def check_job_executor(data, job_id):
job = api.get_job(job_id)
job_type, subtype = edp.split_job_type(job.type)
# Check if cluster contains Oozie service to run job
main_base.check_edp_job_support(data['cluster_id'])
# All types except Java require input and output objects
if job_type == 'Java':
if not _is_main_class_present(data):
raise ex.InvalidDataException('Java job must '
'specify edp.java.main_class')
else:
if not ('input_id' in data and 'output_id' in data):
raise ex.InvalidDataException("%s job requires 'input_id' "
"and 'output_id'" % job.type)
b.check_data_source_exists(data['input_id'])
b.check_data_source_exists(data['output_id'])
b.check_data_sources_are_different(data['input_id'], data['output_id'])
if job_type == 'MapReduce' and (
subtype == 'Streaming' and not _streaming_present(data)):
raise ex.InvalidDataException("%s job "
"must specify streaming mapper "
"and reducer" % job.type)
main_base.check_cluster_exists(data['cluster_id'])
def check_data_sources(data, job):
if not ('input_id' in data and 'output_id' in data):
raise ex.InvalidDataException(_("%s job requires 'input_id' "
"and 'output_id'") % job.type)
b.check_data_source_exists(data['input_id'])
b.check_data_source_exists(data['output_id'])
b.check_data_sources_are_different(data['input_id'], data['output_id'])