def classifyRequests(url, requests, historic, noNameSites, requestType):
for request in requests:
name = request['request_name']
status = 'NoStatus'
reqType = 'NoType'
if 'type' in request.keys():
reqType = request['type']
if reqType != requestType:
continue
if 'status' in request.keys():
status = request['status']
if status in noNameSites.keys():
namefound = 0
for Site in historic.keys():
if name.find(Site) >= 0:
namefound = 1
for stat in historic[Site].keys(
): #stat is the status of the request in the list of requests
if status == stat:
EffectiveLumi = getEffectiveLumiSections(
url, name, requestType)
if EffectiveLumi <= 0:
EffectiveLumi = 0.0000001
TimeEvent = getTimeEventRequest(url, name)
priority = getPriorityWorkflow(url, name)
numevents = dbsTest.getInputEvents(url, name)
checkLumi = False
if float(numevents / EffectiveLumi) > 400:
checkLumi = True
else:
checkLumi = False
maxEvents = maxEventsFileDataset(url, name)
historic[Site][stat].append(
(name, priority, numevents, TimeEvent,
EffectiveLumi, checkLumi, maxEvents))
if namefound == 0:
for stat in noNameSites.keys():
if status == stat:
EffectiveLumi = getEffectiveLumiSections(
url, name, requestType)
if EffectiveLumi <= 0:
EffectiveLumi = 0.0000001
TimeEvent = getTimeEventRequest(url, name)
priority = getPriorityWorkflow(url, name)
numevents = dbsTest.getInputEvents(url, name)
checkLumi = False
if float(numevents / EffectiveLumi) > 400:
checkLumi = True
else:
checkLumi = False
maxEvents = maxEventsFileDataset(url, name)
noNameSites[stat].append(
(name, priority, numevents, TimeEvent,
EffectiveLumi, checkLumi, maxEvents))
def classifyRequests(url, requests):
print '-----------------------------------------------------------------------------------------------------------------------------------------------------------'
print '| Request |req Type |Status Req | Dataset |Status Dataset | Percentage|FilterEfficiency| '
print '-----------------------------------------------------------------------------------------------------------------------------------------------------------'
classifiedRequests = {}
for request in requests:
if 'type' in request:
name = request['request_name']
if request['type'] == 'MonteCarloFromGEN' or request[
'type'] == 'MonteCarlo':
datasetWorkflow = phedexSubscription.outputdatasetsWorkflow(
url, name)
problem = False
percentage = 0
if len(datasetWorkflow) < 1:
continue
dataset = datasetWorkflow[0]
inputEvents = 0.0001
inputEvents = inputEvents + int(
dbsTest.getInputEvents(url, name))
outputEvents = dbsTest.getEventCountDataSet(dataset)
percentage = outputEvents / float(inputEvents)
duplicate = duplicateLumi(dataset)
problem = False
if duplicate:
problem = True
if problem:
FilterEfficiency = getFilterEfficiency(url, name)
datasetStatus = getDatasetStatus(dataset)
print '| %20s | %8s| %8s | %20s | %10s| %10s | %10s| ' % (
name, request['type'], request['status'], dataset,
datasetStatus, str(percentage * 100), FilterEfficiency)
print '---------------------------------------------------------------------------------------------------------------------------'
def getDbsProgress(options,wfs_dict,wfs_dict_skipped):
print cya+"Getting progress from dbs..."+dfa
url = "cmsweb.cern.ch"
for wf in wfs_dict.keys():
wfs_dict[wf]['dbsProgress'] = []
try:
outputDataSets = dbsTest.phedexSubscription.outputdatasetsWorkflow(url, wf)
except:
print "\t%s'%s' cannot be looked up in dbs, skipping.%s" % (red,wf,dfa)
wfs_dict_skipped[wf] = wfs_dict[wf]
del wfs_dict[wf]
continue
try:
inputEvents = dbsTest.getInputEvents(url, wf)
except:
print "\t%s'%s' cannot be looked up in dbs, skipping.%s" % (red,wf,dfa)
wfs_dict_skipped[wf] = wfs_dict[wf]
del wfs_dict[wf]
continue
for dataset in outputDataSets:
outputEvents = dbsTest.getEventCountDataSet(dataset)
wfs_dict[wf]['dbsProgress'].append({"dataset":dataset,"progress":str(outputEvents/float(inputEvents)*100)})
if options.verbose:
print cya+"Added dbs progress info to workflow dictionary."+dfa
appendFile(log_cmst1,"== "+str(datetime.datetime.now())+" == progress queried from dbs ==\n")
return wfs_dict,wfs_dict_skipped
def retrieveSchema(url, workflowName, user, group ):
specURL = os.path.join(reqmgrCouchURL, workflowName, "spec")
helper = WMWorkloadHelper()
#print " retrieving original workflow...",
helper.load(specURL)
#print "done."
schema = {}
#for (key, value) in helper.data.request.schema.dictionary_whole_tree_().iteritems():
for (key, value) in helper.data.request.schema.dictionary_().iteritems():
#print key, value
if key == 'ProcConfigCacheID':
schema['ConfigCacheID'] = value
elif key=='RequestSizeEvents':
schema['RequestSizeEvents'] = value
#schema['RequestNumEvents'] = int(value)
elif key=='Requestor':
schema['Requestor']=user
elif key=='Group':
schema['Group']=group
elif key=='RequestNumEvents':
schema['RequestNumEvents']=getFinalRequestedNumEvents(url, workflowName)
elif key=='FirstLumi':
schema['FirstLumi']=getMaxLumi(url, workflowName)*2
elif key=='FirstEvent':
schema['FirstEvent']=dbsTest.getInputEvents(url, workflowName)*2
elif key=='RequestString':
schema['RequestString']='ACDC_'+value
elif value != None:
schema[key] = value
return schema
def classifyRequests(url, requests):
print '-----------------------------------------------------------------------------------------------------------------------------------------------------------'
print '| Request |req Type |Status Req | Dataset |Status Dataset | Percentage|FilterEfficiency| '
print '-----------------------------------------------------------------------------------------------------------------------------------------------------------'
classifiedRequests={}
for request in requests:
if 'type' in request:
name=request['request_name']
if request['type']=='MonteCarloFromGEN' or request['type']=='MonteCarlo':
datasetWorkflow=phedexSubscription.outputdatasetsWorkflow(url, name)
problem=False
percentage=0
if len(datasetWorkflow)<1:
continue
dataset=datasetWorkflow[0]
inputEvents=0.0001
inputEvents=inputEvents+int(dbsTest.getInputEvents(url, name))
outputEvents=dbsTest.getEventCountDataSet(dataset)
percentage=outputEvents/float(inputEvents)
duplicate=duplicateLumi(dataset)
problem=False
if duplicate:
problem=True
if problem:
FilterEfficiency=getFilterEfficiency(url, name)
datasetStatus=getDatasetStatus(dataset)
print '| %20s | %8s| %8s | %20s | %10s| %10s | %10s| ' % (name, request['type'], request['status'], dataset,datasetStatus, str(percentage*100), FilterEfficiency)
print '---------------------------------------------------------------------------------------------------------------------------'
def main():
url = 'cmsweb.cern.ch'
args = sys.argv[1:]
if not len(args) == 1:
print "usage: python closeOutTaskChainWorkflows.py <inputFile_containing_a_list_of_workflows>"
sys.exit(0)
inputFile = args[0]
f = open(inputFile, 'r')
print '-------------------------------------------------------------------------------------------------------------------------------------'
print '| Request | Closed-out? | Current status |'
print '-------------------------------------------------------------------------------------------------------------------------------------'
closedOut = []
nonClosedOut = []
running = []
for line in f:
workflow = line.rstrip('\n')
outputDatasets = phedexSubscription.outputdatasetsWorkflow(
url, workflow)
inputEvents = dbsTest.getInputEvents(url, workflow)
closeOut = True
for dataset in outputDatasets:
# we cannot calculate completion of ALCARECO samples
if 'ALCARECO' in dataset:
continue
outputEvents = dbsTest.getOutputEvents(url, workflow, dataset)
print "dataset = " + dataset
print "inputEvents = " + str(inputEvents)
print "outputEvents = " + str(outputEvents)
if inputEvents != 0:
if (outputEvents / float(inputEvents) * 100) >= 100.0:
pass
#print dataset+" match: "+str(outputEvents/float(inputEvents)*100) +"%"
else:
#print dataset + " it is less than 99.9999% completed, keeping it in the current status"
closeOut = False
break
else:
print "Input Events 0"
if closeOut:
closedOut.append(workflow)
else:
nonClosedOut.append(workflow)
for workflow in closedOut:
status = getStatus(url, workflow)
#if status == 'completed':
# closeOutTaskChainWorkflows(url, workflow)
#else:
# pass
print "%90s\tYES\t\t%s" % (workflow, status)
for workflow in nonClosedOut:
status = getStatus(url, workflow)
print "%90s\tNO\t\t%s" % (workflow, status)
f.close
print '-------------------------------------------------------------------------------------------------------------------------------------'
sys.exit(0)
def classifyRequests(url, requests, historic, noNameSites, requestType):
for request in requests:
name=request['request_name']
status='NoStatus'
reqType='NoType'
if 'type' in request.keys():
reqType=request['type']
if reqType!=requestType:
continue
if 'status' in request.keys():
status=request['status']
if status in noNameSites.keys():
namefound=0
for Site in historic.keys():
if name.find(Site)>=0:
namefound=1
for stat in historic[Site].keys():#stat is the status of the request in the list of requests
if status==stat:
EffectiveLumi=getEffectiveLumiSections(url, name, requestType)
if EffectiveLumi <=0:
EffectiveLumi=0.0000001
TimeEvent=getTimeEventRequest(url, name)
priority=getPriorityWorkflow(url, name)
numevents=dbsTest.getInputEvents(url, name)
checkLumi=False
if float(numevents/EffectiveLumi)>400:
checkLumi=True
else:
checkLumi=False
maxEvents=maxEventsFileDataset(url, name)
historic[Site][stat].append((name,priority,numevents, TimeEvent,EffectiveLumi, checkLumi, maxEvents))
if namefound==0:
for stat in noNameSites.keys():
if status==stat:
EffectiveLumi=getEffectiveLumiSections(url, name, requestType)
if EffectiveLumi <=0:
EffectiveLumi=0.0000001
TimeEvent=getTimeEventRequest(url, name)
priority=getPriorityWorkflow(url, name)
numevents=dbsTest.getInputEvents(url, name)
checkLumi=False
if float(numevents/EffectiveLumi)>400:
checkLumi=True
else:
checkLumi=False
maxEvents=maxEventsFileDataset(url, name)
noNameSites[stat].append((name,priority,numevents, TimeEvent,EffectiveLumi, checkLumi, maxEvents))
def main():
url='cmsweb.cern.ch'
args=sys.argv[1:]
if not len(args)==1:
print "usage: python closeOutTaskChainWorkflows.py <inputFile_containing_a_list_of_workflows>"
sys.exit(0)
inputFile=args[0]
f = open(inputFile, 'r')
print '-------------------------------------------------------------------------------------------------------------------------------------'
print '| Request | Closed-out? | Current status |'
print '-------------------------------------------------------------------------------------------------------------------------------------'
closedOut = []
nonClosedOut = []
running = []
for line in f:
workflow = line.rstrip('\n')
outputDatasets=phedexSubscription.outputdatasetsWorkflow(url, workflow)
inputEvents=dbsTest.getInputEvents(url, workflow)
closeOut = True
for dataset in outputDatasets:
# we cannot calculate completion of ALCARECO samples
if 'ALCARECO' in dataset:
continue
outputEvents=dbsTest.getOutputEvents(url, workflow, dataset)
print "dataset = " + dataset
print "inputEvents = " + str(inputEvents)
print "outputEvents = " + str(outputEvents)
if inputEvents!=0:
if (outputEvents/float(inputEvents)*100) >= 100.0:
pass
#print dataset+" match: "+str(outputEvents/float(inputEvents)*100) +"%"
else:
#print dataset + " it is less than 99.9999% completed, keeping it in the current status"
closeOut = False
break
else:
print "Input Events 0"
if closeOut:
closedOut.append(workflow)
else:
nonClosedOut.append(workflow)
for workflow in closedOut:
status = getStatus(url, workflow)
#if status == 'completed':
# closeOutTaskChainWorkflows(url, workflow)
#else:
# pass
print "%90s\tYES\t\t%s" % (workflow, status)
for workflow in nonClosedOut:
status = getStatus(url, workflow)
print "%90s\tNO\t\t%s" % (workflow, status)
f.close
print '-------------------------------------------------------------------------------------------------------------------------------------'
sys.exit(0)
def PercentageCompletion(url, workflow, dataset):
inputEvents = 0
inputEvents = inputEvents + int(dbsTest.getInputEvents(url, workflow))
outputEvents = dbsTest.getOutputEvents(url, workflow, dataset)
if inputEvents == 0:
return 0
percentage = outputEvents / float(inputEvents)
return percentage
def PercentageCompletion(url, workflow, dataset): inputEvents=0 inputEvents=inputEvents+int(dbsTest.getInputEvents(url, workflow)) outputEvents=dbsTest.getOutputEvents(url, workflow, dataset) if inputEvents==0: return 0 percentage=outputEvents/float(inputEvents) return percentage
def testEventCountWorkflow(url, workflow): inputEvents=0 inputEvents=inputEvents+dbsTest.getInputEvents(url, workflow) datasets=phedexSubscription.outputdatasetsWorkflow(url, workflow) for dataset in datasets: outputEvents=dbsTest.getEventCountDataSet(dataset) percentage=outputEvents/float(inputEvents) if float(percentage)>float(1): print "Workflow: " + workflow+" duplicate events in outputdataset: "+dataset +" percentage: "+str(outputEvents/float(inputEvents)*100) +"%" return 1
def testEventCountWorkflow(url, workflow):
inputEvents = 0
inputEvents = inputEvents + dbsTest.getInputEvents(url, workflow)
datasets = phedexSubscription.outputdatasetsWorkflow(url, workflow)
for dataset in datasets:
outputEvents = dbsTest.getEventCountDataSet(dataset)
percentage = outputEvents / float(inputEvents)
if float(percentage) > float(1):
print "Workflow: " + workflow + " duplicate events in outputdataset: " + dataset + " percentage: " + str(
outputEvents / float(inputEvents) * 100) + "%"
return 1
def getDbsProgress(options, wfs_dict, wfs_dict_skipped):
print cya + "Getting progress from dbs..." + dfa
url = "cmsweb.cern.ch"
for wf in wfs_dict.keys():
wfs_dict[wf]['dbsProgress'] = []
try:
outputDataSets = dbsTest.phedexSubscription.outputdatasetsWorkflow(
url, wf)
except:
print "\t%s'%s' cannot be looked up in dbs, skipping.%s" % (
red, wf, dfa)
wfs_dict_skipped[wf] = wfs_dict[wf]
del wfs_dict[wf]
continue
try:
inputEvents = dbsTest.getInputEvents(url, wf)
except:
print "\t%s'%s' cannot be looked up in dbs, skipping.%s" % (
red, wf, dfa)
wfs_dict_skipped[wf] = wfs_dict[wf]
del wfs_dict[wf]
continue
for dataset in outputDataSets:
outputEvents = dbsTest.getEventCountDataSet(dataset)
wfs_dict[wf]['dbsProgress'].append({
"dataset":
dataset,
"progress":
str(outputEvents / float(inputEvents) * 100)
})
if options.verbose:
print cya + "Added dbs progress info to workflow dictionary." + dfa
appendFile(
log_cmst1, "== " + str(datetime.datetime.now()) +
" == progress queried from dbs ==\n")
return wfs_dict, wfs_dict_skipped
def main():
url='cmsweb.cern.ch'
args=sys.argv[1:]
if not len(args)==1:
print "usage: python closeOutTaskChainWorkflows.py <inputFile_containing_a_list_of_workflows>"
sys.exit(0)
inputFile=args[0]
f = open(inputFile, 'r')
print '-------------------------------------------------------------------------------------------------------------------------------------'
print '| Request | Closed-out? | Current status |'
print '-------------------------------------------------------------------------------------------------------------------------------------'
closedOut = []
nonClosedOut = []
tooManyEvents = []
running = []
for line in f:
workflow = line.rstrip('\n')
outputDatasets=phedexSubscription.outputdatasetsWorkflow(url, workflow)
inputEvents=dbsTest.getInputEvents(url, workflow)
closeOut = True
tooMany = False
for dataset in outputDatasets:
# we cannot calculate completion of ALCARECO samples
if 'ALCARECO' in dataset:
continue
outputEvents=dbsTest.getOutputEvents(url, workflow, dataset)
if inputEvents!=0:
if outputEvents == inputEvents:
pass
#print dataset+" match: "+str(outputEvents/float(inputEvents)*100) +"%"
elif outputEvents < inputEvents :
#print dataset + " it is less than 99.9999% completed, keeping it in the current status"
closeOut = False
elif outputEvents > inputEvents :
closeOut = False
tooMany = True
break
else:
print "Input Events 0"
if closeOut:
closedOut.append(workflow)
else:
nonClosedOut.append(workflow)
if tooMany:
tooManyEvents.append(workflow)
for workflow in closedOut:
status = getStatus(url, workflow)
if status == 'completed':
closeOutTaskChainWorkflows(url, workflow)
else:
pass
print "%90s\tYES\t\t%s" % (workflow, status)
for workflow in nonClosedOut:
status = getStatus(url, workflow)
print "%90s\tNO\t\t%s" % (workflow, status)
print '-------------------------------------------------------------------------------------------------------------------------------------'
for workflow in tooManyEvents:
outputDatasets=phedexSubscription.outputdatasetsWorkflow(url, workflow)
inputEvents=dbsTest.getInputEvents(url, workflow)
for dataset in outputDatasets:
# we cannot calculate completion of ALCARECO samples
if 'ALCARECO' in dataset:
continue
outputEvents=dbsTest.getOutputEvents(url, workflow, dataset)
if inputEvents!=0:
if outputEvents > inputEvents:
print "WARNING about workflow " + workflow + ": The dataset " + dataset + " contains MORE events than expected. " + str(inputEvents) + " events were expected and " + str(outputEvents) + " were found."
f.close
sys.exit(0)
def getFinalRequestedNumEvents(url, workflow): outputDataSets=phedexSubscription.outputdatasetsWorkflow(url, workflow) obtainedEvents=dbsTest.getOutputEvents(url, workflow, outputDataSets[0]) requestedEvents=dbsTest.getInputEvents(url, workflow) return (requestedEvents-obtainedEvents)
def main():
url = 'cmsweb.cern.ch'
args = sys.argv[1:]
if not len(args) == 1:
print "usage: python closeOutTaskChainWorkflows.py <inputFile_containing_a_list_of_workflows>"
sys.exit(0)
inputFile = args[0]
f = open(inputFile, 'r')
print '-------------------------------------------------------------------------------------------------------------------------------------'
print '| Request | Closed-out? | Current status |'
print '-------------------------------------------------------------------------------------------------------------------------------------'
closedOut = []
nonClosedOut = []
tooManyEvents = []
running = []
for line in f:
workflow = line.rstrip('\n')
outputDatasets = phedexSubscription.outputdatasetsWorkflow(
url, workflow)
inputEvents = dbsTest.getInputEvents(url, workflow)
closeOut = True
tooMany = False
for dataset in outputDatasets:
# we cannot calculate completion of ALCARECO samples
if 'ALCARECO' in dataset:
continue
outputEvents = dbsTest.getOutputEvents(url, workflow, dataset)
if inputEvents != 0:
if outputEvents == inputEvents:
pass
#print dataset+" match: "+str(outputEvents/float(inputEvents)*100) +"%"
elif outputEvents < inputEvents:
#print dataset + " it is less than 99.9999% completed, keeping it in the current status"
closeOut = False
elif outputEvents > inputEvents:
closeOut = False
tooMany = True
break
else:
print "Input Events 0"
if closeOut:
closedOut.append(workflow)
else:
nonClosedOut.append(workflow)
if tooMany:
tooManyEvents.append(workflow)
for workflow in closedOut:
status = getStatus(url, workflow)
if status == 'completed':
closeOutTaskChainWorkflows(url, workflow)
else:
pass
print "%90s\tYES\t\t%s" % (workflow, status)
for workflow in nonClosedOut:
status = getStatus(url, workflow)
print "%90s\tNO\t\t%s" % (workflow, status)
print '-------------------------------------------------------------------------------------------------------------------------------------'
for workflow in tooManyEvents:
outputDatasets = phedexSubscription.outputdatasetsWorkflow(
url, workflow)
inputEvents = dbsTest.getInputEvents(url, workflow)
for dataset in outputDatasets:
# we cannot calculate completion of ALCARECO samples
if 'ALCARECO' in dataset:
continue
outputEvents = dbsTest.getOutputEvents(url, workflow, dataset)
if inputEvents != 0:
if outputEvents > inputEvents:
print "WARNING about workflow " + workflow + ": The dataset " + dataset + " contains MORE events than expected. " + str(
inputEvents) + " events were expected and " + str(
outputEvents) + " were found."
f.close
sys.exit(0)
