def fail_job(self,
run_id,
error_message,
lsf_log_location=None,
input_json_location=None):
lock_id = "run_lock_%s" % run_id
with memcache_task_lock(lock_id, self.app.oid) as acquired:
if acquired:
run = RunObjectFactory.from_db(run_id)
if run.run_obj.is_failed:
logger.info(
format_log("Run Fail already processed", obj=run.run_obj))
return
restart_run = run.run_obj.set_for_restart()
if not restart_run:
run.fail(error_message)
run.to_db()
job_group_notifier = run.job_group_notifier
job_group_notifier_id = str(
job_group_notifier.id) if job_group_notifier else None
ci_review = SetCIReviewEvent(job_group_notifier_id).to_dict()
send_notification.delay(ci_review)
_upload_qc_report(run.run_obj)
_job_finished_notify(run, lsf_log_location,
input_json_location)
else:
run_id, output_directory, execution_id = restart_run
submit_job.delay(run_id, output_directory, execution_id)
else:
logger.warning("Run %s is processing by another worker" % run_id)
def evaluate_sample_errors(self, error_samples):
s = list()
unformatted_s = list()
unformatted_s.append(
"IGO Sample ID\tSample Name / Error\tPatient ID\tSpecimen Type\n")
for sample in error_samples:
s.append("| " + sample["sample_id"] + " | " +
sample["sample_name"] + " |" + sample["patient_id"] +
" |" + sample["specimen_type"] + " |")
unformatted_s.append(sample["sample_id"] + "\t" +
sample["sample_name"] + "\t" +
sample["patient_id"] + "\t" +
sample["specimen_type"] + "\n")
msg = """
Number of samples with error: {number_of_errors}
Error samples (also see error_sample_formatting.txt):
| IGO Sample ID | Sample Name / Error | Patient ID | Specimen Type |
{error_sample_names}
"""
msg = msg.format(number_of_errors=str(len(error_samples)),
error_sample_names="\n".join(s))
self.send_message(msg)
sample_errors_event = UploadAttachmentEvent(
self.job_group_notifier_id, "error_sample_formatting.txt",
"".join(unformatted_s)).to_dict()
send_notification.delay(sample_errors_event)
def evaluate_sample_errors(self, error_samples):
s = list()
unformatted_s = list()
unformatted_s.append(
"IGO Sample ID\tSample Name / Error\tPatient ID\tSpecimen Type\n")
for sample in error_samples:
sample_name = sample.get('SM', "missingSampleName")
sample_id = sample.get('sample_id', 'missingSampleId')
patient_id = sample.get('patient_id', 'missingPatientId')
specimen_type = sample.get('specimen_type', 'missingSpecimenType')
s.append("| " + sample_id + " | " + sample_name + " |" +
patient_id + " |" + specimen_type + " |")
unformatted_s.append(sample_id + "\t" + sample_name + "\t" +
patient_id + "\t" + specimen_type + "\n")
msg = """
Number of samples with error: {number_of_errors}
Error samples (also see error_sample_formatting.txt):
| IGO Sample ID | Sample Name / Error | Patient ID | Specimen Type |
{error_sample_names}
"""
msg = msg.format(number_of_errors=str(len(error_samples)),
error_sample_names='\n'.join(s))
self.send_message(msg)
sample_errors_event = UploadAttachmentEvent(
self.job_group_notifier_id, 'error_sample_formatting.txt',
"".join(unformatted_s)).to_dict()
send_notification.delay(sample_errors_event)
def _job_failed(self, permission_denied=False, recipe=None):
if self.job.run == TYPES["REQUEST"]:
if permission_denied:
cc = settings.PERMISSION_DENIED_CC.get(recipe, "")
permission_denied_event = PermissionDeniedEvent(
self.job.job_group_notifier.id,
"Failed to copy files because of the Permission denied issue",
cc).to_dict()
send_notification.delay(permission_denied_event)
emails = settings.PERMISSION_DENIED_EMAILS.get(recipe,
"").split(",")
for email in emails:
content = (
"Request failed to be imported because some files don't have proper permissions. "
"Check more details on %s/v0/etl/jobs/%s/" %
(settings.BEAGLE_URL, str(self.job.id)))
email = SendEmailEvent(
job_notifier=settings.BEAGLE_NOTIFIER_EMAIL_GROUP,
email_to=email,
email_from=settings.BEAGLE_NOTIFIER_EMAIL_FROM,
subject="Permission Denied for request_id: %s" %
self.job.args.get("request_id"),
content=content,
)
send_notification.delay(email.to_dict())
self._generate_ticket_decription()
def _job_finished_notify(run):
job_group = run.job_group
job_group_notifier = run.job_group_notifier
job_group_notifier_id = str(
job_group_notifier.id) if job_group_notifier else None
pipeline_name = run.run_obj.app.name
pipeline_link = run.run_obj.app.pipeline_link
if run.run_obj.operator_run:
operator_run_id = str(run.run_obj.operator_run.id)
total_runs = run.run_obj.operator_run.total_runs
completed_runs = run.run_obj.operator_run.completed_runs
failed_runs = run.run_obj.operator_run.failed_runs
running_runs = run.run_obj.operator_run.running_runs
else:
operator_run_id = None
total_runs = 1
if run.status == RunStatus.COMPLETED:
completed_runs, failed_runs = 1, 0
else:
completed_runs, failed_runs = 0, 1
running_runs = 0
event = RunFinishedEvent(job_group_notifier_id,
run.tags.get('requestId', 'UNKNOWN REQUEST'),
str(run.run_id), pipeline_name, pipeline_link,
run.run_obj.output_directory,
RunStatus(run.status).name, run.tags,
running_runs, completed_runs, failed_runs,
total_runs, operator_run_id)
e = event.to_dict()
send_notification.delay(e)
def create_request_job(request_id, redelivery=False):
logger.info(
"Searching for job: %s for request_id: %s" % (TYPES['REQUEST'], request_id))
count = Job.objects.filter(run=TYPES['REQUEST'], args__request_id=request_id,
status__in=[JobStatus.CREATED, JobStatus.IN_PROGRESS,
JobStatus.WAITING_FOR_CHILDREN]).count()
request_redelivered = Job.objects.filter(run=TYPES['REQUEST'], args__request_id=request_id).count() > 0
assays = ETLConfiguration.objects.first()
if request_redelivered and not (assays.redelivery and redelivery):
return None, "Request is redelivered, but redelivery deactivated"
if count == 0:
job_group = JobGroup()
job_group.save()
job_group_notifier_id = notifier_start(job_group, request_id)
job_group_notifier = JobGroupNotifier.objects.get(id=job_group_notifier_id)
job = Job(run=TYPES['REQUEST'],
args={'request_id': request_id, 'job_group': str(job_group.id),
'job_group_notifier': job_group_notifier_id, 'redelivery': request_redelivered},
status=JobStatus.CREATED,
max_retry=1,
children=[],
callback=TYPES['REQUEST_CALLBACK'],
callback_args={'request_id': request_id, 'job_group': str(job_group.id),
'job_group_notifier': job_group_notifier_id},
job_group=job_group,
job_group_notifier=job_group_notifier)
job.save()
if request_redelivered:
redelivery_event = RedeliveryEvent(job_group_notifier_id).to_dict()
send_notification.delay(redelivery_event)
return job, "Job Created"
def generate_description(job_group, job_group_notifier, request):
files = FileRepository.filter(metadata={
"requestId": request,
"igocomplete": True
})
if files:
data = files.first().metadata
request_id = data["requestId"]
recipe = data["recipe"]
a_name = data["dataAnalystName"]
a_email = data["dataAnalystEmail"]
i_name = data["investigatorName"]
i_email = data["investigatorEmail"]
l_name = data["labHeadName"]
l_email = data["labHeadEmail"]
p_email = data["piEmail"]
pm_name = data["projectManagerName"]
qc_emails = data["qcAccessEmails"] if "qcAccessEmails" in data else ""
data_access_emails = data[
"dataAccessEmails"] if "dataAccessEmails" in data else ""
other_contact_emails = data[
"otherContactEmails"] if "otherContactEmails" in data else ""
num_samples = len(
files.order_by().values("metadata__cmoSampleName").annotate(
n=Count("pk")))
num_tumors = len(
FileRepository.filter(
queryset=files, metadata={
"tumorOrNormal": "Tumor"
}).order_by().values("metadata__cmoSampleName").annotate(
n=Count("pk")))
num_normals = len(
FileRepository.filter(
queryset=files, metadata={
"tumorOrNormal": "Normal"
}).order_by().values("metadata__cmoSampleName").annotate(
n=Count("pk")))
operator_start_event = OperatorStartEvent(
job_group_notifier,
job_group,
request_id,
num_samples,
recipe,
a_name,
a_email,
i_name,
i_email,
l_name,
l_email,
p_email,
pm_name,
qc_emails,
num_tumors,
num_normals,
data_access_emails,
other_contact_emails,
).to_dict()
send_notification.delay(operator_start_event)
def _job_failed(self):
if self.job.run == TYPES['REQUEST']:
e = ETLJobFailedEvent(
self.job.job_group_notifier.id,
"[CIReviewEvent] ETL Job failed, likely child job import. Check pooled normal import, might already exist in database."
).to_dict()
send_notification.delay(e)
self._generate_ticket_decription()
def _send_as_notification(val, job_group):
uri = val.get('location')
path = FileProcessor.parse_path_from_uri(uri)
file_name = os.path.basename(path)
if job_group:
event = UploadAttachmentEvent(str(job_group.id), file_name, path, download=True)
send_notification.delay(event.to_dict())
logger.info("Can't upload file:%s. JobGroup not specified", path)
return val
def generate_label(job_group_id, request):
files = FileRepository.filter(metadata={
'requestId': request,
'igocomplete': True
})
if files:
data = files.first().metadata
recipe = data['recipe']
recipe_label_event = SetLabelEvent(job_group_id, recipe).to_dict()
send_notification.delay(recipe_label_event)
def fetch_samples(request_id, import_pooled_normals=True, import_samples=True, job_group=None, job_group_notifier=None,
redelivery=False):
logger.info("Fetching sampleIds for requestId:%s" % request_id)
jg = None
jgn = None
try:
jg = JobGroup.objects.get(id=job_group)
logger.debug("JobGroup found")
except JobGroup.DoesNotExist:
logger.debug("No JobGroup Found")
try:
jgn = JobGroupNotifier.objects.get(id=job_group_notifier)
logger.debug("JobGroup found")
except JobGroupNotifier.DoesNotExist:
logger.debug("No JobGroup Found")
children = set()
sample_ids = LIMSClient.get_request_samples(request_id)
if sample_ids['requestId'] != request_id:
raise ErrorInconsistentDataException(
"LIMS returned wrong response for request %s. Got %s instead" % (request_id, sample_ids['requestId']))
request_metadata = {
"dataAnalystEmail": sample_ids['dataAnalystEmail'],
"dataAnalystName": sample_ids['dataAnalystName'],
"investigatorEmail": sample_ids['investigatorEmail'],
"investigatorName": sample_ids['investigatorName'],
"labHeadEmail": sample_ids['labHeadEmail'],
"labHeadName": sample_ids['labHeadName'],
"otherContactEmails": sample_ids['otherContactEmails'],
"dataAccessEmails": sample_ids['dataAccessEmails'],
"qcAccessEmails": sample_ids['qcAccessEmails'],
"projectManagerName": sample_ids['projectManagerName'],
"recipe": sample_ids['recipe'],
"piEmail": sample_ids["piEmail"],
}
set_recipe_event = ETLSetRecipeEvent(job_group_notifier, request_metadata['recipe']).to_dict()
send_notification.delay(set_recipe_event)
pooled_normals = sample_ids.get("pooledNormals", [])
if import_pooled_normals and pooled_normals:
for f in pooled_normals:
job = get_or_create_pooled_normal_job(f, jg)
children.add(str(job.id))
if import_samples:
if not sample_ids.get('samples', False):
raise FailedToFetchSampleException("No samples reported for requestId: %s" % request_id)
for sample in sample_ids.get('samples', []):
job = create_sample_job(sample['igoSampleId'],
sample['igocomplete'],
request_id,
request_metadata,
redelivery,
jg,
jgn)
children.add(str(job.id))
return list(children)
def fail_job(run_id, error_message):
run = RunObject.from_db(run_id)
run.fail(error_message)
run.to_db()
job_group_notifier = run.job_group_notifier
job_group_notifier_id = str(
job_group_notifier.id) if job_group_notifier else None
ci_review = SetCIReviewEvent(job_group_notifier_id).to_dict()
send_notification.delay(ci_review)
_job_finished_notify(run)
def _send_notifications(self, job_group_notifier_id, run):
pipeline_name = run.app.name
pipeline_version = run.app.version
pipeline_link = run.app.pipeline_link
pipeline_description_event = AddPipelineToDescriptionEvent(
job_group_notifier_id, pipeline_name, pipeline_version,
pipeline_link).to_dict()
send_notification.delay(pipeline_description_event)
run_event = RunStartedEvent(job_group_notifier_id, str(run.id),
run.app.name, run.app.pipeline_link,
run.output_directory, run.tags).to_dict()
send_notification.delay(run_event)
def create_jobs_from_request(request_id,
operator_id,
job_group_id,
job_group_notifier_id=None,
pipeline=None):
logger.info(
format_log("Creating operator with %s" % operator_id,
job_group_id=job_group_id,
request_id=request_id))
operator_model = Operator.objects.get(id=operator_id)
if not job_group_notifier_id:
try:
job_group = JobGroup.objects.get(id=job_group_id)
except JobGroup.DoesNotExist:
logger.info(
format_log("Job group does not exist" % operator_id,
job_group_id=job_group_id,
request_id=request_id))
return
try:
job_group_notifier_id = notifier_start(job_group,
request_id,
operator=operator_model)
request_obj = Request.objects.filter(request_id=request_id).first()
if request_obj:
delivery_date_event = SetDeliveryDateFieldEvent(
job_group_notifier_id,
str(request_obj.delivery_date)).to_dict()
send_notification.delay(delivery_date_event)
except Exception as e:
logger.info(
format_log("Failed to instantiate notifier" % operator_id,
job_group_id=job_group_id,
request_id=request_id))
operator = OperatorFactory.get_by_model(
operator_model,
job_group_id=job_group_id,
job_group_notifier_id=job_group_notifier_id,
request_id=request_id,
pipeline=pipeline,
)
_set_link_to_run_ticket(request_id, job_group_notifier_id)
generate_description(job_group_id, job_group_notifier_id, request_id)
generate_label(job_group_notifier_id, request_id)
create_jobs_from_operator(operator, job_group_id, job_group_notifier_id)
def write_to_file(self, fname, s):
"""
Writes file to temporary location, then registers it to the temp file group
Also uploads it to notifier if there is a job group id
"""
output = os.path.join(self.OUTPUT_DIR, fname)
with open(output, "w+") as fh:
fh.write(s)
os.chmod(output, 0o777)
self.register_tmp_file(output)
if self.job_group_notifier_id:
upload_file_event = UploadAttachmentEvent(
self.job_group_notifier_id, fname, s).to_dict()
send_notification.delay(upload_file_event)
return {'class': 'File', 'location': "juno://" + output}
def approve(modeladmin, request, queryset):
for req in queryset:
email = "*****@*****.**" % req.username
User.objects.create(username=req.username, email=email, first_name=req.first_name, last_name=req.last_name)
req.approved = True
req.save()
content = "Your request to access Voyager is approved"
email = SendEmailEvent(
job_notifier=settings.BEAGLE_NOTIFIER_EMAIL_GROUP,
email_to=email,
email_from=settings.BEAGLE_NOTIFIER_EMAIL_FROM,
subject="Registration approved",
content=content,
)
send_notification.delay(email.to_dict())
def save(self, *args, **kwargs):
if self.pk is None:
content = "User %s %s, with email %[email protected] requested Voyager access." % (
self.first_name,
self.last_name,
self.username,
)
for email in settings.BEAGLE_NOTIFIER_EMAIL_ABOUT_NEW_USERS.split(
','):
email = SendEmailEvent(
job_notifier=settings.BEAGLE_NOTIFIER_EMAIL_GROUP,
email_to=email,
email_from=settings.BEAGLE_NOTIFIER_EMAIL_FROM,
subject='Registration access',
content=content)
send_notification.delay(email.to_dict())
super(UserRegistrationRequest, self).save(*args, **kwargs)
def generate_description(job_group, job_group_notifier, request):
files = FileRepository.filter(metadata={
'requestId': request,
'igocomplete': True
})
if files:
data = files.first().metadata
request_id = data['requestId']
recipe = data['recipe']
a_name = data['dataAnalystName']
a_email = data['dataAnalystEmail']
i_name = data['investigatorName']
i_email = data['investigatorEmail']
l_name = data['labHeadName']
l_email = data['labHeadEmail']
p_email = data['piEmail']
pm_name = data['projectManagerName']
qc_emails = data['qcAccessEmails'] if 'qcAccessEmails' in data else ""
num_samples = len(
files.order_by().values('metadata__cmoSampleName').annotate(
n=Count("pk")))
num_tumors = len(
FileRepository.filter(
queryset=files, metadata={
'tumorOrNormal': 'Tumor'
}).order_by().values('metadata__cmoSampleName').annotate(
n=Count("pk")))
num_normals = len(
FileRepository.filter(
queryset=files, metadata={
'tumorOrNormal': 'Normal'
}).order_by().values('metadata__cmoSampleName').annotate(
n=Count("pk")))
operator_start_event = OperatorStartEvent(
job_group_notifier, job_group, request_id, num_samples, recipe,
a_name, a_email, i_name, i_email, l_name, l_email, p_email,
pm_name, qc_emails, num_tumors, num_normals).to_dict()
send_notification.delay(operator_start_event)
def _set_link_to_run_ticket(request_id, job_group_notifier_id):
jira_id = None
import_job = Job.objects.filter(
run=TYPES['REQUEST'],
args__request_id=request_id).order_by('-created_date').first()
if not import_job:
logger.error("Could not find Import JIRA ticket")
return
try:
job_group_notifier_job = JobGroupNotifier.objects.get(
job_group=import_job.job_group.id, notifier_type__default=True)
except JobGroupNotifier.DoesNotExist:
logger.error("Could not find Import JIRA ticket")
return
try:
new_jira = JobGroupNotifier.objects.get(id=job_group_notifier_id)
except JobGroupNotifier.DoesNotExist:
logger.error("Could not find Import JIRA ticket")
return
event = SetRunTicketInImportEvent(job_notifier=str(
job_group_notifier_job.id),
run_jira_id=new_jira.jira_id).to_dict()
send_notification.delay(event)
def on_job_fail(self, run):
cmo_sample_name = run.tags.get("sampleNameTumor")
files = FileRepository.filter(
queryset=self.files, metadata={"cmoSampleName": cmo_sample_name})
if files:
qc_report = files[0].metadata["qcReports"]
sample_id = files[0].metadata["sampleId"]
"""
{
"comments": "Suboptimal quantity",
"qcReportType": "LIBRARY",
"IGORecommendation": "Try",
"investigatorDecision": "Continue processing"
}
"""
report_str = ""
for report in qc_report:
report_str += "{comments}\t{qc_report_type}\t{igo_recommendation}\t{investigator_decision}\n".format(
comments=report["comments"],
qc_report_type=report["qcReportType"],
igo_recommendation=report["IGORecommendation"],
investigator_decision=report["investigatorDecision"],
)
msg = """
cmoSampleId: {cmo_sample_name}
sampleId: {sample_id}
Comments\tQC Report Type\tIGORecommendation\tInvestigator Decision\n
{report_str}
""".format(cmo_sample_name=cmo_sample_name,
sample_id=sample_id,
report_str=report_str)
file_name = "{cmo_sample_name}_igo_qc_report".format(
cmo_sample_name=cmo_sample_name)
sample_errors_event = UploadAttachmentEvent(
self.job_group_notifier_id, file_name, msg).to_dict()
send_notification.delay(sample_errors_event)
def create_or_update_file(path, request_id, file_group_id, file_type, igocomplete, data, library, run, sample,
request_metadata, r, update=False, job_group_notifier=None):
logger.info("Creating file %s " % path)
try:
file_group_obj = FileGroup.objects.get(id=file_group_id)
file_type_obj = FileType.objects.filter(name=file_type).first()
lims_metadata = copy.deepcopy(data)
library_copy = copy.deepcopy(library)
lims_metadata['requestId'] = request_id
lims_metadata['igocomplete'] = igocomplete
lims_metadata['R'] = r
for k, v in library_copy.items():
lims_metadata[k] = v
for k, v in run.items():
lims_metadata[k] = v
for k, v in request_metadata.items():
lims_metadata[k] = v
metadata = format_metadata(lims_metadata)
# validator = MetadataValidator(METADATA_SCHEMA)
except Exception as e:
logger.error("Failed to parse metadata for file %s path" % path)
raise FailedToFetchSampleException("Failed to create file %s. Error %s" % (path, str(e)))
try:
logger.info(lims_metadata)
# validator.validate(metadata)
except MetadataValidationException as e:
logger.error("Failed to create file %s. Error %s" % (path, str(e)))
raise FailedToFetchSampleException("Failed to create file %s. Error %s" % (path, str(e)))
else:
f = FileRepository.filter(path=path).first()
if not f:
create_file_object(path, file_group_obj, lims_metadata, metadata, file_type_obj, sample)
if update:
message = "File registered: %s" % path
update = RedeliveryUpdateEvent(job_group_notifier, message).to_dict()
send_notification.delay(update)
else:
if update:
before = f.file.filemetadata_set.order_by('-created_date').count()
update_file_object(f.file, path, metadata)
after = f.file.filemetadata_set.order_by('-created_date').count()
if after != before:
all_metadata = f.file.filemetadata_set.order_by('-created_date')
ddiff = DeepDiff(all_metadata[1].metadata,
all_metadata[0].metadata,
ignore_order=True)
diff_file_name = "%s_metadata_update.json" % f.file.file_name
message = "Updating file metadata: %s, details in file %s\n" % (path, diff_file_name)
update = RedeliveryUpdateEvent(job_group_notifier, message).to_dict()
diff_details_event = LocalStoreFileEvent(job_group_notifier, diff_file_name, str(ddiff)).to_dict()
send_notification.delay(update)
send_notification.delay(diff_details_event)
else:
raise FailedToFetchSampleException("File %s already exist with id %s" % (path, str(f.id)))
def create_or_update_file(
path,
request_id,
file_group_id,
file_type,
igocomplete,
data,
library,
run,
request_metadata,
r,
update=False,
job_group_notifier=None,
):
logger.info("Creating file %s " % path)
try:
file_group_obj = FileGroup.objects.get(id=file_group_id)
file_type_obj = FileType.objects.filter(name=file_type).first()
lims_metadata = copy.deepcopy(data)
library_copy = copy.deepcopy(library)
lims_metadata["requestId"] = request_id
lims_metadata["igocomplete"] = igocomplete
lims_metadata["R"] = r
for k, v in library_copy.items():
lims_metadata[k] = v
for k, v in run.items():
lims_metadata[k] = v
for k, v in request_metadata.items():
lims_metadata[k] = v
metadata = format_metadata(lims_metadata)
# validator = MetadataValidator(METADATA_SCHEMA)
except Exception as e:
logger.error("Failed to parse metadata for file %s path" % path)
raise FailedToFetchSampleException(
"Failed to create file %s. Error %s" % (path, str(e)))
try:
logger.info(lims_metadata)
# validator.validate(metadata)
except MetadataValidationException as e:
logger.error("Failed to create file %s. Error %s" % (path, str(e)))
raise FailedToFetchSampleException(
"Failed to create file %s. Error %s" % (path, str(e)))
else:
recipe = metadata.get("recipe", "")
new_path = CopyService.remap(recipe, path) # Get copied file path
f = FileRepository.filter(path=new_path).first()
if not f:
try:
if path != new_path:
CopyService.copy(path, new_path)
except Exception as e:
if "Permission denied" in str(e):
raise FailedToCopyFilePermissionDeniedException(
"Failed to copy file %s. Error %s" % (path, str(e)))
else:
raise FailedToCopyFileException(
"Failed to copy file %s. Error %s" % (path, str(e)))
create_file_object(new_path, file_group_obj, lims_metadata,
metadata, file_type_obj)
if update:
message = "File registered: %s" % path
update = RedeliveryUpdateEvent(job_group_notifier,
message).to_dict()
send_notification.delay(update)
else:
if update:
before = f.file.filemetadata_set.order_by(
"-created_date").count()
update_file_object(f.file, f.file.path, metadata)
after = f.file.filemetadata_set.order_by(
"-created_date").count()
if after != before:
all_metadata = f.file.filemetadata_set.order_by(
"-created_date")
ddiff = DeepDiff(all_metadata[1].metadata,
all_metadata[0].metadata,
ignore_order=True)
diff_file_name = "%s_metadata_update_%s.json" % (
f.file.file_name, all_metadata[0].version)
message = "Updating file metadata: %s, details in file %s\n" % (
path, diff_file_name)
update = RedeliveryUpdateEvent(job_group_notifier,
message).to_dict()
diff_details_event = LocalStoreFileEvent(
job_group_notifier, diff_file_name,
str(ddiff)).to_dict()
send_notification.delay(update)
send_notification.delay(diff_details_event)
else:
raise FailedToFetchSampleException(
"File %s already exist with id %s" % (path, str(f.id)))
def send_message(self, msg):
event = OperatorRequestEvent(self.job_group_notifier_id, msg)
e = event.to_dict()
send_notification.delay(e)
def get_jobs(self):
files = FileRepository.filter(queryset=self.files,
metadata={
"requestId": self.request_id,
"igocomplete": True
})
argos_jobs = list()
cnt_tumors = FileRepository.filter(queryset=self.files,
metadata={
"requestId": self.request_id,
"tumorOrNormal": "Tumor",
"igocomplete": True
}).count()
if cnt_tumors == 0:
cant_do = CantDoEvent(self.job_group_notifier_id).to_dict()
send_notification.delay(cant_do)
all_normals_event = SetLabelEvent(self.job_group_notifier_id,
"all_normals").to_dict()
send_notification.delay(all_normals_event)
return argos_jobs
data = list()
for f in files:
sample = dict()
sample["id"] = f.file.id
sample["path"] = f.file.path
sample["file_name"] = f.file.file_name
sample["metadata"] = f.metadata
data.append(sample)
files = list()
samples = list()
# group by igoId
igo_id_group = dict()
for sample in data:
igo_id = sample["metadata"]["sampleId"]
if igo_id not in igo_id_group:
igo_id_group[igo_id] = list()
igo_id_group[igo_id].append(sample)
for igo_id in igo_id_group:
samples.append(build_sample(igo_id_group[igo_id]))
argos_inputs, error_samples = construct_argos_jobs(samples)
number_of_inputs = len(argos_inputs)
sample_pairing = ""
sample_mapping = ""
pipeline = self.get_pipeline_id()
try:
pipeline_obj = Pipeline.objects.get(id=pipeline)
except Pipeline.DoesNotExist:
pass
for i, job in enumerate(argos_inputs):
tumor_sample_name = job["pair"][0]["ID"]
for p in job["pair"][0]["R1"]:
filepath = FileProcessor.parse_path_from_uri(p["location"])
if filepath not in files:
sample_mapping += "\t".join([tumor_sample_name, filepath
]) + "\n"
files.append(filepath)
for p in job["pair"][0]["R2"]:
filepath = FileProcessor.parse_path_from_uri(p["location"])
if filepath not in files:
sample_mapping += "\t".join([tumor_sample_name, filepath
]) + "\n"
files.append(filepath)
for p in job["pair"][0]["zR1"]:
filepath = FileProcessor.parse_path_from_uri(p["location"])
if filepath not in files:
sample_mapping += "\t".join([tumor_sample_name, filepath
]) + "\n"
files.append(filepath)
for p in job["pair"][0]["zR2"]:
filepath = FileProcessor.parse_path_from_uri(p["location"])
if filepath not in files:
sample_mapping += "\t".join([tumor_sample_name, filepath
]) + "\n"
files.append(filepath)
normal_sample_name = job["pair"][1]["ID"]
for p in job["pair"][1]["R1"]:
filepath = FileProcessor.parse_path_from_uri(p["location"])
if filepath not in files:
sample_mapping += "\t".join([normal_sample_name, filepath
]) + "\n"
files.append(filepath)
for p in job["pair"][1]["R2"]:
filepath = FileProcessor.parse_path_from_uri(p["location"])
if filepath not in files:
sample_mapping += "\t".join([normal_sample_name, filepath
]) + "\n"
files.append(filepath)
for p in job["pair"][1]["zR1"]:
filepath = FileProcessor.parse_path_from_uri(p["location"])
if filepath not in files:
sample_mapping += "\t".join([normal_sample_name, filepath
]) + "\n"
files.append(filepath)
for p in job["pair"][1]["zR2"]:
filepath = FileProcessor.parse_path_from_uri(p["location"])
if filepath not in files:
sample_mapping += "\t".join([normal_sample_name, filepath
]) + "\n"
files.append(filepath)
for p in job["pair"][1]["bam"]:
filepath = FileProcessor.parse_path_from_uri(p["location"])
if filepath not in files:
sample_mapping += "\t".join([normal_sample_name, filepath
]) + "\n"
files.append(filepath)
name = "ARGOS %s, %i of %i" % (self.request_id, i + 1,
number_of_inputs)
assay = job["assay"]
pi = job["pi"]
pi_email = job["pi_email"]
sample_pairing += "\t".join(
[normal_sample_name, tumor_sample_name]) + "\n"
tags = {
"requestId": self.request_id,
"sampleNameTumor": tumor_sample_name,
"sampleNameNormal": normal_sample_name,
"labHeadName": pi,
"labHeadEmail": pi_email,
}
argos_jobs.append(
RunCreator(app=pipeline, inputs=job, name=name, tags=tags))
operator_run_summary = UploadAttachmentEvent(
self.job_group_notifier_id, "sample_pairing.txt",
sample_pairing).to_dict()
send_notification.delay(operator_run_summary)
mapping_file_event = UploadAttachmentEvent(self.job_group_notifier_id,
"sample_mapping.txt",
sample_mapping).to_dict()
send_notification.delay(mapping_file_event)
data_clinical = generate_sample_data_content(
files,
pipeline_name=pipeline_obj.name,
pipeline_github=pipeline_obj.github,
pipeline_version=pipeline_obj.version,
)
sample_data_clinical_event = UploadAttachmentEvent(
self.job_group_notifier_id, "sample_data_clinical.txt",
data_clinical).to_dict()
send_notification.delay(sample_data_clinical_event)
self.evaluate_sample_errors(error_samples)
self.summarize_pairing_info(argos_inputs)
return argos_jobs
def request_callback(request_id, job_group=None, job_group_notifier=None):
jg = None
jgn = None
try:
jgn = JobGroupNotifier.objects.get(id=job_group_notifier)
logger.debug("[RequestCallback] JobGroup id: %s", job_group)
except JobGroupNotifier.DoesNotExist:
logger.debug("[RequestCallback] JobGroup not set")
job_group_notifier_id = str(jgn.id) if jgn else None
assays = ETLConfiguration.objects.first()
recipe = LIMSClient.get_request_samples(request_id).get("recipe", None)
if (not all([
JobStatus(job["status"]) == JobStatus.COMPLETED for job in
Job.objects.filter(job_group=job_group,
run=TYPES["SAMPLE"],
args__igocomplete=True).values("status")
]) and recipe in settings.WES_ASSAYS):
wes_job_failed = WESJobFailedEvent(job_group_notifier_id, recipe)
send_notification.delay(wes_job_failed.to_dict())
if not recipe:
raise FailedToSubmitToOperatorException(
"Not enough metadata to choose the operator for requestId:%s" %
request_id)
if not all(item in assays.all_recipes for item in [recipe]):
ci_review_e = SetCIReviewEvent(job_group_notifier_id).to_dict()
send_notification.delay(ci_review_e)
set_unknown_assay_label = SetLabelEvent(
job_group_notifier_id, "unrecognized_assay").to_dict()
send_notification.delay(set_unknown_assay_label)
unknown_assay_event = UnknownAssayEvent(job_group_notifier_id,
recipe).to_dict()
send_notification.delay(unknown_assay_event)
return []
if any(item in assays.hold_recipes for item in [
recipe,
]):
admin_hold_event = AdminHoldEvent(job_group_notifier_id).to_dict()
send_notification.delay(admin_hold_event)
custom_capture_event = CustomCaptureCCEvent(job_group_notifier_id,
recipe).to_dict()
send_notification.delay(custom_capture_event)
return []
if any(item in assays.disabled_recipes for item in [
recipe,
]):
not_for_ci = NotForCIReviewEvent(job_group_notifier_id).to_dict()
send_notification.delay(not_for_ci)
disabled_assay_event = DisabledAssayEvent(job_group_notifier_id,
recipe).to_dict()
send_notification.delay(disabled_assay_event)
return []
if len(
FileRepository.filter(metadata={
"requestId": request_id
},
values_metadata="recipe").all()) == 0:
no_samples_event = AdminHoldEvent(job_group_notifier_id).to_dict()
send_notification.delay(no_samples_event)
return []
if not all([
JobStatus(job["status"]) == JobStatus.COMPLETED
for job in Job.objects.filter(job_group=job_group).values("status")
]):
ci_review_e = SetCIReviewEvent(job_group_notifier_id).to_dict()
send_notification.delay(ci_review_e)
lab_head_email = FileRepository.filter(
metadata={
"requestId": request_id
}, values_metadata="labHeadEmail").first()
try:
if lab_head_email.split("@")[1] != "mskcc.org":
event = ExternalEmailEvent(job_group_notifier_id,
request_id).to_dict()
send_notification.delay(event)
except Exception:
logger.error("Failed to check labHeadEmail")
if len(
FileRepository.filter(metadata={
"requestId": request_id,
"tumorOrNormal": "Tumor"
})) == 0:
only_normal_samples_event = OnlyNormalSamplesEvent(
job_group_notifier_id, request_id).to_dict()
send_notification.delay(only_normal_samples_event)
if recipe in settings.ASSAYS_ADMIN_HOLD_ONLY_NORMALS:
admin_hold_event = AdminHoldEvent(job_group_notifier_id).to_dict()
send_notification.delay(admin_hold_event)
return []
operators = Operator.objects.filter(recipes__overlap=[recipe])
if not operators:
# TODO: Import ticket will have CIReviewNeeded
msg = "No operator defined for requestId %s with recipe %s" % (
request_id, recipe)
logger.error(msg)
e = OperatorRequestEvent(job_group_notifier_id,
"[CIReviewEvent] %s" % msg).to_dict()
send_notification.delay(e)
ci_review_e = SetCIReviewEvent(job_group_notifier_id).to_dict()
send_notification.delay(ci_review_e)
raise FailedToSubmitToOperatorException(msg)
for operator in operators:
if not operator.active:
msg = "Operator not active: %s" % operator.class_name
logger.info(msg)
e = OperatorRequestEvent(job_group_notifier_id,
"[CIReviewEvent] %s" % msg).to_dict()
send_notification.delay(e)
error_label = SetLabelEvent(job_group_notifier_id,
"operator_inactive").to_dict()
send_notification.delay(error_label)
ci_review_e = SetCIReviewEvent(job_group_notifier_id).to_dict()
send_notification.delay(ci_review_e)
else:
logger.info("Submitting request_id %s to %s operator" %
(request_id, operator.class_name))
if Job.objects.filter(job_group=job_group,
args__request_id=request_id,
run=TYPES["SAMPLE"],
status=JobStatus.FAILED).all():
partialy_complete_event = ETLImportPartiallyCompleteEvent(
job_notifier=job_group_notifier_id).to_dict()
send_notification.delay(partialy_complete_event)
else:
complete_event = ETLImportCompleteEvent(
job_notifier=job_group_notifier_id).to_dict()
send_notification.delay(complete_event)
create_jobs_from_request.delay(request_id, operator.id, job_group)
return []
def get_jobs(self):
argos_jobs = list()
if self.request_id:
files = FileRepository.filter(queryset=self.files,
metadata={
'requestId': self.request_id,
'igocomplete': True
},
filter_redact=True)
cnt_tumors = FileRepository.filter(queryset=self.files,
metadata={
'requestId':
self.request_id,
'tumorOrNormal': 'Tumor',
'igocomplete': True
},
filter_redact=True).count()
elif self.pairing:
files, cnt_tumors = self.get_files_for_pairs()
if cnt_tumors == 0:
cant_do = CantDoEvent(self.job_group_notifier_id).to_dict()
send_notification.delay(cant_do)
all_normals_event = SetLabelEvent(self.job_group_notifier_id,
'all_normals').to_dict()
send_notification.delay(all_normals_event)
return argos_jobs
data = list()
for f in files:
sample = dict()
sample['id'] = f.file.id
sample['path'] = f.file.path
sample['file_name'] = f.file.file_name
sample['metadata'] = f.metadata
data.append(sample)
files = list()
samples = list()
# group by igoId
igo_id_group = dict()
for sample in data:
igo_id = sample['metadata']['sampleId']
if igo_id not in igo_id_group:
igo_id_group[igo_id] = list()
igo_id_group[igo_id].append(sample)
for igo_id in igo_id_group:
samples.append(build_sample(igo_id_group[igo_id]))
argos_inputs, error_samples = construct_argos_jobs(
samples, self.pairing)
number_of_inputs = len(argos_inputs)
sample_pairing = ""
sample_mapping = ""
pipeline = self.get_pipeline_id()
try:
pipeline_obj = Pipeline.objects.get(id=pipeline)
except Pipeline.DoesNotExist:
pass
check_for_duplicates = list()
for i, job in enumerate(argos_inputs):
tumor_sample_name = job['pair'][0]['ID']
for p in job['pair'][0]['R1']:
filepath = FileProcessor.parse_path_from_uri(p['location'])
file_str = "\t".join([tumor_sample_name, filepath]) + "\n"
if file_str not in check_for_duplicates:
check_for_duplicates.append(file_str)
sample_mapping += file_str
if filepath not in files:
files.append(filepath)
for p in job['pair'][0]['R2']:
filepath = FileProcessor.parse_path_from_uri(p['location'])
file_str = "\t".join([tumor_sample_name, filepath]) + "\n"
if file_str not in check_for_duplicates:
check_for_duplicates.append(file_str)
sample_mapping += file_str
if filepath not in files:
files.append(filepath)
for p in job['pair'][0]['zR1']:
filepath = FileProcessor.parse_path_from_uri(p['location'])
file_str = "\t".join([tumor_sample_name, filepath]) + "\n"
if file_str not in check_for_duplicates:
check_for_duplicates.append(file_str)
sample_mapping += file_str
if filepath not in files:
files.append(filepath)
for p in job['pair'][0]['zR2']:
filepath = FileProcessor.parse_path_from_uri(p['location'])
file_str = "\t".join([tumor_sample_name, filepath]) + "\n"
if file_str not in check_for_duplicates:
check_for_duplicates.append(file_str)
sample_mapping += file_str
if filepath not in files:
files.append(filepath)
normal_sample_name = job['pair'][1]['ID']
for p in job['pair'][1]['R1']:
filepath = FileProcessor.parse_path_from_uri(p['location'])
file_str = "\t".join([normal_sample_name, filepath]) + "\n"
if file_str not in check_for_duplicates:
check_for_duplicates.append(file_str)
sample_mapping += file_str
if filepath not in files:
files.append(filepath)
for p in job['pair'][1]['R2']:
filepath = FileProcessor.parse_path_from_uri(p['location'])
file_str = "\t".join([normal_sample_name, filepath]) + "\n"
if file_str not in check_for_duplicates:
check_for_duplicates.append(file_str)
sample_mapping += file_str
if filepath not in files:
sample_mapping += "\t".join([normal_sample_name, filepath
]) + "\n"
files.append(filepath)
for p in job['pair'][1]['zR1']:
filepath = FileProcessor.parse_path_from_uri(p['location'])
file_str = "\t".join([normal_sample_name, filepath]) + "\n"
if file_str not in check_for_duplicates:
check_for_duplicates.append(file_str)
sample_mapping += file_str
if filepath not in files:
files.append(filepath)
for p in job['pair'][1]['zR2']:
filepath = FileProcessor.parse_path_from_uri(p['location'])
file_str = "\t".join([normal_sample_name, filepath]) + "\n"
if file_str not in check_for_duplicates:
check_for_duplicates.append(file_str)
sample_mapping += file_str
if filepath not in files:
files.append(filepath)
for p in job['pair'][1]['bam']:
filepath = FileProcessor.parse_path_from_uri(p['location'])
file_str = "\t".join([normal_sample_name, filepath]) + "\n"
if file_str not in check_for_duplicates:
check_for_duplicates.append(file_str)
sample_mapping += file_str
if filepath not in files:
files.append(filepath)
name = "ARGOS %s, %i of %i" % (self.request_id, i + 1,
number_of_inputs)
assay = job['assay']
pi = job['pi']
pi_email = job['pi_email']
sample_pairing += "\t".join(
[normal_sample_name, tumor_sample_name]) + "\n"
argos_jobs.append((APIRunCreateSerializer(
data={
'app': pipeline,
'inputs': argos_inputs,
'name': name,
'tags': {
'requestId': self.request_id,
'sampleNameTumor': tumor_sample_name,
'sampleNameNormal': normal_sample_name,
'labHeadName': pi,
'labHeadEmail': pi_email
}
}), job))
operator_run_summary = UploadAttachmentEvent(
self.job_group_notifier_id, 'sample_pairing.txt',
sample_pairing).to_dict()
send_notification.delay(operator_run_summary)
mapping_file_event = UploadAttachmentEvent(self.job_group_notifier_id,
'sample_mapping.txt',
sample_mapping).to_dict()
send_notification.delay(mapping_file_event)
data_clinical = generate_sample_data_content(
files,
pipeline_name=pipeline_obj.name,
pipeline_github=pipeline_obj.github,
pipeline_version=pipeline_obj.version)
sample_data_clinical_event = UploadAttachmentEvent(
self.job_group_notifier_id, 'sample_data_clinical.txt',
data_clinical).to_dict()
send_notification.delay(sample_data_clinical_event)
self.evaluate_sample_errors(error_samples)
self.summarize_pairing_info(argos_inputs)
return argos_jobs
def fetch_samples(
request_id,
import_pooled_normals=True,
import_samples=True,
job_group=None,
job_group_notifier=None,
redelivery=False,
):
logger.info("Fetching sampleIds for requestId:%s" % request_id)
jg = None
jgn = None
try:
jg = JobGroup.objects.get(id=job_group)
logger.debug("JobGroup found")
except JobGroup.DoesNotExist:
logger.debug("No JobGroup Found")
try:
jgn = JobGroupNotifier.objects.get(id=job_group_notifier)
logger.debug("JobGroup found")
except JobGroupNotifier.DoesNotExist:
logger.debug("No JobGroup Found")
children = set()
sample_ids = LIMSClient.get_request_samples(request_id)
if sample_ids["requestId"] != request_id:
raise ErrorInconsistentDataException(
"LIMS returned wrong response for request %s. Got %s instead" %
(request_id, sample_ids["requestId"]))
request_metadata = {
"dataAnalystEmail": sample_ids["dataAnalystEmail"],
"dataAnalystName": sample_ids["dataAnalystName"],
"investigatorEmail": sample_ids["investigatorEmail"],
"investigatorName": sample_ids["investigatorName"],
"labHeadEmail": sample_ids["labHeadEmail"],
"labHeadName": sample_ids["labHeadName"],
"otherContactEmails": sample_ids["otherContactEmails"],
"dataAccessEmails": sample_ids["dataAccessEmails"],
"qcAccessEmails": sample_ids["qcAccessEmails"],
"projectManagerName": sample_ids["projectManagerName"],
"recipe": sample_ids["recipe"],
"piEmail": sample_ids["piEmail"],
}
set_recipe_event = ETLSetRecipeEvent(job_group_notifier,
request_metadata["recipe"]).to_dict()
send_notification.delay(set_recipe_event)
pooled_normals = sample_ids.get("pooledNormals", [])
if import_pooled_normals and pooled_normals:
for f in pooled_normals:
job = get_or_create_pooled_normal_job(f,
jg,
jgn,
redelivery=redelivery)
children.add(str(job.id))
if import_samples:
if not sample_ids.get("samples", False):
raise FailedToFetchSampleException(
"No samples reported for requestId: %s" % request_id)
for sample in sample_ids.get("samples", []):
sampleMetadata = LIMSClient.get_sample_manifest(
sample["igoSampleId"])
try:
data = sampleMetadata[0]
except Exception as e:
pass
patient_id = format_patient_id(data.get("cmoPatientId"))
if not Patient.objects.filter(patient_id=patient_id):
Patient.objects.create(patient_id=patient_id)
sample_name = data.get("cmoSampleName", None)
specimen_type = data.get("specimenType", None)
cmo_sample_name = format_sample_name(sample_name, specimen_type)
if not Sample.objects.filter(sample_id=sample["igoSampleId"],
sample_name=sample_name,
cmo_sample_name=cmo_sample_name):
Sample.objects.create(sample_id=sample["igoSampleId"],
sample_name=sample_name,
cmo_sample_name=cmo_sample_name)
job = create_sample_job(sample["igoSampleId"],
sample["igoComplete"], request_id,
request_metadata, redelivery, jg, jgn)
children.add(str(job.id))
return list(children)
def create_request_job(request_id, redelivery=False):
logger.info("Searching for job: %s for request_id: %s" %
(TYPES["REQUEST"], request_id))
count = Job.objects.filter(
run=TYPES["REQUEST"],
args__request_id=request_id,
status__in=[
JobStatus.CREATED, JobStatus.IN_PROGRESS,
JobStatus.WAITING_FOR_CHILDREN
],
).count()
request_redelivered = Job.objects.filter(
run=TYPES["REQUEST"], args__request_id=request_id).count() > 0
delivery_date = None
try:
request_from_lims = LIMSClient.get_request_samples(request_id)
delivery_date = datetime.fromtimestamp(
request_from_lims["deliveryDate"] / 1000)
except Exception:
logger.error("Failed to retrieve deliveryDate for request %s" %
request_id)
if not Request.objects.filter(request_id=request_id):
Request.objects.create(request_id=request_id,
delivery_date=delivery_date)
assays = ETLConfiguration.objects.first()
if request_redelivered and not (assays.redelivery and redelivery):
return None, "Request is redelivered, but redelivery deactivated"
if count == 0:
job_group = JobGroup()
job_group.save()
job_group_notifier_id = notifier_start(job_group, request_id)
job_group_notifier = JobGroupNotifier.objects.get(
id=job_group_notifier_id)
job = Job(
run=TYPES["REQUEST"],
args={
"request_id": request_id,
"job_group": str(job_group.id),
"job_group_notifier": job_group_notifier_id,
"redelivery": request_redelivered,
},
status=JobStatus.CREATED,
max_retry=1,
children=[],
callback=TYPES["REQUEST_CALLBACK"],
callback_args={
"request_id": request_id,
"job_group": str(job_group.id),
"job_group_notifier": job_group_notifier_id,
},
job_group=job_group,
job_group_notifier=job_group_notifier,
)
job.save()
if request_redelivered:
redelivery_event = RedeliveryEvent(job_group_notifier_id).to_dict()
send_notification.delay(redelivery_event)
request_obj = Request.objects.filter(request_id=request_id).first()
if request_obj:
delivery_date_event = SetDeliveryDateFieldEvent(
job_group_notifier_id,
str(request_obj.delivery_date)).to_dict()
send_notification.delay(delivery_date_event)
return job, "Job Created"
def create_operator_run_from_jobs(operator,
jobs,
job_group_id=None,
job_group_notifier_id=None):
jg = None
jgn = None
try:
jg = JobGroup.objects.get(id=job_group_id)
logger.info("JobGroup id: %s", job_group_id)
except JobGroup.DoesNotExist:
logger.info("JobGroup not set")
try:
jgn = JobGroupNotifier.objects.get(id=job_group_notifier_id)
except JobGroupNotifier.DoesNotExist:
logger.info("JobGroupNotifier not set")
valid_jobs, invalid_jobs = [], []
for job in jobs:
valid_jobs.append(job) if job[0].is_valid() else invalid_jobs.append(
job)
operator_run = OperatorRun.objects.create(operator=operator.model,
num_total_runs=len(valid_jobs),
job_group=jg,
job_group_notifier=jgn)
run_ids = []
pipeline_id = None
try:
pipeline_id = operator.get_pipeline_id()
p = Pipeline.objects.get(id=pipeline_id)
pipeline_name = p.name
pipeline_version = p.version
pipeline_link = p.pipeline_link
except Pipeline.DoesNotExist:
pipeline_name = ""
pipeline_link = ""
pipeline_version = ""
pipeline_description_event = AddPipelineToDescriptionEvent(
job_group_notifier_id, pipeline_name, pipeline_version,
pipeline_link).to_dict()
send_notification.delay(pipeline_description_event)
set_pipeline_field = SetPipelineFieldEvent(job_group_notifier_id,
pipeline_name).to_dict()
send_notification.delay(set_pipeline_field)
for job in valid_jobs:
logger.info("Creating Run object")
run = job[0].save(operator_run_id=operator_run.id,
job_group_id=job_group_id,
job_group_notifier_id=job_group_notifier_id)
logger.info("Run object created with id: %s" % str(run.id))
run_ids.append({
"run_id": str(run.id),
'tags': run.tags,
'output_directory': run.output_directory
})
output_directory = run.output_directory
if not pipeline_name and not pipeline_link:
logger.info(
"Run [ id: %s ] failed as the pipeline [ id: %s ] was not found",
run.id, pipeline_id)
error_message = dict(details="Pipeline [ id: %s ] was not found.".
format(pipeline_id))
fail_job(run.id, error_message)
else:
create_run_task.delay(str(run.id), job[1], output_directory)
if job_group_id:
event = OperatorRunEvent(job_group_notifier_id, operator.request_id,
pipeline_name, pipeline_link, run_ids,
str(operator_run.id)).to_dict()
send_notification.delay(event)
for job in invalid_jobs:
# TODO: Report this to JIRA ticket also
logger.error("Job invalid: %s" % str(job[0].errors))
operator_run.status = RunStatus.RUNNING
operator_run.save()
def process_triggers():
operator_runs = OperatorRun.objects.prefetch_related(
'runs', 'operator__from_triggers').exclude(
status__in=[RunStatus.COMPLETED, RunStatus.FAILED])
for operator_run in operator_runs:
created_chained_job = False
job_group = operator_run.job_group
job_group_id = str(job_group.id) if job_group else None
job_group_notifier = operator_run.job_group_notifier
job_group_notifier_id = str(
job_group_notifier.id) if job_group_notifier else None
try:
for trigger in operator_run.operator.from_triggers.all():
trigger_type = trigger.run_type
if trigger_type == TriggerRunType.AGGREGATE:
condition = trigger.aggregate_condition
if condition == TriggerAggregateConditionType.ALL_RUNS_SUCCEEDED:
if operator_run.percent_runs_succeeded == 100.0:
created_chained_job = True
create_jobs_from_chaining.delay(
trigger.to_operator_id,
trigger.from_operator_id,
list(
operator_run.runs.order_by(
'id').values_list('id', flat=True)),
job_group_id=job_group_id,
job_group_notifier_id=job_group_notifier_id)
continue
elif condition == TriggerAggregateConditionType.NINTY_PERCENT_SUCCEEDED:
if operator_run.percent_runs_succeeded >= 90.0:
created_chained_job = True
create_jobs_from_chaining.delay(
trigger.to_operator_id,
trigger.from_operator_id,
list(
operator_run.runs.order_by(
'id').values_list('id', flat=True)),
job_group_id=job_group_id,
job_group_notifier_id=job_group_notifier_id)
continue
if operator_run.percent_runs_finished == 100.0:
logger.info("Condition never met for operator run %s" %
operator_run.id)
elif trigger_type == TriggerRunType.INDIVIDUAL:
if operator_run.percent_runs_finished == 100.0:
operator_run.complete()
if operator_run.percent_runs_finished == 100.0:
if operator_run.percent_runs_succeeded == 100.0:
operator_run.complete()
if not created_chained_job and job_group_notifier_id:
completed_event = SetPipelineCompletedEvent(
job_group_notifier_id).to_dict()
send_notification.delay(completed_event)
else:
operator_run.fail()
if job_group_notifier_id:
e = OperatorRequestEvent(
job_group_notifier_id,
"[CIReviewEvent] Operator Run %s failed" %
str(operator_run.id)).to_dict()
send_notification.delay(e)
ci_review_event = SetCIReviewEvent(
job_group_notifier_id).to_dict()
send_notification.delay(ci_review_event)
except Exception as e:
logger.info("Trigger %s Fail. Error %s" %
(operator_run.id, str(e)))
operator_run.fail()
