def delete_data_process(self, data_process_id=""):
# Delete the specified DataProcessDefinition object
data_process_obj = self.read_data_process(data_process_id)
log.debug("delete the association with DataProcessDefinition")
dpd_assn_ids = self.clients.resource_registry.find_associations(subject=data_process_id, predicate=PRED.hasProcessDefinition, id_only=True)
for dpd_assn_id in dpd_assn_ids:
self.clients.resource_registry.delete_association(dpd_assn_id)
self._stop_process(data_process_obj)
log.debug("Finalizing data products by removing streams associated with the dataset and product")
out_products, assocs = self.clients.resource_registry.find_objects(subject=data_process_id, predicate=PRED.hasOutputProduct, id_only=True)
for out_product, assoc in zip(out_products, assocs):
data_product_management = DataProductManagementServiceClient()
data_product_management.remove_streams(out_product)
log.debug("deleting association with output data product '%s'" % out_product)
self.clients.resource_registry.delete_association(assoc)
self.clients.data_acquisition_management.unassign_data_product(data_process_id, out_product)
log.debug("Delete the input product links")
inprod_associations = self.clients.resource_registry.find_associations(data_process_id, PRED.hasInputProduct)
for inprod_association in inprod_associations:
self.clients.resource_registry.delete_association(inprod_association)
try:
self.deactivate_data_process(data_process_id=data_process_id)
log.warn('Deleteing activated data process...')
except BadRequest:
pass
subscription_id = data_process_obj.input_subscription_id
self.clients.pubsub_management.delete_subscription(subscription_id)
data_process_obj.input_subscription_id = None
#unregister the data process in DataAcquisitionMgmtSvc
self.clients.data_acquisition_management.unregister_process(data_process_id)
# Delete the data process
self.clients.resource_registry.retire(data_process_id)
return
class DataProcessManagementService(BaseDataProcessManagementService):
def on_init(self):
IonObject("Resource") # suppress pyflakes error
self.override_clients(self.clients)
self.init_module_uploader()
self.get_unique_id = (lambda : uuid4().hex)
self.data_product_management = DataProductManagementServiceClient()
def init_module_uploader(self):
if self.CFG:
#looking for forms like host=amoeba.ucsd.edu, remotepath=/var/www/release, user=steve
cfg_host = self.CFG.get_safe("service.data_process_management.process_release_host", None)
cfg_remotepath = self.CFG.get_safe("service.data_process_management.process_release_directory", None)
cfg_user = self.CFG.get_safe("service.data_process_management.process_release_user",
pwd.getpwuid(os.getuid())[0])
cfg_wwwprefix = self.CFG.get_safe("service.data_process_management.process_release_wwwprefix", None)
if cfg_host is None or cfg_remotepath is None or cfg_wwwprefix is None:
raise BadRequest("Missing configuration items; host='%s', directory='%s', wwwprefix='%s'" %
(cfg_host, cfg_remotepath, cfg_wwwprefix))
self.module_uploader = RegisterModulePreparerPy(dest_user=cfg_user,
dest_host=cfg_host,
dest_path=cfg_remotepath,
dest_wwwprefix=cfg_wwwprefix)
def override_clients(self, new_clients):
"""
Replaces the service clients with a new set of them... and makes sure they go to the right places
"""
#shortcut names for the import sub-services
if hasattr(self.clients, "resource_registry"):
self.RR = self.clients.resource_registry
#todo: need to know what object will be worked with here
def register_data_process_definition(self, process_code=''):
"""
register a process module by putting it in a web-accessible location
@process_code a base64-encoded python file
"""
# # retrieve the resource
# data_process_definition_obj = self.clients.resource_registry.read(data_process_definition_id)
dest_filename = "process_code_%s.py" % self.get_unique_id() #data_process_definition_obj._id
#process the input file (base64-encoded .py)
uploader_obj, err = self.module_uploader.prepare(process_code, dest_filename)
if None is uploader_obj:
raise BadRequest("Process code failed validation: %s" % err)
# actually upload
up_success, err = uploader_obj.upload()
if not up_success:
raise BadRequest("Upload failed: %s" % err)
# #todo: save module / class?
# data_process_definition_obj.uri = uploader_obj.get_destination_url()
# self.clients.resource_registry.update(data_process_definition_obj)
return uploader_obj.get_destination_url()
@classmethod
def _cmp_transform_function(cls, tf1, tf2):
return tf1.module == tf2.module and \
tf1.cls == tf2.cls and \
tf1.uri == tf2.uri and \
tf1.function_type == tf2.function_type
def create_transform_function(self, transform_function=''):
'''
Creates a new transform function
'''
tfs, _ = self.clients.resource_registry.find_resources(name=transform_function.name, restype=RT.TransformFunction, id_only=False)
if len(tfs):
for tf in tfs:
if self._cmp_transform_function(transform_function, tf):
return tf._id
raise BadRequest('An existing TransformFunction with name %s exists with a different definition' % transform_function.name)
tf_id, rev = self.clients.resource_registry.create(transform_function)
return tf_id
def read_transform_function(self, transform_function_id=''):
tf = self.clients.resource_registry.read(transform_function_id)
validate_is_instance(tf,TransformFunction, 'Resource %s is not a TransformFunction: %s' %(transform_function_id, type(tf)))
return tf
def delete_transform_function(self, transform_function_id=''):
self.read_transform_function(transform_function_id)
self.clients.resource_registry.delete(transform_function_id)
def create_data_process_definition(self, data_process_definition=None):
result, _ = self.clients.resource_registry.find_resources(RT.DataProcessDefinition, None, data_process_definition.name, True)
validate_true( len(result) ==0, "A data process definition named '%s' already exists" % data_process_definition.name)
#todo: determine validation checks for a data process def
data_process_definition_id, version = self.clients.resource_registry.create(data_process_definition)
#-------------------------------
# Process Definition
#-------------------------------
# Create the underlying process definition
process_definition = ProcessDefinition()
process_definition.name = data_process_definition.name
process_definition.description = data_process_definition.description
process_definition.executable = {'module':data_process_definition.module, 'class':data_process_definition.class_name}
process_definition_id = self.clients.process_dispatcher.create_process_definition(process_definition=process_definition)
self.clients.resource_registry.create_association(data_process_definition_id, PRED.hasProcessDefinition, process_definition_id)
return data_process_definition_id
def update_data_process_definition(self, data_process_definition=None):
# TODO: If executable has changed, update underlying ProcessDefinition
# Overwrite DataProcessDefinition object
self.clients.resource_registry.update(data_process_definition)
def read_data_process_definition(self, data_process_definition_id=''):
data_proc_def_obj = self.clients.resource_registry.read(data_process_definition_id)
return data_proc_def_obj
def delete_data_process_definition(self, data_process_definition_id=''):
# Delete all associations for this resource
self._remove_associations(data_process_definition_id)
# Delete the resource
self.clients.resource_registry.retire(data_process_definition_id)
def force_delete_data_process_definition(self, data_process_definition_id=''):
processdef_ids, _ = self.clients.resource_registry.find_objects(subject=data_process_definition_id, predicate=PRED.hasProcessDefinition, object_type=RT.ProcessDefinition, id_only=True)
self._remove_associations(data_process_definition_id)
self.clients.resource_registry.delete(data_process_definition_id)
for processdef_id in processdef_ids:
self.clients.process_dispatcher.delete_process_definition(processdef_id)
def find_data_process_definitions(self, filters=None):
"""
@param filters: dict of parameters to filter down
the list of possible data proc.
@retval
"""
#todo: add filtering
data_process_def_list , _ = self.clients.resource_registry.find_resources(RT.DataProcessDefinition, None, None, True)
return data_process_def_list
def assign_input_stream_definition_to_data_process_definition(self, stream_definition_id='', data_process_definition_id=''):
"""Connect the input stream with a data process definition
"""
# Verify that both ids are valid, RR will throw if not found
stream_definition_obj = self.clients.resource_registry.read(stream_definition_id)
data_process_definition_obj = self.clients.resource_registry.read(data_process_definition_id)
validate_is_not_none(stream_definition_obj, "No stream definition object found for stream definition id: %s" % stream_definition_id)
validate_is_not_none(data_process_definition_obj, "No data process definition object found for data process" \
" definition id: %s" % data_process_definition_id)
self.clients.resource_registry.create_association(data_process_definition_id, PRED.hasInputStreamDefinition, stream_definition_id)
def unassign_input_stream_definition_from_data_process_definition(self, stream_definition_id='', data_process_definition_id=''):
"""
Disconnect the Data Product from the Data Producer
@param stream_definition_id str
@param data_process_definition_id str
@throws NotFound object with specified id does not exist
"""
# Remove the link between the Stream Definition resource and the Data Process Definition resource
associations = self.clients.resource_registry.find_associations(data_process_definition_id, PRED.hasInputStreamDefinition, stream_definition_id, id_only=True)
validate_is_not_none(associations, "No Input Stream Definitions associated with data process definition ID " + str(data_process_definition_id))
for association in associations:
self.clients.resource_registry.delete_association(association)
def assign_stream_definition_to_data_process_definition(self, stream_definition_id='', data_process_definition_id='', binding=''):
"""Connect the output stream with a data process definition
"""
# Verify that both ids are valid, RR will throw if not found
stream_definition_obj = self.clients.resource_registry.read(stream_definition_id)
data_process_definition_obj = self.clients.resource_registry.read(data_process_definition_id)
validate_is_not_none(stream_definition_obj, "No stream definition object found for stream definition id: %s" % stream_definition_id)
validate_is_not_none(data_process_definition_obj, "No data process definition object found for data process"\
" definition id: %s" % data_process_definition_id)
self.clients.resource_registry.create_association(data_process_definition_id, PRED.hasStreamDefinition, stream_definition_id)
data_process_definition_obj.output_bindings[binding] = stream_definition_id
self.clients.resource_registry.update(data_process_definition_obj)
def unassign_stream_definition_from_data_process_definition(self, stream_definition_id='', data_process_definition_id=''):
"""
Disconnect the Data Product from the Data Producer
@param stream_definition_id str
@param data_process_definition_id str
@throws NotFound object with specified id does not exist
"""
# Remove the link between the Stream Definition resource and the Data Process Definition resource
associations = self.clients.resource_registry.find_associations(data_process_definition_id, PRED.hasStreamDefinition, stream_definition_id, id_only=True)
validate_is_not_none(associations, "No Stream Definitions associated with data process definition ID " + str(data_process_definition_id))
for association in associations:
self.clients.resource_registry.delete_association(association)
def create_data_process2(self, data_process_definition_id='', in_data_product_ids=None, out_data_product_ids=None, configuration=None):
'''
Creates a DataProcess resource and launches the process.
A DataProcess is a process that receives one (or more) data products and produces one (or more) data products.
@param data_process_definition_id : The Data Process Definition to use, if none is specified the standard TransformDataProcess is used
@param in_data_product_ids : A list of input data product identifiers
@param out_data_product_ids : A list of output data product identifiers
@param configuration : The configuration dictionary for the process, and the routing table:
The routing table is defined as such:
{ (in_data_product_id, out_data_product_id) : actor }
Routes are specified in the configuration dictionary under the item "routes"
actor is either None (for ParameterFunctions) or a valid TransformFunction identifier
'''
configuration = configuration or DotDict()
routes = configuration.get_safe('process.routes', {})
self.validate_compatibility(in_data_product_ids, out_data_product_ids)
configuration.process.routes = self._manage_routes(routes)
dproc = DataProcess()
dproc.name = 'data_process_%s' % self.get_unique_id()
dproc.configuration = configuration
dproc_id, rev = self.clients.resource_registry.create(dproc)
dproc._id = dproc_id
dproc._rev = rev
self._manage_producers(dproc_id, out_data_product_ids)
self._manage_attachments()
queue_name = self._create_subscription(dproc, in_data_product_ids)
pid = self._launch_data_process(
queue_name=queue_name,
data_process_definition_id=data_process_definition_id,
out_data_product_ids=out_data_product_ids,
configuration=configuration)
self.clients.resource_registry.create_association(subject=dproc_id, predicate=PRED.hasProcess, object=pid)
return dproc_id
# ------------------------------------------------------------------------------------------------
# Working with DataProcess
def create_data_process(self, data_process_definition_id='', in_data_product_ids=None, out_data_products=None, configuration=None):
"""
@param data_process_definition_id: Object with definition of the
process to apply to the input data product
@param in_data_product_ids: ID of the input data products
@param out_data_products: list of IDs of the output data products
@retval data_process_id: ID of the newly created data process object
"""
inform = "Input Data Product: "+str(in_data_product_ids)+\
"\nTransformed by: "+str(data_process_definition_id)+\
"\nTo create output Product: "+str(out_data_products) + "\n"
log.debug("DataProcessManagementService:create_data_process() method called with parameters:\n" +
inform)
#---------------------------------------------------------------------------------------
# Initialize
#---------------------------------------------------------------------------------------
configuration = configuration or DotDict()
#validate_is_not_none( out_data_products, "No output data products passed in")
#---------------------------------------------------------------------------------------
# Read the data process definition
#---------------------------------------------------------------------------------------
data_process_definition = self.read_data_process_definition(data_process_definition_id)
#---------------------------------------------------------------------------------------
# Read the output bindings from the definition
#---------------------------------------------------------------------------------------
output_bindings = data_process_definition.output_bindings
#---------------------------------------------------------------------------------------
# Find the process definition associated with this data process definition.
# From the process definition, we can get the module and class to run....
#---------------------------------------------------------------------------------------
procdef_ids,_ = self.clients.resource_registry.find_objects(data_process_definition_id, PRED.hasProcessDefinition, RT.ProcessDefinition, id_only=True)
if not procdef_ids:
raise BadRequest("Cannot find associated ProcessDefinition for DataProcessDefinition id=%s" % data_process_definition_id)
process_definition_id = procdef_ids[0]
#---------------------------------------------------------------------------------------
# Create a data process object and register it
#---------------------------------------------------------------------------------------
# get the name of the data process and create an IONObject for it
data_process_name = create_unique_identifier("process_" + data_process_definition.name)
data_process_obj = IonObject(RT.DataProcess, name=data_process_name)
# register the data process
data_process_id, version = self.clients.resource_registry.create(data_process_obj)
data_process_obj = self.clients.resource_registry.read(data_process_id)
#---------------------------------------------------------------------------------------
# Make the necessary associations, registering
#---------------------------------------------------------------------------------------
#todo check if this assoc is needed?
# Associate the data process with the data process definition
self.clients.resource_registry.create_association(data_process_id, PRED.hasProcessDefinition, data_process_definition_id)
# Register the data process instance as a data producer with DataAcquisitionMgmtSvc
data_producer_id = self.clients.data_acquisition_management.register_process(data_process_id)
log.debug("DataProcessManagementService:create_data_process register process "
"with DataAcquisitionMgmtSvc: data_producer_id: %s (L4-CI-SA-RQ-181)", str(data_producer_id) )
#---------------------------------------------------------------------------------------
# Register each output data product with DAMS to create DataProducer links
#---------------------------------------------------------------------------------------
output_stream_dict = {}
# if out_data_products is None:
# raise BadRequest("Data Process must have output product(s) specified %s", str(data_process_definition_id) )
if out_data_products:
for binding, output_data_product_id in out_data_products.iteritems():
# check that the product is not already associated with a producer
producer_ids, _ = self.clients.resource_registry.find_objects(output_data_product_id, PRED.hasDataProducer, RT.DataProducer, True)
if producer_ids:
raise BadRequest("Data Product should not already be associated to a DataProducer %s hasDataProducer %s", str(data_process_id), str(producer_ids[0]))
#Assign each output Data Product to this producer resource
output_data_product_obj = self.clients.resource_registry.read(output_data_product_id)
if not output_data_product_obj:
raise NotFound("Output Data Product %s does not exist" % output_data_product_id)
# Associate with DataProcess: register as an output product for this process
log.debug("Link data process %s and output out data product: %s (L4-CI-SA-RQ-260)", str(data_process_id), str(output_data_product_id))
self.clients.data_acquisition_management.assign_data_product(input_resource_id= data_process_id,data_product_id= output_data_product_id)
# Retrieve the id of the OUTPUT stream from the out Data Product
stream_ids, _ = self.clients.resource_registry.find_objects(output_data_product_id, PRED.hasStream, RT.Stream, True)
if not stream_ids:
raise NotFound("No Stream created for output Data Product " + str(output_data_product_id))
if len(stream_ids) != 1:
raise BadRequest("Data Product should only have ONE stream at this time" + str(output_data_product_id))
output_stream_dict[binding] = stream_ids[0]
#------------------------------------------------------------------------------------------------------------------------------------------
#Check for attached objects and put them into the configuration
#------------------------------------------------------------------------------------------------------------------------------------------
# check for attachments in data process definition
configuration = self._find_lookup_tables(data_process_definition_id, configuration)
input_stream_ids = []
if in_data_product_ids:
for in_data_product_id in in_data_product_ids:
self.clients.resource_registry.create_association(data_process_id, PRED.hasInputProduct, in_data_product_id)
log.debug("Associate data process workflows with source data products %s "
"hasInputProducts %s (L4-CI-SA-RQ-260)", str(data_process_id), str(in_data_product_ids))
#check if in data product is attached to an instrument, check instrumentDevice and InstrumentModel for lookup table attachments
instdevice_ids, _ = self.clients.resource_registry.find_subjects(RT.InstrumentDevice, PRED.hasOutputProduct, in_data_product_id, True)
for instdevice_id in instdevice_ids:
log.debug("Instrument device_id assoc to the input data product of this data process: %s (L4-CI-SA-RQ-231)", str(instdevice_id))
# check for attachments in instrument device
configuration = self._find_lookup_tables(instdevice_id, configuration)
instmodel_ids, _ = self.clients.resource_registry.find_objects(instdevice_id, PRED.hasModel, RT.InstrumentModel, True)
for instmodel_id in instmodel_ids:
# check for attachments in instrument model
configuration = self._find_lookup_tables(instmodel_id, configuration)
#------------------------------------------------------------------------------------------------------------------------------------------
# Get the input stream from the input_data_product, which should already be associated with a stream via the Data Producer
#------------------------------------------------------------------------------------------------------------------------------------------
input_stream_ids = self._get_input_stream_ids(in_data_product_ids)
#------------------------------------------------------------------------------------------------------------------------------------------
# Create subscription to the input stream
#------------------------------------------------------------------------------------------------------------------------------------------
input_subscription_id = self.clients.pubsub_management.create_subscription(name=data_process_name, stream_ids=input_stream_ids)
#------------------------------------------------------------------------------------------------------------------------------------------
# Add the subscription id to the data process
#------------------------------------------------------------------------------------------------------------------------------------------
data_process_obj.input_subscription_id = input_subscription_id
log.info("Launching the process")
debug_str = "\n\tQueue Name: %s\n\tOutput Streams: %s\n\tProcess Definition ID: %s\n\tConfiguration: %s" % (data_process_name, output_stream_dict, process_definition_id, configuration)
log.debug(debug_str)
pid = self._launch_process(
queue_name=data_process_name,
out_streams=output_stream_dict,
process_definition_id=process_definition_id,
configuration=configuration)
data_process_obj.process_id = pid
self.clients.resource_registry.update(data_process_obj)
return data_process_id
def _get_input_stream_ids(self, in_data_product_ids = None):
input_stream_ids = []
#------------------------------------------------------------------------------------------------------------------------------------------
# get the streams associated with this IN data products
#------------------------------------------------------------------------------------------------------------------------------------------
for in_data_product_id in in_data_product_ids:
# Get the stream associated with this input data product
stream_ids, _ = self.clients.resource_registry.find_objects(in_data_product_id, PRED.hasStream, RT.Stream, True)
validate_is_not_none( stream_ids, "No Stream created for this input Data Product " + str(in_data_product_id))
validate_is_not_none( len(stream_ids) != 1, "Input Data Product should only have ONE stream" + str(in_data_product_id))
# We take for now one stream_id associated with each input data product
input_stream_ids.append(stream_ids[0])
return input_stream_ids
def _launch_process(self, queue_name='', out_streams=None, process_definition_id='', configuration=None):
"""
Launches the process
"""
# ------------------------------------------------------------------------------------
# Spawn Configuration and Parameters
# ------------------------------------------------------------------------------------
if 'process' not in configuration:
configuration['process'] = {}
configuration['process']['queue_name'] = queue_name
configuration['process']['publish_streams'] = out_streams
# Setting the restart mode
schedule = ProcessSchedule()
schedule.restart_mode = ProcessRestartMode.ABNORMAL
# ------------------------------------------------------------------------------------
# Process Spawning
# ------------------------------------------------------------------------------------
# Spawn the process
pid = self.clients.process_dispatcher.schedule_process(
process_definition_id=process_definition_id,
schedule= schedule,
configuration=configuration
)
validate_is_not_none( pid, "Process could not be spawned")
return pid
def _find_lookup_tables(self, resource_id="", configuration=None):
#check if resource has lookup tables attached
configuration = configuration or DotDict()
attachment_objs, _ = self.clients.resource_registry.find_objects(resource_id, PRED.hasAttachment, RT.Attachment, False)
for attachment_obj in attachment_objs:
words = set(attachment_obj.keywords)
if 'DataProcessInput' in words:
configuration[attachment_obj.name] = attachment_obj.content
log.debug("Lookup table, %s, found in attachment %s" % (attachment_obj.content, attachment_obj.name))
else:
log.debug("NO lookup table in attachment %s" % attachment_obj.name)
return configuration
def update_data_process_inputs(self, data_process_id="", in_stream_ids=None):
#@TODO: INPUT STREAM VALIDATION
log.debug("Updating inputs to data process '%s'", data_process_id)
data_process_obj = self.clients.resource_registry.read(data_process_id)
subscription_id = data_process_obj.input_subscription_id
was_active = False
if subscription_id:
# get rid of all the current streams
try:
log.debug("Deactivating subscription '%s'", subscription_id)
self.clients.pubsub_management.deactivate_subscription(subscription_id)
was_active = True
except BadRequest:
log.info('Subscription was not active')
self.clients.pubsub_management.delete_subscription(subscription_id)
new_subscription_id = self.clients.pubsub_management.create_subscription(data_process_obj.name,
stream_ids=in_stream_ids)
data_process_obj.input_subscription_id = new_subscription_id
self.clients.resource_registry.update(data_process_obj)
if was_active:
log.debug("Activating subscription '%s'", new_subscription_id)
self.clients.pubsub_management.activate_subscription(new_subscription_id)
def update_data_process(self,):
#todo: What are valid ways to update a data process?.
return
def update_data_process2(self):
raise BadRequest('Cannot update an existing data process.')
def read_data_process(self, data_process_id=""):
data_proc_obj = self.clients.resource_registry.read(data_process_id)
return data_proc_obj
def read_data_process2(self, data_process_id=''):
data_proc_obj = self.clients.resource_registry.read(data_process_id)
return data_proc_obj
def delete_data_process(self, data_process_id=""):
# Delete the specified DataProcessDefinition object
data_process_obj = self.read_data_process(data_process_id)
log.debug("delete the association with DataProcessDefinition")
dpd_assn_ids = self.clients.resource_registry.find_associations(subject=data_process_id, predicate=PRED.hasProcessDefinition, id_only=True)
for dpd_assn_id in dpd_assn_ids:
log.debug("Trying to delete the association with this data process definition: %s", dpd_assn_id)
self.clients.resource_registry.delete_association(dpd_assn_id)
self._stop_process(data_process_obj)
log.debug("Finalizing data products by removing streams associated with the dataset and product")
out_products, assocs = self.clients.resource_registry.find_objects(subject=data_process_id, predicate=PRED.hasOutputProduct, id_only=True)
for out_product, assoc in zip(out_products, assocs):
self.data_product_management.remove_streams(out_product)
log.debug("deleting association with output data product '%s'" % out_product)
self.clients.resource_registry.delete_association(assoc)
self.clients.data_acquisition_management.unassign_data_product(data_process_id, out_product)
log.debug("Delete the input product links")
inprod_associations = self.clients.resource_registry.find_associations(data_process_id, PRED.hasInputProduct)
for inprod_association in inprod_associations:
self.clients.resource_registry.delete_association(inprod_association)
try:
self.deactivate_data_process(data_process_id=data_process_id)
log.warn('Deleteing activated data process...')
except BadRequest:
pass
subscription_id = data_process_obj.input_subscription_id
self.clients.pubsub_management.delete_subscription(subscription_id)
data_process_obj.input_subscription_id = None
self.clients.resource_registry.update(data_process_obj)
#unregister the data process in DataAcquisitionMgmtSvc
self.clients.data_acquisition_management.unregister_process(data_process_id)
# Delete the data process
self.clients.resource_registry.retire(data_process_id)
return
def delete_data_process2(self, data_process_id=''):
#Stops processes and deletes the data process associations
#TODO: Delete the processes also?
assocs = self.clients.resource_registry.find_associations(subject=data_process_id, predicate=RT.hasProcess, id_only=False)
for assoc in assocs:
self._stop_process(data_process=assoc.o)
self.clients.resource_registry.delete_association(assoc)
#Delete all subscriptions associations
assocs = self.clients.resource_registry.find_associations(subject=data_process_id, predicate=RT.hasSubscription, id_only=False)
for assoc in assocs:
#Delete subscription
self.clients.pubsub_management.deactivate_subscription(subscription_id=assoc.o._id)
self.clients.pubsub_management.delete_subscription(subscription_id=assoc.o._id)
self.clients.resource_registry.delete_association(assoc)
#Unassign data products
assocs = self.clients.resource_registry.find_associations(subject=data_process_id, predicate=RT.hasOutputProduct, id_only=False)
for assoc in assocs:
self.clients.data_acquisition_management.unassign_data_product(input_resource_id=data_process_id, data_product_id=assoc.o._id)
#Unregister the data process with acquisition
self.clients.data_acquisition_management.unregister_process(data_process_id=data_process_id)
#Delete the data process from the resource registry
self.clients.resource_registry.delete(object_id=data_process_id)
def force_delete_data_process(self, data_process_id=""):
# if not yet deleted, the first execute delete logic
dp_obj = self.read_data_process(data_process_id)
if dp_obj.lcstate != LCS.RETIRED:
self.delete_data_process(data_process_id)
self._remove_associations(data_process_id)
self.clients.resource_registry.delete(data_process_id)
def _stop_process(self, data_process):
log.debug("stopping data process '%s'" % data_process.process_id)
pid = data_process.process_id
self.clients.process_dispatcher.cancel_process(pid)
def find_data_process(self, filters=None):
"""
@param filters: dict of parameters to filter down
the list of possible data proc.
@retval
"""
#todo: add filter processing
data_process_list , _ = self.clients.resource_registry.find_resources(RT.DataProcess, None, None, True)
return data_process_list
def activate_data_process(self, data_process_id=""):
data_process_obj = self.read_data_process(data_process_id)
log.debug("activate_data_process:data_process_obj %s ", str(data_process_obj))
# #update the producer context with the activation time and the configuration
# todo: update the setting of this context with the return vals from process_dispatcher:schedule_process after convert
# todo: process_id, process_definition, schedule, configuration
producer_obj = self._get_process_producer(data_process_id)
producertype = type(producer_obj).__name__
#todo: producer_obj.producer_context.type_ is returning the base type, not the derived type.
if producer_obj.producer_context.type_ == OT.DataProcessProducerContext :
log.debug("activate_data_process:activation_time %s ", str(IonTime().to_string()))
producer_obj.producer_context.activation_time = IonTime().to_string()
producer_obj.producer_context.configuration = data_process_obj.configuration
self.clients.resource_registry.update(producer_obj)
subscription_id = data_process_obj.input_subscription_id
self.clients.pubsub_management.activate_subscription(subscription_id=subscription_id)
def deactivate_data_process(self, data_process_id=""):
data_process_obj = self.read_data_process(data_process_id)
if not data_process_obj.input_subscription_id:
log.warn("data process '%s' has no subscription id to deactivate", data_process_id)
return
subscription_obj = self.clients.pubsub_management.read_subscription(data_process_obj.input_subscription_id)
if subscription_obj.activated:
#update the producer context with the deactivation time
# todo: update the setting of this contect with the return vals from process_dispatcher:schedule_process after convert
producer_obj = self._get_process_producer(data_process_id)
producertype = type(producer_obj).__name__
if producer_obj.producer_context.type_ == OT.DataProcessProducerContext :
log.debug("data_process '%s' (producer '%s'): deactivation_time = %s ",
data_process_id, producer_obj._id, str(IonTime().to_string()))
producer_obj.producer_context.deactivation_time = IonTime().to_string()
self.clients.resource_registry.update(producer_obj)
subscription_id = data_process_obj.input_subscription_id
log.debug("Deactivating subscription '%s'", subscription_id)
self.clients.pubsub_management.deactivate_subscription(subscription_id=subscription_id)
def attach_process(self, process=''):
"""
@param process: Should this be the data_process_id?
@retval
"""
# TODO: Determine the proper input param
pass
def _get_stream_from_dp(self, dp_id):
stream_ids, _ = self.clients.resource_registry.find_objects(subject=dp_id, predicate=PRED.hasStream, id_only=True)
if not stream_ids: raise BadRequest('No streams associated with this data product')
return stream_ids[0]
def _manage_routes(self, routes):
retval = {}
for path,actor in routes.iteritems():
in_data_product_id, out_data_product_id = path
in_stream_id = self._get_stream_from_dp(in_data_product_id)
out_stream_id = self._get_stream_from_dp(out_data_product_id)
if actor:
self.clients.resource_registry.read(actor)
if isinstance(actor,TransformFunction):
actor = 'TODO: figure this out'
retval[(in_stream_id, out_stream_id)] = actor
return retval
def _manage_producers(self, data_process_id, data_product_ids):
self.clients.data_acquisition_management.register_process(data_process_id)
for data_product_id in data_product_ids:
producer_ids, _ = self.clients.resource_registry.find_objects(subject=data_product_id, predicate=PRED.hasDataProducer, object_type=RT.DataProducer, id_only=True)
if len(producer_ids):
raise BadRequest('Only one DataProducer allowed per DataProduct')
# Validate the data product
self.clients.data_product_management.read_data_product(data_product_id)
self.clients.data_acquisition_management.assign_data_product(input_resource_id=data_process_id, data_product_id=data_product_id)
if not self._get_stream_from_dp(data_product_id):
raise BadRequest('No Stream was found for this DataProduct')
def _manage_attachments(self):
pass
def _create_subscription(self, dproc, in_data_product_ids):
stream_ids = [self._get_stream_from_dp(i) for i in in_data_product_ids]
#@TODO Maybe associate a data process with an exchange point but in the mean time:
queue_name = 'sub_%s' % dproc.name
subscription_id = self.clients.pubsub_management.create_subscription(name=queue_name, stream_ids=stream_ids)
self.clients.resource_registry.create_association(subject=dproc._id, predicate=PRED.hasSubscription, object=subscription_id)
return queue_name
def _get_process_definition(self, data_process_definition_id=''):
process_definition_id = ''
if data_process_definition_id:
process_definitions, _ = self.clients.resource_registry.find_objects(subject=data_process_definition_id, predicate=PRED.hasProcessDefinition, id_only=True)
if process_definitions:
process_definition_id = process_definitions[0]
else:
process_definitions, _ = self.clients.resource_registry.find_resources(name='transform_data_process', restype=RT.ProcessDefinition,id_only=True)
if process_definitions:
process_definition_id = process_definitions[0]
else:
process_definition = ProcessDefinition()
process_definition.name = 'transform_data_process'
process_definition.executable['module'] = 'ion.processes.data.transforms.transform_prime'
process_definition.executable['class'] = 'TransformPrime'
process_definition_id = self.clients.process_dispatcher.create_process_definition(process_definition)
return process_definition_id
def _launch_data_process(self, queue_name='', data_process_definition_id='', out_data_product_ids=[], configuration={}):
process_definition_id = self._get_process_definition(data_process_definition_id)
out_streams = {}
for dp_id in out_data_product_ids:
stream_id = self._get_stream_from_dp(dp_id)
out_streams[stream_id] = stream_id
return self._launch_process(queue_name, out_streams, process_definition_id, configuration)
def _validator(self, in_data_product_id, out_data_product_id):
in_stream_id = self._get_stream_from_dp(dp_id=in_data_product_id)
in_stream_defs, _ = self.clients.resource_registry.find_objects(subject=in_stream_id, predicate=PRED.hasStreamDefinition, id_only=True)
if not len(in_stream_defs):
raise BadRequest('No valid stream definition defined for data product stream')
out_stream_id = self._get_stream_from_dp(dp_id=out_data_product_id)
out_stream_defs, _ = self.clients.resource_registry.find_objects(subject=out_stream_id, predicate=PRED.hasStreamDefinition, id_only=True)
if not len(out_stream_defs):
raise BadRequest('No valid stream definition defined for data product stream')
return self.clients.pubsub_management.compatible_stream_definitions(in_stream_definition_id=in_stream_defs[0], out_stream_definition_id=out_stream_defs[0])
def validate_compatibility(self, in_data_product_ids=None, out_data_product_ids=None, routes=None):
'''
Validates compatibility between input and output data products
routes are in this form:
{ (in_data_product_id, out_data_product_id) : actor }
if actor is None then the data process is assumed to use parameter functions.
if actor is a TransformFunction, the validation is done at runtime
'''
if len(out_data_product_ids)>1 and not routes:
raise BadRequest('Multiple output data products but no routes defined')
if len(out_data_product_ids)==1:
return all( [self._validator(i, out_data_product_ids[0]) for i in in_data_product_ids] )
elif len(out_data_product_ids)>1:
for path,actor in routes.iteritems():
in_dp_id, out_dp_id = path
if not self._validator(in_dp_id, out_dp_id):
return False
return True
else:
raise BadRequest('No input data products specified')
def _get_process_producer(self, data_process_id=""):
producer_objs, _ = self.clients.resource_registry.find_objects(subject=data_process_id, predicate=PRED.hasDataProducer, object_type=RT.DataProducer, id_only=False)
if not producer_objs:
raise NotFound("No Producers created for this Data Process " + str(data_process_id))
return producer_objs[0]
############################
#
# EXTENDED RESOURCES
#
############################
def get_data_process_definition_extension(self, data_process_definition_id='', ext_associations=None, ext_exclude=None, user_id=''):
#Returns an DataProcessDefinition Extension object containing additional related information
if not data_process_definition_id:
raise BadRequest("The data_process_definition_id parameter is empty")
extended_resource_handler = ExtendedResourceContainer(self)
extended_data_process_definition = extended_resource_handler.create_extended_resource_container(
extended_resource_type=OT.DataProcessDefinitionExtension,
resource_id=data_process_definition_id,
computed_resource_type=OT.DataProcessDefinitionComputedAttributes,
ext_associations=ext_associations,
ext_exclude=ext_exclude,
user_id=user_id)
#Loop through any attachments and remove the actual content since we don't need
# to send it to the front end this way
#TODO - see if there is a better way to do this in the extended resource frame work.
if hasattr(extended_data_process_definition, 'attachments'):
for att in extended_data_process_definition.attachments:
if hasattr(att, 'content'):
delattr(att, 'content')
# replace list of lists with single list
replacement_data_products = []
for inner_list in extended_data_process_definition.data_products:
if inner_list:
for actual_data_product in inner_list:
if actual_data_product:
replacement_data_products.append(actual_data_product)
extended_data_process_definition.data_products = replacement_data_products
return extended_data_process_definition
def get_data_process_extension(self, data_process_id='', ext_associations=None, ext_exclude=None, user_id=''):
#Returns an DataProcessDefinition Extension object containing additional related information
if not data_process_id:
raise BadRequest("The data_process_definition_id parameter is empty")
extended_resource_handler = ExtendedResourceContainer(self)
extended_data_process = extended_resource_handler.create_extended_resource_container(
extended_resource_type=OT.DataProcessExtension,
resource_id=data_process_id,
computed_resource_type=OT.DataProcessComputedAttributes,
ext_associations=ext_associations,
ext_exclude=ext_exclude,
user_id=user_id)
#Loop through any attachments and remove the actual content since we don't need
# to send it to the front end this way
#TODO - see if there is a better way to do this in the extended resource frame work.
if hasattr(extended_data_process, 'attachments'):
for att in extended_data_process.attachments:
if hasattr(att, 'content'):
delattr(att, 'content')
return extended_data_process
def _remove_associations(self, resource_id=''):
"""
delete all associations to/from a resource
"""
# find all associations where this is the subject
_, obj_assns = self.clients.resource_registry.find_objects(subject=resource_id, id_only=True)
# find all associations where this is the object
_, sbj_assns = self.clients.resource_registry.find_subjects(object=resource_id, id_only=True)
log.debug("pluck will remove %s subject associations and %s object associations",
len(sbj_assns), len(obj_assns))
for assn in obj_assns:
log.debug("pluck deleting object association %s", assn)
self.clients.resource_registry.delete_association(assn)
for assn in sbj_assns:
log.debug("pluck deleting subject association %s", assn)
self.clients.resource_registry.delete_association(assn)
# find all associations where this is the subject
_, obj_assns = self.clients.resource_registry.find_objects(subject=resource_id, id_only=True)
# find all associations where this is the object
_, sbj_assns = self.clients.resource_registry.find_subjects(object=resource_id, id_only=True)
log.debug("post-deletions, pluck found %s subject associations and %s object associations",
len(sbj_assns), len(obj_assns))
