def run_analysis(analysis):
"""Launch analysis (outermost task, calls subtasks that monitor and run
preprocessing, execution, postprocessing)
"""
logger.debug("analysis_manager.tasks run_analysis called")
# updating status of analysis to running
analysis = Analysis.objects.filter(uuid=analysis.uuid)[0]
analysis_status = AnalysisStatus.objects.get(analysis=analysis)
analysis.set_status(Analysis.RUNNING_STATUS)
# DOWNLOADING
# GETTING LIST OF DOWNLOADED REMOTE FILES
datainputs = analysis.workflow_data_input_maps.all()
download_tasks = []
for files in datainputs:
cur_node_uuid = files.data_uuid
cur_fs_uuid = Node.objects.get(uuid=cur_node_uuid).file_uuid
# Adding downloading task if file is not local
if not is_local(cur_fs_uuid):
# getting the current file_uuid from the given node_uuid
task_id = import_file.subtask((cur_fs_uuid, False, ))
download_tasks.append(task_id)
# PREPROCESSING
task_id = run_analysis_preprocessing.subtask((analysis, ))
download_tasks.append(task_id)
result_chord, result_set = progress_chord(download_tasks)(
chord_execution.subtask(analysis=analysis, ))
# saving preprocessing taskset
analysis_status.preprocessing_taskset_id = result_set.task_id
analysis_status.save()
def run_analysis(analysis):
'''Launch analysis (outermost task, calls subtasks that monitor and run
preprocessing, execution, postprocessing)
'''
logger.debug("analysis_manager.tasks run_analysis called")
# updating status of analysis to running
analysis = Analysis.objects.filter(uuid=analysis.uuid)[0]
analysis_status = AnalysisStatus.objects.get(analysis=analysis)
analysis.set_status(Analysis.RUNNING_STATUS)
# DOWNLOADING
# GETTING LIST OF DOWNLOADED REMOTE FILES
datainputs = analysis.workflow_data_input_maps.all()
download_tasks = []
for files in datainputs:
cur_node_uuid = files.data_uuid
cur_fs_uuid = Node.objects.get(uuid=cur_node_uuid).file_uuid
# Adding downloading task if file is not local
if not is_local(cur_fs_uuid):
# getting the current file_uuid from the given node_uuid
task_id = import_file.subtask((
cur_fs_uuid,
False,
))
download_tasks.append(task_id)
# PREPROCESSING
task_id = run_analysis_preprocessing.subtask((analysis, ))
download_tasks.append(task_id)
result_chord, result_set = progress_chord(download_tasks)(
chord_execution.subtask(analysis=analysis, ))
# saving preprocessing taskset
analysis_status.preprocessing_taskset_id = result_set.task_id
analysis_status.save()
def _refinery_file_import(analysis_uuid):
"""
Check on the status of the files being imported into Refinery.
Fail the task appropriately if we cannot retrieve the status.
"""
analysis = _get_analysis(analysis_uuid)
analysis_status = _get_analysis_status(analysis_uuid)
if not analysis_status.refinery_import_task_group_id:
logger.info("Starting analysis '%s'", analysis)
analysis.set_status(Analysis.RUNNING_STATUS)
logger.info("Starting input file import tasks for analysis '%s'",
analysis)
refinery_import_tasks = []
if analysis.is_tool_based:
tool = _get_workflow_tool(analysis_uuid)
input_file_uuid_list = tool.get_input_file_uuid_list()
else:
input_file_uuid_list = analysis.get_input_file_uuid_list()
for input_file_uuid in input_file_uuid_list:
refinery_import_task = import_file.subtask((input_file_uuid, ))
refinery_import_tasks.append(refinery_import_task)
refinery_import_taskset = TaskSet(
tasks=refinery_import_tasks).apply_async()
refinery_import_taskset.save()
analysis_status.refinery_import_task_group_id = \
refinery_import_taskset.taskset_id
analysis_status.save()
run_analysis.retry(countdown=RETRY_INTERVAL)
# check if all files were successfully imported into Refinery
refinery_import_taskset = get_taskset_result(
analysis_status.refinery_import_task_group_id)
if not refinery_import_taskset.ready():
logger.debug("Input file import pending for analysis '%s'", analysis)
run_analysis.retry(countdown=RETRY_INTERVAL)
elif not refinery_import_taskset.successful():
error_msg = "Analysis '{}' failed during file import".format(analysis)
logger.error(error_msg)
analysis.set_status(Analysis.FAILURE_STATUS, error_msg)
analysis.send_email()
refinery_import_taskset.delete()
return
def _get_galaxy_download_tasks(analysis):
"""Get file import tasks for Galaxy analysis results"""
logger.debug("Preparing to download analysis results from Galaxy")
task_list = []
# retrieving list of files to download for workflow
dl_files = analysis.workflow_dl_files
# creating dictionary based on files to download predetermined by workflow
# w/ keep operators
dl_dict = {}
for dl in dl_files.all():
temp_dict = {}
temp_dict['filename'] = dl.filename
temp_dict['pair_id'] = dl.pair_id
dl_dict[str(dl.step_id)] = temp_dict
galaxy_instance = analysis.workflow.workflow_engine.instance
try:
download_list = galaxy_instance.get_history_file_list(
analysis.history_id)
except galaxy.client.ConnectionError as exc:
error_msg = (
"Error downloading Galaxy history files for analysis '%s': %s")
logger.error(error_msg, analysis.name, exc.message)
analysis.set_status(Analysis.FAILURE_STATUS, error_msg)
analysis.galaxy_cleanup()
return task_list
# Iterating through files in current galaxy history
for results in download_list:
# download file if result state is "ok"
if results['state'] == 'ok':
file_type = results["type"]
curr_file_id = results['name']
if curr_file_id in dl_dict:
curr_dl_dict = dl_dict[curr_file_id]
result_name = curr_dl_dict['filename'] + '.' + file_type
# size of file defined by galaxy
file_size = results['file_size']
# Determining tag if galaxy results should be download through
# http or copying files directly to retrieve HTML files as zip
# archives via dataset URL
if galaxy_instance.local_download and file_type != 'html':
download_url = results['file_name']
else:
download_url = urlparse.urljoin(
galaxy_instance.base_url, '/'.join([
'datasets',
str(results['dataset_id']), 'display?to_ext=txt'
]))
# workaround to set the correct file type for zip archives of
# FastQC HTML reports produced by Galaxy dynamically
if file_type == 'html':
file_type = 'zip'
# TODO: when changing permanent=True, fix update of % download
# of file
filestore_uuid = create(source=download_url,
filetype=file_type)
# adding history files to django model
temp_file = AnalysisResult(analysis_uuid=analysis.uuid,
file_store_uuid=filestore_uuid,
file_name=result_name,
file_type=file_type)
temp_file.save()
analysis.results.add(temp_file)
analysis.save()
# downloading analysis results into file_store
# only download files if size is greater than 1
if file_size > 0:
task_id = import_file.subtask(
(filestore_uuid, False, file_size))
task_list.append(task_id)
return task_list
def _get_galaxy_download_task_ids(analysis):
"""Get file import tasks for Galaxy analysis results"""
logger.debug("Preparing to download analysis results from Galaxy")
task_id_list = []
# retrieving list of files to download for workflow
tool = _get_workflow_tool(analysis.uuid)
tool.create_analysis_output_node_connections()
galaxy_instance = analysis.workflow.workflow_engine.instance
try:
download_list = tool.get_galaxy_dataset_download_list()
except galaxy.client.ConnectionError as exc:
error_msg = (
"Error downloading Galaxy history files for analysis '%s': %s")
logger.error(error_msg, analysis.name, exc.message)
analysis.set_status(Analysis.FAILURE_STATUS, error_msg)
analysis.galaxy_cleanup()
return task_id_list
# Iterating through files in current galaxy history
for results in download_list:
# download file if result state is "ok"
if results['state'] == 'ok':
file_extension = results["type"]
result_name = "{}.{}".format(results['name'], file_extension)
# size of file defined by galaxy
file_size = results['file_size']
# Determining tag if galaxy results should be download through
# http or copying files directly to retrieve HTML files as zip
# archives via dataset URL
if galaxy_instance.local_download and file_extension != 'html':
download_url = results['file_name']
else:
download_url = urlparse.urljoin(
galaxy_instance.base_url, '/'.join([
'datasets',
str(results['dataset_id']), 'display?to_ext=txt'
]))
file_store_item = FileStoreItem(source=download_url)
# workaround to set the correct file type for zip archives of
# FastQC HTML reports produced by Galaxy dynamically
if file_extension == 'html':
file_extension = 'zip'
# assign file type manually since it cannot be inferred from source
try:
extension = FileExtension.objects.get(name=file_extension)
except (FileExtension.DoesNotExist,
FileExtension.MultipleObjectsReturned) as exc:
logger.warn(
"Could not assign type to file '%s' using "
"extension '%s': %s", file_store_item, file_extension, exc)
else:
file_store_item.filetype = extension.filetype
file_store_item.save()
# adding history files to django model
temp_file = AnalysisResult(analysis_uuid=analysis.uuid,
file_store_uuid=file_store_item.uuid,
file_name=result_name,
file_type=file_extension)
temp_file.save()
analysis.results.add(temp_file)
analysis.save()
# downloading analysis results into file_store
# only download files if size is greater than 1
if file_size > 0:
task_id = import_file.subtask(
(file_store_item.uuid, False, file_size))
task_id_list.append(task_id)
return task_id_list
def run_analysis(analysis_uuid):
"""Manage analysis execution"""
RETRY_INTERVAL = 5 # seconds
try:
analysis = Analysis.objects.get(uuid=analysis_uuid)
except (Analysis.DoesNotExist, Analysis.MultipleObjectsReturned) as exc:
logger.error("Can not retrieve analysis with UUID '%s': '%s'",
analysis_uuid, exc)
run_analysis.update_state(state=celery.states.FAILURE)
return
# if cancelled by user
if analysis.failed():
return
try:
analysis_status = AnalysisStatus.objects.get(analysis=analysis)
except (AnalysisStatus.DoesNotExist,
AnalysisStatus.MultipleObjectsReturned) as exc:
logger.error("Can not retrieve status for analysis '%s': '%s'",
analysis, exc)
run_analysis.update_state(state=celery.states.FAILURE)
return
if not analysis_status.refinery_import_task_group_id:
logger.info("Starting analysis '%s'", analysis)
analysis.set_status(Analysis.RUNNING_STATUS)
logger.info("Starting input file import tasks for analysis '%s'",
analysis)
refinery_import_tasks = []
for input_file_uuid in analysis.get_input_file_uuid_list():
refinery_import_task = import_file.subtask(
(input_file_uuid, False, ))
refinery_import_tasks.append(refinery_import_task)
refinery_import = TaskSet(tasks=refinery_import_tasks).apply_async()
refinery_import.save()
analysis_status.refinery_import_task_group_id = \
refinery_import.taskset_id
analysis_status.save()
run_analysis.retry(countdown=RETRY_INTERVAL)
# check if all files were successfully imported into Refinery
refinery_import = TaskSetResult.restore(
analysis_status.refinery_import_task_group_id)
if not refinery_import.ready():
logger.debug("Input file import pending for analysis '%s'", analysis)
run_analysis.retry(countdown=RETRY_INTERVAL)
elif not refinery_import.successful():
logger.error("Analysis '%s' failed during file import", analysis)
analysis.set_status(Analysis.FAILURE_STATUS)
analysis.send_email()
refinery_import.delete()
return
# import files into Galaxy and start analysis
if not analysis_status.galaxy_import_task_group_id:
logger.debug("Starting analysis execution in Galaxy")
try:
analysis.prepare_galaxy()
except (requests.exceptions.ConnectionError,
galaxy.client.ConnectionError):
logger.error("Analysis '%s' failed during preparation in Galaxy",
analysis)
analysis.set_status(Analysis.FAILURE_STATUS)
analysis.send_email()
refinery_import.delete()
return
galaxy_import_tasks = [
start_galaxy_analysis.subtask((analysis_uuid, )),
]
galaxy_import = TaskSet(tasks=galaxy_import_tasks).apply_async()
galaxy_import.save()
analysis_status.galaxy_import_task_group_id = \
galaxy_import.taskset_id
analysis_status.set_galaxy_history_state(AnalysisStatus.PROGRESS)
run_analysis.retry(countdown=RETRY_INTERVAL)
# check if data files were successfully imported into Galaxy
galaxy_import = TaskSetResult.restore(
analysis_status.galaxy_import_task_group_id)
if not galaxy_import.ready():
logger.debug("Analysis '%s' pending in Galaxy", analysis)
run_analysis.retry(countdown=RETRY_INTERVAL)
elif not galaxy_import.successful():
logger.error("Analysis '%s' failed in Galaxy", analysis)
analysis.set_status(Analysis.FAILURE_STATUS)
analysis_status.set_galaxy_history_state(AnalysisStatus.ERROR)
analysis.send_email()
refinery_import.delete()
galaxy_import.delete()
analysis.galaxy_cleanup()
return
# check if analysis has finished running in Galaxy
try:
percent_complete = analysis.galaxy_progress()
except RuntimeError:
analysis_status.set_galaxy_history_state(AnalysisStatus.ERROR)
analysis.send_email()
refinery_import.delete()
galaxy_import.delete()
analysis.galaxy_cleanup()
return
except galaxy.client.ConnectionError:
analysis_status.set_galaxy_history_state(AnalysisStatus.UNKNOWN)
run_analysis.retry(countdown=RETRY_INTERVAL)
else:
# workaround to avoid moving the progress bar backward
if analysis_status.galaxy_history_progress < percent_complete:
analysis_status.galaxy_history_progress = percent_complete
analysis_status.save()
if percent_complete < 100:
analysis_status.set_galaxy_history_state(AnalysisStatus.PROGRESS)
run_analysis.retry(countdown=RETRY_INTERVAL)
else:
analysis_status.set_galaxy_history_state(AnalysisStatus.OK)
# retrieve analysis results from Galaxy
if not analysis_status.galaxy_export_task_group_id:
galaxy_export_tasks = get_galaxy_download_tasks(analysis)
logger.info("Starting downloading of results from Galaxy for analysis "
"'%s'", analysis)
galaxy_export = TaskSet(tasks=galaxy_export_tasks).apply_async()
galaxy_export.save()
analysis_status.galaxy_export_task_group_id = galaxy_export.taskset_id
analysis_status.save()
run_analysis.retry(countdown=RETRY_INTERVAL)
# check if analysis results have finished downloading from Galaxy
galaxy_export = TaskSetResult.restore(
analysis_status.galaxy_export_task_group_id)
if not galaxy_export.ready():
logger.debug("Results download pending for analysis '%s'", analysis)
run_analysis.retry(countdown=RETRY_INTERVAL)
# all tasks must have succeeded or failed
elif not galaxy_export.successful():
logger.error("Analysis '%s' failed while downloading results from "
"Galaxy", analysis)
analysis.set_status(Analysis.FAILURE_STATUS)
analysis.send_email()
refinery_import.delete()
galaxy_import.delete()
galaxy_export.delete()
analysis.galaxy_cleanup()
return
# attach workflow outputs back to dataset isatab graph
if analysis.workflow.type == Workflow.ANALYSIS_TYPE:
analysis.attach_outputs_dataset()
elif analysis.workflow.type == Workflow.DOWNLOAD_TYPE:
analysis.attach_outputs_downloads()
else:
logger.warning("Unknown workflow type '%s' in analysis '%s'",
analysis.workflow.type, analysis.name)
analysis.set_status(Analysis.SUCCESS_STATUS)
analysis.rename_results()
analysis.send_email()
logger.info("Analysis '%s' finished successfully", analysis)
analysis.galaxy_cleanup()
refinery_import.delete()
galaxy_import.delete()
galaxy_export.delete()
# Update file count and file size of the corresponding data set
analysis.data_set.file_count = analysis.data_set.get_file_count()
analysis.data_set.file_size = analysis.data_set.get_file_size()
analysis.data_set.save()
def get_galaxy_download_tasks(analysis):
"""Get file import tasks for Galaxy analysis results"""
logger.debug("Preparing to download analysis results from Galaxy")
# retrieving list of files to download for workflow
dl_files = analysis.workflow_dl_files
# creating dictionary based on files to download predetermined by workflow
# w/ keep operators
dl_dict = {}
for dl in dl_files.all():
temp_dict = {}
temp_dict['filename'] = dl.filename
temp_dict['pair_id'] = dl.pair_id
dl_dict[str(dl.step_id)] = temp_dict
task_list = []
galaxy_instance = analysis.workflow.workflow_engine.instance
try:
download_list = galaxy_instance.get_history_file_list(
analysis.history_id)
except galaxy.client.ConnectionError as exc:
error_msg = "Error downloading Galaxy history files for analysis " \
"'%s': %s"
logger.error(error_msg, analysis.name, exc.message)
analysis.set_status(Analysis.FAILURE_STATUS, error_msg)
analysis.galaxy_cleanup()
return task_list
# Iterating through files in current galaxy history
for results in download_list:
# download file if result state is "ok"
if results['state'] == 'ok':
file_type = results["type"]
curr_file_id = results['name']
if curr_file_id in dl_dict:
curr_dl_dict = dl_dict[curr_file_id]
result_name = curr_dl_dict['filename'] + '.' + file_type
# size of file defined by galaxy
file_size = results['file_size']
# Determining tag if galaxy results should be download through
# http or copying files directly to retrieve HTML files as zip
# archives via dataset URL
if galaxy_instance.local_download and file_type != 'html':
download_url = results['file_name']
else:
download_url = urlparse.urljoin(
galaxy_instance.base_url, '/'.join(
['datasets', str(results['dataset_id']),
'display?to_ext=txt']))
# workaround to set the correct file type for zip archives of
# FastQC HTML reports produced by Galaxy dynamically
if file_type == 'html':
file_type = 'zip'
# TODO: when changing permanent=True, fix update of % download
# of file
filestore_uuid = create(
source=download_url, filetype=file_type, permanent=False)
# adding history files to django model
temp_file = AnalysisResult(
analysis_uuid=analysis.uuid,
file_store_uuid=filestore_uuid,
file_name=result_name, file_type=file_type)
temp_file.save()
analysis.results.add(temp_file)
analysis.save()
# downloading analysis results into file_store
# only download files if size is greater than 1
if file_size > 0:
# local download, force copying into the file_store instead
# of symlinking
if galaxy_instance.local_download:
task_id = import_file.subtask(
(filestore_uuid, False, True, file_size,))
else:
task_id = import_file.subtask(
(filestore_uuid, False, False, file_size,))
task_list.append(task_id)
return task_list
def run_analysis(analysis_uuid):
"""Manage analysis execution"""
RETRY_INTERVAL = 5 # seconds
try:
analysis = Analysis.objects.get(uuid=analysis_uuid)
except (Analysis.DoesNotExist, Analysis.MultipleObjectsReturned) as exc:
logger.error("Can not retrieve analysis with UUID '%s': '%s'",
analysis_uuid, exc)
run_analysis.update_state(state=celery.states.FAILURE)
return
# if cancelled by user
if analysis.failed():
return
try:
analysis_status = AnalysisStatus.objects.get(analysis=analysis)
except (AnalysisStatus.DoesNotExist,
AnalysisStatus.MultipleObjectsReturned) as exc:
logger.error("Can not retrieve status for analysis '%s': '%s'",
analysis, exc)
run_analysis.update_state(state=celery.states.FAILURE)
return
if not analysis_status.refinery_import_task_group_id:
logger.info("Starting analysis '%s'", analysis)
analysis.set_status(Analysis.RUNNING_STATUS)
logger.info("Starting input file import tasks for analysis '%s'",
analysis)
refinery_import_tasks = []
for input_file_uuid in analysis.get_input_file_uuid_list():
refinery_import_task = import_file.subtask((input_file_uuid, ))
refinery_import_tasks.append(refinery_import_task)
refinery_import = TaskSet(tasks=refinery_import_tasks).apply_async()
refinery_import.save()
analysis_status.refinery_import_task_group_id = \
refinery_import.taskset_id
analysis_status.save()
run_analysis.retry(countdown=RETRY_INTERVAL)
# check if all files were successfully imported into Refinery
refinery_import = TaskSetResult.restore(
analysis_status.refinery_import_task_group_id)
if not refinery_import.ready():
logger.debug("Input file import pending for analysis '%s'", analysis)
run_analysis.retry(countdown=RETRY_INTERVAL)
elif not refinery_import.successful():
error_msg = "Analysis '{}' failed during file import".format(analysis)
logger.error(error_msg)
analysis.set_status(Analysis.FAILURE_STATUS, error_msg)
analysis.send_email()
refinery_import.delete()
return
# import files into Galaxy and start analysis
if not analysis_status.galaxy_import_task_group_id:
logger.debug("Starting analysis execution in Galaxy")
try:
analysis.prepare_galaxy()
except (requests.exceptions.ConnectionError,
galaxy.client.ConnectionError):
error_msg = "Analysis '{}' failed during preparation in " \
"Galaxy".format(analysis)
logger.error(error_msg)
analysis.set_status(Analysis.FAILURE_STATUS, error_msg)
analysis.send_email()
refinery_import.delete()
return
galaxy_import_tasks = [
start_galaxy_analysis.subtask((analysis_uuid, )),
]
galaxy_import = TaskSet(tasks=galaxy_import_tasks).apply_async()
galaxy_import.save()
analysis_status.galaxy_import_task_group_id = \
galaxy_import.taskset_id
analysis_status.set_galaxy_history_state(AnalysisStatus.PROGRESS)
run_analysis.retry(countdown=RETRY_INTERVAL)
# check if data files were successfully imported into Galaxy
galaxy_import = TaskSetResult.restore(
analysis_status.galaxy_import_task_group_id)
if not galaxy_import.ready():
logger.debug("Analysis '%s' pending in Galaxy", analysis)
run_analysis.retry(countdown=RETRY_INTERVAL)
elif not galaxy_import.successful():
error_msg = "Analysis '{}' failed in Galaxy".format(analysis)
logger.error(error_msg)
analysis.set_status(Analysis.FAILURE_STATUS, error_msg)
analysis_status.set_galaxy_history_state(AnalysisStatus.ERROR)
analysis.send_email()
refinery_import.delete()
galaxy_import.delete()
analysis.galaxy_cleanup()
return
# check if analysis has finished running in Galaxy
try:
percent_complete = analysis.galaxy_progress()
except RuntimeError:
analysis_status.set_galaxy_history_state(AnalysisStatus.ERROR)
analysis.send_email()
refinery_import.delete()
galaxy_import.delete()
analysis.galaxy_cleanup()
return
except galaxy.client.ConnectionError:
analysis_status.set_galaxy_history_state(AnalysisStatus.UNKNOWN)
run_analysis.retry(countdown=RETRY_INTERVAL)
else:
# workaround to avoid moving the progress bar backward
if analysis_status.galaxy_history_progress < percent_complete:
analysis_status.galaxy_history_progress = percent_complete
analysis_status.save()
if percent_complete < 100:
analysis_status.set_galaxy_history_state(AnalysisStatus.PROGRESS)
run_analysis.retry(countdown=RETRY_INTERVAL)
else:
analysis_status.set_galaxy_history_state(AnalysisStatus.OK)
# retrieve analysis results from Galaxy
if not analysis_status.galaxy_export_task_group_id:
galaxy_export_tasks = get_galaxy_download_tasks(analysis)
logger.info(
"Starting downloading of results from Galaxy for analysis "
"'%s'", analysis)
galaxy_export = TaskSet(tasks=galaxy_export_tasks).apply_async()
galaxy_export.save()
analysis_status.galaxy_export_task_group_id = galaxy_export.taskset_id
analysis_status.save()
run_analysis.retry(countdown=RETRY_INTERVAL)
# check if analysis results have finished downloading from Galaxy
galaxy_export = TaskSetResult.restore(
analysis_status.galaxy_export_task_group_id)
if not galaxy_export.ready():
logger.debug("Results download pending for analysis '%s'", analysis)
run_analysis.retry(countdown=RETRY_INTERVAL)
# all tasks must have succeeded or failed
elif not galaxy_export.successful():
error_msg = "Analysis '%s' failed while downloading results from " \
"Galaxy".format(analysis)
logger.error(error_msg)
analysis.set_status(Analysis.FAILURE_STATUS, error_msg)
analysis.send_email()
refinery_import.delete()
galaxy_import.delete()
galaxy_export.delete()
analysis.galaxy_cleanup()
return
# attach workflow outputs back to dataset isatab graph
if analysis.workflow.type == Workflow.ANALYSIS_TYPE:
analysis.attach_outputs_dataset()
elif analysis.workflow.type == Workflow.DOWNLOAD_TYPE:
analysis.attach_outputs_downloads()
else:
logger.warning("Unknown workflow type '%s' in analysis '%s'",
analysis.workflow.type, analysis.name)
analysis.set_status(Analysis.SUCCESS_STATUS)
analysis.rename_results()
analysis.send_email()
logger.info("Analysis '%s' finished successfully", analysis)
analysis.galaxy_cleanup()
refinery_import.delete()
galaxy_import.delete()
galaxy_export.delete()
# Update file count and file size of the corresponding data set
analysis.data_set.file_count = analysis.data_set.get_file_count()
# FIXME: line below is causing analyses to be marked as failed
# analysis.data_set.file_size = analysis.data_set.get_file_size()
analysis.data_set.save()
def download_history_files(analysis):
"""Download entire histories from galaxy.
Getting files out of history to file store.
"""
logger.debug("analysis_manger.download_history_files called")
# retrieving list of files to download for workflow
#TODO: handle Django exceptions
analysis = Analysis.objects.get(uuid=analysis.uuid)
dl_files = analysis.workflow_dl_files
### creating dictionary based on files to download predetermined by workflow w/ keep operators
dl_dict = {}
for dl in dl_files.all():
temp_dict = {}
temp_dict['filename'] = dl.filename
temp_dict['pair_id'] = dl.pair_id
dl_dict[str(dl.step_id)] = temp_dict
task_list = []
# gets current galaxy connection
connection = analysis.get_galaxy_connection()
try:
download_list = connection.get_history_file_list(analysis.history_id)
except RuntimeError as exc:
error_msg = "Post-processing failed: " + \
"error downloading Galaxy history files for analysis '{}': {}" \
.format(analysis.name, exc.message)
logger.error(error_msg)
if not isinstance(exc, (ConnectionError, TimeoutError, AuthError)):
analysis.set_status(Analysis.FAILURE_STATUS, error_msg)
try:
analysis.delete_galaxy_library()
analysis.delete_galaxy_workflow()
analysis.delete_galaxy_history()
except RuntimeError:
logger.error("Cleanup failed for analysis '{}'".format(
analysis.name))
return task_list
# Iterating through files in current galaxy history
for results in download_list:
# download file if result state is "ok"
if results['state'] == 'ok':
file_type = results["type"]
curr_file_id = results['name']
if curr_file_id in dl_dict:
curr_dl_dict = dl_dict[curr_file_id]
result_name = curr_dl_dict['filename'] + '.' + file_type
# size of file defined by galaxy
file_size = results['file_size']
# Determing tag if galaxy results should be download through http or copying files directly
local_download = analysis.workflow.workflow_engine.instance.local_download
# to retrieve HTML files as zip archives via dataset URL
if local_download and file_type != 'html':
download_url = results['file_name']
else:
download_url = connection.make_url(str(
results['dataset_id']),
is_data=True,
key=False)
# workaround to set the correct file type for zip archives of
# reports produced by FASTQC
if file_type == 'html':
file_type = 'zip'
# getting file_store_uuid,
# TODO: when changing permanent=True, fix update of % download of file
filestore_uuid = create(source=download_url,
filetype=file_type,
permanent=False)
# adding history files to django model
temp_file = AnalysisResult(analysis_uuid=analysis.uuid,
file_store_uuid=filestore_uuid,
file_name=result_name,
file_type=file_type)
temp_file.save()
analysis.results.add(temp_file)
analysis.save()
# downloading analysis results into file_store
# only download files if size is greater than 1
if file_size > 0:
#task_id = import_file.subtask((filestore_uuid, True, False, file_size,))
# local download, force copying into the file_store instead of symlinking
if local_download:
task_id = import_file.subtask((
filestore_uuid,
False,
True,
file_size,
))
else:
task_id = import_file.subtask((
filestore_uuid,
False,
False,
file_size,
))
task_list.append(task_id)
return task_list
def download_history_files(analysis) :
"""Download entire histories from galaxy.
Getting files out of history to file store.
"""
logger.debug("analysis_manger.download_history_files called")
# retrieving list of files to download for workflow
#TODO: handle Django exceptions
analysis = Analysis.objects.get(uuid=analysis.uuid)
dl_files = analysis.workflow_dl_files
### creating dictionary based on files to download predetermined by workflow w/ keep operators
dl_dict = {}
for dl in dl_files.all():
temp_dict = {}
temp_dict['filename'] = dl.filename
temp_dict['pair_id'] = dl.pair_id
dl_dict[str(dl.step_id)] = temp_dict
task_list = []
# gets current galaxy connection
connection = analysis.get_galaxy_connection()
try:
download_list = connection.get_history_file_list(analysis.history_id)
except RuntimeError as exc:
error_msg = "Post-processing failed: " + \
"error downloading Galaxy history files for analysis '{}': {}" \
.format(analysis.name, exc.message)
logger.error(error_msg)
if not isinstance(exc, (ConnectionError, TimeoutError, AuthError)):
analysis.set_status(Analysis.FAILURE_STATUS, error_msg)
try:
analysis.delete_galaxy_library()
analysis.delete_galaxy_workflow()
analysis.delete_galaxy_history()
except RuntimeError:
logger.error(
"Cleanup failed for analysis '{}'".format(analysis.name))
return task_list
# Iterating through files in current galaxy history
for results in download_list:
# download file if result state is "ok"
if results['state'] == 'ok':
file_type = results["type"]
curr_file_id = results['name']
if curr_file_id in dl_dict:
curr_dl_dict = dl_dict[curr_file_id]
result_name = curr_dl_dict['filename'] + '.' + file_type
# size of file defined by galaxy
file_size = results['file_size']
# Determing tag if galaxy results should be download through http or copying files directly
local_download = analysis.workflow.workflow_engine.instance.local_download
# to retrieve HTML files as zip archives via dataset URL
if local_download and file_type != 'html':
download_url = results['file_name']
else:
download_url = connection.make_url(
str(results['dataset_id']), is_data=True, key=False)
# workaround to set the correct file type for zip archives of
# reports produced by FASTQC
if file_type == 'html':
file_type = 'zip'
# getting file_store_uuid,
# TODO: when changing permanent=True, fix update of % download of file
filestore_uuid = create(
source=download_url,
filetype=file_type,
permanent=False
)
# adding history files to django model
temp_file = AnalysisResult(
analysis_uuid=analysis.uuid, file_store_uuid=filestore_uuid,
file_name=result_name, file_type=file_type)
temp_file.save()
analysis.results.add(temp_file)
analysis.save()
# downloading analysis results into file_store
# only download files if size is greater than 1
if file_size > 0:
#task_id = import_file.subtask((filestore_uuid, True, False, file_size,))
# local download, force copying into the file_store instead of symlinking
if local_download:
task_id = import_file.subtask(
(filestore_uuid, False, True, file_size,))
else:
task_id = import_file.subtask(
(filestore_uuid, False, False, file_size,))
task_list.append(task_id)
return task_list
