ViewVC Help

View File | Revision Log | Show Annotations | Root Listing

root/cvsroot/COMP/CRAB/python/cms_cmssw.py

Comparing COMP/CRAB/python/cms_cmssw.py (file contents):
Revision 1.169 by spiga, Tue Apr 1 14:53:36 2008 UTC vs.
Revision 1.247 by fanzago, Mon Sep 29 17:19:16 2008 UTC

#	Line 2 | Line 2 | from JobType import JobType
2		from crab_logger import Logger
3		from crab_exceptions import *
4		from crab_util import *
5	<	from BlackWhiteListParser import BlackWhiteListParser
5	>	from BlackWhiteListParser import SEBlackWhiteListParser
6		import common
7		import Scram
8
9		import os, string, glob
10
11		class Cmssw(JobType):
12	<	def __init__(self, cfg_params, ncjobs):
12	>	def __init__(self, cfg_params, ncjobs,skip_blocks, isNew):
13		JobType.__init__(self, 'CMSSW')
14		common.logger.debug(3,'CMSSW::__init__')
15	+	self.skip_blocks = skip_blocks
16
17		self.argsList = []
18
19		self._params = {}
20		self.cfg_params = cfg_params
21		# init BlackWhiteListParser
22	<	self.blackWhiteListParser = BlackWhiteListParser(cfg_params)
22	>	self.blackWhiteListParser = SEBlackWhiteListParser(cfg_params)
23
24	<	self.MaxTarBallSize = float(self.cfg_params.get('EDG.maxtarballsize',9.5))
24	>	### Temporary patch to automatically skip the ISB size check:
25	>	server=self.cfg_params.get('CRAB.server_name',None)
26	>	size = 9.5
27	>	if server: size = 99999
28	>	### D.S.
29	>	self.MaxTarBallSize = float(self.cfg_params.get('EDG.maxtarballsize',size))
30
31		# number of jobs requested to be created, limit obj splitting
32		self.ncjobs = ncjobs
#	Line 33 | Line 39 | class Cmssw(JobType):
39		self.executable = ''
40		self.executable_arch = self.scram.getArch()
41		self.tgz_name = 'default.tgz'
36	–	self.additional_tgz_name = 'additional.tgz'
42		self.scriptName = 'CMSSW.sh'
43	<	self.pset = ''      #scrip use case Da
44	<	self.datasetPath = '' #scrip use case Da
43	>	self.pset = ''
44	>	self.datasetPath = ''
45
46		# set FJR file name
47		self.fjrFileName = 'crab_fjr.xml'
48
49		self.version = self.scram.getSWVersion()
50	<
51	<	#
52	<	# Try to block creation in case of arch/version mismatch
53	<	#
54	<
55	<	#        a = string.split(self.version, "_")
56	<	#
57	<	#        if int(a[1]) == 1 and (int(a[2]) < 5 and self.executable_arch.find('slc4') == 0):
58	<	#            msg = "Warning: You are using %s version of CMSSW  with %s architecture. \n--> Did you compile your libraries with SLC3? Otherwise you can find some problems running on SLC4 Grid nodes.\n"%(self.version, self.executable_arch)
59	<	#            common.logger.message(msg)
60	<	#        if int(a[1]) == 1 and (int(a[2]) >= 5 and self.executable_arch.find('slc3') == 0):
56	<	#            msg = "Error: CMS does not support %s with %s architecture"%(self.version, self.executable_arch)
57	<	#            raise CrabException(msg)
58	<	#
59	<
60	<	self.setParam_('application', self.version)
50	>	version_array = self.version.split('_')
51	>	self.CMSSW_major = 0
52	>	self.CMSSW_minor = 0
53	>	self.CMSSW_patch = 0
54	>	try:
55	>	self.CMSSW_major = int(version_array[1])
56	>	self.CMSSW_minor = int(version_array[2])
57	>	self.CMSSW_patch = int(version_array[3])
58	>	except:
59	>	msg = "Cannot parse CMSSW version string: " + self.version + " for major and minor release number!"
60	>	raise CrabException(msg)
61
62		### collect Data cards
63
64		if not cfg_params.has_key('CMSSW.datasetpath'):
65		msg = "Error: datasetpath not defined "
66		raise CrabException(msg)
67	+
68	+	### Temporary: added to remove input file control in the case of PU
69	+	self.dataset_pu = cfg_params.get('CMSSW.dataset_pu', None)
70	+
71		tmp =  cfg_params['CMSSW.datasetpath']
72		log.debug(6, "CMSSW::CMSSW(): datasetPath = "+tmp)
73	<	if string.lower(tmp)=='none':
73	>
74	>	if tmp =='':
75	>	msg = "Error: datasetpath not defined "
76	>	raise CrabException(msg)
77	>	elif string.lower(tmp)=='none':
78		self.datasetPath = None
79		self.selectNoInput = 1
80		else:
81		self.datasetPath = tmp
82		self.selectNoInput = 0
83
76	–	# ML monitoring
77	–	# split dataset path style: /PreProdR3Minbias/SIM/GEN-SIM
78	–	if not self.datasetPath:
79	–	self.setParam_('dataset', 'None')
80	–	self.setParam_('owner', 'None')
81	–	else:
82	–	## SL what is supposed to fail here?
83	–	try:
84	–	datasetpath_split = self.datasetPath.split("/")
85	–	# standard style
86	–	self.setParam_('datasetFull', self.datasetPath)
87	–	self.setParam_('dataset', datasetpath_split[1])
88	–	self.setParam_('owner', datasetpath_split[2])
89	–	except:
90	–	self.setParam_('dataset', self.datasetPath)
91	–	self.setParam_('owner', self.datasetPath)
92	–
93	–	self.setParam_('taskId', common._db.queryTask('name')) ## new BL--DS
94	–
84		self.dataTiers = []
85	<
85	>	self.debugWrap = ''
86	>	self.debug_wrapper = cfg_params.get('USER.debug_wrapper',False)
87	>	if self.debug_wrapper: self.debugWrap='--debug'
88		## now the application
89		self.executable = cfg_params.get('CMSSW.executable','cmsRun')
99	–	self.setParam_('exe', self.executable)
90		log.debug(6, "CMSSW::CMSSW(): executable = "+self.executable)
91
92		if not cfg_params.has_key('CMSSW.pset'):
#	Line 117 | Line 107 | class Cmssw(JobType):
107		self.output_file_sandbox.append(self.fjrFileName)
108
109		# other output files to be returned via sandbox or copied to SE
110	+	outfileflag = False
111		self.output_file = []
112		tmp = cfg_params.get('CMSSW.output_file',None)
113		if tmp :
114	<	tmpOutFiles = string.split(tmp,',')
115	<	log.debug(7, 'cmssw::cmssw(): output files '+str(tmpOutFiles))
116	<	for tmp in tmpOutFiles:
117	<	tmp=string.strip(tmp)
127	<	self.output_file.append(tmp)
128	<	pass
129	<	else:
130	<	log.message("No output file defined: only stdout/err and the CRAB Framework Job Report will be available\n")
131	<	pass
114	>	self.output_file = [x.strip() for x in tmp.split(',')]
115	>	outfileflag = True #output found
116	>	#else:
117	>	#    log.message("No output file defined: only stdout/err and the CRAB Framework Job Report will be available\n")
118
119		# script_exe file as additional file in inputSandbox
120		self.scriptExe = cfg_params.get('USER.script_exe',None)
121		if self.scriptExe :
122	<	if not os.path.isfile(self.scriptExe):
123	<	msg ="ERROR. file "+self.scriptExe+" not found"
124	<	raise CrabException(msg)
125	<	self.additional_inbox_files.append(string.strip(self.scriptExe))
122	>	if not os.path.isfile(self.scriptExe):
123	>	msg ="ERROR. file "+self.scriptExe+" not found"
124	>	raise CrabException(msg)
125	>	self.additional_inbox_files.append(string.strip(self.scriptExe))
126
141	–	#CarlosDaniele
127		if self.datasetPath == None and self.pset == None and self.scriptExe == '' :
128	<	msg ="Error. script_exe  not defined"
129	<	raise CrabException(msg)
128	>	msg ="Error. script_exe  not defined"
129	>	raise CrabException(msg)
130	>
131	>	# use parent files...
132	>	self.useParent = self.cfg_params.get('CMSSW.use_parent',False)
133
134		## additional input files
135		if cfg_params.has_key('USER.additional_input_files'):
#	Line 161 | Line 149 | class Cmssw(JobType):
149		if not os.path.exists(file):
150		raise CrabException("Additional input file not found: "+file)
151		pass
164	–	# fname = string.split(file, '/')[-1]
165	–	# storedFile = common.work_space.pathForTgz()+'share/'+fname
166	–	# shutil.copyfile(file, storedFile)
152		self.additional_inbox_files.append(string.strip(file))
153		pass
154		pass
#	Line 189 | Line 174 | class Cmssw(JobType):
174		if cfg_params.has_key('CMSSW.total_number_of_events'):
175		self.total_number_of_events = int(cfg_params['CMSSW.total_number_of_events'])
176		self.selectTotalNumberEvents = 1
177	+	if self.selectNumberOfJobs  == 1:
178	+	if (self.total_number_of_events != -1) and int(self.total_number_of_events) < int(self.theNumberOfJobs):
179	+	msg = 'Must specify at least one event per job. total_number_of_events > number_of_jobs '
180	+	raise CrabException(msg)
181		else:
182		self.total_number_of_events = 0
183		self.selectTotalNumberEvents = 0
184
185	<	if self.pset != None: #CarlosDaniele
185	>	if self.pset != None:
186		if ( (self.selectTotalNumberEvents + self.selectEventsPerJob + self.selectNumberOfJobs) != 2 ):
187		msg = 'Must define exactly two of total_number_of_events, events_per_job, or number_of_jobs.'
188		raise CrabException(msg)
#	Line 216 | Line 205 | class Cmssw(JobType):
205		tmp.strip()
206		self.incrementSeeds.append(tmp)
207
208	<	## Old method of dealing with seeds
209	<	## FUTURE: This is for old CMSSW and old CRAB. Can throw exceptions after a couple of CRAB releases and then
221	<	## remove
222	<	self.sourceSeed = cfg_params.get('CMSSW.pythia_seed',None)
223	<	if self.sourceSeed:
224	<	print "pythia_seed is a deprecated parameter. Use preserve_seeds or increment_seeds in the future.\n","Added to increment_seeds."
225	<	self.incrementSeeds.append('sourceSeed')
226	<
208	>	## FUTURE: Can remove in CRAB 2.4.0
209	>	self.sourceSeed    = cfg_params.get('CMSSW.pythia_seed',None)
210		self.sourceSeedVtx = cfg_params.get('CMSSW.vtx_seed',None)
211	<	if self.sourceSeedVtx:
229	<	print "vtx_seed is a deprecated parameter. Use preserve_seeds or increment_seeds in the future.\n","Added to increment_seeds."
230	<	self.incrementSeeds.append('VtxSmeared')
231	<
232	<	self.sourceSeedG4 = cfg_params.get('CMSSW.g4_seed',None)
233	<	if self.sourceSeedG4:
234	<	print "g4_seed is a deprecated parameter. Use preserve_seeds or increment_seeds in the future.\n","Added to increment_seeds."
235	<	self.incrementSeeds.append('g4SimHits')
236	<
211	>	self.sourceSeedG4  = cfg_params.get('CMSSW.g4_seed',None)
212		self.sourceSeedMix = cfg_params.get('CMSSW.mix_seed',None)
213	<	if self.sourceSeedMix:
214	<	print "mix_seed is a deprecated parameter. Use preserve_seeds or increment_seeds in the future.\n","Added to increment_seeds."
215	<	self.incrementSeeds.append('mix')
213	>	if self.sourceSeed or self.sourceSeedVtx or self.sourceSeedG4 or self.sourceSeedMix:
214	>	msg = 'pythia_seed, vtx_seed, g4_seed, and mix_seed are no longer valid settings. You must use increment_seeds or preserve_seeds'
215	>	raise CrabException(msg)
216
217		self.firstRun = cfg_params.get('CMSSW.first_run',None)
218
244	–	if self.pset != None: #CarlosDaniele
245	–	import PsetManipulator as pp
246	–	PsetEdit = pp.PsetManipulator(self.pset) #Daniele Pset
247	–
219		# Copy/return
249	–
220		self.copy_data = int(cfg_params.get('USER.copy_data',0))
221		self.return_data = int(cfg_params.get('USER.return_data',0))
222
#	Line 262 | Line 232 | class Cmssw(JobType):
232		blockSites = self.DataDiscoveryAndLocation(cfg_params)
233		#DBSDLS-end
234
265	–	self.tgzNameWithPath = self.getTarBall(self.executable)
266	–
235		## Select Splitting
236		if self.selectNoInput:
237	<	if self.pset == None: #CarlosDaniele
237	>	if self.pset == None:
238		self.jobSplittingForScript()
239		else:
240		self.jobSplittingNoInput()
241	+	elif (cfg_params.get('CMSSW.noblockboundary',0)):
242	+	self.jobSplittingNoBlockBoundary(blockSites)
243		else:
244		self.jobSplittingByBlocks(blockSites)
245
246	<	# modify Pset
247	<	if self.pset != None: #CarlosDaniele
248	<	try:
249	<	# Add FrameworkJobReport to parameter-set, set max events.
250	<	# Reset later for data jobs by writeCFG which does all modifications
251	<	PsetEdit.addCrabFJR(self.fjrFileName)
252	<	PsetEdit.maxEvent(self.eventsPerJob)
253	<	PsetEdit.psetWriter(self.configFilename())
254	<	except:
255	<	msg='Error while manipuliating ParameterSet: exiting...'
256	<	raise CrabException(msg)
246	>	# modify Pset only the first time
247	>	if isNew:
248	>	if self.pset != None:
249	>	import PsetManipulator as pp
250	>	PsetEdit = pp.PsetManipulator(self.pset)
251	>	try:
252	>	# Add FrameworkJobReport to parameter-set, set max events.
253	>	# Reset later for data jobs by writeCFG which does all modifications
254	>	PsetEdit.addCrabFJR(self.fjrFileName) # FUTURE: Job report addition not needed by CMSSW>1.5
255	>	PsetEdit.maxEvent(self.eventsPerJob)
256	>	PsetEdit.psetWriter(self.configFilename())
257	>	## If present, add TFileService to output files
258	>	if not int(cfg_params.get('CMSSW.skip_TFileService_output',0)):
259	>	tfsOutput = PsetEdit.getTFileService()
260	>	if tfsOutput:
261	>	if tfsOutput in self.output_file:
262	>	common.logger.debug(5,"Output from TFileService "+tfsOutput+" already in output files")
263	>	else:
264	>	outfileflag = True #output found
265	>	self.output_file.append(tfsOutput)
266	>	common.logger.message("Adding "+tfsOutput+" to output files (from TFileService)")
267	>	pass
268	>	pass
269	>	## If present and requested, add PoolOutputModule to output files
270	>	if int(cfg_params.get('CMSSW.get_edm_output',0)):
271	>	edmOutput = PsetEdit.getPoolOutputModule()
272	>	if edmOutput:
273	>	if edmOutput in self.output_file:
274	>	common.logger.debug(5,"Output from PoolOutputModule "+edmOutput+" already in output files")
275	>	else:
276	>	self.output_file.append(edmOutput)
277	>	common.logger.message("Adding "+edmOutput+" to output files (from PoolOutputModule)")
278	>	pass
279	>	pass
280	>	except CrabException:
281	>	msg='Error while manipulating ParameterSet: exiting...'
282	>	raise CrabException(msg)
283	>	## Prepare inputSandbox TarBall (only the first time)
284	>	self.tgzNameWithPath = self.getTarBall(self.executable)
285
286		def DataDiscoveryAndLocation(self, cfg_params):
287
#	Line 296 | Line 294 | class Cmssw(JobType):
294		## Contact the DBS
295		common.logger.message("Contacting Data Discovery Services ...")
296		try:
297	<	self.pubdata=DataDiscovery.DataDiscovery(datasetPath, cfg_params)
297	>	self.pubdata=DataDiscovery.DataDiscovery(datasetPath, cfg_params,self.skip_blocks)
298		self.pubdata.fetchDBSInfo()
299
300		except DataDiscovery.NotExistingDatasetError, ex :
#	Line 312 | Line 310 | class Cmssw(JobType):
310		self.filesbyblock=self.pubdata.getFiles()
311		self.eventsbyblock=self.pubdata.getEventsPerBlock()
312		self.eventsbyfile=self.pubdata.getEventsPerFile()
313	+	self.parentFiles=self.pubdata.getParent()
314
315		## get max number of events
316	<	self.maxEvents=self.pubdata.getMaxEvents() ##  self.maxEvents used in Creator.py
316	>	self.maxEvents=self.pubdata.getMaxEvents()
317
318		## Contact the DLS and build a list of sites hosting the fileblocks
319		try:
#	Line 338 | Line 337 | class Cmssw(JobType):
337
338		return sites
339
341	–	# to Be Removed  DS -- BL
342	–	#  def setArgsList(self, argsList):
343	–	#      self.argsList = argsList
344	–
340		def jobSplittingByBlocks(self, blockSites):
341		"""
342		Perform job splitting. Jobs run over an integer number of files
#	Line 392 | Line 387 | class Cmssw(JobType):
387		else :
388		totalNumberOfJobs = self.ncjobs
389
395	–
390		blocks = blockSites.keys()
391		blockCount = 0
392		# Backup variable in case self.maxEvents counted events in a non-included block
#	Line 433 | Line 427 | class Cmssw(JobType):
427
428		# ---- Iterate over the files in the block until we've met the requested ---- #
429		# ---- total # of events or we've gone over all the files in this block  ---- #
430	+	pString=''
431		while ( (eventsRemaining > 0) and (fileCount < numFilesInBlock) and (jobCount < totalNumberOfJobs) ):
432		file = files[fileCount]
433	+	if self.useParent:
434	+	parent = self.parentFiles[file]
435	+	for f in parent :
436	+	pString += '\\\"' + f + '\\\"\,'
437	+	common.logger.debug(6, "File "+str(file)+" has the following parents: "+str(parent))
438	+	common.logger.write("File "+str(file)+" has the following parents: "+str(parent))
439		if newFile :
440		try:
441		numEventsInFile = self.eventsbyfile[file]
#	Line 447 | Line 448 | class Cmssw(JobType):
448		except KeyError:
449		common.logger.message("File "+str(file)+" has unknown number of events: skipping")
450
451	<
451	>	eventsPerJobRequested = min(eventsPerJobRequested, eventsRemaining)
452		# if less events in file remain than eventsPerJobRequested
453	<	if ( filesEventCount - jobSkipEventCount < eventsPerJobRequested ) :
453	>	if ( filesEventCount - jobSkipEventCount < eventsPerJobRequested):
454		# if last file in block
455		if ( fileCount == numFilesInBlock-1 ) :
456		# end job using last file, use remaining events in block
457		# close job and touch new file
458		fullString = parString[:-2]
459	<	list_of_lists.append([fullString,str(-1),str(jobSkipEventCount)])
459	>	if self.useParent:
460	>	fullParentString = pString[:-2]
461	>	list_of_lists.append([fullString,fullParentString,str(-1),str(jobSkipEventCount)])
462	>	else:
463	>	list_of_lists.append([fullString,str(-1),str(jobSkipEventCount)])
464		common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(filesEventCount - jobSkipEventCount)+" events (last file in block).")
465		self.jobDestination.append(blockSites[block])
466		common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
#	Line 467 | Line 472 | class Cmssw(JobType):
472		eventsRemaining = eventsRemaining - filesEventCount + jobSkipEventCount
473		jobSkipEventCount = 0
474		# reset file
475	+	pString = ""
476		parString = ""
477		filesEventCount = 0
478		newFile = 1
#	Line 479 | Line 485 | class Cmssw(JobType):
485		elif ( filesEventCount - jobSkipEventCount == eventsPerJobRequested ) :
486		# close job and touch new file
487		fullString = parString[:-2]
488	<	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
488	>	if self.useParent:
489	>	fullParentString = pString[:-2]
490	>	list_of_lists.append([fullString,fullParentString,str(eventsPerJobRequested),str(jobSkipEventCount)])
491	>	else:
492	>	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
493		common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.")
494		self.jobDestination.append(blockSites[block])
495		common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
#	Line 490 | Line 500 | class Cmssw(JobType):
500		eventsRemaining = eventsRemaining - eventsPerJobRequested
501		jobSkipEventCount = 0
502		# reset file
503	+	pString = ""
504		parString = ""
505		filesEventCount = 0
506		newFile = 1
#	Line 499 | Line 510 | class Cmssw(JobType):
510		else :
511		# close job but don't touch new file
512		fullString = parString[:-2]
513	<	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
513	>	if self.useParent:
514	>	fullParentString = pString[:-2]
515	>	list_of_lists.append([fullString,fullParentString,str(eventsPerJobRequested),str(jobSkipEventCount)])
516	>	else:
517	>	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
518		common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.")
519		self.jobDestination.append(blockSites[block])
520		common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
#	Line 513 | Line 528 | class Cmssw(JobType):
528		jobSkipEventCount = eventsPerJobRequested - (filesEventCount - jobSkipEventCount - self.eventsbyfile[file])
529		# remove all but the last file
530		filesEventCount = self.eventsbyfile[file]
531	+	if self.useParent:
532	+	for f in parent : pString += '\\\"' + f + '\\\"\,'
533		parString = '\\\"' + file + '\\\"\,'
534		pass # END if
535		pass # END while (iterate over files in the block)
#	Line 533 | Line 550 | class Cmssw(JobType):
550		for block in blocks:
551		if block in jobsOfBlock.keys() :
552		blockCounter += 1
553	<	screenOutput += "Block %5i: jobs %20s: sites: %s\n" % (blockCounter,spanRanges(jobsOfBlock[block]),','.join(self.blackWhiteListParser.checkWhiteList(self.blackWhiteListParser.checkBlackList(blockSites[block],block),block)))
553	>	screenOutput += "Block %5i: jobs %20s: sites: %s\n" % (blockCounter,spanRanges(jobsOfBlock[block]),
554	>	','.join(self.blackWhiteListParser.checkWhiteList(self.blackWhiteListParser.checkBlackList(blockSites[block],block),block)))
555		if len(self.blackWhiteListParser.checkWhiteList(self.blackWhiteListParser.checkBlackList(blockSites[block],block),block)) == 0:
556		noSiteBlock.append( spanRanges(jobsOfBlock[block]) )
557		bloskNoSite.append( blockCounter )
#	Line 567 | Line 585 | class Cmssw(JobType):
585		self.list_of_args = list_of_lists
586		return
587
588	+	def jobSplittingNoBlockBoundary(self,blockSites):
589	+	"""
590	+	"""
591	+	# ---- Handle the possible job splitting configurations ---- #
592	+	if (self.selectTotalNumberEvents):
593	+	totalEventsRequested = self.total_number_of_events
594	+	if (self.selectEventsPerJob):
595	+	eventsPerJobRequested = self.eventsPerJob
596	+	if (self.selectNumberOfJobs):
597	+	totalEventsRequested = self.theNumberOfJobs * self.eventsPerJob
598	+
599	+	# If user requested all the events in the dataset
600	+	if (totalEventsRequested == -1):
601	+	eventsRemaining=self.maxEvents
602	+	# If user requested more events than are in the dataset
603	+	elif (totalEventsRequested > self.maxEvents):
604	+	eventsRemaining = self.maxEvents
605	+	common.logger.message("Requested "+str(self.total_number_of_events)+ " events, but only "+str(self.maxEvents)+" events are available.")
606	+	# If user requested less events than are in the dataset
607	+	else:
608	+	eventsRemaining = totalEventsRequested
609	+
610	+	# If user requested more events per job than are in the dataset
611	+	if (self.selectEventsPerJob and eventsPerJobRequested > self.maxEvents):
612	+	eventsPerJobRequested = self.maxEvents
613	+
614	+	# For user info at end
615	+	totalEventCount = 0
616	+
617	+	if (self.selectTotalNumberEvents and self.selectNumberOfJobs):
618	+	eventsPerJobRequested = int(eventsRemaining/self.theNumberOfJobs)
619	+
620	+	if (self.selectNumberOfJobs):
621	+	common.logger.message("May not create the exact number_of_jobs requested.")
622	+
623	+	if ( self.ncjobs == 'all' ) :
624	+	totalNumberOfJobs = 999999999
625	+	else :
626	+	totalNumberOfJobs = self.ncjobs
627	+
628	+	blocks = blockSites.keys()
629	+	blockCount = 0
630	+	# Backup variable in case self.maxEvents counted events in a non-included block
631	+	numBlocksInDataset = len(blocks)
632	+
633	+	jobCount = 0
634	+	list_of_lists = []
635	+
636	+	#AF
637	+	#AF do not reset input files and event count on block boundary
638	+	#AF
639	+	parString=""
640	+	filesEventCount = 0
641	+	#AF
642	+
643	+	# list tracking which jobs are in which jobs belong to which block
644	+	jobsOfBlock = {}
645	+	while ( (eventsRemaining > 0) and (blockCount < numBlocksInDataset) and (jobCount < totalNumberOfJobs)):
646	+	block = blocks[blockCount]
647	+	blockCount += 1
648	+	if block not in jobsOfBlock.keys() :
649	+	jobsOfBlock[block] = []
650	+
651	+	if self.eventsbyblock.has_key(block) :
652	+	numEventsInBlock = self.eventsbyblock[block]
653	+	common.logger.debug(5,'Events in Block File '+str(numEventsInBlock))
654	+	files = self.filesbyblock[block]
655	+	numFilesInBlock = len(files)
656	+	if (numFilesInBlock <= 0):
657	+	continue
658	+	fileCount = 0
659	+	#AF
660	+	#AF do not reset input files and event count of block boundary
661	+	#AF
662	+	## ---- New block => New job ---- #
663	+	#parString = ""
664	+	# counter for number of events in files currently worked on
665	+	#filesEventCount = 0
666	+	#AF
667	+	# flag if next while loop should touch new file
668	+	newFile = 1
669	+	# job event counter
670	+	jobSkipEventCount = 0
671	+
672	+	# ---- Iterate over the files in the block until we've met the requested ---- #
673	+	# ---- total # of events or we've gone over all the files in this block  ---- #
674	+	pString=''
675	+	while ( (eventsRemaining > 0) and (fileCount < numFilesInBlock) and (jobCount < totalNumberOfJobs) ):
676	+	file = files[fileCount]
677	+	if self.useParent:
678	+	parent = self.parentFiles[file]
679	+	for f in parent :
680	+	pString += '\\\"' + f + '\\\"\,'
681	+	common.logger.debug(6, "File "+str(file)+" has the following parents: "+str(parent))
682	+	common.logger.write("File "+str(file)+" has the following parents: "+str(parent))
683	+	if newFile :
684	+	try:
685	+	numEventsInFile = self.eventsbyfile[file]
686	+	common.logger.debug(6, "File "+str(file)+" has "+str(numEventsInFile)+" events")
687	+	# increase filesEventCount
688	+	filesEventCount += numEventsInFile
689	+	# Add file to current job
690	+	parString += '\\\"' + file + '\\\"\,'
691	+	newFile = 0
692	+	except KeyError:
693	+	common.logger.message("File "+str(file)+" has unknown number of events: skipping")
694	+	eventsPerJobRequested = min(eventsPerJobRequested, eventsRemaining)
695	+	#common.logger.message("AF filesEventCount %s - jobSkipEventCount %s "%(filesEventCount,jobSkipEventCount))
696	+	# if less events in file remain than eventsPerJobRequested
697	+	if ( filesEventCount - jobSkipEventCount < eventsPerJobRequested):
698	+	#AF
699	+	#AF skip fileboundary part
700	+	#AF
701	+	# go to next file
702	+	newFile = 1
703	+	fileCount += 1
704	+	# if events in file equal to eventsPerJobRequested
705	+	elif ( filesEventCount - jobSkipEventCount == eventsPerJobRequested ) :
706	+	# close job and touch new file
707	+	fullString = parString[:-2]
708	+	if self.useParent:
709	+	fullParentString = pString[:-2]
710	+	list_of_lists.append([fullString,fullParentString,str(eventsPerJobRequested),str(jobSkipEventCount)])
711	+	else:
712	+	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
713	+	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.")
714	+	self.jobDestination.append(blockSites[block])
715	+	common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
716	+	jobsOfBlock[block].append(jobCount+1)
717	+	# reset counter
718	+	jobCount = jobCount + 1
719	+	totalEventCount = totalEventCount + eventsPerJobRequested
720	+	eventsRemaining = eventsRemaining - eventsPerJobRequested
721	+	jobSkipEventCount = 0
722	+	# reset file
723	+	pString = ""
724	+	parString = ""
725	+	filesEventCount = 0
726	+	newFile = 1
727	+	fileCount += 1
728	+
729	+	# if more events in file remain than eventsPerJobRequested
730	+	else :
731	+	# close job but don't touch new file
732	+	fullString = parString[:-2]
733	+	if self.useParent:
734	+	fullParentString = pString[:-2]
735	+	list_of_lists.append([fullString,fullParentString,str(eventsPerJobRequested),str(jobSkipEventCount)])
736	+	else:
737	+	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
738	+	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.")
739	+	self.jobDestination.append(blockSites[block])
740	+	common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
741	+	jobsOfBlock[block].append(jobCount+1)
742	+	# increase counter
743	+	jobCount = jobCount + 1
744	+	totalEventCount = totalEventCount + eventsPerJobRequested
745	+	eventsRemaining = eventsRemaining - eventsPerJobRequested
746	+	# calculate skip events for last file
747	+	# use filesEventCount (contains several files), jobSkipEventCount and eventsPerJobRequest
748	+	jobSkipEventCount = eventsPerJobRequested - (filesEventCount - jobSkipEventCount - self.eventsbyfile[file])
749	+	# remove all but the last file
750	+	filesEventCount = self.eventsbyfile[file]
751	+	if self.useParent:
752	+	for f in parent : pString += '\\\"' + f + '\\\"\,'
753	+	parString = '\\\"' + file + '\\\"\,'
754	+	pass # END if
755	+	pass # END while (iterate over files in the block)
756	+	pass # END while (iterate over blocks in the dataset)
757	+	self.ncjobs = self.total_number_of_jobs = jobCount
758	+	if (eventsRemaining > 0 and jobCount < totalNumberOfJobs ):
759	+	common.logger.message("eventsRemaining "+str(eventsRemaining))
760	+	common.logger.message("jobCount "+str(jobCount))
761	+	common.logger.message(" totalNumberOfJobs "+str(totalNumberOfJobs))
762	+	common.logger.message("Could not run on all requested events because some blocks not hosted at allowed sites.")
763	+	common.logger.message(str(jobCount)+" job(s) can run on "+str(totalEventCount)+" events.\n")
764	+
765	+	# screen output
766	+	screenOutput = "List of jobs and available destination sites:\n\n"
767	+
768	+	#AF
769	+	#AF   skip check on  block with no sites
770	+	#AF
771	+	self.list_of_args = list_of_lists
772	+
773	+	return
774	+
775	+
776	+
777		def jobSplittingNoInput(self):
778		"""
779		Perform job splitting based on number of event per job
#	Line 620 | Line 827 | class Cmssw(JobType):
827		return
828
829
830	<	def jobSplittingForScript(self):#CarlosDaniele
830	>	def jobSplittingForScript(self):
831		"""
832		Perform job splitting based on number of job
833		"""
#	Line 636 | Line 843 | class Cmssw(JobType):
843		# argument is seed number.$i
844		self.list_of_args = []
845		for i in range(self.total_number_of_jobs):
639	–	## Since there is no input, any site is good
640	–	# self.jobDestination.append(["Any"])
846		self.jobDestination.append([""])
642	–	## no random seed
847		self.list_of_args.append([str(i)])
848		return
849
850	<	def split(self, jobParams):
850	>	def split(self, jobParams,firstJobID):
851
648	–	#### Fabio
852		njobs = self.total_number_of_jobs
853		arglist = self.list_of_args
854		# create the empty structure
#	Line 654 | Line 857 | class Cmssw(JobType):
857
858		listID=[]
859		listField=[]
860	<	for job in range(njobs):
861	<	jobParams[job] = arglist[job]
860	>	for id in range(njobs):
861	>	job = id + int(firstJobID)
862	>	jobParams[id] = arglist[id]
863		listID.append(job+1)
864		job_ToSave ={}
865		concString = ' '
866		argu=''
867	<	if len(jobParams[job]):
868	<	argu +=   concString.join(jobParams[job] )
869	<	job_ToSave['arguments']= str(job+1)+' '+argu## new BL--DS
870	<	job_ToSave['dlsDestination']= self.jobDestination[job]## new BL--DS
667	<	#common._db.updateJob_(job,job_ToSave)## new BL--DS
867	>	if len(jobParams[id]):
868	>	argu +=   concString.join(jobParams[id] )
869	>	job_ToSave['arguments']= str(job+1)+' '+argu
870	>	job_ToSave['dlsDestination']= self.jobDestination[id]
871		listField.append(job_ToSave)
872		msg="Job "+str(job)+" Arguments:   "+str(job+1)+" "+argu+"\n"  \
873	<	+"                     Destination: "+str(self.jobDestination[job])
873	>	+"                     Destination: "+str(self.jobDestination[id])
874		common.logger.debug(5,msg)
875	<	#common.logger.debug(5,"Job "+str(job)+" Destination: "+str(self.jobDestination[job]))
876	<	common._db.updateJob_(listID,listField)## new BL--DS
674	<	## Pay Attention Here....DS--BL
675	<	self.argsList = (len(jobParams[1])+1)
875	>	common._db.updateJob_(listID,listField)
876	>	self.argsList = (len(jobParams[0])+1)
877
878		return
678	–	#
679	–	#    def getJobTypeArguments(self, nj, sched):
680	–	#        result = ''
681	–	#        jobs=[]
682	–	#        jobs.append(nj)
683	–	#        for i in common._db.queryJob('arguments',jobs):##  BL--DS
684	–	#            result=result+str(i)+" "
685	–	#        return result
879
880		def numberOfJobs(self):
688	–	# Fabio
881		return self.total_number_of_jobs
882
883		def getTarBall(self, exe):
884		"""
885		Return the TarBall with lib and exe
886		"""
887	<
696	<	# if it exist, just return it
697	<	#
698	<	# Marco. Let's start to use relative path for Boss XML files
699	<	#
700	<	self.tgzNameWithPath = common.work_space.pathForTgz()+'share/'+self.tgz_name
887	>	self.tgzNameWithPath = common.work_space.pathForTgz()+self.tgz_name
888		if os.path.exists(self.tgzNameWithPath):
889		return self.tgzNameWithPath
890
#	Line 710 | Line 897 | class Cmssw(JobType):
897
898		# First of all declare the user Scram area
899		swArea = self.scram.getSWArea_()
713	–	#print "swArea = ", swArea
714	–	# swVersion = self.scram.getSWVersion()
715	–	# print "swVersion = ", swVersion
900		swReleaseTop = self.scram.getReleaseTop_()
717	–	#print "swReleaseTop = ", swReleaseTop
901
902		## check if working area is release top
903		if swReleaseTop == '' or swArea == swReleaseTop:
904	+	common.logger.debug(3,"swArea = "+swArea+" swReleaseTop ="+swReleaseTop)
905		return
906
907		import tarfile
#	Line 759 | Line 943 | class Cmssw(JobType):
943		tar.add(module,moduleDir)
944
945		## Now check if any data dir(s) is present
946	<	swAreaLen=len(swArea)
947	<	for root, dirs, files in os.walk(swArea):
948	<	if "data" in dirs:
949	<	common.logger.debug(5,"data "+root+"/data"+" to be tarred")
950	<	tar.add(root+"/data",root[swAreaLen:]+"/data")
946	>	self.dataExist = False
947	>	todo_list = [(i, i) for i in  os.listdir(swArea+"/src")]
948	>	while len(todo_list):
949	>	entry, name = todo_list.pop()
950	>	if name.startswith('crab_0_') or  name.startswith('.') or name == 'CVS':
951	>	continue
952	>	if os.path.isdir(swArea+"/src/"+entry):
953	>	entryPath = entry + '/'
954	>	todo_list += [(entryPath + i, i) for i in  os.listdir(swArea+"/src/"+entry)]
955	>	if name == 'data':
956	>	self.dataExist=True
957	>	common.logger.debug(5,"data "+entry+" to be tarred")
958	>	tar.add(swArea+"/src/"+entry,"src/"+entry)
959	>	pass
960	>	pass
961	>
962	>	### CMSSW ParameterSet
963	>	if not self.pset is None:
964	>	cfg_file = common.work_space.jobDir()+self.configFilename()
965	>	tar.add(cfg_file,self.configFilename())
966	>	common.logger.debug(5,"File added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
967
968
969		## Add ProdCommon dir to tar
970	<	prodcommonDir = 'ProdCommon'
971	<	prodcommonPath = os.environ['CRABDIR'] + '/' + 'ProdCommon'
972	<	if os.path.isdir(prodcommonPath):
973	<	tar.add(prodcommonPath,prodcommonDir)
970	>	prodcommonDir = './'
971	>	prodcommonPath = os.environ['CRABDIR'] + '/' + 'external/'
972	>	neededStuff = ['ProdCommon/__init__.py','ProdCommon/FwkJobRep', 'ProdCommon/CMSConfigTools', \
973	>	'ProdCommon/Core', 'ProdCommon/MCPayloads', 'IMProv', 'ProdCommon/Storage']
974	>	for file in neededStuff:
975	>	tar.add(prodcommonPath+file,prodcommonDir+file)
976	>	common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
977
978	+	##### ML stuff
979	+	ML_file_list=['report.py', 'DashboardAPI.py', 'Logger.py', 'ProcInfo.py', 'apmon.py']
980	+	path=os.environ['CRABDIR'] + '/python/'
981	+	for file in ML_file_list:
982	+	tar.add(path+file,file)
983		common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
984	+
985	+	##### Utils
986	+	Utils_file_list=['parseCrabFjr.py','writeCfg.py', 'fillCrabFjr.py','cmscp.py']
987	+	for file in Utils_file_list:
988	+	tar.add(path+file,file)
989	+	common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
990	+
991	+	##### AdditionalFiles
992	+	for file in self.additional_inbox_files:
993	+	tar.add(file,string.split(file,'/')[-1])
994	+	common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
995	+
996		tar.close()
997	<	except :
998	<	raise CrabException('Could not create tar-ball')
997	>	except IOError, exc:
998	>	common.logger.write(str(exc))
999	>	raise CrabException('Could not create tar-ball '+self.tgzNameWithPath)
1000	>	except tarfile.TarError, exc:
1001	>	common.logger.write(str(exc))
1002	>	raise CrabException('Could not create tar-ball '+self.tgzNameWithPath)
1003
1004		## check for tarball size
1005		tarballinfo = os.stat(self.tgzNameWithPath)
1006		if ( tarballinfo.st_size > self.MaxTarBallSize*1024*1024 ) :
1007	<	raise CrabException('Input sandbox size of ' + str(float(tarballinfo.st_size)/1024.0/1024.0) + ' MB is larger than the allowed ' + str(self.MaxTarBallSize) + ' MB input sandbox limit and not supported by the used GRID submission system. Please make sure that no unnecessary files are in all data directories in your local CMSSW project area as they are automatically packed into the input sandbox.')
1007	>	msg  = 'Input sandbox size of ' + str(float(tarballinfo.st_size)/1024.0/1024.0) + ' MB is larger than the allowed ' + str(self.MaxTarBallSize) \
1008	>	+'MB input sandbox limit \n'
1009	>	msg += '      and not supported by the direct GRID submission system.\n'
1010	>	msg += '      Please use the CRAB server mode by setting server_name=<NAME> in section [CRAB] of your crab.cfg.\n'
1011	>	msg += '      For further infos please see https://twiki.cern.ch/twiki/bin/view/CMS/CrabServer#CRABSERVER_for_Users'
1012	>	raise CrabException(msg)
1013
1014		## create tar-ball with ML stuff
786	–	self.MLtgzfile =  common.work_space.pathForTgz()+'share/MLfiles.tgz'
787	–	try:
788	–	tar = tarfile.open(self.MLtgzfile, "w:gz")
789	–	path=os.environ['CRABDIR'] + '/python/'
790	–	#for file in ['report.py', 'DashboardAPI.py', 'Logger.py', 'ProcInfo.py', 'apmon.py', 'parseCrabFjr.py','writeCfg.py']:
791	–	### FEDE ####
792	–	for file in ['report.py', 'DashboardAPI.py', 'Logger.py', 'ProcInfo.py', 'apmon.py', 'parseCrabFjr.py','writeCfg.py', 'JobReportErrorCode.py']:
793	–	###############
794	–	tar.add(path+file,file)
795	–	common.logger.debug(5,"Files added to "+self.MLtgzfile+" : "+str(tar.getnames()))
796	–	tar.close()
797	–	except :
798	–	raise CrabException('Could not create ML files tar-ball')
799	–
800	–	return
801	–
802	–	def additionalInputFileTgz(self):
803	–	"""
804	–	Put all additional files into a tar ball and return its name
805	–	"""
806	–	import tarfile
807	–	tarName=  common.work_space.pathForTgz()+'share/'+self.additional_tgz_name
808	–	tar = tarfile.open(tarName, "w:gz")
809	–	for file in self.additional_inbox_files:
810	–	tar.add(file,string.split(file,'/')[-1])
811	–	common.logger.debug(5,"Files added to "+self.additional_tgz_name+" : "+str(tar.getnames()))
812	–	tar.close()
813	–	return tarName
1015
1016		def wsSetupEnvironment(self, nj=0):
1017		"""
1018		Returns part of a job script which prepares
1019		the execution environment for the job 'nj'.
1020		"""
1021	+	if (self.CMSSW_major >= 2 and self.CMSSW_minor >= 1) or (self.CMSSW_major >= 3):
1022	+	psetName = 'pset.py'
1023	+	else:
1024	+	psetName = 'pset.cfg'
1025		# Prepare JobType-independent part
1026		txt = '\n#Written by cms_cmssw::wsSetupEnvironment\n'
1027		txt += 'echo ">>> setup environment"\n'
#	Line 825 | Line 1030 | class Cmssw(JobType):
1030		txt += 'elif [ $middleware == OSG ]; then\n'
1031		txt += '    WORKING_DIR=`/bin/mktemp  -d $OSG_WN_TMP/cms_XXXXXXXXXXXX`\n'
1032		txt += '    if [ ! $? == 0 ] ;then\n'
828	–	#txt += '        echo "SET_CMS_ENV 10016 ==> OSG $WORKING_DIR could not be created on WN `hostname`"\n'
829	–	#txt += '        echo "JOB_EXIT_STATUS = 10016"\n'
830	–	#txt += '        echo "JobExitCode=10016" | tee -a $RUNTIME_AREA/$repo\n'
831	–	#txt += '        dumpStatus $RUNTIME_AREA/$repo\n'
832	–	#txt += '        exit 1\n'
1033		txt += '        echo "ERROR ==> OSG $WORKING_DIR could not be created on WN `hostname`"\n'
1034		txt += '        job_exit_code=10016\n'
1035		txt += '        func_exit\n'
#	Line 840 | Line 1040 | class Cmssw(JobType):
1040		txt += '    cd $WORKING_DIR\n'
1041		txt += '    echo ">>> current directory (WORKING_DIR): $WORKING_DIR"\n'
1042		txt += self.wsSetupCMSOSGEnvironment_()
843	–	#txt += '    echo "### Set SCRAM ARCH to ' + self.executable_arch + ' ###"\n'
844	–	#txt += '    export SCRAM_ARCH='+self.executable_arch+'\n'
1043		txt += 'fi\n'
1044
1045		# Prepare JobType-specific part
#	Line 852 | Line 1050 | class Cmssw(JobType):
1050		txt += scram+' project CMSSW '+self.version+'\n'
1051		txt += 'status=$?\n'
1052		txt += 'if [ $status != 0 ] ; then\n'
855	–	#txt += '    echo "SET_EXE_ENV 10034 ==>ERROR CMSSW '+self.version+' not found on `hostname`" \n'
856	–	#txt += '    echo "JOB_EXIT_STATUS = 10034"\n'
857	–	#txt += '    echo "JobExitCode=10034" | tee -a $RUNTIME_AREA/$repo\n'
858	–	#txt += '    dumpStatus $RUNTIME_AREA/$repo\n'
1053		txt += '    echo "ERROR ==> CMSSW '+self.version+' not found on `hostname`" \n'
1054		txt += '    job_exit_code=10034\n'
861	–	#txt += '    if [ $middleware == OSG ]; then \n'
862	–	#txt += '        cd $RUNTIME_AREA\n'
863	–	#txt += '        echo ">>> current directory (RUNTIME_AREA): $RUNTIME_AREA"\n'
864	–	#txt += '        echo ">>> Remove working directory: $WORKING_DIR"\n'
865	–	#txt += '        /bin/rm -rf $WORKING_DIR\n'
866	–	#txt += '        if [ -d $WORKING_DIR ] ;then\n'
867	–	#txt += '            echo "SET_CMS_ENV 10018 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after CMSSW CMSSW_0_6_1 not found on `hostname`"\n'
868	–	#txt += '            echo "JOB_EXIT_STATUS = 10018"\n'
869	–	#txt += '            echo "JobExitCode=10018" | tee -a $RUNTIME_AREA/$repo\n'
870	–	#txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
871	–	#txt += '            echo "ERROR ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after CMSSW CMSSW_0_6_1 not found on `hostname`"\n'
872	–	#txt += '            job_exit_code=10017\n'
873	–	#txt += '        fi\n'
874	–	#txt += '    fi \n'
875	–	#txt += '    exit 1 \n'
1055		txt += '    func_exit\n'
1056		txt += 'fi \n'
1057		txt += 'cd '+self.version+'\n'
879	–	########## FEDE FOR DBS2 ######################
1058		txt += 'SOFTWARE_DIR=`pwd`\n'
1059		txt += 'echo ">>> current directory (SOFTWARE_DIR): $SOFTWARE_DIR" \n'
882	–	###############################################
883	–	### needed grep for bug in scramv1 ###
1060		txt += 'eval `'+scram+' runtime -sh | grep -v SCRAMRT_LSB_JOBNAME`\n'
1061	+	txt += 'if [ $? != 0 ] ; then\n'
1062	+	txt += '    echo "ERROR ==> Problem with the command: "\n'
1063	+	txt += '    echo "eval \`'+scram+' runtime -sh | grep -v SCRAMRT_LSB_JOBNAME \` at `hostname`"\n'
1064	+	txt += '    job_exit_code=10034\n'
1065	+	txt += '    func_exit\n'
1066	+	txt += 'fi \n'
1067		# Handle the arguments:
1068		txt += "\n"
1069		txt += "## number of arguments (first argument always jobnumber)\n"
1070		txt += "\n"
889	–	# txt += "if [ $nargs -lt "+str(len(self.argsList[nj].split()))+" ]\n"
1071		txt += "if [ $nargs -lt "+str(self.argsList)+" ]\n"
1072		txt += "then\n"
892	–	#txt += "    echo 'SET_EXE_ENV 1 ==> ERROR Too few arguments' +$nargs+ \n"
893	–	#txt += '    echo "JOB_EXIT_STATUS = 50113"\n'
894	–	#txt += '    echo "JobExitCode=50113" | tee -a $RUNTIME_AREA/$repo\n'
895	–	#txt += '    dumpStatus $RUNTIME_AREA/$repo\n'
1073		txt += "    echo 'ERROR ==> Too few arguments' +$nargs+ \n"
1074		txt += '    job_exit_code=50113\n'
898	–	#txt += '    if [ $middleware == OSG ]; then \n'
899	–	#txt += '        cd $RUNTIME_AREA\n'
900	–	#txt += '        echo ">>> current directory (RUNTIME_AREA): $RUNTIME_AREA"\n'
901	–	#txt += '        echo ">>> Remove working directory: $WORKING_DIR"\n'
902	–	#txt += '        /bin/rm -rf $WORKING_DIR\n'
903	–	#txt += '        if [ -d $WORKING_DIR ] ;then\n'
904	–	#txt += '            echo "SET_EXE_ENV 50114 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after Too few arguments for CRAB job wrapper"\n'
905	–	#txt += '            echo "JOB_EXIT_STATUS = 50114"\n'
906	–	#txt += '            echo "JobExitCode=50114" | tee -a $RUNTIME_AREA/$repo\n'
907	–	#txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
908	–	#txt += '            echo "ERROR ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after Too few arguments for CRAB job wrapper"\n'
909	–	#txt += '            job_exit_code=10017\n'
910	–	#txt += '        fi\n'
911	–	#txt += '    fi\n'
912	–	#txt += "    exit 1\n"
1075		txt += "    func_exit\n"
1076		txt += "fi\n"
1077		txt += "\n"
1078
1079		# Prepare job-specific part
1080		job = common.job_list[nj]
919	–	### FEDE FOR DBS OUTPUT PUBLICATION
1081		if (self.datasetPath):
1082	+	self.primaryDataset = self.datasetPath.split("/")[1]
1083	+	DataTier = self.datasetPath.split("/")[2]
1084		txt += '\n'
1085		txt += 'DatasetPath='+self.datasetPath+'\n'
1086
1087	<	datasetpath_split = self.datasetPath.split("/")
1088	<
926	<	txt += 'PrimaryDataset='+datasetpath_split[1]+'\n'
927	<	txt += 'DataTier='+datasetpath_split[2]+'\n'
1087	>	txt += 'PrimaryDataset='+self.primaryDataset +'\n'
1088	>	txt += 'DataTier='+DataTier+'\n'
1089		txt += 'ApplicationFamily=cmsRun\n'
1090
1091		else:
1092	+	self.primaryDataset = 'null'
1093		txt += 'DatasetPath=MCDataTier\n'
1094		txt += 'PrimaryDataset=null\n'
1095		txt += 'DataTier=null\n'
1096		txt += 'ApplicationFamily=MCDataTier\n'
1097	<	if self.pset != None:
1097	>	if self.pset != None:
1098		pset = os.path.basename(job.configFilename())
1099		txt += '\n'
1100		txt += 'cp  $RUNTIME_AREA/'+pset+' .\n'
1101		if (self.datasetPath): # standard job
1102		txt += 'InputFiles=${args[1]}; export InputFiles\n'
1103	<	txt += 'MaxEvents=${args[2]}; export MaxEvents\n'
1104	<	txt += 'SkipEvents=${args[3]}; export SkipEvents\n'
1103	>	if (self.useParent):
1104	>	txt += 'ParentFiles=${args[2]}; export ParentFiles\n'
1105	>	txt += 'MaxEvents=${args[3]}; export MaxEvents\n'
1106	>	txt += 'SkipEvents=${args[4]}; export SkipEvents\n'
1107	>	else:
1108	>	txt += 'MaxEvents=${args[2]}; export MaxEvents\n'
1109	>	txt += 'SkipEvents=${args[3]}; export SkipEvents\n'
1110		txt += 'echo "Inputfiles:<$InputFiles>"\n'
1111	+	if (self.useParent): txt += 'echo "ParentFiles:<$ParentFiles>"\n'
1112		txt += 'echo "MaxEvents:<$MaxEvents>"\n'
1113		txt += 'echo "SkipEvents:<$SkipEvents>"\n'
1114		else:  # pythia like job
#	Line 952 | Line 1120 | class Cmssw(JobType):
1120		txt += 'FirstRun=${args[1]}; export FirstRun\n'
1121		txt += 'echo "FirstRun: <$FirstRun>"\n'
1122
1123	<	txt += 'mv -f '+pset+' pset.cfg\n'
1123	>	txt += 'mv -f ' + pset + ' ' + psetName + '\n'
1124
957	–	if len(self.additional_inbox_files) > 0:
958	–	txt += 'if [ -e $RUNTIME_AREA/'+self.additional_tgz_name+' ] ; then\n'
959	–	txt += '  tar xzvf $RUNTIME_AREA/'+self.additional_tgz_name+'\n'
960	–	txt += 'fi\n'
961	–	pass
1125
1126		if self.pset != None:
1127	+	# FUTURE: Can simply for 2_1_x and higher
1128		txt += '\n'
1129	<	txt += 'echo "***** cat pset.cfg *********"\n'
1130	<	txt += 'cat pset.cfg\n'
1131	<	txt += 'echo "****** end pset.cfg ********"\n'
1132	<	txt += '\n'
1133	<	txt += 'PSETHASH=`EdmConfigHash < pset.cfg` \n'
1129	>	if self.debug_wrapper==True:
1130	>	txt += 'echo "***** cat ' + psetName + ' *********"\n'
1131	>	txt += 'cat ' + psetName + '\n'
1132	>	txt += 'echo "****** end ' + psetName + ' ********"\n'
1133	>	txt += '\n'
1134	>	if (self.CMSSW_major >= 2 and self.CMSSW_minor >= 1) or (self.CMSSW_major >= 3):
1135	>	txt += 'PSETHASH=`edmConfigHash ' + psetName + '` \n'
1136	>	else:
1137	>	txt += 'PSETHASH=`edmConfigHash < ' + psetName + '` \n'
1138		txt += 'echo "PSETHASH = $PSETHASH" \n'
1139		txt += '\n'
1140		return txt
1141	<	#### FEDE #####
1141	>
1142		def wsUntarSoftware(self, nj=0):
1143		"""
1144		Put in the script the commands to build an executable
#	Line 982 | Line 1150 | class Cmssw(JobType):
1150		if os.path.isfile(self.tgzNameWithPath):
1151		txt += 'echo ">>> tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+' :" \n'
1152		txt += 'tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+'\n'
1153	+	if  self.debug_wrapper:
1154	+	txt += 'ls -Al \n'
1155		txt += 'untar_status=$? \n'
1156		txt += 'if [ $untar_status -ne 0 ]; then \n'
1157		txt += '   echo "ERROR ==> Untarring .tgz file failed"\n'
#	Line 991 | Line 1161 | class Cmssw(JobType):
1161		txt += '   echo "Successful untar" \n'
1162		txt += 'fi \n'
1163		txt += '\n'
1164	<	txt += 'echo ">>> Include ProdCommon in PYTHONPATH:"\n'
1164	>	txt += 'echo ">>> Include $RUNTIME_AREA in PYTHONPATH:"\n'
1165		txt += 'if [ -z "$PYTHONPATH" ]; then\n'
1166	<	txt += '   export PYTHONPATH=$RUNTIME_AREA/ProdCommon\n'
1166	>	txt += '   export PYTHONPATH=$RUNTIME_AREA/\n'
1167		txt += 'else\n'
1168	<	txt += '   export PYTHONPATH=$RUNTIME_AREA/ProdCommon:${PYTHONPATH}\n'
1168	>	txt += '   export PYTHONPATH=$RUNTIME_AREA/:${PYTHONPATH}\n'
1169		txt += 'echo "PYTHONPATH=$PYTHONPATH"\n'
1170		txt += 'fi\n'
1171		txt += '\n'
#	Line 1003 | Line 1173 | class Cmssw(JobType):
1173		pass
1174
1175		return txt
1176	<
1176	>
1177		def wsBuildExe(self, nj=0):
1178		"""
1179		Put in the script the commands to build an executable
#	Line 1013 | Line 1183 | class Cmssw(JobType):
1183		txt = '\n#Written by cms_cmssw::wsBuildExe\n'
1184		txt += 'echo ">>> moving CMSSW software directories in `pwd`" \n'
1185
1186	<	txt += 'mv $RUNTIME_AREA/lib . \n'
1187	<	txt += 'mv $RUNTIME_AREA/module . \n'
1188	<	txt += 'mv $RUNTIME_AREA/ProdCommon . \n'
1189	<
1190	<
1191	<	#if os.path.isfile(self.tgzNameWithPath):
1192	<	#    txt += 'echo ">>> tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+' :" \n'
1193	<	#    txt += 'tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+'\n'
1194	<	#    txt += 'untar_status=$? \n'
1195	<	#    txt += 'if [ $untar_status -ne 0 ]; then \n'
1196	<	#    txt += '   echo "ERROR ==> Untarring .tgz file failed"\n'
1197	<	#    txt += '   job_exit_code=$untar_status\n'
1198	<	#    txt += '   func_exit\n'
1029	<	#    txt += 'else \n'
1030	<	#    txt += '   echo "Successful untar" \n'
1031	<	#    txt += 'fi \n'
1032	<	#    txt += '\n'
1033	<	#    txt += 'echo ">>> Include ProdCommon in PYTHONPATH:"\n'
1186	>	txt += 'rm -r lib/ module/ \n'
1187	>	txt += 'mv $RUNTIME_AREA/lib/ . \n'
1188	>	txt += 'mv $RUNTIME_AREA/module/ . \n'
1189	>	if self.dataExist == True:
1190	>	txt += 'rm -r src/ \n'
1191	>	txt += 'mv $RUNTIME_AREA/src/ . \n'
1192	>	if len(self.additional_inbox_files)>0:
1193	>	for file in self.additional_inbox_files:
1194	>	txt += 'mv $RUNTIME_AREA/'+os.path.basename(file)+' . \n'
1195	>	# txt += 'mv $RUNTIME_AREA/ProdCommon/ . \n'
1196	>	# txt += 'mv $RUNTIME_AREA/IMProv/ . \n'
1197	>
1198	>	txt += 'echo ">>> Include $RUNTIME_AREA in PYTHONPATH:"\n'
1199		txt += 'if [ -z "$PYTHONPATH" ]; then\n'
1200	<	txt += '   export PYTHONPATH=$SOFTWARE_DIR/ProdCommon\n'
1200	>	txt += '   export PYTHONPATH=$RUNTIME_AREA/\n'
1201		txt += 'else\n'
1202	<	txt += '   export PYTHONPATH=$SOFTWARE_DIR/ProdCommon:${PYTHONPATH}\n'
1202	>	txt += '   export PYTHONPATH=$RUNTIME_AREA/:${PYTHONPATH}\n'
1203		txt += 'echo "PYTHONPATH=$PYTHONPATH"\n'
1204		txt += 'fi\n'
1205		txt += '\n'
1206
1207		return txt
1043	–	############################################################################
1208
1045	–	def modifySteeringCards(self, nj):
1046	–	"""
1047	–	modify the card provided by the user,
1048	–	writing a new card into share dir
1049	–	"""
1209
1210		def executableName(self):
1211	<	if self.scriptExe: #CarlosDaniele
1211	>	if self.scriptExe:
1212		return "sh "
1213		else:
1214		return self.executable
#	Line 1059 | Line 1218 | class Cmssw(JobType):
1218		if self.scriptExe:#CarlosDaniele
1219		return   self.scriptExe + " $NJob"
1220		else:
1062	–	version_array = self.scram.getSWVersion().split('_')
1063	–	major = 0
1064	–	minor = 0
1065	–	try:
1066	–	major = int(version_array[1])
1067	–	minor = int(version_array[2])
1068	–	except:
1069	–	msg = "Cannot parse CMSSW version string: " + "_".join(version_array) + " for major and minor release number!"
1070	–	raise CrabException(msg)
1071	–
1221		ex_args = ""
1222	<
1222	>	# FUTURE: This tests the CMSSW version. Can remove code as versions deprecated
1223		# Framework job report
1224	<	if major >= 1 and minor >= 5 :
1076	<	#ex_args += " -j " + self.fjrFileName
1077	<	### FEDE it could be improved!!! ####
1224	>	if (self.CMSSW_major >= 1 and self.CMSSW_minor >= 5) or (self.CMSSW_major >= 2):
1225		ex_args += " -j $RUNTIME_AREA/crab_fjr_$NJob.xml"
1226	<	#######################################
1227	<	# Type of cfg file
1228	<	if major >= 2 :
1082	<	ex_args += " -p pset.pycfg"
1226	>	# Type of config file
1227	>	if self.CMSSW_major >= 2 :
1228	>	ex_args += " -p pset.py"
1229		else:
1230		ex_args += " -p pset.cfg"
1231		return ex_args
#	Line 1089 | Line 1235 | class Cmssw(JobType):
1235		Returns a list of filenames to be put in JDL input sandbox.
1236		"""
1237		inp_box = []
1092	–	# # dict added to delete duplicate from input sandbox file list
1093	–	# seen = {}
1094	–	## code
1238		if os.path.isfile(self.tgzNameWithPath):
1239		inp_box.append(self.tgzNameWithPath)
1240	<	if os.path.isfile(self.MLtgzfile):
1098	<	inp_box.append(self.MLtgzfile)
1099	<	## config
1100	<	if not self.pset is None:
1101	<	inp_box.append(common.work_space.pathForTgz() + 'job/' + self.configFilename())
1102	<	## additional input files
1103	<	tgz = self.additionalInputFileTgz()
1104	<	inp_box.append(tgz)
1105	<	## executable
1106	<	wrapper = os.path.basename(str(common._db.queryTask('scriptName')))
1107	<	inp_box.append(common.work_space.pathForTgz() +'job/'+ wrapper)
1240	>	inp_box.append(common.work_space.jobDir() + self.scriptName)
1241		return inp_box
1242
1243		def outputSandbox(self, nj):
#	Line 1116 | Line 1249 | class Cmssw(JobType):
1249		## User Declared output files
1250		for out in (self.output_file+self.output_file_sandbox):
1251		n_out = nj + 1
1252	<	out_box.append(self.numberFile_(out,str(n_out)))
1252	>	out_box.append(numberFile(out,str(n_out)))
1253		return out_box
1254
1122	–	def prepareSteeringCards(self):
1123	–	"""
1124	–	Make initial modifications of the user's steering card file.
1125	–	"""
1126	–	return
1255
1256		def wsRenameOutput(self, nj):
1257		"""
#	Line 1133 | Line 1261 | class Cmssw(JobType):
1261		txt = '\n#Written by cms_cmssw::wsRenameOutput\n'
1262		txt += 'echo ">>> current directory (SOFTWARE_DIR): $SOFTWARE_DIR" \n'
1263		txt += 'echo ">>> current directory content:"\n'
1264	<	txt += 'ls \n'
1264	>	if self.debug_wrapper:
1265	>	txt += 'ls -Al\n'
1266		txt += '\n'
1267
1139	–	#txt += 'output_exit_status=0\n'
1140	–
1141	–	### FEDE #######
1142	–	#for fileWithSuffix in (self.output_file_sandbox):
1143	–	#    output_file_num = self.numberFile_(fileWithSuffix, '$NJob')
1144	–	#    txt += '\n'
1145	–	#    txt += '# check output file\n'
1146	–	#    txt += 'if [ -e ./'+fileWithSuffix+' ] ; then\n'
1147	–	#    txt += '    mv '+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n'
1148	–	#    txt += '    ln -s $RUNTIME_AREA/'+output_file_num+' $RUNTIME_AREA/'+fileWithSuffix+'\n'
1149	–	#    txt += 'else\n'
1150	–	#    txt += '    echo "WARNING: Output file '+fileWithSuffix+' not found"\n'
1151	–	#    txt += '    job_exit_code=60302\n'
1152	–	#    if common.scheduler.name().upper() == 'CONDOR_G':
1153	–	#        txt += '    if [ $middleware == OSG ]; then \n'
1154	–	#        txt += '        echo "prepare dummy output file"\n'
1155	–	#        txt += '        echo "Processing of job output failed" > $RUNTIME_AREA/'+output_file_num+'\n'
1156	–	#        txt += '    fi \n'
1157	–	#    txt += 'fi\n'
1158	–
1268		for fileWithSuffix in (self.output_file):
1269	<	output_file_num = self.numberFile_(fileWithSuffix, '$NJob')
1269	>	output_file_num = numberFile(fileWithSuffix, '$NJob')
1270		txt += '\n'
1271		txt += '# check output file\n'
1272		txt += 'if [ -e ./'+fileWithSuffix+' ] ; then\n'
#	Line 1168 | Line 1277 | class Cmssw(JobType):
1277		txt += '    mv '+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n'
1278		txt += '    ln -s $RUNTIME_AREA/'+output_file_num+' $RUNTIME_AREA/'+fileWithSuffix+'\n'
1279		txt += 'else\n'
1171	–	#txt += '    exit_status=60302\n'
1172	–	#txt += '    echo "ERROR: Output file '+fileWithSuffix+' not found"\n'
1173	–	#txt += '    echo "JOB_EXIT_STATUS = $exit_status"\n'
1174	–	#txt += '    output_exit_status=$exit_status\n'
1280		txt += '    job_exit_code=60302\n'
1281		txt += '    echo "WARNING: Output file '+fileWithSuffix+' not found"\n'
1282		if common.scheduler.name().upper() == 'CONDOR_G':
#	Line 1182 | Line 1287 | class Cmssw(JobType):
1287		txt += 'fi\n'
1288		file_list = []
1289		for fileWithSuffix in (self.output_file):
1290	<	file_list.append(self.numberFile_(fileWithSuffix, '$NJob'))
1290	>	file_list.append(numberFile('$SOFTWARE_DIR/'+fileWithSuffix, '$NJob'))
1291
1292	<	txt += 'file_list="'+string.join(file_list,' ')+'"\n'
1292	>	txt += 'file_list="'+string.join(file_list,',')+'"\n'
1293		txt += '\n'
1294		txt += 'echo ">>> current directory (SOFTWARE_DIR): $SOFTWARE_DIR" \n'
1295		txt += 'echo ">>> current directory content:"\n'
1296	<	txt += 'ls \n'
1296	>	if self.debug_wrapper:
1297	>	txt += 'ls -Al\n'
1298		txt += '\n'
1299		txt += 'cd $RUNTIME_AREA\n'
1300		txt += 'echo ">>> current directory (RUNTIME_AREA):  $RUNTIME_AREA"\n'
1301		return txt
1302
1197	–	def numberFile_(self, file, txt):
1198	–	"""
1199	–	append _'txt' before last extension of a file
1200	–	"""
1201	–	p = string.split(file,".")
1202	–	# take away last extension
1203	–	name = p[0]
1204	–	for x in p[1:-1]:
1205	–	name=name+"."+x
1206	–	# add "_txt"
1207	–	if len(p)>1:
1208	–	ext = p[len(p)-1]
1209	–	result = name + '_' + txt + "." + ext
1210	–	else:
1211	–	result = name + '_' + txt
1212	–
1213	–	return result
1214	–
1303		def getRequirements(self, nj=[]):
1304		"""
1305		return job requirements to add to jdl files
#	Line 1221 | Line 1309 | class Cmssw(JobType):
1309		req='Member("VO-cms-' + \
1310		self.version + \
1311		'", other.GlueHostApplicationSoftwareRunTimeEnvironment)'
1312	<	## SL add requirement for OS version only if SL4
1225	<	#reSL4 = re.compile( r'slc4' )
1226	<	if self.executable_arch: # and reSL4.search(self.executable_arch):
1312	>	if self.executable_arch:
1313		req+=' && Member("VO-cms-' + \
1314		self.executable_arch + \
1315		'", other.GlueHostApplicationSoftwareRunTimeEnvironment)'
1316
1317		req = req + ' && (other.GlueHostNetworkAdapterOutboundIP)'
1318	<	if common.scheduler.name() == "glitecoll":
1318	>	if ( common.scheduler.name() == "glitecoll" ) or ( common.scheduler.name() == "glite"):
1319		req += ' && other.GlueCEStateStatus == "Production" '
1320
1321		return req
1322
1323		def configFilename(self):
1324		""" return the config filename """
1325	<	return self.name()+'.cfg'
1325	>	# FUTURE: Can remove cfg mode for CMSSW >= 2_1_x
1326	>	if (self.CMSSW_major >= 2 and self.CMSSW_minor >= 1) or (self.CMSSW_major >= 3):
1327	>	return self.name()+'.py'
1328	>	else:
1329	>	return self.name()+'.cfg'
1330
1331		def wsSetupCMSOSGEnvironment_(self):
1332		"""
#	Line 1252 | Line 1342 | class Cmssw(JobType):
1342		txt += '      # Use $OSG_APP/cmssoft/cms/cmsset_default.sh to setup cms software\n'
1343		txt += '        source $OSG_APP/cmssoft/cms/cmsset_default.sh '+self.version+'\n'
1344		txt += '    else\n'
1255	–	#txt += '        echo "SET_CMS_ENV 10020 ==> ERROR $OSG_APP/cmssoft/cms/cmsset_default.sh file not found"\n'
1256	–	#txt += '        echo "JOB_EXIT_STATUS = 10020"\n'
1257	–	#txt += '        echo "JobExitCode=10020" | tee -a $RUNTIME_AREA/$repo\n'
1258	–	#txt += '        dumpStatus $RUNTIME_AREA/$repo\n'
1345		txt += '        echo "ERROR ==> $OSG_APP/cmssoft/cms/cmsset_default.sh file not found"\n'
1346		txt += '        job_exit_code=10020\n'
1261	–	#txt += '        cd $RUNTIME_AREA\n'
1262	–	#txt += '        echo ">>> current directory (RUNTIME_AREA): $RUNTIME_AREA"\n'
1263	–	#txt += '        echo ">>> Remove working directory: $WORKING_DIR"\n'
1264	–	#txt += '        /bin/rm -rf $WORKING_DIR\n'
1265	–	#txt += '        if [ -d $WORKING_DIR ] ;then\n'
1266	–	#txt += '            echo "SET_CMS_ENV 10017 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after $OSG_APP/cmssoft/cms/cmsset_default.sh file not found"\n'
1267	–	#txt += '            echo "JOB_EXIT_STATUS = 10017"\n'
1268	–	#txt += '            echo "JobExitCode=10017" | tee -a $RUNTIME_AREA/$repo\n'
1269	–	#txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
1270	–	#txt += '            echo "ERROR ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after $OSG_APP/cmssoft/cms/cmsset_default.sh file not found"\n'
1271	–	#txt += '            job_exit_code=10017\n'
1272	–	#txt += '        fi\n'
1273	–	txt += '\n'
1274	–	#txt += '        exit 1\n'
1347		txt += '        func_exit\n'
1348		txt += '    fi\n'
1349		txt += '\n'
#	Line 1280 | Line 1352 | class Cmssw(JobType):
1352
1353		return txt
1354
1283	–	### OLI_DANIELE
1355		def wsSetupCMSLCGEnvironment_(self):
1356		"""
1357		Returns part of a job script which is prepares
#	Line 1292 | Line 1363 | class Cmssw(JobType):
1363		txt += '    export SCRAM_ARCH='+self.executable_arch+'\n'
1364		txt += '    export BUILD_ARCH='+self.executable_arch+'\n'
1365		txt += '    if [ ! $VO_CMS_SW_DIR ] ;then\n'
1295	–	#txt += '        echo "SET_CMS_ENV 10031 ==> ERROR CMS software dir not found on WN `hostname`"\n'
1296	–	#txt += '        echo "JOB_EXIT_STATUS = 10031" \n'
1297	–	#txt += '        echo "JobExitCode=10031" | tee -a $RUNTIME_AREA/$repo\n'
1298	–	#txt += '        dumpStatus $RUNTIME_AREA/$repo\n'
1299	–	#txt += '        exit 1\n'
1366		txt += '        echo "ERROR ==> CMS software dir not found on WN `hostname`"\n'
1367		txt += '        job_exit_code=10031\n'
1368		txt += '        func_exit\n'
1369		txt += '    else\n'
1370		txt += '        echo "Sourcing environment... "\n'
1371		txt += '        if [ ! -s $VO_CMS_SW_DIR/cmsset_default.sh ] ;then\n'
1306	–	#txt += '            echo "SET_CMS_ENV 10020 ==> ERROR cmsset_default.sh file not found into dir $VO_CMS_SW_DIR"\n'
1307	–	#txt += '            echo "JOB_EXIT_STATUS = 10020"\n'
1308	–	#txt += '            echo "JobExitCode=10020" | tee -a $RUNTIME_AREA/$repo\n'
1309	–	#txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
1310	–	#txt += '            exit 1\n'
1372		txt += '            echo "ERROR ==> cmsset_default.sh file not found into dir $VO_CMS_SW_DIR"\n'
1373		txt += '            job_exit_code=10020\n'
1374		txt += '            func_exit\n'
#	Line 1316 | Line 1377 | class Cmssw(JobType):
1377		txt += '        source $VO_CMS_SW_DIR/cmsset_default.sh\n'
1378		txt += '        result=$?\n'
1379		txt += '        if [ $result -ne 0 ]; then\n'
1319	–	#txt += '            echo "SET_CMS_ENV 10032 ==> ERROR problem sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
1320	–	#txt += '            echo "JOB_EXIT_STATUS = 10032"\n'
1321	–	#txt += '            echo "JobExitCode=10032" | tee -a $RUNTIME_AREA/$repo\n'
1322	–	#txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
1323	–	#txt += '            exit 1\n'
1380		txt += '            echo "ERROR ==> problem sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
1381		txt += '            job_exit_code=10032\n'
1382		txt += '            func_exit\n'
#	Line 1330 | Line 1386 | class Cmssw(JobType):
1386		txt += '    echo "==> setup cms environment ok"\n'
1387		return txt
1388
1389	<	### FEDE FOR DBS OUTPUT PUBLICATION
1334	<	def modifyReport(self, nj):
1389	>	def wsModifyReport(self, nj):
1390		"""
1391		insert the part of the script that modifies the FrameworkJob Report
1392		"""
1393	<
1394	<	txt = '\n#Written by cms_cmssw::modifyReport\n'
1340	<	try:
1341	<	publish_data = int(self.cfg_params['USER.publish_data'])
1342	<	except KeyError:
1343	<	publish_data = 0
1393	>	txt = '\n#Written by cms_cmssw::wsModifyReport\n'
1394	>	publish_data = int(self.cfg_params.get('USER.publish_data',0))
1395		if (publish_data == 1):
1396	<
1397	<	txt += 'if [ $copy_exit_status -eq 0 ]; then\n'
1398	<	txt += '    echo ">>> Modify Job Report:" \n'
1399	<	txt += '    chmod a+x $SOFTWARE_DIR/ProdCommon/ProdCommon/FwkJobRep/ModifyJobReport.py\n'
1400	<	#txt += '    if [ -z "$SE" ]; then\n'
1401	<	#txt += '        SE="" \n'
1402	<	#txt += '    fi \n'
1403	<	#txt += '    if [ -z "$SE_PATH" ]; then\n'
1404	<	#txt += '        SE_PATH="" \n'
1405	<	#txt += '    fi \n'
1406	<	txt += '    echo "SE = $SE"\n'
1407	<	txt += '    echo "SE_PATH = $SE_PATH"\n'
1408	<
1409	<	processedDataset = self.cfg_params['USER.publish_data_name']
1410	<	txt += '    ProcessedDataset='+processedDataset+'\n'
1411	<	#txt += '    if [ "$SE_PATH" == "" ]; then\n'
1412	<	#txt += '        FOR_LFN=/copy_problems/ \n'
1413	<	#txt += '    else \n'
1414	<	#txt += '        tmp=`echo $SE_PATH | awk -F \'store\' \'{print$2}\'` \n'
1415	<	#txt += '        FOR_LFN=/store$tmp \n'
1416	<	#txt += '    fi \n'
1417	<	txt += '    tmp=`echo $SE_PATH | awk -F \'store\' \'{print$2}\'` \n'
1418	<	txt += '    FOR_LFN=/store$tmp \n'
1419	<	txt += '    echo "ProcessedDataset = $ProcessedDataset"\n'
1420	<	txt += '    echo "FOR_LFN = $FOR_LFN" \n'
1421	<	txt += '    echo "CMSSW_VERSION = $CMSSW_VERSION"\n\n'
1422	<	#txt += '    echo "$SOFTWARE_DIR/ProdCommon/ProdCommon/FwkJobRep/ModifyJobReport.py crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH"\n'
1423	<	#txt += '    $SOFTWARE_DIR/ProdCommon/ProdCommon/FwkJobRep/ModifyJobReport.py crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH\n'
1424	<	### FEDE ####
1425	<	txt += '    echo "$SOFTWARE_DIR/ProdCommon/ProdCommon/FwkJobRep/ModifyJobReport.py $RUNTIME_AREA/crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH"\n'
1375	<	txt += '    $SOFTWARE_DIR/ProdCommon/ProdCommon/FwkJobRep/ModifyJobReport.py $RUNTIME_AREA/crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH\n'
1376	<	####################################
1377	<	txt += '    modifyReport_result=$?\n'
1378	<	txt += '    if [ $modifyReport_result -ne 0 ]; then\n'
1379	<	txt += '        modifyReport_result=70500\n'
1380	<	txt += '        job_exit_code=$modifyReport_result\n'
1381	<	txt += '        echo "ModifyReportResult=$modifyReport_result" | tee -a $RUNTIME_AREA/$repo\n'
1382	<	txt += '        echo "WARNING: Problem with ModifyJobReport"\n'
1383	<	txt += '    else\n'
1384	<	### FEDE #####
1385	<	#txt += '        mv NewFrameworkJobReport.xml crab_fjr_$NJob.xml\n'
1386	<	#######################
1387	<	txt += '        mv NewFrameworkJobReport.xml $RUNTIME_AREA/crab_fjr_$NJob.xml\n'
1388	<	txt += '    fi\n'
1396	>
1397	>	txt += 'if [ $StageOutExitStatus -eq 0 ]; then\n'
1398	>	txt += '    FOR_LFN=$LFNBaseName/${PSETHASH}/\n'
1399	>	txt += 'else\n'
1400	>	txt += '    FOR_LFN=/copy_problems/ \n'
1401	>	txt += '    SE=""\n'
1402	>	txt += '    SE_PATH=""\n'
1403	>	txt += 'fi\n'
1404	>
1405	>	txt += 'echo ">>> Modify Job Report:" \n'
1406	>	txt += 'chmod a+x $RUNTIME_AREA/ProdCommon/FwkJobRep/ModifyJobReport.py\n'
1407	>	txt += 'ProcessedDataset= $procDataset \n'
1408	>	txt += 'echo "ProcessedDataset = $ProcessedDataset"\n'
1409	>	txt += 'echo "SE = $SE"\n'
1410	>	txt += 'echo "SE_PATH = $SE_PATH"\n'
1411	>	txt += 'echo "FOR_LFN = $FOR_LFN" \n'
1412	>	txt += 'echo "CMSSW_VERSION = $CMSSW_VERSION"\n\n'
1413	>	args = '$RUNTIME_AREA/crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier ' \
1414	>	'$User -$ProcessedDataset-$PSETHASH $ApplicationFamily '+ \
1415	>	'  $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH'
1416	>	txt += 'echo "$RUNTIME_AREA/ProdCommon/FwkJobRep/ModifyJobReport.py '+str(args)+'"\n'
1417	>	txt += '$RUNTIME_AREA/ProdCommon/FwkJobRep/ModifyJobReport.py '+str(args)+'\n'
1418	>	txt += 'modifyReport_result=$?\n'
1419	>	txt += 'if [ $modifyReport_result -ne 0 ]; then\n'
1420	>	txt += '    modifyReport_result=70500\n'
1421	>	txt += '    job_exit_code=$modifyReport_result\n'
1422	>	txt += '    echo "ModifyReportResult=$modifyReport_result" | tee -a $RUNTIME_AREA/$repo\n'
1423	>	txt += '    echo "WARNING: Problem with ModifyJobReport"\n'
1424	>	txt += 'else\n'
1425	>	txt += '    mv NewFrameworkJobReport.xml $RUNTIME_AREA/crab_fjr_$NJob.xml\n'
1426		txt += 'fi\n'
1427		return txt
1428
1429	<	def cleanEnv(self):
1430	<	txt = '\n#Written by cms_cmssw::cleanEnv\n'
1431	<	txt += 'if [ $middleware == OSG ]; then\n'
1432	<	txt += '    cd $RUNTIME_AREA\n'
1433	<	txt += '    echo ">>> current directory (RUNTIME_AREA): $RUNTIME_AREA"\n'
1434	<	txt += '    echo ">>> Remove working directory: $WORKING_DIR"\n'
1435	<	txt += '    /bin/rm -rf $WORKING_DIR\n'
1436	<	txt += '    if [ -d $WORKING_DIR ] ;then\n'
1437	<	txt += '        echo "ERROR ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after cleanup of WN"\n'
1438	<	txt += '        job_exit_code=10017\n'
1439	<	txt += '        func_exit\n'
1429	>	def wsParseFJR(self):
1430	>	"""
1431	>	Parse the FrameworkJobReport to obtain useful infos
1432	>	"""
1433	>	txt = '\n#Written by cms_cmssw::wsParseFJR\n'
1434	>	txt += 'echo ">>> Parse FrameworkJobReport crab_fjr.xml"\n'
1435	>	txt += 'if [ -s $RUNTIME_AREA/crab_fjr_$NJob.xml ]; then\n'
1436	>	txt += '    if [ -s $RUNTIME_AREA/parseCrabFjr.py ]; then\n'
1437	>	txt += '        cmd_out=`python $RUNTIME_AREA/parseCrabFjr.py --input $RUNTIME_AREA/crab_fjr_$NJob.xml --dashboard $MonitorID,$MonitorJobID '+self.debugWrap+'`\n'
1438	>	if self.debug_wrapper :
1439	>	txt += '        echo "Result of parsing the FrameworkJobReport crab_fjr.xml: $cmd_out"\n'
1440	>	txt += '        executable_exit_status=`python $RUNTIME_AREA/parseCrabFjr.py --input $RUNTIME_AREA/crab_fjr_$NJob.xml --exitcode`\n'
1441	>	txt += '        if [ $executable_exit_status -eq 50115 ];then\n'
1442	>	txt += '            echo ">>> crab_fjr.xml contents: "\n'
1443	>	txt += '            cat $RUNTIME_AREA/crab_fjr_$NJob.xml\n'
1444	>	txt += '            echo "Wrong FrameworkJobReport --> does not contain useful info. ExitStatus: $executable_exit_status"\n'
1445	>	txt += '        elif [ $executable_exit_status -eq -999 ];then\n'
1446	>	txt += '            echo "ExitStatus from FrameworkJobReport not available. not available. Using exit code of executable from command line."\n'
1447	>	txt += '        else\n'
1448	>	txt += '            echo "Extracted ExitStatus from FrameworkJobReport parsing output: $executable_exit_status"\n'
1449	>	txt += '        fi\n'
1450	>	txt += '    else\n'
1451	>	txt += '        echo "CRAB python script to parse CRAB FrameworkJobReport crab_fjr.xml is not available, using exit code of executable from command line."\n'
1452		txt += '    fi\n'
1453	+	#### Patch to check input data reading for CMSSW16x Hopefully we-ll remove it asap
1454	+	txt += '    if [ $executable_exit_status -eq 0 ];then\n'
1455	+	txt += '      echo ">>> Executable succeded  $executable_exit_status"\n'
1456	+	if (self.datasetPath and not (self.dataset_pu or self.useParent)) :
1457	+	# VERIFY PROCESSED DATA
1458	+	txt += '      echo ">>> Verify list of processed files:"\n'
1459	+	txt += '      echo $InputFiles |tr -d \'\\\\\' |tr \',\' \'\\n\'|tr -d \'"\' > input-files.txt\n'
1460	+	txt += '      python $RUNTIME_AREA/parseCrabFjr.py --input $RUNTIME_AREA/crab_fjr_$NJob.xml --lfn > processed-files.txt\n'
1461	+	txt += '      cat input-files.txt  | sort | uniq > tmp.txt\n'
1462	+	txt += '      mv tmp.txt input-files.txt\n'
1463	+	txt += '      echo "cat input-files.txt"\n'
1464	+	txt += '      echo "----------------------"\n'
1465	+	txt += '      cat input-files.txt\n'
1466	+	txt += '      cat processed-files.txt | sort | uniq > tmp.txt\n'
1467	+	txt += '      mv tmp.txt processed-files.txt\n'
1468	+	txt += '      echo "----------------------"\n'
1469	+	txt += '      echo "cat processed-files.txt"\n'
1470	+	txt += '      echo "----------------------"\n'
1471	+	txt += '      cat processed-files.txt\n'
1472	+	txt += '      echo "----------------------"\n'
1473	+	txt += '      diff -q input-files.txt processed-files.txt\n'
1474	+	txt += '      fileverify_status=$?\n'
1475	+	txt += '      if [ $fileverify_status -ne 0 ]; then\n'
1476	+	txt += '         executable_exit_status=30001\n'
1477	+	txt += '         echo "ERROR ==> not all input files processed"\n'
1478	+	txt += '         echo "      ==> list of processed files from crab_fjr.xml differs from list in pset.cfg"\n'
1479	+	txt += '         echo "      ==> diff input-files.txt processed-files.txt"\n'
1480	+	txt += '      fi\n'
1481	+	txt += '    elif [ $executable_exit_status -ne 0 ] || [ $executable_exit_status -ne 50015 ] || [ $executable_exit_status -ne 50017 ];then\n'
1482	+	txt += '      echo ">>> Executable failed  $executable_exit_status"\n'
1483	+	txt += '      func_exit\n'
1484	+	txt += '    fi\n'
1485	+	txt += '\n'
1486	+	txt += 'else\n'
1487	+	txt += '    echo "CRAB FrameworkJobReport crab_fjr.xml is not available, using exit code of executable from command line."\n'
1488		txt += 'fi\n'
1489		txt += '\n'
1490	+	txt += 'echo "ExeExitCode=$executable_exit_status" | tee -a $RUNTIME_AREA/$repo\n'
1491	+	txt += 'echo "EXECUTABLE_EXIT_STATUS = $executable_exit_status"\n'
1492	+	txt += 'job_exit_code=$executable_exit_status\n'
1493	+
1494		return txt
1495
1496		def setParam_(self, param, value):
#	Line 1427 | Line 1515 | class Cmssw(JobType):
1515		txt = ''
1516		txt += 'echo ">>> list of expected files on output sandbox"\n'
1517		listOutFiles = []
1518	<	stdout = 'CMSSW_$NJob.stdout'
1518	>	stdout = 'CMSSW_$NJob.stdout'
1519		stderr = 'CMSSW_$NJob.stderr'
1520		if (self.return_data == 1):
1521		for file in (self.output_file+self.output_file_sandbox):
1522	<	listOutFiles.append(self.numberFile_(file, '$NJob'))
1522	>	listOutFiles.append(numberFile(file, '$NJob'))
1523		listOutFiles.append(stdout)
1524		listOutFiles.append(stderr)
1525		else:
1526		for file in (self.output_file_sandbox):
1527	<	listOutFiles.append(self.numberFile_(file, '$NJob'))
1527	>	listOutFiles.append(numberFile(file, '$NJob'))
1528		listOutFiles.append(stdout)
1529		listOutFiles.append(stderr)
1530		txt += 'echo "output files: '+string.join(listOutFiles,' ')+'"\n'
1531		txt += 'filesToCheck="'+string.join(listOutFiles,' ')+'"\n'
1532		txt += 'export filesToCheck\n'
1533	<	return txt
1533	>	return txt

Diff Legend

–	Removed lines
+	Added lines
<	Changed lines
>	Changed lines