def put_object(self, bucket_name, key, data):
"""
Put an object in COS. Override the object if the key already exists.
:param key: key of the object.
:param data: data of the object
:type data: str/bytes
:return: None
"""
retries = 0
status = None
while status is None:
try:
res = self.cos_client.put_object(Bucket=bucket_name, Key=key, Body=data)
status = 'OK' if res['ResponseMetadata']['HTTPStatusCode'] == 200 else 'Error'
try:
logger.debug('PUT Object {} - Size: {} - {}'.format(key, sizeof_fmt(len(data)), status))
except Exception:
logger.debug('PUT Object {} {}'.format(key, status))
except ibm_botocore.exceptions.ClientError as e:
if e.response['Error']['Code'] == "NoSuchKey":
raise StorageNoSuchKeyError(bucket_name, key)
else:
raise e
except ibm_botocore.exceptions.ReadTimeoutError as e:
if retries == OBJ_REQ_RETRIES:
raise e
logger.debug('PUT Object timeout. Retrying request')
retries += 1
def _send_status_os(self):
"""
Send the status event to the Object Storage
"""
executor_id = self.response['executor_id']
job_id = self.response['job_id']
call_id = self.response['call_id']
if self.response['type'] == '__init__':
init_key = create_init_key(JOBS_PREFIX, executor_id, job_id,
call_id)
self.internal_storage.put_data(init_key, '')
elif self.response['type'] == '__end__':
status_key = create_status_key(JOBS_PREFIX, executor_id, job_id,
call_id)
dmpd_response_status = json.dumps(self.response)
drs = sizeof_fmt(len(dmpd_response_status))
logger.info(
"Storing execution stats - status.json - Size: {}".format(drs))
self.internal_storage.put_data(status_key, dmpd_response_status)
def _send_status_rabbitmq(self):
"""
Send the status event to RabbitMQ
"""
dmpd_response_status = json.dumps(self.response)
drs = sizeof_fmt(len(dmpd_response_status))
executor_id = self.response['executor_id']
job_id = self.response['job_id']
rabbit_amqp_url = self.config['rabbitmq'].get('amqp_url')
status_sent = False
output_query_count = 0
params = pika.URLParameters(rabbit_amqp_url)
exchange = 'pywren-{}-{}'.format(executor_id, job_id)
while not status_sent and output_query_count < 5:
output_query_count = output_query_count + 1
try:
connection = pika.BlockingConnection(params)
channel = connection.channel()
channel.exchange_declare(exchange=exchange,
exchange_type='fanout',
auto_delete=True)
channel.basic_publish(exchange=exchange,
routing_key='',
body=dmpd_response_status)
connection.close()
logger.info(
"Execution status sent to rabbitmq - Size: {}".format(drs))
status_sent = True
except Exception as e:
logger.error("Unable to send status to rabbitmq")
logger.error(str(e))
logger.info('Retrying to send status to rabbitmq...')
time.sleep(0.2)
def _create_job(config,
internal_storage,
executor_id,
job_id,
func,
data,
runtime_meta,
runtime_memory=None,
extra_env=None,
invoke_pool_threads=128,
include_modules=[],
exclude_modules=[],
original_func_name=None,
execution_timeout=EXECUTION_TIMEOUT):
"""
:param func: the function to map over the data
:param iterdata: An iterable of input data
:param extra_env: Additional environment variables for CF environment. Default None.
:param extra_meta: Additional metadata to pass to CF. Default None.
:param remote_invocation: Enable remote invocation. Default False.
:param invoke_pool_threads: Number of threads to use to invoke.
:param data_all_as_one: upload the data as a single object. Default True
:param overwrite_invoke_args: Overwrite other args. Mainly used for testing.
:param exclude_modules: Explicitly keep these modules from pickled dependencies.
:param original_func_name: Name of the function to invoke.
:return: A list with size `len(iterdata)` of futures for each job
:rtype: list of futures.
"""
log_level = os.getenv('PYWREN_LOGLEVEL')
runtime_name = config['pywren']['runtime']
if runtime_memory is None:
runtime_memory = config['pywren']['runtime_memory']
if original_func_name:
func_name = original_func_name
else:
func_name = func.__name__
extra_env = {} if extra_env is None else extra_env
if extra_env:
extra_env = utils.convert_bools_to_string(extra_env)
logger.debug("Extra environment vars {}".format(extra_env))
if not data:
return []
host_job_meta = {}
job_description = {}
job_description['runtime_name'] = runtime_name
job_description['runtime_memory'] = int(runtime_memory)
job_description['execution_timeout'] = execution_timeout
job_description['func_name'] = func_name
job_description['extra_env'] = extra_env
job_description['total_calls'] = len(data)
job_description['invoke_pool_threads'] = invoke_pool_threads
job_description['executor_id'] = executor_id
job_description['job_id'] = job_id
exclude_modules_cfg = config['pywren'].get('exclude_modules', [])
include_modules_cfg = config['pywren'].get('include_modules', [])
exc_modules = set()
inc_modules = set()
if exclude_modules_cfg:
exc_modules.update(exclude_modules_cfg)
if exclude_modules:
exc_modules.update(exclude_modules)
if include_modules_cfg is not None:
inc_modules.update(include_modules_cfg)
if include_modules_cfg is None and not include_modules:
inc_modules = None
if include_modules is not None and include_modules:
inc_modules.update(include_modules)
if include_modules is None:
inc_modules = None
logger.debug(
'ExecutorID {} | JobID {} - Serializing function and data'.format(
executor_id, job_id))
# pickle func and all data (to capture module dependencies)
serializer = SerializeIndependent(runtime_meta['preinstalls'])
func_and_data_ser, mod_paths = serializer([func] + data, inc_modules,
exc_modules)
func_str = func_and_data_ser[0]
data_strs = func_and_data_ser[1:]
data_size_bytes = sum(len(x) for x in data_strs)
host_job_meta['agg_data'] = False
host_job_meta['data_size_bytes'] = data_size_bytes
log_msg = 'ExecutorID {} | JobID {} - Uploading function and data'.format(
executor_id, job_id)
logger.info(log_msg)
if not log_level:
print(log_msg, end=' ')
if data_size_bytes < MAX_AGG_DATA_SIZE:
agg_data_key = create_agg_data_key(JOBS_PREFIX, executor_id, job_id)
job_description['data_key'] = agg_data_key
agg_data_bytes, agg_data_ranges = _agg_data(data_strs)
job_description['data_ranges'] = agg_data_ranges
agg_upload_time = time.time()
internal_storage.put_data(agg_data_key, agg_data_bytes)
host_job_meta['agg_data'] = True
host_job_meta['data_upload_time'] = time.time() - agg_upload_time
host_job_meta['data_upload_timestamp'] = time.time()
else:
log_msg = ('ExecutorID {} | JobID {} - Total data exceeded '
'maximum size of {} bytes'.format(executor_id, job_id,
MAX_AGG_DATA_SIZE))
raise Exception(log_msg)
module_data = create_module_data(mod_paths)
# Create func and upload
host_job_meta['func_name'] = func_name
func_module_str = pickle.dumps(
{
'func': func_str,
'module_data': module_data
}, -1)
host_job_meta['func_module_bytes'] = len(func_module_str)
func_upload_time = time.time()
func_key = create_func_key(JOBS_PREFIX, executor_id, job_id)
job_description['func_key'] = func_key
internal_storage.put_func(func_key, func_module_str)
host_job_meta['func_upload_time'] = time.time() - func_upload_time
host_job_meta['func_upload_timestamp'] = time.time()
if not log_level:
func_and_data_size = utils.sizeof_fmt(
host_job_meta['func_module_bytes'] +
host_job_meta['data_size_bytes'])
log_msg = '- Total: {}'.format(func_and_data_size)
print(log_msg)
job_description['metadata'] = host_job_meta
return job_description
def run(self):
"""
Runs the function
"""
logger.info("Started")
result = None
exception = False
try:
self.internal_storage = InternalStorage(self.storage_config)
self.internal_storage.tmp_obj_prefix = self.output_key.rsplit(
'/', 1)[0]
loaded_func_all = self._get_function_and_modules()
self._save_modules(loaded_func_all['module_data'])
function = self._unpickle_function(loaded_func_all['func'])
data = self._load_data()
logger.info("data_obj {}".format(data))
if is_object_processing_function(function):
self._create_data_stream(data)
self._fill_optional_args(function, data)
if self.show_memory:
logger.debug(
"Memory usage before call the function: {}".format(
get_current_memory_usage()))
logger.info("Going to execute '{}()'".format(str(
function.__name__)))
print('---------------------- FUNCTION LOG ----------------------',
flush=True)
func_exec_time_t1 = time.time()
result = function(**data)
func_exec_time_t2 = time.time()
print('----------------------------------------------------------',
flush=True)
logger.info("Success function execution")
if self.show_memory:
logger.debug("Memory usage after call the function: {}".format(
get_current_memory_usage()))
self.stats.write('function_exec_time',
round(func_exec_time_t2 - func_exec_time_t1, 8))
# Check for new futures
if result is not None:
self.stats.write("result", True)
if isinstance(result, ResponseFuture) or \
(type(result) == list and len(result) > 0 and isinstance(result[0], ResponseFuture)):
self.stats.write('new_futures', True)
logger.debug("Pickling result")
output_dict = {'result': result}
pickled_output = pickle.dumps(output_dict)
if self.show_memory:
logger.debug(
"Memory usage after output serialization: {}".format(
get_current_memory_usage()))
else:
logger.debug("No result to store")
self.stats.write("result", False)
except Exception:
exception = True
self.stats.write("exception", True)
exc_type, exc_value, exc_traceback = sys.exc_info()
print('----------------------- EXCEPTION !-----------------------',
flush=True)
traceback.print_exc(file=sys.stdout)
print('----------------------------------------------------------',
flush=True)
if self.show_memory:
logger.debug("Memory usage after call the function: {}".format(
get_current_memory_usage()))
try:
logger.debug("Pickling exception")
pickled_exc = pickle.dumps(
(exc_type, exc_value, exc_traceback))
pickle.loads(
pickled_exc
) # this is just to make sure they can be unpickled
self.stats.write("exc_info", str(pickled_exc))
except Exception as pickle_exception:
# Shockingly often, modules like subprocess don't properly
# call the base Exception.__init__, which results in them
# being unpickleable. As a result, we actually wrap this in a try/catch block
# and more-carefully handle the exceptions if any part of this save / test-reload
# fails
self.stats.write("exc_pickle_fail", True)
pickled_exc = pickle.dumps({
'exc_type': str(exc_type),
'exc_value': str(exc_value),
'exc_traceback': exc_traceback,
'pickle_exception': pickle_exception
})
pickle.loads(
pickled_exc
) # this is just to make sure they can be unpickled
self.stats.write("exc_info", str(pickled_exc))
finally:
store_result = strtobool(os.environ.get('STORE_RESULT', 'True'))
if result is not None and store_result and not exception:
output_upload_timestamp_t1 = time.time()
logger.info(
"Storing function result - output.pickle - Size: {}".
format(sizeof_fmt(len(pickled_output))))
self.internal_storage.put_data(self.output_key, pickled_output)
output_upload_timestamp_t2 = time.time()
self.stats.write(
"output_upload_time",
round(
output_upload_timestamp_t2 -
output_upload_timestamp_t1, 8))
self.jobrunner_conn.send("Finished")
logger.info("Finished")
def function_handler(event):
start_time = time.time()
logger.debug("Action handler started")
response_status = {'exception': False}
response_status['host_submit_time'] = event['host_submit_time']
response_status['start_time'] = start_time
context_dict = {
'ibm_cf_request_id': os.environ.get("__OW_ACTIVATION_ID"),
'ibm_cf_python_version': os.environ.get("PYTHON_VERSION"),
}
config = event['config']
storage_config = wrenconfig.extract_storage_config(config)
log_level = event['log_level']
ibm_cf_logging_config(log_level)
call_id = event['call_id']
callgroup_id = event['callgroup_id']
executor_id = event['executor_id']
logger.info("Execution ID: {}/{}/{}".format(executor_id, callgroup_id, call_id))
job_max_runtime = event.get("job_max_runtime", 590) # default for CF
status_key = event['status_key']
func_key = event['func_key']
data_key = event['data_key']
data_byte_range = event['data_byte_range']
output_key = event['output_key']
extra_env = event.get('extra_env', {})
response_status['call_id'] = call_id
response_status['callgroup_id'] = callgroup_id
response_status['executor_id'] = executor_id
# response_status['func_key'] = func_key
# response_status['data_key'] = data_key
# response_status['output_key'] = output_key
# response_status['status_key'] = status_key
try:
if version.__version__ != event['pywren_version']:
raise Exception("WRONGVERSION", "PyWren version mismatch",
version.__version__, event['pywren_version'])
# response_status['free_disk_bytes'] = free_disk_space("/tmp")
custom_env = {'PYWREN_CONFIG': json.dumps(config),
'PYWREN_EXECUTOR_ID': executor_id,
'PYTHONPATH': "{}:{}".format(os.getcwd(), PYWREN_LIBS_PATH),
'PYTHONUNBUFFERED': 'True'}
os.environ.update(custom_env)
os.environ.update(extra_env)
# pass a full json blob
jobrunner_config = {'func_key': func_key,
'data_key': data_key,
'log_level': log_level,
'data_byte_range': data_byte_range,
'python_module_path': PYTHON_MODULE_PATH,
'output_key': output_key,
'stats_filename': JOBRUNNER_STATS_FILENAME}
if os.path.exists(JOBRUNNER_STATS_FILENAME):
os.remove(JOBRUNNER_STATS_FILENAME)
setup_time = time.time()
response_status['setup_time'] = round(setup_time - start_time, 8)
result_queue = multiprocessing.Queue()
jr = jobrunner(jobrunner_config, result_queue)
jr.daemon = True
logger.info("Starting jobrunner process")
jr.start()
jr.join(job_max_runtime)
response_status['exec_time'] = round(time.time() - setup_time, 8)
if jr.is_alive():
# If process is still alive after jr.join(job_max_runtime), kill it
logger.error("Process exceeded maximum runtime of {} seconds".format(job_max_runtime))
# Send the signal to all the process groups
jr.terminate()
raise Exception("OUTATIME", "Process executed for too long and was killed")
try:
# Only 1 message is returned by jobrunner
result_queue.get(block=False)
except Exception:
# If no message, this means that the process was killed due an exception pickling an exception
raise Exception("EXCPICKLEERROR", "PyWren was unable to pickle the exception, check function logs")
# print(subprocess.check_output("find {}".format(PYTHON_MODULE_PATH), shell=True))
# print(subprocess.check_output("find {}".format(os.getcwd()), shell=True))
if os.path.exists(JOBRUNNER_STATS_FILENAME):
with open(JOBRUNNER_STATS_FILENAME, 'r') as fid:
for l in fid.readlines():
key, value = l.strip().split(" ", 1)
try:
response_status[key] = float(value)
except Exception:
response_status[key] = value
if key == 'exception' or key == 'exc_pickle_fail' \
or key == 'result':
response_status[key] = eval(value)
# response_status['server_info'] = get_server_info()
response_status.update(context_dict)
response_status['end_time'] = time.time()
except Exception as e:
# internal runtime exceptions
logger.error("There was an exception: {}".format(str(e)))
response_status['end_time'] = time.time()
response_status['exception'] = True
pickled_exc = pickle.dumps(sys.exc_info())
pickle.loads(pickled_exc) # this is just to make sure they can be unpickled
response_status['exc_info'] = str(pickled_exc)
finally:
store_status = strtobool(os.environ.get('STORE_STATUS', 'True'))
rabbit_amqp_url = config['rabbitmq'].get('amqp_url')
dmpd_response_status = json.dumps(response_status)
drs = sizeof_fmt(len(dmpd_response_status))
if rabbit_amqp_url and store_status:
status_sent = False
output_query_count = 0
while not status_sent and output_query_count < 5:
output_query_count = output_query_count + 1
try:
params = pika.URLParameters(rabbit_amqp_url)
connection = pika.BlockingConnection(params)
channel = connection.channel()
channel.queue_declare(queue=executor_id, auto_delete=True)
channel.basic_publish(exchange='', routing_key=executor_id,
body=dmpd_response_status)
connection.close()
logger.info("Execution stats sent to rabbitmq - Size: {}".format(drs))
status_sent = True
except Exception as e:
logger.error("Unable to send status to rabbitmq")
logger.error(str(e))
logger.info('Retrying to send stats to rabbitmq...')
time.sleep(0.2)
if store_status:
internal_storage = InternalStorage(storage_config)
logger.info("Storing execution stats - status.json - Size: {}".format(drs))
internal_storage.put_data(status_key, dmpd_response_status)