def run(self):
url = self.url
dataset = self.dataset
reasons = []
statuses = [
'assignment-approved', 'assigned', 'acquired', 'running-open',
'running-closed', 'completed', 'force-complete', 'closed-out'
]
##print "reqmgr check on ",dataset
actors = getWorkflowByInput(url, dataset, details=True)
using_actors = [
actor for actor in actors if actor['RequestStatus'] in statuses
]
if len(using_actors):
reasons.append('input')
actors = getWorkflowByOutput(url, dataset, details=True)
using_actors = [
actor for actor in actors if actor['RequestStatus'] in statuses
]
if len(using_actors):
reasons.append('output')
actors = getWorkflowByMCPileup(url, dataset, details=True)
using_actors = [
actor for actor in actors if actor['RequestStatus'] in statuses
]
if len(using_actors):
reasons.append('pilup')
self.reasons = reasons
def run(self):
url = self.url
dataset= self.dataset
reasons = []
statuses = ['assignment-approved','assigned','acquired','running-open','running-closed','completed','force-complete','closed-out']
##print "reqmgr check on ",dataset
actors = getWorkflowByInput( url, dataset , details=True)
using_actors = [actor for actor in actors if actor['RequestStatus'] in statuses]
if len(using_actors):
reasons.append('input')
actors = getWorkflowByOutput( url, dataset , details=True)
using_actors = [actor for actor in actors if actor['RequestStatus'] in statuses]
if len(using_actors):
reasons.append('output')
actors = getWorkflowByMCPileup( url, dataset , details=True)
using_actors = [actor for actor in actors if actor['RequestStatus'] in statuses]
if len(using_actors):
reasons.append('pilup')
self.reasons = reasons
usors = getWorkflowByMCPileup(url, dataset, details=True)
## find the latest request date using that dataset in secondary
for usor in usors:
d =time.mktime(time.strptime("-".join(map(str,usor['RequestDate'])), "%Y-%m-%d-%H-%M-%S"))
secondary_timeout[dataset] = max(secondary_timeout[dataset],d)
if secondary_timeout[dataset]: ## different than zero
delay_days = 30
delay = delay_days*24*60*60 # 30 days
if (now-secondary_timeout[dataset])>delay:
print "unlocking secondary input after",delay_days,"days"
unlock = True
tier = dataset.split('/')[-1]
creators = getWorkflowByOutput( url, dataset , details=True)
if not creators and not tier == 'RAW':
ds_status = getDatasetStatus( dataset )
if not '-v0/' in dataset and ds_status!=None:
sendEmail('failing get by output','%s has not been produced by anything?'%dataset)
newly_locking.add(dataset)
continue
else:
# does not matter, cannot be an OK dataset
unlock = True
bad_ds = True
creators_status = [r['RequestStatus'] for r in creators]
print "Statuses of workflow that made the dataset",dataset,"are",creators_status
if all([status in ['failed','aborted','rejected','aborted-archived','rejected-archived'] for status in creators_status]):
## crap
print "\tunlocking",dataset,"for bad workflow statuses"
now - odb.date
) / 24 * 60 * 60, "[days] since announcement, limit is", delay_days, "[days]"
else:
unlock = False
print "re-locking", dataset, "because ", delay_days, "[days] expiration date is not passed, now:", now, "announced", odb.date, ":", (
now - odb.date) / 24 * 60 * 60, "[days]"
else:
print "re-Locking", dataset, "because of special tier needing double check"
unlock = False
time_point("Checked to keep on disk for 30 days", sub_lap=True)
if unlock:
print "\tunlocking", dataset
LI.release(dataset)
##would like to pass to *-unlock, or even destroy from local db
creators = getWorkflowByOutput(url, dataset, details=True)
for creator in creators:
for wfo in session.query(Workflow).filter(
Workflow.name == creator['RequestName']).all():
if not 'unlock' in wfo.status and any([
wfo.status.startswith(key)
for key in ['done', 'forget']
]):
wfo.status += '-unlock'
print "setting", wfo.name, "to", wfo.status
session.commit()
else:
print "\nrelocking", dataset
newly_locking.add(dataset)
time_point("Checked all")
input_json = defaultdict(list)
input_rl = []
output_json = {}
output_rl = defaultdict(list)
missing_rl = defaultdict(list)
errors_by_lb = defaultdict(lambda : defaultdict(set))
dbsapi = DbsApi(url='https://cmsweb.cern.ch/dbs/prod/global/DBSReader')
ecode_ban = []#99999,139,134,92]
## try to get more workflows by their outputs
outs=set()
for wf in wfs:
outs.update( wf['OutputDatasets'] )
for out in outs:
o_wfs = getWorkflowByOutput( url , out, details=True )
wfns = [wf['RequestName'] for wf in wfs]
for o_wf in o_wfs:
if not o_wf['RequestName'] in wfns:
print "got also",o_wf['RequestName']
wfs.append( o_wf )
for wf in wfs:
wfi = workflowInfo( url, wf['RequestName'], request=wf)
date = ''.join(map(lambda s : '%02d'%s ,wf['RequestDate']))
if wf['RequestStatus'] not in ['completed','announced','normal-archived']: continue
print "Checking on",wf['RequestName']
## create the input json
input_json = defaultdict(list)
input_rl = []
output_json = {}
output_rl = defaultdict(list)
missing_rl = defaultdict(list)
errors_by_lb = defaultdict(lambda: defaultdict(set))
dbsapi = DbsApi(url='https://cmsweb.cern.ch/dbs/prod/global/DBSReader')
ecode_ban = [] #99999,139,134,92]
## try to get more workflows by their outputs
outs = set()
for wf in wfs:
outs.update(wf['OutputDatasets'])
for out in outs:
o_wfs = getWorkflowByOutput(url, out, details=True)
wfns = [wf['RequestName'] for wf in wfs]
for o_wf in o_wfs:
if not o_wf['RequestName'] in wfns:
print "got also", o_wf['RequestName']
wfs.append(o_wf)
for wf in wfs:
wfi = workflowInfo(url, wf['RequestName'], request=wf)
date = ''.join(map(lambda s: '%02d' % s, wf['RequestDate']))
if wf['RequestStatus'] not in [
'completed', 'announced', 'normal-archived'
]:
continue
