def init_job_queue():
if RuntimeConfig.WORK_MODE == WorkMode.STANDALONE:
job_queue = InProcessQueue()
RuntimeConfig.init_config(JOB_QUEUE=job_queue)
elif RuntimeConfig.WORK_MODE == WorkMode.CLUSTER:
job_queue = RedisQueue(queue_name='fate_flow_job_queue', host=REDIS['host'], port=REDIS['port'],
password=REDIS['password'], max_connections=REDIS['max_connections'])
RuntimeConfig.init_config(JOB_QUEUE=job_queue)
else:
raise Exception('init queue failed.')
def __init__(self):
database_config = DATABASE.copy()
db_name = database_config.pop("name")
if WORK_MODE == WorkMode.STANDALONE:
self.database_connection = APSWDatabase('fate_flow_sqlite.db')
RuntimeConfig.init_config(USE_LOCAL_DATABASE=True)
stat_logger.info('init sqlite database on standalone mode successfully')
elif WORK_MODE == WorkMode.CLUSTER:
self.database_connection = PooledMySQLDatabase(db_name, **database_config)
stat_logger.info('init mysql database on cluster mode successfully')
RuntimeConfig.init_config(USE_LOCAL_DATABASE=False)
else:
raise Exception('can not init database')
def save_pipelined_model(cls, job_id, role, party_id):
schedule_logger(job_id).info(
'job {} on {} {} start to save pipeline'.format(
job_id, role, party_id))
job_dsl, job_runtime_conf, train_runtime_conf = job_utils.get_job_configuration(
job_id=job_id, role=role, party_id=party_id)
job_parameters = job_runtime_conf.get('job_parameters', {})
model_id = job_parameters['model_id']
model_version = job_parameters['model_version']
job_type = job_parameters.get('job_type', '')
if job_type == 'predict':
return
dag = schedule_utils.get_job_dsl_parser(
dsl=job_dsl,
runtime_conf=job_runtime_conf,
train_runtime_conf=train_runtime_conf)
predict_dsl = dag.get_predict_dsl(role=role)
pipeline = pipeline_pb2.Pipeline()
pipeline.inference_dsl = json_dumps(predict_dsl, byte=True)
pipeline.train_dsl = json_dumps(job_dsl, byte=True)
pipeline.train_runtime_conf = json_dumps(job_runtime_conf, byte=True)
pipeline.fate_version = RuntimeConfig.get_env("FATE")
pipeline.model_id = model_id
pipeline.model_version = model_version
tracker = Tracker(job_id=job_id,
role=role,
party_id=party_id,
model_id=model_id,
model_version=model_version)
tracker.save_pipelined_model(pipelined_buffer_object=pipeline)
if role != 'local':
tracker.save_machine_learning_model_info()
schedule_logger(job_id).info(
'job {} on {} {} save pipeline successfully'.format(
job_id, role, party_id))
def save_pipelined_model(cls, job_id, role, party_id):
schedule_logger(job_id).info(
'job {} on {} {} start to save pipeline'.format(
job_id, role, party_id))
job_dsl, job_runtime_conf, runtime_conf_on_party, train_runtime_conf = job_utils.get_job_configuration(
job_id=job_id, role=role, party_id=party_id)
job_parameters = runtime_conf_on_party.get('job_parameters', {})
if role in job_parameters.get("assistant_role", []):
return
model_id = job_parameters['model_id']
model_version = job_parameters['model_version']
job_type = job_parameters.get('job_type', '')
work_mode = job_parameters['work_mode']
roles = runtime_conf_on_party['role']
initiator_role = runtime_conf_on_party['initiator']['role']
initiator_party_id = runtime_conf_on_party['initiator']['party_id']
if job_type == 'predict':
return
dag = schedule_utils.get_job_dsl_parser(
dsl=job_dsl,
runtime_conf=job_runtime_conf,
train_runtime_conf=train_runtime_conf)
predict_dsl = dag.get_predict_dsl(role=role)
pipeline = pipeline_pb2.Pipeline()
pipeline.inference_dsl = json_dumps(predict_dsl, byte=True)
pipeline.train_dsl = json_dumps(job_dsl, byte=True)
pipeline.train_runtime_conf = json_dumps(job_runtime_conf, byte=True)
pipeline.fate_version = RuntimeConfig.get_env("FATE")
pipeline.model_id = model_id
pipeline.model_version = model_version
pipeline.parent = True
pipeline.loaded_times = 0
pipeline.roles = json_dumps(roles, byte=True)
pipeline.work_mode = work_mode
pipeline.initiator_role = initiator_role
pipeline.initiator_party_id = initiator_party_id
pipeline.runtime_conf_on_party = json_dumps(runtime_conf_on_party,
byte=True)
pipeline.parent_info = json_dumps({}, byte=True)
tracker = Tracker(job_id=job_id,
role=role,
party_id=party_id,
model_id=model_id,
model_version=model_version)
tracker.save_pipelined_model(pipelined_buffer_object=pipeline)
if role != 'local':
tracker.save_machine_learning_model_info()
schedule_logger(job_id).info(
'job {} on {} {} save pipeline successfully'.format(
job_id, role, party_id))
def run_task(cls):
task_info = {}
try:
parser = argparse.ArgumentParser()
parser.add_argument('-j',
'--job_id',
required=True,
type=str,
help="job id")
parser.add_argument('-n',
'--component_name',
required=True,
type=str,
help="component name")
parser.add_argument('-t',
'--task_id',
required=True,
type=str,
help="task id")
parser.add_argument('-v',
'--task_version',
required=True,
type=int,
help="task version")
parser.add_argument('-r',
'--role',
required=True,
type=str,
help="role")
parser.add_argument('-p',
'--party_id',
required=True,
type=int,
help="party id")
parser.add_argument('-c',
'--config',
required=True,
type=str,
help="task parameters")
parser.add_argument('--run_ip', help="run ip", type=str)
parser.add_argument('--job_server', help="job server", type=str)
args = parser.parse_args()
schedule_logger(args.job_id).info('enter task process')
schedule_logger(args.job_id).info(args)
# init function args
if args.job_server:
RuntimeConfig.init_config(
JOB_SERVER_HOST=args.job_server.split(':')[0],
HTTP_PORT=args.job_server.split(':')[1])
RuntimeConfig.set_process_role(ProcessRole.EXECUTOR)
job_id = args.job_id
component_name = args.component_name
task_id = args.task_id
task_version = args.task_version
role = args.role
party_id = args.party_id
executor_pid = os.getpid()
task_info.update({
"job_id": job_id,
"component_name": component_name,
"task_id": task_id,
"task_version": task_version,
"role": role,
"party_id": party_id,
"run_ip": args.run_ip,
"run_pid": executor_pid
})
start_time = current_timestamp()
job_conf = job_utils.get_job_conf(job_id, role)
job_dsl = job_conf["job_dsl_path"]
job_runtime_conf = job_conf["job_runtime_conf_path"]
dsl_parser = schedule_utils.get_job_dsl_parser(
dsl=job_dsl,
runtime_conf=job_runtime_conf,
train_runtime_conf=job_conf["train_runtime_conf_path"],
pipeline_dsl=job_conf["pipeline_dsl_path"])
party_index = job_runtime_conf["role"][role].index(party_id)
job_args_on_party = TaskExecutor.get_job_args_on_party(
dsl_parser, job_runtime_conf, role, party_id)
component = dsl_parser.get_component_info(
component_name=component_name)
component_parameters = component.get_role_parameters()
component_parameters_on_party = component_parameters[role][
party_index] if role in component_parameters else {}
module_name = component.get_module()
task_input_dsl = component.get_input()
task_output_dsl = component.get_output()
component_parameters_on_party[
'output_data_name'] = task_output_dsl.get('data')
task_parameters = RunParameters(
**file_utils.load_json_conf(args.config))
job_parameters = task_parameters
if job_parameters.assistant_role:
TaskExecutor.monkey_patch()
except Exception as e:
traceback.print_exc()
schedule_logger().exception(e)
task_info["party_status"] = TaskStatus.FAILED
return
try:
job_log_dir = os.path.join(
job_utils.get_job_log_directory(job_id=job_id), role,
str(party_id))
task_log_dir = os.path.join(job_log_dir, component_name)
log.LoggerFactory.set_directory(directory=task_log_dir,
parent_log_dir=job_log_dir,
append_to_parent_log=True,
force=True)
tracker = Tracker(job_id=job_id,
role=role,
party_id=party_id,
component_name=component_name,
task_id=task_id,
task_version=task_version,
model_id=job_parameters.model_id,
model_version=job_parameters.model_version,
component_module_name=module_name,
job_parameters=job_parameters)
tracker_client = TrackerClient(
job_id=job_id,
role=role,
party_id=party_id,
component_name=component_name,
task_id=task_id,
task_version=task_version,
model_id=job_parameters.model_id,
model_version=job_parameters.model_version,
component_module_name=module_name,
job_parameters=job_parameters)
run_class_paths = component_parameters_on_party.get(
'CodePath').split('/')
run_class_package = '.'.join(
run_class_paths[:-2]) + '.' + run_class_paths[-2].replace(
'.py', '')
run_class_name = run_class_paths[-1]
task_info["party_status"] = TaskStatus.RUNNING
cls.report_task_update_to_driver(task_info=task_info)
# init environment, process is shared globally
RuntimeConfig.init_config(
WORK_MODE=job_parameters.work_mode,
COMPUTING_ENGINE=job_parameters.computing_engine,
FEDERATION_ENGINE=job_parameters.federation_engine,
FEDERATED_MODE=job_parameters.federated_mode)
if RuntimeConfig.COMPUTING_ENGINE == ComputingEngine.EGGROLL:
session_options = task_parameters.eggroll_run.copy()
else:
session_options = {}
sess = session.Session(
computing_type=job_parameters.computing_engine,
federation_type=job_parameters.federation_engine)
computing_session_id = job_utils.generate_session_id(
task_id, task_version, role, party_id)
sess.init_computing(computing_session_id=computing_session_id,
options=session_options)
federation_session_id = job_utils.generate_task_version_id(
task_id, task_version)
component_parameters_on_party[
"job_parameters"] = job_parameters.to_dict()
sess.init_federation(
federation_session_id=federation_session_id,
runtime_conf=component_parameters_on_party,
service_conf=job_parameters.engines_address.get(
EngineType.FEDERATION, {}))
sess.as_default()
schedule_logger().info('Run {} {} {} {} {} task'.format(
job_id, component_name, task_id, role, party_id))
schedule_logger().info("Component parameters on party {}".format(
component_parameters_on_party))
schedule_logger().info("Task input dsl {}".format(task_input_dsl))
task_run_args = cls.get_task_run_args(
job_id=job_id,
role=role,
party_id=party_id,
task_id=task_id,
task_version=task_version,
job_args=job_args_on_party,
job_parameters=job_parameters,
task_parameters=task_parameters,
input_dsl=task_input_dsl,
)
if module_name in {"Upload", "Download", "Reader", "Writer"}:
task_run_args["job_parameters"] = job_parameters
run_object = getattr(importlib.import_module(run_class_package),
run_class_name)()
run_object.set_tracker(tracker=tracker_client)
run_object.set_task_version_id(
task_version_id=job_utils.generate_task_version_id(
task_id, task_version))
# add profile logs
profile.profile_start()
run_object.run(component_parameters_on_party, task_run_args)
profile.profile_ends()
output_data = run_object.save_data()
if not isinstance(output_data, list):
output_data = [output_data]
for index in range(0, len(output_data)):
data_name = task_output_dsl.get(
'data')[index] if task_output_dsl.get(
'data') else '{}'.format(index)
persistent_table_namespace, persistent_table_name = tracker.save_output_data(
computing_table=output_data[index],
output_storage_engine=job_parameters.storage_engine,
output_storage_address=job_parameters.engines_address.get(
EngineType.STORAGE, {}))
if persistent_table_namespace and persistent_table_name:
tracker.log_output_data_info(
data_name=data_name,
table_namespace=persistent_table_namespace,
table_name=persistent_table_name)
output_model = run_object.export_model()
# There is only one model output at the current dsl version.
tracker.save_output_model(
output_model, task_output_dsl['model'][0]
if task_output_dsl.get('model') else 'default')
task_info["party_status"] = TaskStatus.SUCCESS
except Exception as e:
task_info["party_status"] = TaskStatus.FAILED
schedule_logger().exception(e)
finally:
try:
task_info["end_time"] = current_timestamp()
task_info["elapsed"] = task_info["end_time"] - start_time
cls.report_task_update_to_driver(task_info=task_info)
except Exception as e:
task_info["party_status"] = TaskStatus.FAILED
traceback.print_exc()
schedule_logger().exception(e)
schedule_logger().info('task {} {} {} start time: {}'.format(
task_id, role, party_id, timestamp_to_date(start_time)))
schedule_logger().info('task {} {} {} end time: {}'.format(
task_id, role, party_id, timestamp_to_date(task_info["end_time"])))
schedule_logger().info('task {} {} {} takes {}s'.format(
task_id, role, party_id,
int(task_info["elapsed"]) / 1000))
schedule_logger().info('Finish {} {} {} {} {} {} task {}'.format(
job_id, component_name, task_id, task_version, role, party_id,
task_info["party_status"]))
print('Finish {} {} {} {} {} {} task {}'.format(
job_id, component_name, task_id, task_version, role, party_id,
task_info["party_status"]))
return task_info
})
# init
signal.signal(signal.SIGTERM, job_utils.cleaning)
signal.signal(signal.SIGCHLD, job_utils.wait_child_process)
# init db
init_flow_db()
init_arch_db()
# init runtime config
import argparse
parser = argparse.ArgumentParser()
parser.add_argument('--standalone_node',
default=False,
help="if standalone node mode or not ",
action='store_true')
args = parser.parse_args()
RuntimeConfig.init_env()
RuntimeConfig.set_process_role(ProcessRole.DRIVER)
PrivilegeAuth.init()
ServiceUtils.register()
ResourceManager.initialize()
Detector(interval=5 * 1000).start()
DAGScheduler(interval=2 * 1000).start()
server = grpc.server(
futures.ThreadPoolExecutor(max_workers=10),
options=[(cygrpc.ChannelArgKey.max_send_message_length, -1),
(cygrpc.ChannelArgKey.max_receive_message_length, -1)])
proxy_pb2_grpc.add_DataTransferServiceServicer_to_server(
UnaryService(), server)
server.add_insecure_port("{}:{}".format(IP, GRPC_PORT))
server.start()
'host': 'mysql',
'port': 3306,
'max_connections': 100,
'stale_timeout': 30,
}
REDIS = {
'host': 'redis',
'port': 6379,
'password': '******',
'max_connections': 500
}
REDIS_QUEUE_DB_INDEX = 0
JOB_MODULE_CONF = file_utils.load_json_conf("fate_flow/job_module_conf.json")
"""
Services
"""
server_conf = file_utils.load_json_conf("arch/conf/server_conf.json")
PROXY_HOST = server_conf.get(SERVERS).get('proxy').get('host')
PROXY_PORT = server_conf.get(SERVERS).get('proxy').get('port')
BOARD_HOST = server_conf.get(SERVERS).get('fateboard').get('host')
if BOARD_HOST == 'localhost':
BOARD_HOST = get_lan_ip()
BOARD_PORT = server_conf.get(SERVERS).get('fateboard').get('port')
SERVINGS = server_conf.get(SERVERS).get('servings')
BOARD_DASHBOARD_URL = 'http://%s:%d/index.html#/dashboard?job_id={}&role={}&party_id={}' % (BOARD_HOST, BOARD_PORT)
RuntimeConfig.init_config(WORK_MODE=WORK_MODE)
RuntimeConfig.init_config(HTTP_PORT=HTTP_PORT)
def run_task():
task = Task()
task.f_create_time = current_timestamp()
try:
parser = argparse.ArgumentParser()
parser.add_argument('-j', '--job_id', required=True, type=str, help="job id")
parser.add_argument('-n', '--component_name', required=True, type=str,
help="component name")
parser.add_argument('-t', '--task_id', required=True, type=str, help="task id")
parser.add_argument('-r', '--role', required=True, type=str, help="role")
parser.add_argument('-p', '--party_id', required=True, type=str, help="party id")
parser.add_argument('-c', '--config', required=True, type=str, help="task config")
parser.add_argument('--processors_per_node', help="processors_per_node", type=int)
parser.add_argument('--job_server', help="job server", type=str)
args = parser.parse_args()
schedule_logger(args.job_id).info('enter task process')
schedule_logger(args.job_id).info(args)
# init function args
if args.job_server:
RuntimeConfig.init_config(HTTP_PORT=args.job_server.split(':')[1])
RuntimeConfig.set_process_role(ProcessRole.EXECUTOR)
job_id = args.job_id
component_name = args.component_name
task_id = args.task_id
role = args.role
party_id = int(args.party_id)
executor_pid = os.getpid()
task_config = file_utils.load_json_conf(args.config)
job_parameters = task_config['job_parameters']
job_initiator = task_config['job_initiator']
job_args = task_config['job_args']
task_input_dsl = task_config['input']
task_output_dsl = task_config['output']
component_parameters = TaskExecutor.get_parameters(job_id, component_name, role, party_id)
task_parameters = task_config['task_parameters']
module_name = task_config['module_name']
TaskExecutor.monkey_patch()
except Exception as e:
traceback.print_exc()
schedule_logger().exception(e)
task.f_status = TaskStatus.FAILED
return
try:
job_log_dir = os.path.join(job_utils.get_job_log_directory(job_id=job_id), role, str(party_id))
task_log_dir = os.path.join(job_log_dir, component_name)
log_utils.LoggerFactory.set_directory(directory=task_log_dir, parent_log_dir=job_log_dir,
append_to_parent_log=True, force=True)
task.f_job_id = job_id
task.f_component_name = component_name
task.f_task_id = task_id
task.f_role = role
task.f_party_id = party_id
task.f_operator = 'python_operator'
tracker = Tracking(job_id=job_id, role=role, party_id=party_id, component_name=component_name,
task_id=task_id,
model_id=job_parameters['model_id'],
model_version=job_parameters['model_version'],
component_module_name=module_name)
task.f_start_time = current_timestamp()
task.f_run_ip = get_lan_ip()
task.f_run_pid = executor_pid
run_class_paths = component_parameters.get('CodePath').split('/')
run_class_package = '.'.join(run_class_paths[:-2]) + '.' + run_class_paths[-2].replace('.py', '')
run_class_name = run_class_paths[-1]
task.f_status = TaskStatus.RUNNING
TaskExecutor.sync_task_status(job_id=job_id, component_name=component_name, task_id=task_id, role=role,
party_id=party_id, initiator_party_id=job_initiator.get('party_id', None),
initiator_role=job_initiator.get('role', None),
task_info=task.to_json())
# init environment, process is shared globally
RuntimeConfig.init_config(WORK_MODE=job_parameters['work_mode'],
BACKEND=job_parameters.get('backend', 0))
if args.processors_per_node and args.processors_per_node > 0 and RuntimeConfig.BACKEND == Backend.EGGROLL:
session_options = {"eggroll.session.processors.per.node": args.processors_per_node}
else:
session_options = {}
session.init(job_id=job_utils.generate_session_id(task_id, role, party_id),
mode=RuntimeConfig.WORK_MODE,
backend=RuntimeConfig.BACKEND,
options=session_options)
federation.init(job_id=task_id, runtime_conf=component_parameters)
schedule_logger().info('run {} {} {} {} {} task'.format(job_id, component_name, task_id, role, party_id))
schedule_logger().info(component_parameters)
schedule_logger().info(task_input_dsl)
task_run_args = TaskExecutor.get_task_run_args(job_id=job_id, role=role, party_id=party_id,
task_id=task_id,
job_args=job_args,
job_parameters=job_parameters,
task_parameters=task_parameters,
input_dsl=task_input_dsl,
if_save_as_task_input_data=job_parameters.get("save_as_task_input_data", SAVE_AS_TASK_INPUT_DATA_SWITCH)
)
run_object = getattr(importlib.import_module(run_class_package), run_class_name)()
run_object.set_tracker(tracker=tracker)
run_object.set_taskid(taskid=task_id)
run_object.run(component_parameters, task_run_args)
output_data = run_object.save_data()
tracker.save_output_data_table(output_data, task_output_dsl.get('data')[0] if task_output_dsl.get('data') else 'component')
output_model = run_object.export_model()
# There is only one model output at the current dsl version.
tracker.save_output_model(output_model, task_output_dsl['model'][0] if task_output_dsl.get('model') else 'default')
task.f_status = TaskStatus.COMPLETE
except Exception as e:
task.f_status = TaskStatus.FAILED
schedule_logger().exception(e)
finally:
sync_success = False
try:
task.f_end_time = current_timestamp()
task.f_elapsed = task.f_end_time - task.f_start_time
task.f_update_time = current_timestamp()
TaskExecutor.sync_task_status(job_id=job_id, component_name=component_name, task_id=task_id, role=role,
party_id=party_id,
initiator_party_id=job_initiator.get('party_id', None),
initiator_role=job_initiator.get('role', None),
task_info=task.to_json())
sync_success = True
except Exception as e:
traceback.print_exc()
schedule_logger().exception(e)
schedule_logger().info('task {} {} {} start time: {}'.format(task_id, role, party_id, timestamp_to_date(task.f_start_time)))
schedule_logger().info('task {} {} {} end time: {}'.format(task_id, role, party_id, timestamp_to_date(task.f_end_time)))
schedule_logger().info('task {} {} {} takes {}s'.format(task_id, role, party_id, int(task.f_elapsed)/1000))
schedule_logger().info(
'finish {} {} {} {} {} {} task'.format(job_id, component_name, task_id, role, party_id, task.f_status if sync_success else TaskStatus.FAILED))
print('finish {} {} {} {} {} {} task'.format(job_id, component_name, task_id, role, party_id, task.f_status if sync_success else TaskStatus.FAILED))
def get_fate_version_info():
version = RuntimeConfig.get_env(request.json.get('module', 'FATE'))
return get_json_result(data={request.json.get('module'): version})
'/{}/forward'.format(API_VERSION): forward_app_manager
})
# init
signal.signal(signal.SIGTERM, job_utils.cleaning)
signal.signal(signal.SIGCHLD, job_utils.wait_child_process)
init_database_tables()
# init runtime config
import argparse
parser = argparse.ArgumentParser()
parser.add_argument('--standalone_node',
default=False,
help="if standalone node mode or not ",
action='store_true')
args = parser.parse_args()
if args.standalone_node:
RuntimeConfig.init_config(WORK_MODE=WorkMode.STANDALONE)
RuntimeConfig.init_config(HTTP_PORT=CLUSTER_STANDALONE_JOB_SERVER_PORT)
session_utils.init_session_for_flow_server()
RuntimeConfig.init_env()
RuntimeConfig.set_process_role(ProcessRole.SERVER)
queue_manager.init_job_queue()
job_controller.JobController.init()
history_job_clean = job_controller.JobClean()
history_job_clean.start()
PrivilegeAuth.init()
ServiceUtils.register()
# start job detector
job_detector.JobDetector(interval=5 * 1000).start()
# start scheduler
scheduler = dag_scheduler.DAGScheduler(
queue=RuntimeConfig.JOB_QUEUE, concurrent_num=MAX_CONCURRENT_JOB_RUN)
def run_task():
task = Task()
task.f_create_time = current_timestamp()
try:
parser = argparse.ArgumentParser()
parser.add_argument('-j',
'--job_id',
required=True,
type=str,
help="job id")
parser.add_argument('-n',
'--component_name',
required=True,
type=str,
help="component name")
parser.add_argument('-t',
'--task_id',
required=True,
type=str,
help="task id")
parser.add_argument('-r',
'--role',
required=True,
type=str,
help="role")
parser.add_argument('-p',
'--party_id',
required=True,
type=str,
help="party id")
parser.add_argument('-c',
'--config',
required=True,
type=str,
help="task config")
parser.add_argument('--job_server', help="job server", type=str)
args = parser.parse_args()
schedule_logger.info('enter task process')
schedule_logger.info(args)
# init function args
if args.job_server:
RuntimeConfig.init_config(
HTTP_PORT=args.job_server.split(':')[1])
job_id = args.job_id
component_name = args.component_name
task_id = args.task_id
role = args.role
party_id = int(args.party_id)
task_config = file_utils.load_json_conf(args.config)
job_parameters = task_config['job_parameters']
job_initiator = task_config['job_initiator']
job_args = task_config['job_args']
task_input_dsl = task_config['input']
task_output_dsl = task_config['output']
parameters = task_config['parameters']
module_name = task_config['module_name']
except Exception as e:
schedule_logger.exception(e)
task.f_status = TaskStatus.FAILED
return
try:
# init environment, process is shared globally
RuntimeConfig.init_config(WORK_MODE=job_parameters['work_mode'])
storage.init_storage(job_id=task_id,
work_mode=RuntimeConfig.WORK_MODE)
federation.init(job_id=task_id, runtime_conf=parameters)
job_log_dir = os.path.join(
job_utils.get_job_log_directory(job_id=job_id), role,
str(party_id))
task_log_dir = os.path.join(job_log_dir, component_name)
log_utils.LoggerFactory.set_directory(directory=task_log_dir,
parent_log_dir=job_log_dir,
append_to_parent_log=True,
force=True)
task.f_job_id = job_id
task.f_component_name = component_name
task.f_task_id = task_id
task.f_role = role
task.f_party_id = party_id
task.f_operator = 'python_operator'
tracker = Tracking(job_id=job_id,
role=role,
party_id=party_id,
component_name=component_name,
task_id=task_id,
model_id=job_parameters['model_id'],
model_version=job_parameters['model_version'],
module_name=module_name)
task.f_start_time = current_timestamp()
task.f_run_ip = get_lan_ip()
task.f_run_pid = os.getpid()
run_class_paths = parameters.get('CodePath').split('/')
run_class_package = '.'.join(
run_class_paths[:-2]) + '.' + run_class_paths[-2].replace(
'.py', '')
run_class_name = run_class_paths[-1]
task_run_args = TaskExecutor.get_task_run_args(
job_id=job_id,
role=role,
party_id=party_id,
job_parameters=job_parameters,
job_args=job_args,
input_dsl=task_input_dsl)
run_object = getattr(importlib.import_module(run_class_package),
run_class_name)()
run_object.set_tracker(tracker=tracker)
run_object.set_taskid(taskid=task_id)
task.f_status = TaskStatus.RUNNING
TaskExecutor.sync_task_status(job_id=job_id,
component_name=component_name,
task_id=task_id,
role=role,
party_id=party_id,
initiator_party_id=job_initiator.get(
'party_id', None),
task_info=task.to_json())
schedule_logger.info('run {} {} {} {} {} task'.format(
job_id, component_name, task_id, role, party_id))
schedule_logger.info(parameters)
schedule_logger.info(task_input_dsl)
run_object.run(parameters, task_run_args)
if task_output_dsl:
if task_output_dsl.get('data', []):
output_data = run_object.save_data()
tracker.save_output_data_table(
output_data,
task_output_dsl.get('data')[0])
if task_output_dsl.get('model', []):
output_model = run_object.export_model()
# There is only one model output at the current dsl version.
tracker.save_output_model(output_model,
task_output_dsl['model'][0])
task.f_status = TaskStatus.SUCCESS
except Exception as e:
schedule_logger.exception(e)
task.f_status = TaskStatus.FAILED
finally:
try:
task.f_end_time = current_timestamp()
task.f_elapsed = task.f_end_time - task.f_start_time
task.f_update_time = current_timestamp()
TaskExecutor.sync_task_status(
job_id=job_id,
component_name=component_name,
task_id=task_id,
role=role,
party_id=party_id,
initiator_party_id=job_initiator.get('party_id', None),
task_info=task.to_json())
except Exception as e:
schedule_logger.exception(e)
schedule_logger.info('finish {} {} {} {} {} {} task'.format(
job_id, component_name, task_id, role, party_id, task.f_status))
print('finish {} {} {} {} {} {} task'.format(job_id, component_name,
task_id, role, party_id,
task.f_status))
'/{}/pipeline'.format(API_VERSION): pipeline_app_manager,
'/{}/schedule'.format(API_VERSION): schedule_app_manager
})
# init
signal.signal(signal.SIGCHLD, job_utils.wait_child_process)
init_database_tables()
# init runtime config
import argparse
parser = argparse.ArgumentParser()
parser.add_argument('--standalone_node',
default=False,
help="if standalone node mode or not ",
action='store_true')
args = parser.parse_args()
if args.standalone_node:
RuntimeConfig.init_config(WORK_MODE=WorkMode.STANDALONE)
RuntimeConfig.init_config(HTTP_PORT=CLUSTER_STANDALONE_JOB_SERVER_PORT)
storage.init_storage(work_mode=RuntimeConfig.WORK_MODE)
queue_manager.init_job_queue()
job_controller.JobController.init()
# start job detector
job_detector.JobDetector(interval=5 * 1000).start()
# start scheduler
scheduler = dag_scheduler.DAGScheduler(
queue=RuntimeConfig.JOB_QUEUE, concurrent_num=MAX_CONCURRENT_JOB_RUN)
scheduler.start()
# start grpc server
server = grpc.server(
futures.ThreadPoolExecutor(max_workers=10),
options=[(cygrpc.ChannelArgKey.max_send_message_length, -1),
# Endpoint
FATE_FLOW_MODEL_TRANSFER_ENDPOINT = "/v1/model/transfer"
FATE_MANAGER_GET_NODE_INFO_ENDPOINT = "/fate-manager/api/site/secretinfo"
FATE_MANAGER_NODE_CHECK_ENDPOINT = "/fate-manager/api/site/checksite"
FATE_BOARD_DASHBOARD_ENDPOINT = "/index.html#/dashboard?job_id={}&role={}&party_id={}"
# Logger
log.LoggerFactory.LEVEL = 10
# {CRITICAL: 50, FATAL:50, ERROR:40, WARNING:30, WARN:30, INFO:20, DEBUG:10, NOTSET:0}
log.LoggerFactory.set_directory(
os.path.join(file_utils.get_project_base_directory(), "logs", "fate_flow"))
stat_logger = log.getLogger("fate_flow_stat")
detect_logger = log.getLogger("fate_flow_detect")
access_logger = log.getLogger("fate_flow_access")
data_manager_logger = log.getLogger("fate_flow_data_manager")
peewee_logger = log.getLogger("peewee")
# Switch
UPLOAD_DATA_FROM_CLIENT = True
USE_AUTHENTICATION = False
PRIVILEGE_COMMAND_WHITELIST = []
CHECK_NODES_IDENTITY = False
DEFAULT_FEDERATED_STATUS_COLLECT_TYPE = get_base_config(
FATEFLOW_SERVICE_NAME, {}).get("default_federated_status_collect_type",
"PUSH")
# Init
RuntimeConfig.init_config(WORK_MODE=WORK_MODE)
RuntimeConfig.init_config(JOB_SERVER_HOST=IP, HTTP_PORT=HTTP_PORT)
if __name__ == '__main__':
manager.url_map.strict_slashes = False
app = DispatcherMiddleware(
manager, {
'/{}/data'.format(API_VERSION): data_access_app_manager,
'/{}/model'.format(API_VERSION): model_app_manager,
'/{}/job'.format(API_VERSION): job_app_manager,
'/{}/table'.format(API_VERSION): table_app_manager,
'/{}/tracking'.format(API_VERSION): tracking_app_manager,
'/{}/pipeline'.format(API_VERSION): pipeline_app_manager,
})
# init
signal.signal(signal.SIGCHLD, job_utils.wait_child_process)
init_database_tables()
RuntimeConfig.init_config(WORK_MODE=WORK_MODE)
storage.init_storage(work_mode=RuntimeConfig.WORK_MODE)
queue_manager.init_job_queue()
job_controller.JobController.init()
# start job detector
job_detector.JobDetector(interval=5 * 1000).start()
# start scheduler
scheduler = dag_scheduler.DAGScheduler(
queue=RuntimeConfig.JOB_QUEUE, concurrent_num=MAX_CONCURRENT_JOB_RUN)
scheduler.start()
# start grpc server
server = grpc.server(
futures.ThreadPoolExecutor(max_workers=10),
options=[(cygrpc.ChannelArgKey.max_send_message_length, -1),
(cygrpc.ChannelArgKey.max_receive_message_length, -1)])
