def wait_job(self, name, namespace, delete=False, sleep=5):
jobs_path = self.jobs_path.format(namespace)
url = 'https://{0}:{1}{2}/{3}'.format(self.kubernetes_service_host,
self.kubernetes_service_port,
jobs_path,
name)
while True:
try:
r = requests.get(url,
verify=self.cert_verify,
headers=self.auth_header)
if r.status_code != 200:
raise Exception(f'Error obtaining {name} info - {str(r.status_code)}\n{str(r.content)}')
job = r.json()
if (utils.is_value_in_dict(job['status'], 'succeeded') and
utils.is_value_in_dict(job['spec'], 'completions')):
if job['status']['succeeded'] >= job['spec']['completions']:
# Delete succeeded jobs if delete=True
if delete:
self.delete_job(name, namespace)
break
if (utils.is_value_in_dict(job['status'], 'failed') and
utils.is_value_in_dict(job['spec'], 'backoffLimit')):
if job['status']['failed'] >= job['spec']['backoffLimit']:
logger.error(f'{name} failed! See pod logs for details')
break
time.sleep(sleep)
except Exception as e:
logger.error(e)
break
def add_s3_environment_vars(self):
if utils.is_value_in_dict(self.aws_properties, 's3'):
s3_props = self.aws_properties['s3']
if utils.is_value_in_dict(s3_props, 'input_bucket'):
self.add_lambda_environment_variable('INPUT_BUCKET',
s3_props['input_bucket'])
if utils.is_value_in_dict(s3_props, 'output_bucket'):
self.add_lambda_environment_variable('OUTPUT_BUCKET',
s3_props['output_bucket'])
if utils.is_value_in_dict(s3_props, 'output_folder'):
self.add_lambda_environment_variable('OUTPUT_FOLDER',
s3_props['output_folder'])
def save_tmp_udocker_env(cls):
#Avoid override global variables
if utils.is_value_in_dict(os.environ, 'UDOCKER_TARBALL'):
cls.udocker_tarball = os.environ['UDOCKER_TARBALL']
if utils.is_value_in_dict(os.environ, 'UDOCKER_DIR'):
cls.udocker_dir = os.environ['UDOCKER_DIR']
# Set temporal global vars
udocker_tarball = utils.resource_path(
utils.join_paths(cls.lambda_code_files_path, "udocker",
"udocker-1.1.3.tar.gz"))
utils.set_environment_variable('UDOCKER_TARBALL', udocker_tarball)
utils.set_environment_variable(
'UDOCKER_DIR', utils.join_paths(cls.scar_temporal_folder,
"udocker"))
def create_command(self):
self.add_container_volumes()
self.add_container_environment_variables()
# Container running script
if utils.is_value_in_dict(lambda_instance.event, 'script'):
self.add_script_as_entrypoint()
# Container with args
elif utils.is_value_in_dict(lambda_instance.event,'cmd_args'):
self.add_args()
# Script to be executed every time (if defined)
elif utils.is_variable_in_environment('INIT_SCRIPT_PATH'):
self.add_init_script()
# Only container
else:
self.cmd_container_execution += [self.container_name]
def get_s3_record(self):
if len(lambda_instance.event['Records']) > 1:
logger.warning("Multiple records detected. Only processing the first one.")
record = lambda_instance.event['Records'][0]
if utils.is_value_in_dict(record, 's3'):
return record['s3']
def __init__(self):
if utils.is_value_in_dict(lambda_instance.event, 'Records'):
self.record = self.get_s3_record()
self.input_bucket = self.record['bucket']['name']
self.file_key = unquote_plus(self.record['object']['key'])
self.file_name = os.path.basename(self.file_key).replace(' ', '')
self.file_download_path = '{0}/{1}'.format(lambda_instance.input_folder, self.file_name)
def invoke_function_batch(self,scar_input_file,variables):
self.register_job_definition(lambda_instance.function_name, utils.get_environment_variable("IMAGE_ID"))
data=""
if utils.is_variable_in_environment('INIT_SCRIPT_PATH') :
with open(str(os.environ['INIT_SCRIPT_PATH'])) as myfile:
data = myfile.read()
if utils.is_value_in_dict(lambda_instance.event, 'script'):
data=utils.base64_to_utf8_string(lambda_instance.event['script'])
if lambda_instance.has_input_bucket():
self.bucket_name_input = lambda_instance.input_bucket
if lambda_instance.has_output_bucket():
self.bucket_name_output = lambda_instance.output_bucket
if (data!=""):
self.register_job_definition("scarfiles",self.scarfile_id_image)
response = self.submit_job("scarfiles",1,"INIT",data,self.bucket_name_input,self.bucket_name_output,"script.sh",scar_input_file,variables)
idJob = self.submit_job(lambda_instance.function_name,1,"MED",data,self.bucket_name_input,self.bucket_name_output,lambda_instance.input_folder+"/script.sh",scar_input_file,variables,idJob=response["jobId"])["jobId"]
response = self.submit_job("scarfiles",1,"FINISH",data,self.bucket_name_input,self.bucket_name_output,"script.sh",scar_input_file,variables,idJob=idJob)
elif(self.bucket_name_output!="NO" or self.bucket_name_input!="NO"):
self.register_job_definition("scarfiles",self.scarfile_id_image)
response = self.submit_job("scarfiles",1,"INIT",data,self.bucket_name_input,self.bucket_name_output,"",scar_input_file,variables)
idJob = self.submit_job(lambda_instance.function_name,1,"MED",data,self.bucket_name_input,self.bucket_name_output,"",scar_input_file,variables,idJob=response["jobId"])["jobId"]
response = self.submit_job("scarfiles",1,"FINISH",data,self.bucket_name_input,self.bucket_name_output,"",scar_input_file,variables,idJob=idJob)
else:
idJob = self.submit_job(lambda_instance.function_name,1,"MED",data,self.bucket_name_input,self.bucket_name_output,"",scar_input_file,variables)["jobId"]
return idJob
def get_user_script(self):
script = ""
if utils.is_variable_in_environment('INIT_SCRIPT_PATH'):
file_content = utils.read_file(utils.get_environment_variable('INIT_SCRIPT_PATH'), 'rb')
script = utils.utf8_to_base64_string(file_content)
if utils.is_value_in_dict(self.lambda_instance.event, 'script'):
script = self.lambda_instance.event['script']
return script
def set_required_environment_variables(self):
self.add_lambda_environment_variable(
'TIMEOUT_THRESHOLD', str(self.properties['timeout_threshold']))
self.add_lambda_environment_variable('LOG_LEVEL',
self.properties['log_level'])
if utils.is_value_in_dict(self.properties, 'image'):
self.add_lambda_environment_variable('IMAGE_ID',
self.properties['image'])
self.add_s3_environment_vars()
if 'api_gateway' in self.aws_properties:
self.add_lambda_environment_variable(
'API_GATEWAY_ID', self.aws_properties['api_gateway']['id'])
def create_command(self):
self.add_container_volumes()
self.add_container_environment_variables()
# Container running script
if utils.is_value_in_dict(self.lambda_instance.event, 'script'):
# Add script in memory as entrypoint
script_path = "{0}/script.sh".format(self.lambda_instance.temporal_folder)
script_content = utils.base64_to_utf8_string(self.lambda_instance.event['script'])
utils.create_file_with_content(script_path, script_content)
self.cmd_container_execution += ["--entrypoint={0} {1}".format(self.script_exec, script_path), self.container_name]
# Container with args
elif utils.is_value_in_dict(self.lambda_instance.event,'cmd_args'):
# Add args
self.cmd_container_execution += [self.container_name]
self.cmd_container_execution += json.loads(self.lambda_instance.event['cmd_args'])
# Script to be executed every time (if defined)
elif utils.is_variable_in_environment('INIT_SCRIPT_PATH'):
# Add init script
init_script_path = "{0}/init_script.sh".format(self.lambda_instance.temporal_folder)
shutil.copyfile(utils.get_environment_variable("INIT_SCRIPT_PATH"), init_script_path)
self.cmd_container_execution += ["--entrypoint={0} {1}".format(self.script_exec, init_script_path), self.container_name]
# Only container
else:
self.cmd_container_execution += [self.container_name]
def set_required_environment_variables(self):
self.add_lambda_environment_variable(
'TIMEOUT_THRESHOLD', str(self.properties['timeout_threshold']))
self.add_lambda_environment_variable('LOG_LEVEL',
self.properties['log_level'])
###
self.add_lambda_environment_variable('EXECUTION_MODE',
self.properties['execution_mode'])
self.add_lambda_environment_variable(
'ROLE', self.aws_properties['iam']['role'])
self.add_lambda_environment_variable('MEMORY',
str(self.properties['memory']))
###
if utils.is_value_in_dict(self.properties, 'image'):
self.add_lambda_environment_variable('IMAGE_ID',
self.properties['image'])
self.add_s3_environment_vars()
if 'api_gateway' in self.aws_properties:
self.add_lambda_environment_variable(
'API_GATEWAY_ID', self.aws_properties['api_gateway']['id'])
def is_s3_event(self):
if utils.is_value_in_dict(self.lambda_instance.event, 'Records'):
return self.lambda_instance.event['Records'][0][
'eventSource'] == "aws:s3"
return False
def is_s3_event(self):
if utils.is_value_in_dict(lambda_instance.event, 'Records'):
# Check if the event is an S3 event
return lambda_instance.event['Records'][0]['eventSource'] == "aws:s3"
return False
