def run(self):
time.sleep(5)
jobs = job_utils.query_job(status='running', is_initiator=1)
job_ids = set([job.f_job_id for job in jobs])
for job_id in job_ids:
schedule_logger(job_id).info('fate flow server start clean job')
TaskScheduler.stop(job_id, JobStatus.FAILED)
def stop_job():
job_id = request.json.get('job_id')
response = TaskScheduler.stop(job_id=job_id, end_status=JobStatus.CANCELED)
if not response:
TaskScheduler.stop(job_id=request.json.get('job_id', ''),
end_status=JobStatus.FAILED)
return get_json_result(retcode=0, retmsg='kill job success')
return get_json_result(retcode=0, retmsg='cancel job success')
def submit_job(job_data):
job_id = generate_job_id()
schedule_logger.info('submit job, job_id {}, body {}'.format(job_id, job_data))
job_dsl = job_data.get('job_dsl', {})
job_runtime_conf = job_data.get('job_runtime_conf', {})
job_utils.check_pipeline_job_runtime_conf(job_runtime_conf)
job_parameters = job_runtime_conf['job_parameters']
job_initiator = job_runtime_conf['initiator']
job_type = job_parameters.get('job_type', '')
if job_type != 'predict':
# generate job model info
job_parameters['model_id'] = '#'.join([dtable_utils.all_party_key(job_runtime_conf['role']), 'model'])
job_parameters['model_version'] = job_id
train_runtime_conf = {}
else:
detect_utils.check_config(job_parameters, ['model_id', 'model_version'])
# get inference dsl from pipeline model as job dsl
job_tracker = Tracking(job_id=job_id, role=job_initiator['role'], party_id=job_initiator['party_id'],
model_id=job_parameters['model_id'], model_version=job_parameters['model_version'])
pipeline_model = job_tracker.get_output_model('pipeline')
job_dsl = json_loads(pipeline_model['Pipeline'].inference_dsl)
train_runtime_conf = json_loads(pipeline_model['Pipeline'].train_runtime_conf)
job_dsl_path, job_runtime_conf_path = save_job_conf(job_id=job_id,
job_dsl=job_dsl,
job_runtime_conf=job_runtime_conf)
job = Job()
job.f_job_id = job_id
job.f_roles = json_dumps(job_runtime_conf['role'])
job.f_work_mode = job_parameters['work_mode']
job.f_initiator_party_id = job_initiator['party_id']
job.f_dsl = json_dumps(job_dsl)
job.f_runtime_conf = json_dumps(job_runtime_conf)
job.f_train_runtime_conf = json_dumps(train_runtime_conf)
job.f_run_ip = ''
job.f_status = JobStatus.WAITING
job.f_progress = 0
job.f_create_time = current_timestamp()
# save job info
TaskScheduler.distribute_job(job=job, roles=job_runtime_conf['role'], job_initiator=job_initiator)
# push into queue
RuntimeConfig.JOB_QUEUE.put_event({
'job_id': job_id,
"initiator_role": job_initiator['role'],
"initiator_party_id": job_initiator['party_id']
}
)
schedule_logger.info(
'submit job successfully, job id is {}, model id is {}'.format(job.f_job_id, job_parameters['model_id']))
board_url = BOARD_DASHBOARD_URL.format(job_id, job_initiator['role'], job_initiator['party_id'])
return job_id, job_dsl_path, job_runtime_conf_path, {'model_id': job_parameters['model_id'],
'model_version': job_parameters[
'model_version']}, board_url
def update_job():
job_info = request.json
jobs = job_utils.query_job(job_id=job_info['job_id'], party_id=job_info['party_id'], role=job_info['role'])
if not jobs:
return get_json_result(retcode=101, retmsg='find job failed')
else:
job = jobs[0]
TaskScheduler.sync_job_status(job_id=job.f_job_id, roles={job.f_role: [job.f_party_id]},
work_mode=job.f_work_mode, initiator_party_id=job.f_party_id,
initiator_role=job.f_role, job_info={'f_description': job_info.get('notes', '')})
return get_json_result(retcode=0, retmsg='success')
def run_do(self):
try:
running_tasks = job_utils.query_task(status='running',
run_ip=get_lan_ip())
stop_job_ids = set()
# detect_logger.info('start to detect running job..')
for task in running_tasks:
try:
process_exist = job_utils.check_job_process(
int(task.f_run_pid))
if not process_exist:
detect_logger.info(
'job {} component {} on {} {} task {} {} process does not exist'
.format(task.f_job_id, task.f_component_name,
task.f_role, task.f_party_id,
task.f_task_id, task.f_run_pid))
stop_job_ids.add(task.f_job_id)
except Exception as e:
detect_logger.exception(e)
if stop_job_ids:
schedule_logger().info(
'start to stop jobs: {}'.format(stop_job_ids))
for job_id in stop_job_ids:
jobs = job_utils.query_job(job_id=job_id)
if jobs:
initiator_party_id = jobs[0].f_initiator_party_id
job_work_mode = jobs[0].f_work_mode
if len(jobs) > 1:
# i am initiator
my_party_id = initiator_party_id
else:
my_party_id = jobs[0].f_party_id
initiator_party_id = jobs[0].f_initiator_party_id
api_utils.federated_api(
job_id=job_id,
method='POST',
endpoint='/{}/job/stop'.format(API_VERSION),
src_party_id=my_party_id,
dest_party_id=initiator_party_id,
src_role=None,
json_body={
'job_id': job_id,
'operate': 'kill'
},
work_mode=job_work_mode)
TaskScheduler.finish_job(job_id=job_id,
job_runtime_conf=json_loads(
jobs[0].f_runtime_conf),
stop=True)
except Exception as e:
detect_logger.exception(e)
finally:
detect_logger.info('finish detect running job')
def stop_job():
job_id = request.json.get('job_id')
operate = request.json.get('operate')
if operate == 'kill':
TaskScheduler.stop(job_id=job_id, end_status=JobStatus.FAILED)
else:
response = TaskScheduler.stop(job_id=job_id,
end_status=JobStatus.CANCELED)
operate = 'cancel'
if not response:
TaskScheduler.stop(job_id=job_id, end_status=JobStatus.FAILED)
operate = 'kill'
return get_json_result(retcode=0, retmsg='{} job success'.format(operate))
def run_task(job_id, component_name, task_id, role, party_id):
TaskScheduler.start_task(job_id, component_name, task_id, role, party_id,
request.json)
return get_json_result(retcode=0, retmsg='success')
def submit_job(job_data):
job_id = generate_job_id()
schedule_logger(job_id).info('submit job, job_id {}, body {}'.format(job_id, job_data))
job_dsl = job_data.get('job_dsl', {})
job_runtime_conf = job_data.get('job_runtime_conf', {})
job_utils.check_pipeline_job_runtime_conf(job_runtime_conf)
job_parameters = job_runtime_conf['job_parameters']
job_initiator = job_runtime_conf['initiator']
job_type = job_parameters.get('job_type', '')
if job_type != 'predict':
# generate job model info
job_parameters['model_id'] = '#'.join([dtable_utils.all_party_key(job_runtime_conf['role']), 'model'])
job_parameters['model_version'] = job_id
train_runtime_conf = {}
else:
detect_utils.check_config(job_parameters, ['model_id', 'model_version'])
# get inference dsl from pipeline model as job dsl
job_tracker = Tracking(job_id=job_id, role=job_initiator['role'], party_id=job_initiator['party_id'],
model_id=job_parameters['model_id'], model_version=job_parameters['model_version'])
pipeline_model = job_tracker.get_output_model('pipeline')
job_dsl = json_loads(pipeline_model['Pipeline'].inference_dsl)
train_runtime_conf = json_loads(pipeline_model['Pipeline'].train_runtime_conf)
path_dict = save_job_conf(job_id=job_id,
job_dsl=job_dsl,
job_runtime_conf=job_runtime_conf,
train_runtime_conf=train_runtime_conf,
pipeline_dsl=None)
job = Job()
job.f_job_id = job_id
job.f_roles = json_dumps(job_runtime_conf['role'])
job.f_work_mode = job_parameters['work_mode']
job.f_initiator_party_id = job_initiator['party_id']
job.f_dsl = json_dumps(job_dsl)
job.f_runtime_conf = json_dumps(job_runtime_conf)
job.f_train_runtime_conf = json_dumps(train_runtime_conf)
job.f_run_ip = ''
job.f_status = JobStatus.WAITING
job.f_progress = 0
job.f_create_time = current_timestamp()
initiator_role = job_initiator['role']
initiator_party_id = job_initiator['party_id']
if initiator_party_id not in job_runtime_conf['role'][initiator_role]:
schedule_logger(job_id).info("initiator party id error:{}".format(initiator_party_id))
raise Exception("initiator party id error {}".format(initiator_party_id))
get_job_dsl_parser(dsl=job_dsl,
runtime_conf=job_runtime_conf,
train_runtime_conf=train_runtime_conf)
TaskScheduler.distribute_job(job=job, roles=job_runtime_conf['role'], job_initiator=job_initiator)
# push into queue
job_event = job_utils.job_event(job_id, initiator_role, initiator_party_id)
try:
RuntimeConfig.JOB_QUEUE.put_event(job_event)
except Exception as e:
raise Exception('push job into queue failed')
schedule_logger(job_id).info(
'submit job successfully, job id is {}, model id is {}'.format(job.f_job_id, job_parameters['model_id']))
board_url = BOARD_DASHBOARD_URL.format(job_id, job_initiator['role'], job_initiator['party_id'])
logs_directory = get_job_log_directory(job_id)
return job_id, path_dict['job_dsl_path'], path_dict['job_runtime_conf_path'], logs_directory, \
{'model_id': job_parameters['model_id'],'model_version': job_parameters['model_version']}, board_url
def start_stop_job():
job_id = request.json.get('job_id')
response = TaskScheduler.start_stop(job_id=job_id)
return get_json_result(retcode=response.get('retcode'),
retmsg=response.get('retmsg'))
def clean_queue():
TaskScheduler.clean_queue()
return get_json_result(retcode=0, retmsg='success')
def stop_job():
TaskScheduler.stop_job(job_id=request.json.get('job_id', ''))
return get_json_result(retcode=0, retmsg='success')