, np.array([self.event_data_process_id]))
log.debug("DeviceStatusAlertEvent: %s" , str(status_alert_event.__dict__))
self.event_verified.set()
def start_event_transform_listener(self):
es = EventSubscriber(event_type=OT.DeviceStatusAlertEvent, callback=self.validate_transform_event)
es.start()
self.addCleanup(es.stop)
def test_download(self):
egg_url = 'http://sddevrepo.oceanobservatories.org/releases/ion_example-0.1-py2.7.egg'
egg_path = TransformWorker.download_egg(egg_url)
import pkg_resources
pkg_resources.working_set.add_entry(egg_path)
from ion_example.add_arrays import add_arrays
a = add_arrays(1,2)
self.assertEquals(a,3)
class TestTransformWorker(IonIntegrationTestCase):
def setUp(self):
self._start_container()
self.container.start_rel_from_url('res/deploy/r2deploy.yml')
# Instantiate a process to represent the test
process = TransformWorkerTestProcess()
self.dataset_management_client = DatasetManagementServiceClient(
node=self.container.node)
self.pubsub_client = PubsubManagementServiceClient(
node=self.container.node)
self.dataproductclient = DataProductManagementServiceClient(
node=self.container.node)
self.dataprocessclient = DataProcessManagementServiceClient(
node=self.container.node)
self.processdispatchclient = ProcessDispatcherServiceClient(
node=self.container.node)
self.damsclient = DataAcquisitionManagementServiceClient(
node=self.container.node)
self.rrclient = ResourceRegistryServiceClient(node=self.container.node)
self.imsclient = InstrumentManagementServiceProcessClient(
node=self.container.node, process=process)
self.time_dom, self.spatial_dom = time_series_domain()
self.ph = ParameterHelper(self.dataset_management_client,
self.addCleanup)
self.wait_time = CFG.get_safe('endpoint.receive.timeout', 10)
def push_granule(self, data_product_id):
'''
Publishes and monitors that the granule arrived
'''
datasets, _ = self.rrclient.find_objects(data_product_id,
PRED.hasDataset,
id_only=True)
dataset_monitor = DatasetMonitor(datasets[0])
rdt = self.ph.rdt_for_data_product(data_product_id)
self.ph.fill_parsed_rdt(rdt)
self.ph.publish_rdt_to_data_product(data_product_id, rdt)
assert dataset_monitor.wait()
dataset_monitor.stop()
@attr('LOCOINT')
@unittest.skipIf(os.getenv('CEI_LAUNCH_TEST', False),
'Skip test while in CEI LAUNCH mode')
def test_transform_worker(self):
# test that a data process (type: data-product-in / data-product-out) can be defined and launched.
# verify that the output granule fields are correctly populated
# test that the input and output data products are linked to facilitate provenance
self.dp_list = []
self.data_process_objs = []
self._output_stream_ids = []
self.granule_verified = Event()
self.worker_assigned_event_verified = Event()
self.dp_created_event_verified = Event()
self.heartbeat_event_verified = Event()
self.parameter_dict_id = self.dataset_management_client.read_parameter_dictionary_by_name(
name='ctd_parsed_param_dict', id_only=True)
# create the StreamDefinition
self.stream_def_id = self.pubsub_client.create_stream_definition(
name='stream_def', parameter_dictionary_id=self.parameter_dict_id)
self.addCleanup(self.pubsub_client.delete_stream_definition,
self.stream_def_id)
# create the DataProduct that is the input to the data processes
input_dp_obj = IonObject(RT.DataProduct,
name='input_data_product',
description='input test stream')
self.input_dp_id = self.dataproductclient.create_data_product(
data_product=input_dp_obj, stream_definition_id=self.stream_def_id)
# retrieve the Stream for this data product
stream_ids, assoc_ids = self.rrclient.find_objects(
self.input_dp_id, PRED.hasStream, RT.Stream, True)
self.stream_id = stream_ids[0]
self.start_event_listener()
# create the DPD, DataProcess and output DataProduct
dataprocessdef_id, dataprocess_id, dataproduct_id = self.create_data_process(
)
self.dp_list.append(dataprocess_id)
# validate the repository for data product algorithms persists the new resources NEW SA-1
# create_data_process call created one of each
dpd_ids, _ = self.rrclient.find_resources(
restype=OT.DataProcessDefinition, id_only=False)
# there will be more than one becuase of the DPDs that reperesent the PFs in the data product above
self.assertTrue(dpd_ids is not None)
dp_ids, _ = self.rrclient.find_resources(restype=OT.DataProcess,
id_only=False)
# only one DP becuase the PFs that are in the code dataproduct above are not activated yet.
self.assertEquals(len(dp_ids), 1)
# validate the name and version label NEW SA - 2
dataprocessdef_obj = self.dataprocessclient.read_data_process_definition(
dataprocessdef_id)
self.assertEqual(dataprocessdef_obj.version_label, '1.0a')
self.assertEqual(dataprocessdef_obj.name, 'add_arrays')
# validate that the DPD has an attachment NEW SA - 21
attachment_ids, assoc_ids = self.rrclient.find_objects(
dataprocessdef_id, PRED.hasAttachment, RT.Attachment, True)
self.assertEqual(len(attachment_ids), 1)
attachment_obj = self.rrclient.read_attachment(attachment_ids[0])
log.debug('attachment: %s', attachment_obj)
# validate that the data process resource has input and output data products associated
# L4-CI-SA-RQ-364 and NEW SA-3
outproduct_ids, assoc_ids = self.rrclient.find_objects(
dataprocess_id, PRED.hasOutputProduct, RT.DataProduct, True)
self.assertEqual(len(outproduct_ids), 1)
inproduct_ids, assoc_ids = self.rrclient.find_objects(
dataprocess_id, PRED.hasInputProduct, RT.DataProduct, True)
self.assertEqual(len(inproduct_ids), 1)
# Test for provenance. Get Data product produced by the data processes
output_data_product_id, _ = self.rrclient.find_objects(
subject=dataprocess_id,
object_type=RT.DataProduct,
predicate=PRED.hasOutputProduct,
id_only=True)
output_data_product_provenance = self.dataproductclient.get_data_product_provenance(
output_data_product_id[0])
# Do a basic check to see if there were 3 entries in the provenance graph. Parent and Child and the
# DataProcessDefinition creating the child from the parent.
self.assertTrue(len(output_data_product_provenance) == 2)
self.assertTrue(self.input_dp_id in output_data_product_provenance[
output_data_product_id[0]]['parents'])
self.assertTrue(output_data_product_provenance[
output_data_product_id[0]]['parents'][self.input_dp_id]
['data_process_definition_id'] == dataprocessdef_id)
# NEW SA - 4 | Data processing shall include the appropriate data product algorithm name and version number in
# the metadata of each output data product created by the data product algorithm.
output_data_product_obj, _ = self.rrclient.find_objects(
subject=dataprocess_id,
object_type=RT.DataProduct,
predicate=PRED.hasOutputProduct,
id_only=False)
self.assertTrue(output_data_product_obj[0].name != None)
self.assertTrue(output_data_product_obj[0]._rev != None)
# retrieve subscription from data process
subscription_objs, _ = self.rrclient.find_objects(
subject=dataprocess_id,
predicate=PRED.hasSubscription,
object_type=RT.Subscription,
id_only=False)
log.debug('test_transform_worker subscription_obj: %s',
subscription_objs[0])
# create a queue to catch the published granules
self.subscription_id = self.pubsub_client.create_subscription(
name='parsed_subscription',
stream_ids=[self.stream_id],
exchange_name=subscription_objs[0].exchange_name)
self.addCleanup(self.pubsub_client.delete_subscription,
self.subscription_id)
self.pubsub_client.activate_subscription(self.subscription_id)
self.addCleanup(self.pubsub_client.deactivate_subscription,
self.subscription_id)
stream_route = self.pubsub_client.read_stream_route(self.stream_id)
self.publisher = StandaloneStreamPublisher(stream_id=self.stream_id,
stream_route=stream_route)
for n in range(1, 101):
rdt = RecordDictionaryTool(stream_definition_id=self.stream_def_id)
rdt['time'] = [0] # time should always come first
rdt['conductivity'] = [1]
rdt['pressure'] = [2]
rdt['salinity'] = [8]
self.publisher.publish(rdt.to_granule())
# validate that the output granule is received and the updated value is correct
self.assertTrue(self.granule_verified.wait(self.wait_time))
# validate that the data process loaded into worker event is received (L4-CI-SA-RQ-182)
self.assertTrue(
self.worker_assigned_event_verified.wait(self.wait_time))
# validate that the data process create (with data product ids) event is received (NEW SA -42)
self.assertTrue(self.dp_created_event_verified.wait(self.wait_time))
# validate that the data process heartbeat event is received (for every hundred granules processed) (L4-CI-SA-RQ-182)
#this takes a while so set wait limit to large value
self.assertTrue(self.heartbeat_event_verified.wait(200))
# validate that the code from the transform function can be retrieve via inspect_data_process_definition
src = self.dataprocessclient.inspect_data_process_definition(
dataprocessdef_id)
self.assertIn('def add_arrays(a, b)', src)
# now delete the DPD and DP then verify that the resources are retired so that information required for provenance are still available
self.dataprocessclient.delete_data_process(dataprocess_id)
self.dataprocessclient.delete_data_process_definition(
dataprocessdef_id)
in_dp_objs, _ = self.rrclient.find_objects(
subject=dataprocess_id,
predicate=PRED.hasInputProduct,
object_type=RT.DataProduct,
id_only=True)
self.assertTrue(in_dp_objs is not None)
dpd_objs, _ = self.rrclient.find_subjects(
subject_type=RT.DataProcessDefinition,
predicate=PRED.hasDataProcess,
object=dataprocess_id,
id_only=True)
self.assertTrue(dpd_objs is not None)
@attr('LOCOINT')
@unittest.skipIf(os.getenv('CEI_LAUNCH_TEST', False),
'Skip test while in CEI LAUNCH mode')
def test_transform_worker_with_instrumentdevice(self):
# test that a data process (type: data-product-in / data-product-out) can be defined and launched.
# verify that the output granule fields are correctly populated
# test that the input and output data products are linked to facilitate provenance
self.data_process_objs = []
self._output_stream_ids = []
self.event_verified = Event()
# Create CTD Parsed as the initial data product
# create a stream definition for the data from the ctd simulator
self.parameter_dict_id = self.dataset_management_client.read_parameter_dictionary_by_name(
'ctd_parsed_param_dict', id_only=True)
self.stream_def_id = self.pubsub_client.create_stream_definition(
name='stream_def', parameter_dictionary_id=self.parameter_dict_id)
# create the DataProduct that is the input to the data processes
input_dp_obj = IonObject(RT.DataProduct,
name='input_data_product',
description='input test stream')
self.input_dp_id = self.dataproductclient.create_data_product(
data_product=input_dp_obj, stream_definition_id=self.stream_def_id)
# retrieve the Stream for this data product
stream_ids, assoc_ids = self.rrclient.find_objects(
self.input_dp_id, PRED.hasStream, RT.Stream, True)
self.stream_id = stream_ids[0]
log.debug('new ctd_parsed_data_product_id = %s' % self.input_dp_id)
# only ever need one device for testing purposes.
instDevice_obj, _ = self.rrclient.find_resources(
restype=RT.InstrumentDevice, name='test_ctd_device')
if instDevice_obj:
instDevice_id = instDevice_obj[0]._id
else:
instDevice_obj = IonObject(RT.InstrumentDevice,
name='test_ctd_device',
description="test_ctd_device",
serial_number="12345")
instDevice_id = self.imsclient.create_instrument_device(
instrument_device=instDevice_obj)
self.damsclient.assign_data_product(input_resource_id=instDevice_id,
data_product_id=self.input_dp_id)
# create the DPD, DataProcess and output DataProduct
dataprocessdef_id, dataprocess_id, dataproduct_id = self.create_data_process(
)
self.addCleanup(self.dataprocessclient.delete_data_process,
dataprocess_id)
self.addCleanup(self.dataprocessclient.delete_data_process_definition,
dataprocessdef_id)
# Test for provenance. Get Data product produced by the data processes
output_data_product_id, _ = self.rrclient.find_objects(
subject=dataprocess_id,
object_type=RT.DataProduct,
predicate=PRED.hasOutputProduct,
id_only=True)
output_data_product_provenance = self.dataproductclient.get_data_product_provenance(
output_data_product_id[0])
# Do a basic check to see if there were 3 entries in the provenance graph. Parent and Child and the
# DataProcessDefinition creating the child from the parent.
self.assertTrue(len(output_data_product_provenance) == 3)
self.assertTrue(self.input_dp_id in output_data_product_provenance[
output_data_product_id[0]]['parents'])
self.assertTrue(instDevice_id in output_data_product_provenance[
self.input_dp_id]['parents'])
self.assertTrue(output_data_product_provenance[instDevice_id]['type']
== 'InstrumentDevice')
@attr('LOCOINT')
@unittest.skipIf(os.getenv('CEI_LAUNCH_TEST', False),
'Skip test while in CEI LAUNCH mode')
def test_transform_worker_with_platformdevice(self):
# test that a data process (type: data-product-in / data-product-out) can be defined and launched.
# verify that the output granule fields are correctly populated
# test that the input and output data products are linked to facilitate provenance
self.data_process_objs = []
self._output_stream_ids = []
self.event_verified = Event()
# Create CTD Parsed as the initial data product
# create a stream definition for the data from the ctd simulator
self.parameter_dict_id = self.dataset_management_client.read_parameter_dictionary_by_name(
'ctd_parsed_param_dict', id_only=True)
self.stream_def_id = self.pubsub_client.create_stream_definition(
name='stream_def', parameter_dictionary_id=self.parameter_dict_id)
# create the DataProduct that is the input to the data processes
input_dp_obj = IonObject(RT.DataProduct,
name='input_data_product',
description='input test stream')
self.input_dp_id = self.dataproductclient.create_data_product(
data_product=input_dp_obj, stream_definition_id=self.stream_def_id)
# retrieve the Stream for this data product
stream_ids, assoc_ids = self.rrclient.find_objects(
self.input_dp_id, PRED.hasStream, RT.Stream, True)
self.stream_id = stream_ids[0]
log.debug('new ctd_parsed_data_product_id = %s' % self.input_dp_id)
# only ever need one device for testing purposes.
platform_device_obj, _ = self.rrclient.find_resources(
restype=RT.PlatformDevice, name='TestPlatform')
if platform_device_obj:
platform_device_id = platform_device_obj[0]._id
else:
platform_device_obj = IonObject(RT.PlatformDevice,
name='TestPlatform',
description="TestPlatform",
serial_number="12345")
platform_device_id = self.imsclient.create_platform_device(
platform_device=platform_device_obj)
self.damsclient.assign_data_product(
input_resource_id=platform_device_id,
data_product_id=self.input_dp_id)
# create the DPD, DataProcess and output DataProduct
dataprocessdef_id, dataprocess_id, dataproduct_id = self.create_data_process(
)
self.addCleanup(self.dataprocessclient.delete_data_process,
dataprocess_id)
self.addCleanup(self.dataprocessclient.delete_data_process_definition,
dataprocessdef_id)
# Test for provenance. Get Data product produced by the data processes
output_data_product_id, _ = self.rrclient.find_objects(
subject=dataprocess_id,
object_type=RT.DataProduct,
predicate=PRED.hasOutputProduct,
id_only=True)
output_data_product_provenance = self.dataproductclient.get_data_product_provenance(
output_data_product_id[0])
# Do a basic check to see if there were 3 entries in the provenance graph. Parent and Child and the
# DataProcessDefinition creating the child from the parent.
self.assertTrue(len(output_data_product_provenance) == 3)
self.assertTrue(self.input_dp_id in output_data_product_provenance[
output_data_product_id[0]]['parents'])
self.assertTrue(platform_device_id in output_data_product_provenance[
self.input_dp_id]['parents'])
self.assertTrue(output_data_product_provenance[platform_device_id]
['type'] == 'PlatformDevice')
@attr('LOCOINT')
@unittest.skipIf(os.getenv('CEI_LAUNCH_TEST', False),
'Skip test while in CEI LAUNCH mode')
def test_event_transform_worker(self):
self.data_process_objs = []
self._output_stream_ids = []
self.event_verified = Event()
# test that a data process (type: data-product-in / event-out) can be defined and launched.
# verify that event fields are correctly populated
self.parameter_dict_id = self.dataset_management_client.read_parameter_dictionary_by_name(
name='ctd_parsed_param_dict', id_only=True)
# create the StreamDefinition
self.stream_def_id = self.pubsub_client.create_stream_definition(
name='stream_def', parameter_dictionary_id=self.parameter_dict_id)
self.addCleanup(self.pubsub_client.delete_stream_definition,
self.stream_def_id)
# create the DataProduct
input_dp_obj = IonObject(RT.DataProduct,
name='input_data_product',
description='input test stream')
self.input_dp_id = self.dataproductclient.create_data_product(
data_product=input_dp_obj, stream_definition_id=self.stream_def_id)
# retrieve the Stream for this data product
stream_ids, assoc_ids = self.rrclient.find_objects(
self.input_dp_id, PRED.hasStream, RT.Stream, True)
self.stream_id = stream_ids[0]
# create the DPD and two DPs
self.event_data_process_id = self.create_event_data_processes()
# retrieve subscription from data process
subscription_objs, _ = self.rrclient.find_objects(
subject=self.event_data_process_id,
predicate=PRED.hasSubscription,
object_type=RT.Subscription,
id_only=False)
log.debug('test_event_transform_worker subscription_obj: %s',
subscription_objs[0])
# create a queue to catch the published granules
self.subscription_id = self.pubsub_client.create_subscription(
name='parsed_subscription',
stream_ids=[self.stream_id],
exchange_name=subscription_objs[0].exchange_name)
self.addCleanup(self.pubsub_client.delete_subscription,
self.subscription_id)
self.pubsub_client.activate_subscription(self.subscription_id)
self.addCleanup(self.pubsub_client.deactivate_subscription,
self.subscription_id)
stream_route = self.pubsub_client.read_stream_route(self.stream_id)
self.publisher = StandaloneStreamPublisher(stream_id=self.stream_id,
stream_route=stream_route)
self.start_event_transform_listener()
self.data_modified = Event()
rdt = RecordDictionaryTool(stream_definition_id=self.stream_def_id)
rdt['time'] = [0] # time should always come first
rdt['conductivity'] = [1]
rdt['pressure'] = [2]
rdt['salinity'] = [8]
self.publisher.publish(rdt.to_granule())
self.assertTrue(self.event_verified.wait(self.wait_time))
@attr('LOCOINT')
@unittest.skipIf(os.getenv('CEI_LAUNCH_TEST', False),
'Skip test while in CEI LAUNCH mode')
def test_bad_argument_map(self):
self._output_stream_ids = []
# test that a data process (type: data-product-in / data-product-out) parameter mapping it validated during
# data process creation and that the correct exception is raised for both input and output.
self.parameter_dict_id = self.dataset_management_client.read_parameter_dictionary_by_name(
name='ctd_parsed_param_dict', id_only=True)
# create the StreamDefinition
self.stream_def_id = self.pubsub_client.create_stream_definition(
name='stream_def', parameter_dictionary_id=self.parameter_dict_id)
self.addCleanup(self.pubsub_client.delete_stream_definition,
self.stream_def_id)
# create the DataProduct that is the input to the data processes
input_dp_obj = IonObject(RT.DataProduct,
name='input_data_product',
description='input test stream')
self.input_dp_id = self.dataproductclient.create_data_product(
data_product=input_dp_obj, stream_definition_id=self.stream_def_id)
# two data processes using one transform and one DPD
dp1_func_output_dp_id = self.create_output_data_product()
# Set up DPD and DP #2 - array add function
tf_obj = IonObject(
RT.TransformFunction,
name='add_array_func',
description='adds values in an array',
function='add_arrays',
module="ion_example.add_arrays",
arguments=['arr1', 'arr2'],
function_type=TransformFunctionType.TRANSFORM,
uri=
'http://sddevrepo.oceanobservatories.org/releases/ion_example-0.1-py2.7.egg'
)
add_array_func_id, rev = self.rrclient.create(tf_obj)
dpd_obj = IonObject(
RT.DataProcessDefinition,
name='add_arrays',
description='adds the values of two arrays',
data_process_type=DataProcessTypeEnum.TRANSFORM_PROCESS)
add_array_dpd_id = self.dataprocessclient.create_data_process_definition(
data_process_definition=dpd_obj, function_id=add_array_func_id)
self.dataprocessclient.assign_stream_definition_to_data_process_definition(
self.stream_def_id, add_array_dpd_id, binding='add_array_func')
# create the data process with invalid argument map
argument_map = {"arr1": "foo", "arr2": "bar"}
output_param = "salinity"
with self.assertRaises(BadRequest) as cm:
dp1_data_process_id = self.dataprocessclient.create_data_process(
data_process_definition_id=add_array_dpd_id,
inputs=[self.input_dp_id],
outputs=[dp1_func_output_dp_id],
argument_map=argument_map,
out_param_name=output_param)
ex = cm.exception
log.debug(' exception raised: %s', cm)
self.assertEqual(
ex.message,
"Input data product does not contain the parameters defined in argument map"
)
# create the data process with invalid output parameter name
argument_map = {"arr1": "conductivity", "arr2": "pressure"}
output_param = "foo"
with self.assertRaises(BadRequest) as cm:
dp1_data_process_id = self.dataprocessclient.create_data_process(
data_process_definition_id=add_array_dpd_id,
inputs=[self.input_dp_id],
outputs=[dp1_func_output_dp_id],
argument_map=argument_map,
out_param_name=output_param)
ex = cm.exception
log.debug(' exception raised: %s', cm)
self.assertEqual(
ex.message,
"Output data product does not contain the output parameter name provided"
)
def create_event_data_processes(self):
# two data processes using one transform and one DPD
argument_map = {"a": "salinity"}
# set up DPD and DP #2 - array add function
tf_obj = IonObject(
RT.TransformFunction,
name='validate_salinity_array',
description='validate_salinity_array',
function='validate_salinity_array',
module="ion.processes.data.transforms.test.test_transform_worker",
arguments=['a'],
function_type=TransformFunctionType.TRANSFORM)
add_array_func_id, rev = self.rrclient.create(tf_obj)
dpd_obj = IonObject(
RT.DataProcessDefinition,
name='validate_salinity_array',
description='validate_salinity_array',
data_process_type=DataProcessTypeEnum.TRANSFORM_PROCESS,
)
add_array_dpd_id = self.dataprocessclient.create_data_process_definition(
data_process_definition=dpd_obj, function_id=add_array_func_id)
self.dataprocessclient.assign_stream_definition_to_data_process_definition(
self.stream_def_id,
add_array_dpd_id,
binding='validate_salinity_array')
# create the data process
dp1_data_process_id = self.dataprocessclient.create_data_process(
data_process_definition_id=add_array_dpd_id,
inputs=[self.input_dp_id],
outputs=None,
argument_map=argument_map)
self.damsclient.register_process(dp1_data_process_id)
self.addCleanup(self.dataprocessclient.delete_data_process,
dp1_data_process_id)
return dp1_data_process_id
def create_data_process(self):
# two data processes using one transform and one DPD
dp1_func_output_dp_id = self.create_output_data_product()
argument_map = {"arr1": "conductivity", "arr2": "pressure"}
output_param = "salinity"
# set up DPD and DP #2 - array add function
tf_obj = IonObject(
RT.TransformFunction,
name='add_array_func',
description='adds values in an array',
function='add_arrays',
module="ion_example.add_arrays",
arguments=['arr1', 'arr2'],
function_type=TransformFunctionType.TRANSFORM,
uri=
'http://sddevrepo.oceanobservatories.org/releases/ion_example-0.1-py2.7.egg'
)
add_array_func_id, rev = self.rrclient.create(tf_obj)
dpd_obj = IonObject(
RT.DataProcessDefinition,
name='add_arrays',
description='adds the values of two arrays',
data_process_type=DataProcessTypeEnum.TRANSFORM_PROCESS,
version_label='1.0a')
add_array_dpd_id = self.dataprocessclient.create_data_process_definition(
data_process_definition=dpd_obj, function_id=add_array_func_id)
self.dataprocessclient.assign_stream_definition_to_data_process_definition(
self.stream_def_id, add_array_dpd_id, binding='add_array_func')
# create the data process
dp1_data_process_id = self.dataprocessclient.create_data_process(
data_process_definition_id=add_array_dpd_id,
inputs=[self.input_dp_id],
outputs=[dp1_func_output_dp_id],
argument_map=argument_map,
out_param_name=output_param)
self.damsclient.register_process(dp1_data_process_id)
#self.addCleanup(self.dataprocessclient.delete_data_process, dp1_data_process_id)
# add an attachment object to this DPD to test new SA-21
import msgpack
attachment_content = 'foo bar'
attachment_obj = IonObject(RT.Attachment,
name='test_attachment',
attachment_type=AttachmentType.ASCII,
content_type='text/plain',
content=msgpack.packb(attachment_content))
att_id = self.rrclient.create_attachment(add_array_dpd_id,
attachment_obj)
self.addCleanup(self.rrclient.delete_attachment, att_id)
return add_array_dpd_id, dp1_data_process_id, dp1_func_output_dp_id
def create_output_data_product(self):
dp1_outgoing_stream_id = self.pubsub_client.create_stream_definition(
name='dp1_stream', parameter_dictionary_id=self.parameter_dict_id)
dp1_output_dp_obj = IonObject(RT.DataProduct,
name='data_process1_data_product',
description='output of add array func')
dp1_func_output_dp_id = self.dataproductclient.create_data_product(
dp1_output_dp_obj, dp1_outgoing_stream_id)
self.addCleanup(self.dataproductclient.delete_data_product,
dp1_func_output_dp_id)
# retrieve the id of the OUTPUT stream from the out Data Product and add to granule logger
stream_ids, _ = self.rrclient.find_objects(dp1_func_output_dp_id,
PRED.hasStream, None, True)
self._output_stream_ids.append(stream_ids[0])
subscription_id = self.pubsub_client.create_subscription(
'validator', data_product_ids=[dp1_func_output_dp_id])
self.addCleanup(self.pubsub_client.delete_subscription,
subscription_id)
def on_granule(msg, route, stream_id):
log.debug('recv_packet stream_id: %s route: %s msg: %s',
stream_id, route, msg)
self.validate_output_granule(msg, route, stream_id)
self.granule_verified.set()
validator = StandaloneStreamSubscriber('validator',
callback=on_granule)
validator.start()
self.addCleanup(validator.stop)
self.pubsub_client.activate_subscription(subscription_id)
self.addCleanup(self.pubsub_client.deactivate_subscription,
subscription_id)
return dp1_func_output_dp_id
def validate_event(self, *args, **kwargs):
"""
This method is a callback function for receiving DataProcessStatusEvent.
"""
data_process_event = args[0]
log.debug("DataProcessStatusEvent: %s",
str(data_process_event.__dict__))
# if data process already created, check origin
if self.dp_list:
self.assertIn(data_process_event.origin, self.dp_list)
# if this is a heartbeat event then 100 granules have been processed
if 'data process status update.' in data_process_event.description:
self.heartbeat_event_verified.set()
else:
# else check that this is the assign event
if 'Data process assigned to transform worker' in data_process_event.description:
self.worker_assigned_event_verified.set()
elif 'Data process created for data product' in data_process_event.description:
self.dp_created_event_verified.set()
def validate_output_granule(self, msg, route, stream_id):
self.assertIn(stream_id, self._output_stream_ids)
rdt = RecordDictionaryTool.load_from_granule(msg)
log.debug('validate_output_granule rdt: %s', rdt)
sal_val = rdt['salinity']
np.testing.assert_array_equal(sal_val, np.array([3]))
def start_event_listener(self):
es = EventSubscriber(event_type=OT.DataProcessStatusEvent,
callback=self.validate_event)
es.start()
self.addCleanup(es.stop)
def validate_transform_event(self, *args, **kwargs):
"""
This method is a callback function for receiving DataProcessStatusEvent.
"""
status_alert_event = args[0]
np.testing.assert_array_equal(status_alert_event.origin,
self.stream_id)
np.testing.assert_array_equal(status_alert_event.values,
np.array([self.event_data_process_id]))
log.debug("DeviceStatusAlertEvent: %s",
str(status_alert_event.__dict__))
self.event_verified.set()
def start_event_transform_listener(self):
es = EventSubscriber(event_type=OT.DeviceStatusAlertEvent,
callback=self.validate_transform_event)
es.start()
self.addCleanup(es.stop)
def test_download(self):
egg_url = 'http://sddevrepo.oceanobservatories.org/releases/ion_example-0.1-py2.7.egg'
egg_path = TransformWorker.download_egg(egg_url)
import pkg_resources
pkg_resources.working_set.add_entry(egg_path)
from ion_example.add_arrays import add_arrays
a = add_arrays(1, 2)
self.assertEquals(a, 3)