def dsub_start_job(command,
job_name=None,
envs=None,
labels=None,
inputs=None,
inputs_recursive=None,
outputs=None,
outputs_recursive=None,
wait=False):
envs = envs or {}
labels = labels or {}
inputs = inputs or {}
inputs_recursive = inputs_recursive or {}
outputs = outputs or {}
outputs_recursive = outputs_recursive or {}
labels['test-token'] = test_setup.TEST_TOKEN
labels['test-name'] = test_setup.TEST_NAME
logging = param_util.build_logging_param(test.LOGGING)
job_resources = job_model.Resources(image='ubuntu',
logging=logging,
zones=['us-central1-*'])
env_data = {job_model.EnvParam(k, v) for (k, v) in envs.items()}
label_data = {job_model.LabelParam(k, v) for (k, v) in labels.items()}
input_file_param_util = param_util.InputFileParamUtil('input')
input_data = set()
for (recursive, items) in ((False, inputs.items()),
(True, inputs_recursive.items())):
for (name, value) in items:
name = input_file_param_util.get_variable_name(name)
input_data.add(
input_file_param_util.make_param(name, value, recursive))
output_file_param_util = param_util.OutputFileParamUtil('output')
output_data = set()
for (recursive, items) in ((False, outputs.items()),
(True, outputs_recursive.items())):
for (name, value) in items:
name = output_file_param_util.get_variable_name(name)
output_data.add(
output_file_param_util.make_param(name, value, recursive))
job_params = {
'envs': env_data,
'inputs': input_data,
'outputs': output_data,
'labels': label_data,
}
task_descriptors = [
job_model.TaskDescriptor({'task-id': None}, {
'envs': set(),
'labels': set(),
'inputs': set(),
'outputs': set(),
}, job_model.Resources())
]
return dsub.run(get_dsub_provider(),
job_resources,
job_params,
task_descriptors,
name=job_name,
command=command,
wait=wait,
disable_warning=True)
def start_job(self,
command,
name=None,
envs={},
labels={},
inputs={},
inputs_recursive={},
outputs={},
outputs_recursive={},
task_count=1,
wait=False):
logging = param_util.build_logging_param(self.log_path)
resources = job_model.Resources(image=DOCKER_IMAGE,
logging=logging,
zones=['us-central1*'])
env_data = {param_util.EnvParam(k, v) for (k, v) in envs.items()}
label_data = {
job_model.LabelParam(k, v)
for (k, v) in labels.items()
}
# This is mostly an extraction dsubs argument parsing here:
# https://github.com/googlegenomics/dsub/blob/master/dsub/lib/param_util.py#L720
# Reworked it to handle dictionaries rather than a list of items
# of the form 'key=val'
input_file_param_util = param_util.InputFileParamUtil('input')
input_data = set()
for (recursive, items) in ((False, inputs.items()),
(True, inputs_recursive.items())):
for (name, value) in items:
name = input_file_param_util.get_variable_name(name)
input_data.add(
input_file_param_util.make_param(
name, value, recursive))
output_file_param_util = param_util.OutputFileParamUtil('output')
output_data = set()
for (recursive, items) in ((False, outputs.items()),
(True, outputs_recursive.items())):
for (name, value) in items:
name = output_file_param_util.get_variable_name(name)
output_data.add(
output_file_param_util.make_param(
name, value, recursive))
job_params = {
'envs': env_data,
'inputs': input_data,
'outputs': output_data,
'labels': label_data,
}
if task_count > 1:
task_descriptors = [
job_model.TaskDescriptor({'task-id': i + 1}, {
'envs': env_data,
'inputs': input_data,
'outputs': output_data,
'labels': label_data,
}, job_model.Resources()) for i in xrange(task_count)
]
all_task_data = [{
'task-id': i + 1
} for i in xrange(task_count)]
else:
task_descriptors = [
job_model.TaskDescriptor({'task-id': None}, {
'labels': set(),
'envs': set(),
'inputs': set(),
'outputs': set()
}, job_model.Resources())
]
return execute_redirect_stdout(
lambda: dsub.run(self.provider,
resources,
job_params,
task_descriptors,
name=name,
command=command,
wait=wait,
disable_warning=True))