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 1.145 by fanzago, Wed Dec 5 14:40:06 2007 UTC vs.
Revision 1.231 by spiga, Mon Aug 25 16:13:05 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	+	from LFNBaseName import *
9
10		import os, string, glob
11
12		class Cmssw(JobType):
13	<	def __init__(self, cfg_params, ncjobs):
13	>	def __init__(self, cfg_params, ncjobs,skip_blocks, isNew):
14		JobType.__init__(self, 'CMSSW')
15		common.logger.debug(3,'CMSSW::__init__')
16	+	self.skip_blocks = skip_blocks
17
18		self.argsList = []
19
20		self._params = {}
21		self.cfg_params = cfg_params
22		# init BlackWhiteListParser
23	<	self.blackWhiteListParser = BlackWhiteListParser(cfg_params)
23	>	self.blackWhiteListParser = SEBlackWhiteListParser(cfg_params)
24
25	<	try:
24	<	self.MaxTarBallSize = float(self.cfg_params['EDG.maxtarballsize'])
25	<	except KeyError:
26	<	self.MaxTarBallSize = 9.5
25	>	self.MaxTarBallSize = float(self.cfg_params.get('EDG.maxtarballsize',9.5))
26
27		# number of jobs requested to be created, limit obj splitting
28		self.ncjobs = ncjobs
#	Line 36 | Line 35 | class Cmssw(JobType):
35		self.executable = ''
36		self.executable_arch = self.scram.getArch()
37		self.tgz_name = 'default.tgz'
39	–	self.additional_tgz_name = 'additional.tgz'
38		self.scriptName = 'CMSSW.sh'
39	<	self.pset = ''      #scrip use case Da
40	<	self.datasetPath = '' #scrip use case Da
39	>	self.pset = ''
40	>	self.datasetPath = ''
41
42		# set FJR file name
43		self.fjrFileName = 'crab_fjr.xml'
44
45		self.version = self.scram.getSWVersion()
46	<
47	<	#
48	<	# Try to block creation in case of arch/version mismatch
49	<	#
50	<
51	<	a = string.split(self.version, "_")
52	<
53	<	if int(a[1]) == 1 and (int(a[2]) < 5 and self.executable_arch.find('slc4') == 0):
54	<	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)
55	<	common.logger.message(msg)
58	<	if int(a[1]) == 1 and (int(a[2]) >= 5 and self.executable_arch.find('slc3') == 0):
59	<	msg = "Error: CMS does not support %s with %s architecture"%(self.version, self.executable_arch)
46	>	version_array = self.version.split('_')
47	>	self.CMSSW_major = 0
48	>	self.CMSSW_minor = 0
49	>	self.CMSSW_patch = 0
50	>	try:
51	>	self.CMSSW_major = int(version_array[1])
52	>	self.CMSSW_minor = int(version_array[2])
53	>	self.CMSSW_patch = int(version_array[3])
54	>	except:
55	>	msg = "Cannot parse CMSSW version string: " + self.version + " for major and minor release number!"
56		raise CrabException(msg)
57
62	–	common.taskDB.setDict('codeVersion',self.version)
63	–	self.setParam_('application', self.version)
64	–
58		### collect Data cards
59
60	<	try:
68	<	tmp =  cfg_params['CMSSW.datasetpath']
69	<	log.debug(6, "CMSSW::CMSSW(): datasetPath = "+tmp)
70	<	if string.lower(tmp)=='none':
71	<	self.datasetPath = None
72	<	self.selectNoInput = 1
73	<	else:
74	<	self.datasetPath = tmp
75	<	self.selectNoInput = 0
76	<	except KeyError:
60	>	if not cfg_params.has_key('CMSSW.datasetpath'):
61		msg = "Error: datasetpath not defined "
62		raise CrabException(msg)
63
64	<	# ML monitoring
65	<	# split dataset path style: /PreProdR3Minbias/SIM/GEN-SIM
82	<	if not self.datasetPath:
83	<	self.setParam_('dataset', 'None')
84	<	self.setParam_('owner', 'None')
85	<	else:
86	<	try:
87	<	datasetpath_split = self.datasetPath.split("/")
88	<	# standard style
89	<	self.setParam_('datasetFull', self.datasetPath)
90	<	self.setParam_('dataset', datasetpath_split[1])
91	<	self.setParam_('owner', datasetpath_split[2])
92	<	except:
93	<	self.setParam_('dataset', self.datasetPath)
94	<	self.setParam_('owner', self.datasetPath)
64	>	### Temporary: added to remove input file control in the case of PU
65	>	self.dataset_pu = cfg_params.get('CMSSW.dataset_pu', None)
66
67	<	self.setTaskid_()
68	<	self.setParam_('taskId', self.cfg_params['taskId'])
67	>	tmp =  cfg_params['CMSSW.datasetpath']
68	>	log.debug(6, "CMSSW::CMSSW(): datasetPath = "+tmp)
69	>	if string.lower(tmp)=='none':
70	>	self.datasetPath = None
71	>	self.selectNoInput = 1
72	>	else:
73	>	self.datasetPath = tmp
74	>	self.selectNoInput = 0
75
76		self.dataTiers = []
77	<
77	>	self.debugWrap = ''
78	>	self.debug_wrapper = cfg_params.get('USER.debug_wrapper',False)
79	>	if self.debug_wrapper: self.debugWrap='--debug'
80		## now the application
81	<	try:
82	<	self.executable = cfg_params['CMSSW.executable']
104	<	self.setParam_('exe', self.executable)
105	<	log.debug(6, "CMSSW::CMSSW(): executable = "+self.executable)
106	<	msg = "Default executable cmsRun overridden. Switch to " + self.executable
107	<	log.debug(3,msg)
108	<	except KeyError:
109	<	self.executable = 'cmsRun'
110	<	self.setParam_('exe', self.executable)
111	<	msg = "User executable not defined. Use cmsRun"
112	<	log.debug(3,msg)
113	<	pass
81	>	self.executable = cfg_params.get('CMSSW.executable','cmsRun')
82	>	log.debug(6, "CMSSW::CMSSW(): executable = "+self.executable)
83
84	<	try:
116	<	self.pset = cfg_params['CMSSW.pset']
117	<	log.debug(6, "Cmssw::Cmssw(): PSet file = "+self.pset)
118	<	if self.pset.lower() != 'none' :
119	<	if (not os.path.exists(self.pset)):
120	<	raise CrabException("User defined PSet file "+self.pset+" does not exist")
121	<	else:
122	<	self.pset = None
123	<	except KeyError:
84	>	if not cfg_params.has_key('CMSSW.pset'):
85		raise CrabException("PSet file missing. Cannot run cmsRun ")
86	+	self.pset = cfg_params['CMSSW.pset']
87	+	log.debug(6, "Cmssw::Cmssw(): PSet file = "+self.pset)
88	+	if self.pset.lower() != 'none' :
89	+	if (not os.path.exists(self.pset)):
90	+	raise CrabException("User defined PSet file "+self.pset+" does not exist")
91	+	else:
92	+	self.pset = None
93
94		# output files
95		## stuff which must be returned always via sandbox
#	Line 131 | Line 99 | class Cmssw(JobType):
99		self.output_file_sandbox.append(self.fjrFileName)
100
101		# other output files to be returned via sandbox or copied to SE
102	<	try:
103	<	self.output_file = []
104	<	tmp = cfg_params['CMSSW.output_file']
105	<	if tmp != '':
106	<	tmpOutFiles = string.split(cfg_params['CMSSW.output_file'],',')
107	<	log.debug(7, 'cmssw::cmssw(): output files '+str(tmpOutFiles))
108	<	for tmp in tmpOutFiles:
109	<	tmp=string.strip(tmp)
142	<	self.output_file.append(tmp)
143	<	pass
144	<	else:
145	<	log.message("No output file defined: only stdout/err and the CRAB Framework Job Report will be available\n")
146	<	pass
147	<	pass
148	<	except KeyError:
149	<	log.message("No output file defined: only stdout/err and the CRAB Framework Job Report will be available\n")
150	<	pass
102	>	outfileflag = False
103	>	self.output_file = []
104	>	tmp = cfg_params.get('CMSSW.output_file',None)
105	>	if tmp :
106	>	self.output_file = [x.strip() for x in tmp.split(',')]
107	>	outfileflag = True #output found
108	>	#else:
109	>	#    log.message("No output file defined: only stdout/err and the CRAB Framework Job Report will be available\n")
110
111		# script_exe file as additional file in inputSandbox
112	<	try:
113	<	self.scriptExe = cfg_params['USER.script_exe']
114	<	if self.scriptExe != '':
115	<	if not os.path.isfile(self.scriptExe):
116	<	msg ="ERROR. file "+self.scriptExe+" not found"
117	<	raise CrabException(msg)
159	<	self.additional_inbox_files.append(string.strip(self.scriptExe))
160	<	except KeyError:
161	<	self.scriptExe = ''
112	>	self.scriptExe = cfg_params.get('USER.script_exe',None)
113	>	if self.scriptExe :
114	>	if not os.path.isfile(self.scriptExe):
115	>	msg ="ERROR. file "+self.scriptExe+" not found"
116	>	raise CrabException(msg)
117	>	self.additional_inbox_files.append(string.strip(self.scriptExe))
118
163	–	#CarlosDaniele
119		if self.datasetPath == None and self.pset == None and self.scriptExe == '' :
120	<	msg ="Error. script_exe  not defined"
121	<	raise CrabException(msg)
120	>	msg ="Error. script_exe  not defined"
121	>	raise CrabException(msg)
122	>
123	>	# use parent files...
124	>	self.useParent = self.cfg_params.get('CMSSW.use_parent',False)
125
126		## additional input files
127	<	try:
127	>	if cfg_params.has_key('USER.additional_input_files'):
128		tmpAddFiles = string.split(cfg_params['USER.additional_input_files'],',')
129		for tmp in tmpAddFiles:
130		tmp = string.strip(tmp)
#	Line 183 | Line 141 | class Cmssw(JobType):
141		if not os.path.exists(file):
142		raise CrabException("Additional input file not found: "+file)
143		pass
186	–	# fname = string.split(file, '/')[-1]
187	–	# storedFile = common.work_space.pathForTgz()+'share/'+fname
188	–	# shutil.copyfile(file, storedFile)
144		self.additional_inbox_files.append(string.strip(file))
145		pass
146		pass
147		common.logger.debug(5,"Additional input files: "+str(self.additional_inbox_files))
148	<	except KeyError:
194	<	pass
195	<
196	<	# files per job
197	<	try:
198	<	if (cfg_params['CMSSW.files_per_jobs']):
199	<	raise CrabException("files_per_jobs no longer supported.  Quitting.")
200	<	except KeyError:
201	<	pass
148	>	pass
149
150		## Events per job
151	<	try:
151	>	if cfg_params.has_key('CMSSW.events_per_job'):
152		self.eventsPerJob =int( cfg_params['CMSSW.events_per_job'])
153		self.selectEventsPerJob = 1
154	<	except KeyError:
154	>	else:
155		self.eventsPerJob = -1
156		self.selectEventsPerJob = 0
157
158		## number of jobs
159	<	try:
159	>	if cfg_params.has_key('CMSSW.number_of_jobs'):
160		self.theNumberOfJobs =int( cfg_params['CMSSW.number_of_jobs'])
161		self.selectNumberOfJobs = 1
162	<	except KeyError:
162	>	else:
163		self.theNumberOfJobs = 0
164		self.selectNumberOfJobs = 0
165
166	<	try:
166	>	if cfg_params.has_key('CMSSW.total_number_of_events'):
167		self.total_number_of_events = int(cfg_params['CMSSW.total_number_of_events'])
168		self.selectTotalNumberEvents = 1
169	<	except KeyError:
169	>	if self.selectNumberOfJobs  == 1:
170	>	if (self.total_number_of_events != -1) and int(self.total_number_of_events) < int(self.theNumberOfJobs):
171	>	msg = 'Must specify at least one event per job. total_number_of_events > number_of_jobs '
172	>	raise CrabException(msg)
173	>	else:
174		self.total_number_of_events = 0
175		self.selectTotalNumberEvents = 0
176
177	<	if self.pset != None: #CarlosDaniele
177	>	if self.pset != None:
178		if ( (self.selectTotalNumberEvents + self.selectEventsPerJob + self.selectNumberOfJobs) != 2 ):
179		msg = 'Must define exactly two of total_number_of_events, events_per_job, or number_of_jobs.'
180		raise CrabException(msg)
#	Line 232 | Line 183 | class Cmssw(JobType):
183		msg = 'Must specify  number_of_jobs.'
184		raise CrabException(msg)
185
186	<	## source seed for pythia
187	<	try:
188	<	self.sourceSeed = int(cfg_params['CMSSW.pythia_seed'])
189	<	except KeyError:
190	<	self.sourceSeed = None
191	<	common.logger.debug(5,"No seed given")
192	<
193	<	try:
194	<	self.sourceSeedVtx = int(cfg_params['CMSSW.vtx_seed'])
195	<	except KeyError:
196	<	self.sourceSeedVtx = None
197	<	common.logger.debug(5,"No vertex seed given")
186	>	## New method of dealing with seeds
187	>	self.incrementSeeds = []
188	>	self.preserveSeeds = []
189	>	if cfg_params.has_key('CMSSW.preserve_seeds'):
190	>	tmpList = cfg_params['CMSSW.preserve_seeds'].split(',')
191	>	for tmp in tmpList:
192	>	tmp.strip()
193	>	self.preserveSeeds.append(tmp)
194	>	if cfg_params.has_key('CMSSW.increment_seeds'):
195	>	tmpList = cfg_params['CMSSW.increment_seeds'].split(',')
196	>	for tmp in tmpList:
197	>	tmp.strip()
198	>	self.incrementSeeds.append(tmp)
199	>
200	>	## FUTURE: Can remove in CRAB 2.4.0
201	>	self.sourceSeed    = cfg_params.get('CMSSW.pythia_seed',None)
202	>	self.sourceSeedVtx = cfg_params.get('CMSSW.vtx_seed',None)
203	>	self.sourceSeedG4  = cfg_params.get('CMSSW.g4_seed',None)
204	>	self.sourceSeedMix = cfg_params.get('CMSSW.mix_seed',None)
205	>	if self.sourceSeed or self.sourceSeedVtx or self.sourceSeedG4 or self.sourceSeedMix:
206	>	msg = 'pythia_seed, vtx_seed, g4_seed, and mix_seed are no longer valid settings. You must use increment_seeds or preserve_seeds'
207	>	raise CrabException(msg)
208
209	<	try:
249	<	self.sourceSeedG4 = int(cfg_params['CMSSW.g4_seed'])
250	<	except KeyError:
251	<	self.sourceSeedG4 = None
252	<	common.logger.debug(5,"No g4 sim hits seed given")
209	>	self.firstRun = cfg_params.get('CMSSW.first_run',None)
210
211	<	try:
212	<	self.sourceSeedMix = int(cfg_params['CMSSW.mix_seed'])
213	<	except KeyError:
257	<	self.sourceSeedMix = None
258	<	common.logger.debug(5,"No mix seed given")
259	<
260	<	try:
261	<	self.firstRun = int(cfg_params['CMSSW.first_run'])
262	<	except KeyError:
263	<	self.firstRun = None
264	<	common.logger.debug(5,"No first run given")
265	<	if self.pset != None: #CarlosDaniele
266	<	import PsetManipulator as pp
267	<	PsetEdit = pp.PsetManipulator(self.pset) #Daniele Pset
211	>	# Copy/return
212	>	self.copy_data = int(cfg_params.get('USER.copy_data',0))
213	>	self.return_data = int(cfg_params.get('USER.return_data',0))
214
215		#DBSDLS-start
216		## Initialize the variables that are extracted from DBS/DLS and needed in other places of the code
#	Line 278 | Line 224 | class Cmssw(JobType):
224		blockSites = self.DataDiscoveryAndLocation(cfg_params)
225		#DBSDLS-end
226
281	–	self.tgzNameWithPath = self.getTarBall(self.executable)
282	–
227		## Select Splitting
228		if self.selectNoInput:
229	<	if self.pset == None: #CarlosDaniele
229	>	if self.pset == None:
230		self.jobSplittingForScript()
231		else:
232		self.jobSplittingNoInput()
233		else:
234		self.jobSplittingByBlocks(blockSites)
235
236	<	# modify Pset
237	<	if self.pset != None: #CarlosDaniele
238	<	try:
239	<	if (self.datasetPath): # standard job
240	<	# allow to processa a fraction of events in a file
241	<	PsetEdit.inputModule("INPUTFILE")
242	<	PsetEdit.maxEvent(0)
243	<	PsetEdit.skipEvent(0)
244	<	else:  # pythia like job
236	>	# modify Pset only the first time
237	>	if isNew:
238	>	if self.pset != None:
239	>	import PsetManipulator as pp
240	>	PsetEdit = pp.PsetManipulator(self.pset)
241	>	try:
242	>	# Add FrameworkJobReport to parameter-set, set max events.
243	>	# Reset later for data jobs by writeCFG which does all modifications
244	>	PsetEdit.addCrabFJR(self.fjrFileName) # FUTURE: Job report addition not needed by CMSSW>1.5
245		PsetEdit.maxEvent(self.eventsPerJob)
246	<	if (self.firstRun):
247	<	PsetEdit.pythiaFirstRun(0)  #First Run
248	<	if (self.sourceSeed) :
249	<	PsetEdit.pythiaSeed(0)
250	<	if (self.sourceSeedVtx) :
251	<	PsetEdit.vtxSeed(0)
252	<	if (self.sourceSeedG4) :
253	<	PsetEdit.g4Seed(0)
254	<	if (self.sourceSeedMix) :
255	<	PsetEdit.mixSeed(0)
256	<	# add FrameworkJobReport to parameter-set
257	<	PsetEdit.addCrabFJR(self.fjrFileName)
258	<	PsetEdit.psetWriter(self.configFilename())
259	<	except:
260	<	msg='Error while manipuliating ParameterSet: exiting...'
261	<	raise CrabException(msg)
246	>	PsetEdit.psetWriter(self.configFilename())
247	>	## If present, add TFileService to output files
248	>	if not int(cfg_params.get('CMSSW.skip_TFileService_output',0)):
249	>	tfsOutput = PsetEdit.getTFileService()
250	>	if tfsOutput:
251	>	if tfsOutput in self.output_file:
252	>	common.logger.debug(5,"Output from TFileService "+tfsOutput+" already in output files")
253	>	else:
254	>	outfileflag = True #output found
255	>	self.output_file.append(tfsOutput)
256	>	common.logger.message("Adding "+tfsOutput+" to output files (from TFileService)")
257	>	pass
258	>	pass
259	>	## If present and requested, add PoolOutputModule to output files
260	>	if int(cfg_params.get('CMSSW.get_edm_output',0)):
261	>	edmOutput = PsetEdit.getPoolOutputModule()
262	>	if edmOutput:
263	>	if edmOutput in self.output_file:
264	>	common.logger.debug(5,"Output from PoolOutputModule "+edmOutput+" already in output files")
265	>	else:
266	>	self.output_file.append(edmOutput)
267	>	common.logger.message("Adding "+edmOutput+" to output files (from PoolOutputModule)")
268	>	pass
269	>	pass
270	>	except CrabException:
271	>	msg='Error while manipulating ParameterSet: exiting...'
272	>	raise CrabException(msg)
273	>	## Prepare inputSandbox TarBall (only the first time)
274	>	self.tgzNameWithPath = self.getTarBall(self.executable)
275
276		def DataDiscoveryAndLocation(self, cfg_params):
277
#	Line 327 | Line 284 | class Cmssw(JobType):
284		## Contact the DBS
285		common.logger.message("Contacting Data Discovery Services ...")
286		try:
287	<
331	<	self.pubdata=DataDiscovery.DataDiscovery(datasetPath, cfg_params)
287	>	self.pubdata=DataDiscovery.DataDiscovery(datasetPath, cfg_params,self.skip_blocks)
288		self.pubdata.fetchDBSInfo()
289
290		except DataDiscovery.NotExistingDatasetError, ex :
#	Line 344 | Line 300 | class Cmssw(JobType):
300		self.filesbyblock=self.pubdata.getFiles()
301		self.eventsbyblock=self.pubdata.getEventsPerBlock()
302		self.eventsbyfile=self.pubdata.getEventsPerFile()
303	+	self.parentFiles=self.pubdata.getParent()
304
305		## get max number of events
306	<	self.maxEvents=self.pubdata.getMaxEvents() ##  self.maxEvents used in Creator.py
306	>	self.maxEvents=self.pubdata.getMaxEvents()
307
308		## Contact the DLS and build a list of sites hosting the fileblocks
309		try:
#	Line 370 | Line 327 | class Cmssw(JobType):
327
328		return sites
329
373	–	def setArgsList(self, argsList):
374	–	self.argsList = argsList
375	–
330		def jobSplittingByBlocks(self, blockSites):
331		"""
332		Perform job splitting. Jobs run over an integer number of files
#	Line 423 | Line 377 | class Cmssw(JobType):
377		else :
378		totalNumberOfJobs = self.ncjobs
379
426	–
380		blocks = blockSites.keys()
381		blockCount = 0
382		# Backup variable in case self.maxEvents counted events in a non-included block
#	Line 464 | Line 417 | class Cmssw(JobType):
417
418		# ---- Iterate over the files in the block until we've met the requested ---- #
419		# ---- total # of events or we've gone over all the files in this block  ---- #
420	+	pString=''
421		while ( (eventsRemaining > 0) and (fileCount < numFilesInBlock) and (jobCount < totalNumberOfJobs) ):
422		file = files[fileCount]
423	+	if self.useParent:
424	+	parent = self.parentFiles[file]
425	+	for f in parent :
426	+	pString += '\\\"' + f + '\\\"\,'
427	+	common.logger.debug(6, "File "+str(file)+" has the following parents: "+str(parent))
428	+	common.logger.write("File "+str(file)+" has the following parents: "+str(parent))
429		if newFile :
430		try:
431		numEventsInFile = self.eventsbyfile[file]
#	Line 478 | Line 438 | class Cmssw(JobType):
438		except KeyError:
439		common.logger.message("File "+str(file)+" has unknown number of events: skipping")
440
441	<
441	>	eventsPerJobRequested = min(eventsPerJobRequested, eventsRemaining)
442		# if less events in file remain than eventsPerJobRequested
443	<	if ( filesEventCount - jobSkipEventCount < eventsPerJobRequested ) :
443	>	if ( filesEventCount - jobSkipEventCount < eventsPerJobRequested):
444		# if last file in block
445		if ( fileCount == numFilesInBlock-1 ) :
446		# end job using last file, use remaining events in block
447		# close job and touch new file
448		fullString = parString[:-2]
449	<	list_of_lists.append([fullString,str(-1),str(jobSkipEventCount)])
449	>	if self.useParent:
450	>	fullParentString = pString[:-2]
451	>	list_of_lists.append([fullString,fullParentString,str(-1),str(jobSkipEventCount)])
452	>	else:
453	>	list_of_lists.append([fullString,str(-1),str(jobSkipEventCount)])
454		common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(filesEventCount - jobSkipEventCount)+" events (last file in block).")
455		self.jobDestination.append(blockSites[block])
456		common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
#	Line 498 | Line 462 | class Cmssw(JobType):
462		eventsRemaining = eventsRemaining - filesEventCount + jobSkipEventCount
463		jobSkipEventCount = 0
464		# reset file
465	+	pString = ""
466		parString = ""
467		filesEventCount = 0
468		newFile = 1
#	Line 510 | Line 475 | class Cmssw(JobType):
475		elif ( filesEventCount - jobSkipEventCount == eventsPerJobRequested ) :
476		# close job and touch new file
477		fullString = parString[:-2]
478	<	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
478	>	if self.useParent:
479	>	fullParentString = pString[:-2]
480	>	list_of_lists.append([fullString,fullParentString,str(eventsPerJobRequested),str(jobSkipEventCount)])
481	>	else:
482	>	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
483		common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.")
484		self.jobDestination.append(blockSites[block])
485		common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
#	Line 521 | Line 490 | class Cmssw(JobType):
490		eventsRemaining = eventsRemaining - eventsPerJobRequested
491		jobSkipEventCount = 0
492		# reset file
493	+	pString = ""
494		parString = ""
495		filesEventCount = 0
496		newFile = 1
#	Line 530 | Line 500 | class Cmssw(JobType):
500		else :
501		# close job but don't touch new file
502		fullString = parString[:-2]
503	<	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
503	>	if self.useParent:
504	>	fullParentString = pString[:-2]
505	>	list_of_lists.append([fullString,fullParentString,str(eventsPerJobRequested),str(jobSkipEventCount)])
506	>	else:
507	>	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
508		common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.")
509		self.jobDestination.append(blockSites[block])
510		common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
#	Line 544 | Line 518 | class Cmssw(JobType):
518		jobSkipEventCount = eventsPerJobRequested - (filesEventCount - jobSkipEventCount - self.eventsbyfile[file])
519		# remove all but the last file
520		filesEventCount = self.eventsbyfile[file]
521	<	parString = ""
522	<	parString += '\\\"' + file + '\\\"\,'
521	>	if self.useParent:
522	>	for f in parent : pString += '\\\"' + f + '\\\"\,'
523	>	parString = '\\\"' + file + '\\\"\,'
524		pass # END if
525		pass # END while (iterate over files in the block)
526		pass # END while (iterate over blocks in the dataset)
#	Line 565 | Line 540 | class Cmssw(JobType):
540		for block in blocks:
541		if block in jobsOfBlock.keys() :
542		blockCounter += 1
543	<	screenOutput += "Block %5i: jobs %20s: sites: %s\n" % (blockCounter,spanRanges(jobsOfBlock[block]),','.join(self.blackWhiteListParser.checkWhiteList(self.blackWhiteListParser.checkBlackList(blockSites[block],block),block)))
543	>	screenOutput += "Block %5i: jobs %20s: sites: %s\n" % (blockCounter,spanRanges(jobsOfBlock[block]),
544	>	','.join(self.blackWhiteListParser.checkWhiteList(self.blackWhiteListParser.checkBlackList(blockSites[block],block),block)))
545		if len(self.blackWhiteListParser.checkWhiteList(self.blackWhiteListParser.checkBlackList(blockSites[block],block),block)) == 0:
546		noSiteBlock.append( spanRanges(jobsOfBlock[block]) )
547		bloskNoSite.append( blockCounter )
#	Line 585 | Line 561 | class Cmssw(JobType):
561		for range_jobs in noSiteBlock:
562		msg += str(range_jobs) + virgola
563		msg += '\n               will not be submitted and this block of data can not be analyzed!\n'
564	+	if self.cfg_params.has_key('EDG.se_white_list'):
565	+	msg += 'WARNING: SE White List: '+self.cfg_params['EDG.se_white_list']+'\n'
566	+	msg += '(Hint: By whitelisting you force the job to run at this particular site(s).\n'
567	+	msg += 'Please check if the dataset is available at this site!)\n'
568	+	if self.cfg_params.has_key('EDG.ce_white_list'):
569	+	msg += 'WARNING: CE White List: '+self.cfg_params['EDG.ce_white_list']+'\n'
570	+	msg += '(Hint: By whitelisting you force the job to run at this particular site(s).\n'
571	+	msg += 'Please check if the dataset is available at this site!)\n'
572	+
573		common.logger.message(msg)
574
575		self.list_of_args = list_of_lists
#	Line 638 | Line 623 | class Cmssw(JobType):
623		if (self.firstRun):
624		## pythia first run
625		args.append(str(self.firstRun)+str(i))
641	–	if (self.sourceSeed):
642	–	args.append(str(self.sourceSeed)+str(i))
643	–	if (self.sourceSeedVtx):
644	–	## + vtx random seed
645	–	args.append(str(self.sourceSeedVtx)+str(i))
646	–	if (self.sourceSeedG4):
647	–	## + G4 random seed
648	–	args.append(str(self.sourceSeedG4)+str(i))
649	–	if (self.sourceSeedMix):
650	–	## + Mix random seed
651	–	args.append(str(self.sourceSeedMix)+str(i))
652	–	pass
653	–	pass
626		self.list_of_args.append(args)
655	–	pass
627
628		return
629
630
631	<	def jobSplittingForScript(self):#CarlosDaniele
631	>	def jobSplittingForScript(self):
632		"""
633		Perform job splitting based on number of job
634		"""
#	Line 673 | Line 644 | class Cmssw(JobType):
644		# argument is seed number.$i
645		self.list_of_args = []
646		for i in range(self.total_number_of_jobs):
676	–	## Since there is no input, any site is good
677	–	# self.jobDestination.append(["Any"])
647		self.jobDestination.append([""])
679	–	## no random seed
648		self.list_of_args.append([str(i)])
649		return
650
651	<	def split(self, jobParams):
651	>	def split(self, jobParams,firstJobID):
652
685	–	common.jobDB.load()
686	–	#### Fabio
653		njobs = self.total_number_of_jobs
654		arglist = self.list_of_args
655		# create the empty structure
656		for i in range(njobs):
657		jobParams.append("")
658
659	<	for job in range(njobs):
660	<	jobParams[job] = arglist[job]
661	<	# print str(arglist[job])
662	<	# print jobParams[job]
663	<	common.jobDB.setArguments(job, jobParams[job])
664	<	common.logger.debug(5,"Job "+str(job)+" Destination: "+str(self.jobDestination[job]))
665	<	common.jobDB.setDestination(job, self.jobDestination[job])
659	>	listID=[]
660	>	listField=[]
661	>	for id in range(njobs):
662	>	job = id + int(firstJobID)
663	>	jobParams[id] = arglist[id]
664	>	listID.append(job+1)
665	>	job_ToSave ={}
666	>	concString = ' '
667	>	argu=''
668	>	if len(jobParams[id]):
669	>	argu +=   concString.join(jobParams[id] )
670	>	job_ToSave['arguments']= str(job+1)+' '+argu
671	>	job_ToSave['dlsDestination']= self.jobDestination[id]
672	>	listField.append(job_ToSave)
673	>	msg="Job "+str(job)+" Arguments:   "+str(job+1)+" "+argu+"\n"  \
674	>	+"                     Destination: "+str(self.jobDestination[id])
675	>	common.logger.debug(5,msg)
676	>	common._db.updateJob_(listID,listField)
677	>	self.argsList = (len(jobParams[0])+1)
678
701	–	common.jobDB.save()
679		return
680
704	–	def getJobTypeArguments(self, nj, sched):
705	–	result = ''
706	–	for i in common.jobDB.arguments(nj):
707	–	result=result+str(i)+" "
708	–	return result
709	–
681		def numberOfJobs(self):
711	–	# Fabio
682		return self.total_number_of_jobs
683
684		def getTarBall(self, exe):
685		"""
686		Return the TarBall with lib and exe
687		"""
718	–
719	–	# if it exist, just return it
720	–	#
721	–	# Marco. Let's start to use relative path for Boss XML files
722	–	#
688		self.tgzNameWithPath = common.work_space.pathForTgz()+'share/'+self.tgz_name
689		if os.path.exists(self.tgzNameWithPath):
690		return self.tgzNameWithPath
#	Line 733 | Line 698 | class Cmssw(JobType):
698
699		# First of all declare the user Scram area
700		swArea = self.scram.getSWArea_()
736	–	#print "swArea = ", swArea
737	–	# swVersion = self.scram.getSWVersion()
738	–	# print "swVersion = ", swVersion
701		swReleaseTop = self.scram.getReleaseTop_()
740	–	#print "swReleaseTop = ", swReleaseTop
702
703		## check if working area is release top
704		if swReleaseTop == '' or swArea == swReleaseTop:
705	+	common.logger.debug(3,"swArea = "+swArea+" swReleaseTop ="+swReleaseTop)
706		return
707
708		import tarfile
#	Line 782 | Line 744 | class Cmssw(JobType):
744		tar.add(module,moduleDir)
745
746		## Now check if any data dir(s) is present
747	<	swAreaLen=len(swArea)
748	<	for root, dirs, files in os.walk(swArea):
749	<	if "data" in dirs:
750	<	common.logger.debug(5,"data "+root+"/data"+" to be tarred")
751	<	tar.add(root+"/data",root[swAreaLen:]+"/data")
752	<
753	<	## Add ProdAgent dir to tar
754	<	paDir = 'ProdAgentApi'
755	<	pa = os.environ['CRABDIR'] + '/' + 'ProdAgentApi'
756	<	if os.path.isdir(pa):
757	<	tar.add(pa,paDir)
758	<
759	<	### FEDE FOR DBS PUBLICATION
760	<	## Add PRODCOMMON dir to tar
761	<	prodcommonDir = 'ProdCommon'
800	<	prodcommonPath = os.environ['CRABDIR'] + '/' + 'ProdCommon'
801	<	if os.path.isdir(prodcommonPath):
802	<	tar.add(prodcommonPath,prodcommonDir)
803	<	#############################
747	>	self.dataExist = False
748	>	todo_list = [(i, i) for i in  os.listdir(swArea+"/src")]
749	>	while len(todo_list):
750	>	entry, name = todo_list.pop()
751	>	if name.startswith('crab_0_') or  name.startswith('.') or name == 'CVS':
752	>	continue
753	>	if os.path.isdir(swArea+"/src/"+entry):
754	>	entryPath = entry + '/'
755	>	todo_list += [(entryPath + i, i) for i in  os.listdir(swArea+"/src/"+entry)]
756	>	if name == 'data':
757	>	self.dataExist=True
758	>	common.logger.debug(5,"data "+entry+" to be tarred")
759	>	tar.add(swArea+"/src/"+entry,"src/"+entry)
760	>	pass
761	>	pass
762
763	+	### CMSSW ParameterSet
764	+	if not self.pset is None:
765	+	cfg_file = common.work_space.jobDir()+self.configFilename()
766	+	tar.add(cfg_file,self.configFilename())
767	+	common.logger.debug(5,"File added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
768	+
769	+
770	+	## Add ProdCommon dir to tar
771	+	prodcommonDir = './'
772	+	prodcommonPath = os.environ['CRABDIR'] + '/' + 'external/'
773	+	neededStuff = ['ProdCommon/__init__.py','ProdCommon/FwkJobRep', 'ProdCommon/CMSConfigTools','ProdCommon/Core','ProdCommon/MCPayloads', 'IMProv']
774	+	for file in neededStuff:
775	+	tar.add(prodcommonPath+file,prodcommonDir+file)
776		common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
777	+
778	+	##### ML stuff
779	+	ML_file_list=['report.py', 'DashboardAPI.py', 'Logger.py', 'ProcInfo.py', 'apmon.py']
780	+	path=os.environ['CRABDIR'] + '/python/'
781	+	for file in ML_file_list:
782	+	tar.add(path+file,file)
783	+	common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
784	+
785	+	##### Utils
786	+	Utils_file_list=['parseCrabFjr.py','writeCfg.py', 'fillCrabFjr.py']
787	+	for file in Utils_file_list:
788	+	tar.add(path+file,file)
789	+	common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
790	+
791	+	##### AdditionalFiles
792	+	for file in self.additional_inbox_files:
793	+	tar.add(file,string.split(file,'/')[-1])
794	+	common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
795	+
796		tar.close()
797	<	except :
798	<	raise CrabException('Could not create tar-ball')
797	>	except IOError:
798	>	raise CrabException('Could not create tar-ball '+self.tgzNameWithPath)
799	>	except tarfile.TarError:
800	>	raise CrabException('Could not create tar-ball '+self.tgzNameWithPath)
801
802		## check for tarball size
803		tarballinfo = os.stat(self.tgzNameWithPath)
#	Line 813 | Line 805 | class Cmssw(JobType):
805		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.')
806
807		## create tar-ball with ML stuff
816	–	self.MLtgzfile =  common.work_space.pathForTgz()+'share/MLfiles.tgz'
817	–	try:
818	–	tar = tarfile.open(self.MLtgzfile, "w:gz")
819	–	path=os.environ['CRABDIR'] + '/python/'
820	–	for file in ['report.py', 'DashboardAPI.py', 'Logger.py', 'ProcInfo.py', 'apmon.py', 'parseCrabFjr.py']:
821	–	tar.add(path+file,file)
822	–	common.logger.debug(5,"Files added to "+self.MLtgzfile+" : "+str(tar.getnames()))
823	–	tar.close()
824	–	except :
825	–	raise CrabException('Could not create ML files tar-ball')
808
809	<	return
828	<
829	<	def additionalInputFileTgz(self):
830	<	"""
831	<	Put all additional files into a tar ball and return its name
832	<	"""
833	<	import tarfile
834	<	tarName=  common.work_space.pathForTgz()+'share/'+self.additional_tgz_name
835	<	tar = tarfile.open(tarName, "w:gz")
836	<	for file in self.additional_inbox_files:
837	<	tar.add(file,string.split(file,'/')[-1])
838	<	common.logger.debug(5,"Files added to "+self.additional_tgz_name+" : "+str(tar.getnames()))
839	<	tar.close()
840	<	return tarName
841	<
842	<	def wsSetupEnvironment(self, nj):
809	>	def wsSetupEnvironment(self, nj=0):
810		"""
811		Returns part of a job script which prepares
812		the execution environment for the job 'nj'.
813		"""
814	+	if (self.CMSSW_major >= 2 and self.CMSSW_minor >= 1) or (self.CMSSW_major >= 3):
815	+	psetName = 'pset.py'
816	+	else:
817	+	psetName = 'pset.cfg'
818		# Prepare JobType-independent part
819	<	txt = ''
819	>	txt = '\n#Written by cms_cmssw::wsSetupEnvironment\n'
820		txt += 'echo ">>> setup environment"\n'
821		txt += 'if [ $middleware == LCG ]; then \n'
822		txt += self.wsSetupCMSLCGEnvironment_()
823		txt += 'elif [ $middleware == OSG ]; then\n'
824		txt += '    WORKING_DIR=`/bin/mktemp  -d $OSG_WN_TMP/cms_XXXXXXXXXXXX`\n'
825		txt += '    if [ ! $? == 0 ] ;then\n'
826	<	txt += '        echo "SET_CMS_ENV 10016 ==> OSG $WORKING_DIR could not be created on WN `hostname`"\n'
827	<	txt += '        echo "JOB_EXIT_STATUS = 10016"\n'
828	<	txt += '        echo "JobExitCode=10016" | tee -a $RUNTIME_AREA/$repo\n'
858	<	txt += '        dumpStatus $RUNTIME_AREA/$repo\n'
859	<	txt += '        exit 1\n'
826	>	txt += '        echo "ERROR ==> OSG $WORKING_DIR could not be created on WN `hostname`"\n'
827	>	txt += '        job_exit_code=10016\n'
828	>	txt += '        func_exit\n'
829		txt += '    fi\n'
830		txt += '    echo ">>> Created working directory: $WORKING_DIR"\n'
831		txt += '\n'
#	Line 864 | Line 833 | class Cmssw(JobType):
833		txt += '    cd $WORKING_DIR\n'
834		txt += '    echo ">>> current directory (WORKING_DIR): $WORKING_DIR"\n'
835		txt += self.wsSetupCMSOSGEnvironment_()
867	–	#txt += '    echo "### Set SCRAM ARCH to ' + self.executable_arch + ' ###"\n'
868	–	#txt += '    export SCRAM_ARCH='+self.executable_arch+'\n'
836		txt += 'fi\n'
837
838		# Prepare JobType-specific part
#	Line 876 | Line 843 | class Cmssw(JobType):
843		txt += scram+' project CMSSW '+self.version+'\n'
844		txt += 'status=$?\n'
845		txt += 'if [ $status != 0 ] ; then\n'
846	<	txt += '    echo "SET_EXE_ENV 10034 ==>ERROR CMSSW '+self.version+' not found on `hostname`" \n'
847	<	txt += '    echo "JOB_EXIT_STATUS = 10034"\n'
848	<	txt += '    echo "JobExitCode=10034" | tee -a $RUNTIME_AREA/$repo\n'
882	<	txt += '    dumpStatus $RUNTIME_AREA/$repo\n'
883	<	txt += '    if [ $middleware == OSG ]; then \n'
884	<	txt += '        cd $RUNTIME_AREA\n'
885	<	txt += '        echo ">>> current directory (RUNTIME_AREA): $RUNTIME_AREA"\n'
886	<	txt += '        echo ">>> Remove working directory: $WORKING_DIR"\n'
887	<	txt += '        /bin/rm -rf $WORKING_DIR\n'
888	<	txt += '        if [ -d $WORKING_DIR ] ;then\n'
889	<	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'
890	<	txt += '            echo "JOB_EXIT_STATUS = 10018"\n'
891	<	txt += '            echo "JobExitCode=10018" | tee -a $RUNTIME_AREA/$repo\n'
892	<	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
893	<	txt += '        fi\n'
894	<	txt += '    fi \n'
895	<	txt += '    exit 1 \n'
846	>	txt += '    echo "ERROR ==> CMSSW '+self.version+' not found on `hostname`" \n'
847	>	txt += '    job_exit_code=10034\n'
848	>	txt += '    func_exit\n'
849		txt += 'fi \n'
850		txt += 'cd '+self.version+'\n'
898	–	########## FEDE FOR DBS2 ######################
851		txt += 'SOFTWARE_DIR=`pwd`\n'
852		txt += 'echo ">>> current directory (SOFTWARE_DIR): $SOFTWARE_DIR" \n'
901	–	###############################################
902	–	### needed grep for bug in scramv1 ###
903	–	txt += scram+' runtime -sh\n'
853		txt += 'eval `'+scram+' runtime -sh | grep -v SCRAMRT_LSB_JOBNAME`\n'
854	<	txt += 'echo $PATH\n'
854	>	txt += 'if [ $? != 0 ] ; then\n'
855	>	txt += '    echo "ERROR ==> Problem with the command: "\n'
856	>	txt += '    echo "eval \`'+scram+' runtime -sh | grep -v SCRAMRT_LSB_JOBNAME \` at `hostname`"\n'
857	>	txt += '    job_exit_code=10034\n'
858	>	txt += '    func_exit\n'
859	>	txt += 'fi \n'
860		# Handle the arguments:
861		txt += "\n"
862		txt += "## number of arguments (first argument always jobnumber)\n"
863		txt += "\n"
864	<	txt += "if [ $nargs -lt "+str(len(self.argsList[nj].split()))+" ]\n"
864	>	txt += "if [ $nargs -lt "+str(self.argsList)+" ]\n"
865		txt += "then\n"
866	<	txt += "    echo 'SET_EXE_ENV 1 ==> ERROR Too few arguments' +$nargs+ \n"
867	<	txt += '    echo "JOB_EXIT_STATUS = 50113"\n'
868	<	txt += '    echo "JobExitCode=50113" | tee -a $RUNTIME_AREA/$repo\n'
915	<	txt += '    dumpStatus $RUNTIME_AREA/$repo\n'
916	<	txt += '    if [ $middleware == OSG ]; then \n'
917	<	txt += '        cd $RUNTIME_AREA\n'
918	<	txt += '        echo ">>> current directory (RUNTIME_AREA): $RUNTIME_AREA"\n'
919	<	txt += '        echo ">>> Remove working directory: $WORKING_DIR"\n'
920	<	txt += '        /bin/rm -rf $WORKING_DIR\n'
921	<	txt += '        if [ -d $WORKING_DIR ] ;then\n'
922	<	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'
923	<	txt += '            echo "JOB_EXIT_STATUS = 50114"\n'
924	<	txt += '            echo "JobExitCode=50114" | tee -a $RUNTIME_AREA/$repo\n'
925	<	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
926	<	txt += '        fi\n'
927	<	txt += '    fi \n'
928	<	txt += "    exit 1\n"
866	>	txt += "    echo 'ERROR ==> Too few arguments' +$nargs+ \n"
867	>	txt += '    job_exit_code=50113\n'
868	>	txt += "    func_exit\n"
869		txt += "fi\n"
870		txt += "\n"
871
872		# Prepare job-specific part
873		job = common.job_list[nj]
934	–	### FEDE FOR DBS OUTPUT PUBLICATION
874		if (self.datasetPath):
875		txt += '\n'
876		txt += 'DatasetPath='+self.datasetPath+'\n'
877
878		datasetpath_split = self.datasetPath.split("/")
879	<
879	>	### FEDE FOR NEW LFN ###
880	>	self.primaryDataset = datasetpath_split[1]
881	>	########################
882		txt += 'PrimaryDataset='+datasetpath_split[1]+'\n'
883		txt += 'DataTier='+datasetpath_split[2]+'\n'
884		txt += 'ApplicationFamily=cmsRun\n'
885
886		else:
887		txt += 'DatasetPath=MCDataTier\n'
888	+	### FEDE FOR NEW LFN ###
889	+	self.primaryDataset = 'null'
890	+	########################
891		txt += 'PrimaryDataset=null\n'
892		txt += 'DataTier=null\n'
893		txt += 'ApplicationFamily=MCDataTier\n'
894	<	if self.pset != None: #CarlosDaniele
894	>	if self.pset != None:
895		pset = os.path.basename(job.configFilename())
896		txt += '\n'
897		txt += 'cp  $RUNTIME_AREA/'+pset+' .\n'
898		if (self.datasetPath): # standard job
899	<	txt += 'InputFiles=${args[1]}\n'
900	<	txt += 'MaxEvents=${args[2]}\n'
901	<	txt += 'SkipEvents=${args[3]}\n'
899	>	txt += 'InputFiles=${args[1]}; export InputFiles\n'
900	>	if (self.useParent):
901	>	txt += 'ParentFiles=${args[2]}; export ParentFiles\n'
902	>	txt += 'MaxEvents=${args[3]}; export MaxEvents\n'
903	>	txt += 'SkipEvents=${args[4]}; export SkipEvents\n'
904	>	else:
905	>	txt += 'MaxEvents=${args[2]}; export MaxEvents\n'
906	>	txt += 'SkipEvents=${args[3]}; export SkipEvents\n'
907		txt += 'echo "Inputfiles:<$InputFiles>"\n'
908	<	txt += 'sed "s#\'INPUTFILE\'#$InputFiles#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
908	>	if (self.useParent): txt += 'echo "ParentFiles:<$ParentFiles>"\n'
909		txt += 'echo "MaxEvents:<$MaxEvents>"\n'
961	–	txt += 'sed "s#int32 input = 0#int32 input = $MaxEvents#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
910		txt += 'echo "SkipEvents:<$SkipEvents>"\n'
963	–	txt += 'sed "s#uint32 skipEvents = 0#uint32 skipEvents = $SkipEvents#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
911		else:  # pythia like job
912	<	seedIndex=1
912	>	txt += 'PreserveSeeds='  + ','.join(self.preserveSeeds)  + '; export PreserveSeeds\n'
913	>	txt += 'IncrementSeeds=' + ','.join(self.incrementSeeds) + '; export IncrementSeeds\n'
914	>	txt += 'echo "PreserveSeeds: <$PreserveSeeds>"\n'
915	>	txt += 'echo "IncrementSeeds:<$IncrementSeeds>"\n'
916		if (self.firstRun):
917	<	txt += 'FirstRun=${args['+str(seedIndex)+']}\n'
917	>	txt += 'FirstRun=${args[1]}; export FirstRun\n'
918		txt += 'echo "FirstRun: <$FirstRun>"\n'
969	–	txt += 'sed "s#uint32 firstRun = 0#uint32 firstRun = $FirstRun#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
970	–	seedIndex=seedIndex+1
919
920	<	if (self.sourceSeed):
973	<	txt += 'Seed=${args['+str(seedIndex)+']}\n'
974	<	txt += 'sed "s#uint32 sourceSeed = 0#uint32 sourceSeed = $Seed#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
975	<	seedIndex=seedIndex+1
976	<	## the following seeds are not always present
977	<	if (self.sourceSeedVtx):
978	<	txt += 'VtxSeed=${args['+str(seedIndex)+']}\n'
979	<	txt += 'echo "VtxSeed: <$VtxSeed>"\n'
980	<	txt += 'sed "s#uint32 VtxSmeared = 0#uint32 VtxSmeared = $VtxSeed#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
981	<	seedIndex += 1
982	<	if (self.sourceSeedG4):
983	<	txt += 'G4Seed=${args['+str(seedIndex)+']}\n'
984	<	txt += 'echo "G4Seed: <$G4Seed>"\n'
985	<	txt += 'sed "s#uint32 g4SimHits = 0#uint32 g4SimHits = $G4Seed#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
986	<	seedIndex += 1
987	<	if (self.sourceSeedMix):
988	<	txt += 'mixSeed=${args['+str(seedIndex)+']}\n'
989	<	txt += 'echo "MixSeed: <$mixSeed>"\n'
990	<	txt += 'sed "s#uint32 mix = 0#uint32 mix = $mixSeed#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
991	<	seedIndex += 1
992	<	pass
993	<	pass
994	<	txt += 'mv -f '+pset+' pset.cfg\n'
920	>	txt += 'mv -f ' + pset + ' ' + psetName + '\n'
921
996	–	if len(self.additional_inbox_files) > 0:
997	–	txt += 'if [ -e $RUNTIME_AREA/'+self.additional_tgz_name+' ] ; then\n'
998	–	txt += '  tar xzvf $RUNTIME_AREA/'+self.additional_tgz_name+'\n'
999	–	txt += 'fi\n'
1000	–	pass
922
923	<	if self.pset != None: #CarlosDaniele
923	>	if self.pset != None:
924	>	# FUTURE: Can simply for 2_1_x and higher
925		txt += '\n'
926	<	txt += 'echo "***** cat pset.cfg *********"\n'
927	<	txt += 'cat pset.cfg\n'
928	<	txt += 'echo "****** end pset.cfg ********"\n'
929	<	txt += '\n'
930	<	### FEDE FOR DBS OUTPUT PUBLICATION
931	<	txt += 'PSETHASH=`EdmConfigHash < pset.cfg` \n'
926	>	if self.debug_wrapper==True:
927	>	txt += 'echo "***** cat ' + psetName + ' *********"\n'
928	>	txt += 'cat ' + psetName + '\n'
929	>	txt += 'echo "****** end ' + psetName + ' ********"\n'
930	>	txt += '\n'
931	>	if (self.CMSSW_major >= 2 and self.CMSSW_minor >= 1) or (self.CMSSW_major >= 3):
932	>	txt += 'PSETHASH=`edmConfigHash ' + psetName + '` \n'
933	>	else:
934	>	txt += 'PSETHASH=`edmConfigHash < ' + psetName + '` \n'
935		txt += 'echo "PSETHASH = $PSETHASH" \n'
1011	–	##############
936		txt += '\n'
937		return txt
938
939	<	def wsBuildExe(self, nj=0):
939	>	def wsUntarSoftware(self, nj=0):
940		"""
941		Put in the script the commands to build an executable
942		or a library.
943		"""
944
945	<	txt = ""
945	>	txt = '\n#Written by cms_cmssw::wsUntarSoftware\n'
946
947		if os.path.isfile(self.tgzNameWithPath):
948		txt += 'echo ">>> tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+' :" \n'
949		txt += 'tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+'\n'
950	+	if  self.debug_wrapper:
951	+	txt += 'ls -Al \n'
952		txt += 'untar_status=$? \n'
953		txt += 'if [ $untar_status -ne 0 ]; then \n'
954	<	txt += '   echo "SET_EXE 1 ==> ERROR Untarring .tgz file failed"\n'
955	<	txt += '   echo "JOB_EXIT_STATUS = $untar_status" \n'
956	<	txt += '   echo "JobExitCode=$untar_status" | tee -a $RUNTIME_AREA/$repo\n'
1031	<	txt += '   if [ $middleware == OSG ]; then \n'
1032	<	txt += '       cd $RUNTIME_AREA\n'
1033	<	txt += '        echo ">>> current directory (RUNTIME_AREA): $RUNTIME_AREA"\n'
1034	<	txt += '        echo ">>> Remove working directory: $WORKING_DIR"\n'
1035	<	txt += '       /bin/rm -rf $WORKING_DIR\n'
1036	<	txt += '       if [ -d $WORKING_DIR ] ;then\n'
1037	<	txt += '           echo "SET_EXE 50999 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after Untarring .tgz file failed"\n'
1038	<	txt += '           echo "JOB_EXIT_STATUS = 50999"\n'
1039	<	txt += '           echo "JobExitCode=50999" | tee -a $RUNTIME_AREA/$repo\n'
1040	<	txt += '           dumpStatus $RUNTIME_AREA/$repo\n'
1041	<	txt += '       fi\n'
1042	<	txt += '   fi \n'
1043	<	txt += '   \n'
1044	<	txt += '   exit 1 \n'
954	>	txt += '   echo "ERROR ==> Untarring .tgz file failed"\n'
955	>	txt += '   job_exit_code=$untar_status\n'
956	>	txt += '   func_exit\n'
957		txt += 'else \n'
958		txt += '   echo "Successful untar" \n'
959		txt += 'fi \n'
960		txt += '\n'
961	<	txt += 'echo ">>> Include ProdAgentApi and PRODCOMMON in PYTHONPATH:"\n'
961	>	txt += 'echo ">>> Include $RUNTIME_AREA in PYTHONPATH:"\n'
962		txt += 'if [ -z "$PYTHONPATH" ]; then\n'
963	<	#### FEDE FOR DBS OUTPUT PUBLICATION
1052	<	txt += '   export PYTHONPATH=$SOFTWARE_DIR/ProdAgentApi:$SOFTWARE_DIR/ProdCommon\n'
963	>	txt += '   export PYTHONPATH=$RUNTIME_AREA/\n'
964		txt += 'else\n'
965	<	txt += '   export PYTHONPATH=$SOFTWARE_DIR/ProdAgentApi:$SOFTWARE_DIR/ProdCommon:${PYTHONPATH}\n'
965	>	txt += '   export PYTHONPATH=$RUNTIME_AREA/:${PYTHONPATH}\n'
966		txt += 'echo "PYTHONPATH=$PYTHONPATH"\n'
1056	–	###################
967		txt += 'fi\n'
968		txt += '\n'
969
#	Line 1061 | Line 971 | class Cmssw(JobType):
971
972		return txt
973
974	<	def modifySteeringCards(self, nj):
974	>	def wsBuildExe(self, nj=0):
975		"""
976	<	modify the card provided by the user,
977	<	writing a new card into share dir
976	>	Put in the script the commands to build an executable
977	>	or a library.
978		"""
979
980	+	txt = '\n#Written by cms_cmssw::wsBuildExe\n'
981	+	txt += 'echo ">>> moving CMSSW software directories in `pwd`" \n'
982	+
983	+	txt += 'rm -r lib/ module/ \n'
984	+	txt += 'mv $RUNTIME_AREA/lib/ . \n'
985	+	txt += 'mv $RUNTIME_AREA/module/ . \n'
986	+	if self.dataExist == True:
987	+	txt += 'rm -r src/ \n'
988	+	txt += 'mv $RUNTIME_AREA/src/ . \n'
989	+	if len(self.additional_inbox_files)>0:
990	+	for file in self.additional_inbox_files:
991	+	txt += 'mv $RUNTIME_AREA/'+os.path.basename(file)+' . \n'
992	+	# txt += 'mv $RUNTIME_AREA/ProdCommon/ . \n'
993	+	# txt += 'mv $RUNTIME_AREA/IMProv/ . \n'
994	+
995	+	txt += 'echo ">>> Include $RUNTIME_AREA in PYTHONPATH:"\n'
996	+	txt += 'if [ -z "$PYTHONPATH" ]; then\n'
997	+	txt += '   export PYTHONPATH=$RUNTIME_AREA/\n'
998	+	txt += 'else\n'
999	+	txt += '   export PYTHONPATH=$RUNTIME_AREA/:${PYTHONPATH}\n'
1000	+	txt += 'echo "PYTHONPATH=$PYTHONPATH"\n'
1001	+	txt += 'fi\n'
1002	+	txt += '\n'
1003	+
1004	+	return txt
1005	+
1006	+
1007		def executableName(self):
1008	<	if self.scriptExe: #CarlosDaniele
1008	>	if self.scriptExe:
1009		return "sh "
1010		else:
1011		return self.executable
1012
1013		def executableArgs(self):
1014	+	# FUTURE: This function tests the CMSSW version. Can be simplified as we drop support for old versions
1015		if self.scriptExe:#CarlosDaniele
1016		return   self.scriptExe + " $NJob"
1017		else:
1018	<	# if >= CMSSW_1_5_X, add -j crab_fjr.xml
1019	<	version_array = self.scram.getSWVersion().split('_')
1020	<	major = 0
1021	<	minor = 0
1022	<	try:
1023	<	major = int(version_array[1])
1024	<	minor = int(version_array[2])
1025	<	except:
1088	<	msg = "Cannot parse CMSSW version string: " + "_".join(version_array) + " for major and minor release number!"
1089	<	raise CrabException(msg)
1090	<	if major >= 1 and minor >= 5 :
1091	<	return " -j " + self.fjrFileName + " -p pset.cfg"
1018	>	ex_args = ""
1019	>	# FUTURE: This tests the CMSSW version. Can remove code as versions deprecated
1020	>	# Framework job report
1021	>	if (self.CMSSW_major >= 1 and self.CMSSW_minor >= 5) or (self.CMSSW_major >= 2):
1022	>	ex_args += " -j $RUNTIME_AREA/crab_fjr_$NJob.xml"
1023	>	# Type of config file
1024	>	if self.CMSSW_major >= 2 :
1025	>	ex_args += " -p pset.py"
1026		else:
1027	<	return " -p pset.cfg"
1027	>	ex_args += " -p pset.cfg"
1028	>	return ex_args
1029
1030		def inputSandbox(self, nj):
1031		"""
1032		Returns a list of filenames to be put in JDL input sandbox.
1033		"""
1034		inp_box = []
1100	–	# # dict added to delete duplicate from input sandbox file list
1101	–	# seen = {}
1102	–	## code
1035		if os.path.isfile(self.tgzNameWithPath):
1036		inp_box.append(self.tgzNameWithPath)
1037	<	if os.path.isfile(self.MLtgzfile):
1038	<	inp_box.append(self.MLtgzfile)
1107	<	## config
1108	<	if not self.pset is None:
1109	<	inp_box.append(common.work_space.pathForTgz() + 'job/' + self.configFilename())
1110	<	## additional input files
1111	<	tgz = self.additionalInputFileTgz()
1112	<	inp_box.append(tgz)
1037	>	wrapper = os.path.basename(str(common._db.queryTask('scriptName')))
1038	>	inp_box.append(common.work_space.pathForTgz() +'job/'+ wrapper)
1039		return inp_box
1040
1041		def outputSandbox(self, nj):
#	Line 1121 | Line 1047 | class Cmssw(JobType):
1047		## User Declared output files
1048		for out in (self.output_file+self.output_file_sandbox):
1049		n_out = nj + 1
1050	<	out_box.append(self.numberFile_(out,str(n_out)))
1050	>	out_box.append(numberFile(out,str(n_out)))
1051		return out_box
1052
1127	–	def prepareSteeringCards(self):
1128	–	"""
1129	–	Make initial modifications of the user's steering card file.
1130	–	"""
1131	–	return
1053
1054		def wsRenameOutput(self, nj):
1055		"""
1056		Returns part of a job script which renames the produced files.
1057		"""
1058
1059	<	txt = '\n'
1060	<	txt += 'echo" >>> directory content:"\n'
1061	<	txt += 'ls \n'
1059	>	txt = '\n#Written by cms_cmssw::wsRenameOutput\n'
1060	>	txt += 'echo ">>> current directory (SOFTWARE_DIR): $SOFTWARE_DIR" \n'
1061	>	txt += 'echo ">>> current directory content:"\n'
1062	>	if self.debug_wrapper:
1063	>	txt += 'ls -Al\n'
1064		txt += '\n'
1065
1143	–	txt += 'output_exit_status=0\n'
1144	–
1145	–	for fileWithSuffix in (self.output_file_sandbox):
1146	–	output_file_num = self.numberFile_(fileWithSuffix, '$NJob')
1147	–	txt += '\n'
1148	–	txt += '# check output file\n'
1149	–	txt += 'if [ -e ./'+fileWithSuffix+' ] ; then\n'
1150	–	#txt += '    mv '+fileWithSuffix+' $RUNTIME_AREA\n'
1151	–	txt += '    mv '+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n'
1152	–	#txt += '    cp $RUNTIME_AREA/'+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n'
1153	–	txt += '    ln -s $RUNTIME_AREA/'+output_file_num+' $RUNTIME_AREA/'+fileWithSuffix+'\n'
1154	–	txt += 'else\n'
1155	–	txt += '    exit_status=60302\n'
1156	–	txt += '    echo "ERROR: Problem with output file '+fileWithSuffix+'"\n'
1157	–	if common.scheduler.boss_scheduler_name == 'condor_g':
1158	–	txt += '    if [ $middleware == OSG ]; then \n'
1159	–	txt += '        echo "prepare dummy output file"\n'
1160	–	txt += '        echo "Processing of job output failed" > $RUNTIME_AREA/'+output_file_num+'\n'
1161	–	txt += '    fi \n'
1162	–	txt += 'fi\n'
1163	–
1066		for fileWithSuffix in (self.output_file):
1067	<	output_file_num = self.numberFile_(fileWithSuffix, '$NJob')
1067	>	output_file_num = numberFile(fileWithSuffix, '$NJob')
1068		txt += '\n'
1069		txt += '# check output file\n'
1070		txt += 'if [ -e ./'+fileWithSuffix+' ] ; then\n'
1071	<	#txt += '    mv '+fileWithSuffix+' $RUNTIME_AREA\n'
1072	<	txt += '    mv '+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n'
1073	<	#txt += '    cp $RUNTIME_AREA/'+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n'
1074	<	txt += '    ln -s $RUNTIME_AREA/'+output_file_num+' $RUNTIME_AREA/'+fileWithSuffix+'\n'
1071	>	if (self.copy_data == 1):  # For OSG nodes, file is in $WORKING_DIR, should not be moved to $RUNTIME_AREA
1072	>	txt += '    mv '+fileWithSuffix+' '+output_file_num+'\n'
1073	>	txt += '    ln -s `pwd`/'+output_file_num+' $RUNTIME_AREA/'+fileWithSuffix+'\n'
1074	>	else:
1075	>	txt += '    mv '+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n'
1076	>	txt += '    ln -s $RUNTIME_AREA/'+output_file_num+' $RUNTIME_AREA/'+fileWithSuffix+'\n'
1077		txt += 'else\n'
1078	<	txt += '    exit_status=60302\n'
1079	<	txt += '    echo "ERROR: Problem with output file '+fileWithSuffix+'"\n'
1080	<	txt += '    echo "JOB_EXIT_STATUS = $exit_status"\n'
1177	<	txt += '    output_exit_status=$exit_status\n'
1178	<	if common.scheduler.boss_scheduler_name == 'condor_g':
1078	>	txt += '    job_exit_code=60302\n'
1079	>	txt += '    echo "WARNING: Output file '+fileWithSuffix+' not found"\n'
1080	>	if common.scheduler.name().upper() == 'CONDOR_G':
1081		txt += '    if [ $middleware == OSG ]; then \n'
1082		txt += '        echo "prepare dummy output file"\n'
1083		txt += '        echo "Processing of job output failed" > $RUNTIME_AREA/'+output_file_num+'\n'
#	Line 1183 | Line 1085 | class Cmssw(JobType):
1085		txt += 'fi\n'
1086		file_list = []
1087		for fileWithSuffix in (self.output_file):
1088	<	file_list.append(self.numberFile_(fileWithSuffix, '$NJob'))
1088	>	file_list.append(numberFile(fileWithSuffix, '$NJob'))
1089
1090		txt += 'file_list="'+string.join(file_list,' ')+'"\n'
1091	+	txt += '\n'
1092	+	txt += 'echo ">>> current directory (SOFTWARE_DIR): $SOFTWARE_DIR" \n'
1093	+	txt += 'echo ">>> current directory content:"\n'
1094	+	if self.debug_wrapper:
1095	+	txt += 'ls -Al\n'
1096	+	txt += '\n'
1097		txt += 'cd $RUNTIME_AREA\n'
1098		txt += 'echo ">>> current directory (RUNTIME_AREA):  $RUNTIME_AREA"\n'
1099		return txt
1100
1193	–	def numberFile_(self, file, txt):
1194	–	"""
1195	–	append _'txt' before last extension of a file
1196	–	"""
1197	–	p = string.split(file,".")
1198	–	# take away last extension
1199	–	name = p[0]
1200	–	for x in p[1:-1]:
1201	–	name=name+"."+x
1202	–	# add "_txt"
1203	–	if len(p)>1:
1204	–	ext = p[len(p)-1]
1205	–	result = name + '_' + txt + "." + ext
1206	–	else:
1207	–	result = name + '_' + txt
1208	–
1209	–	return result
1210	–
1101		def getRequirements(self, nj=[]):
1102		"""
1103		return job requirements to add to jdl files
#	Line 1217 | Line 1107 | class Cmssw(JobType):
1107		req='Member("VO-cms-' + \
1108		self.version + \
1109		'", other.GlueHostApplicationSoftwareRunTimeEnvironment)'
1110	<	## SL add requirement for OS version only if SL4
1221	<	#reSL4 = re.compile( r'slc4' )
1222	<	if self.executable_arch: # and reSL4.search(self.executable_arch):
1110	>	if self.executable_arch:
1111		req+=' && Member("VO-cms-' + \
1112		self.executable_arch + \
1113		'", other.GlueHostApplicationSoftwareRunTimeEnvironment)'
1114
1115		req = req + ' && (other.GlueHostNetworkAdapterOutboundIP)'
1116	+	if ( common.scheduler.name() == "glitecoll" ) or ( common.scheduler.name() == "glite"):
1117	+	req += ' && other.GlueCEStateStatus == "Production" '
1118
1119		return req
1120
1121		def configFilename(self):
1122		""" return the config filename """
1123	<	return self.name()+'.cfg'
1123	>	# FUTURE: Can remove cfg mode for CMSSW >= 2_1_x
1124	>	if (self.CMSSW_major >= 2 and self.CMSSW_minor >= 1) or (self.CMSSW_major >= 3):
1125	>	return self.name()+'.py'
1126	>	else:
1127	>	return self.name()+'.cfg'
1128
1129		def wsSetupCMSOSGEnvironment_(self):
1130		"""
1131		Returns part of a job script which is prepares
1132		the execution environment and which is common for all CMS jobs.
1133		"""
1134	<	txt = '    echo ">>> setup CMS OSG environment:"\n'
1134	>	txt = '\n#Written by cms_cmssw::wsSetupCMSOSGEnvironment_\n'
1135	>	txt += '    echo ">>> setup CMS OSG environment:"\n'
1136		txt += '    echo "set SCRAM ARCH to ' + self.executable_arch + '"\n'
1137		txt += '    export SCRAM_ARCH='+self.executable_arch+'\n'
1138		txt += '    echo "SCRAM_ARCH = $SCRAM_ARCH"\n'
#	Line 1245 | Line 1140 | class Cmssw(JobType):
1140		txt += '      # Use $OSG_APP/cmssoft/cms/cmsset_default.sh to setup cms software\n'
1141		txt += '        source $OSG_APP/cmssoft/cms/cmsset_default.sh '+self.version+'\n'
1142		txt += '    else\n'
1143	<	txt += '        echo "SET_CMS_ENV 10020 ==> ERROR $OSG_APP/cmssoft/cms/cmsset_default.sh file not found"\n'
1144	<	txt += '        echo "JOB_EXIT_STATUS = 10020"\n'
1145	<	txt += '        echo "JobExitCode=10020" | tee -a $RUNTIME_AREA/$repo\n'
1251	<	txt += '        dumpStatus $RUNTIME_AREA/$repo\n'
1252	<	txt += '\n'
1253	<	txt += '        cd $RUNTIME_AREA\n'
1254	<	txt += '        echo ">>> current directory (RUNTIME_AREA): $RUNTIME_AREA"\n'
1255	<	txt += '        echo ">>> Remove working directory: $WORKING_DIR"\n'
1256	<	txt += '        /bin/rm -rf $WORKING_DIR\n'
1257	<	txt += '        if [ -d $WORKING_DIR ] ;then\n'
1258	<	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'
1259	<	txt += '            echo "JOB_EXIT_STATUS = 10017"\n'
1260	<	txt += '            echo "JobExitCode=10017" | tee -a $RUNTIME_AREA/$repo\n'
1261	<	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
1262	<	txt += '        fi\n'
1263	<	txt += '\n'
1264	<	txt += '        exit 1\n'
1143	>	txt += '        echo "ERROR ==> $OSG_APP/cmssoft/cms/cmsset_default.sh file not found"\n'
1144	>	txt += '        job_exit_code=10020\n'
1145	>	txt += '        func_exit\n'
1146		txt += '    fi\n'
1147		txt += '\n'
1148	<	txt += '    echo "SET_CMS_ENV 0 ==> setup cms environment ok"\n'
1148	>	txt += '    echo "==> setup cms environment ok"\n'
1149		txt += '    echo "SCRAM_ARCH = $SCRAM_ARCH"\n'
1150
1151		return txt
1152
1272	–	### OLI_DANIELE
1153		def wsSetupCMSLCGEnvironment_(self):
1154		"""
1155		Returns part of a job script which is prepares
1156		the execution environment and which is common for all CMS jobs.
1157		"""
1158	<	txt = '    echo ">>> setup CMS LCG environment:"\n'
1158	>	txt = '\n#Written by cms_cmssw::wsSetupCMSLCGEnvironment_\n'
1159	>	txt += '    echo ">>> setup CMS LCG environment:"\n'
1160		txt += '    echo "set SCRAM ARCH and BUILD_ARCH to ' + self.executable_arch + ' ###"\n'
1161		txt += '    export SCRAM_ARCH='+self.executable_arch+'\n'
1162		txt += '    export BUILD_ARCH='+self.executable_arch+'\n'
1163		txt += '    if [ ! $VO_CMS_SW_DIR ] ;then\n'
1164	<	txt += '        echo "SET_CMS_ENV 10031 ==> ERROR CMS software dir not found on WN `hostname`"\n'
1165	<	txt += '        echo "JOB_EXIT_STATUS = 10031" \n'
1166	<	txt += '        echo "JobExitCode=10031" | tee -a $RUNTIME_AREA/$repo\n'
1286	<	txt += '        dumpStatus $RUNTIME_AREA/$repo\n'
1287	<	txt += '        exit 1\n'
1164	>	txt += '        echo "ERROR ==> CMS software dir not found on WN `hostname`"\n'
1165	>	txt += '        job_exit_code=10031\n'
1166	>	txt += '        func_exit\n'
1167		txt += '    else\n'
1168		txt += '        echo "Sourcing environment... "\n'
1169		txt += '        if [ ! -s $VO_CMS_SW_DIR/cmsset_default.sh ] ;then\n'
1170	<	txt += '            echo "SET_CMS_ENV 10020 ==> ERROR cmsset_default.sh file not found into dir $VO_CMS_SW_DIR"\n'
1171	<	txt += '            echo "JOB_EXIT_STATUS = 10020"\n'
1172	<	txt += '            echo "JobExitCode=10020" | tee -a $RUNTIME_AREA/$repo\n'
1294	<	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
1295	<	txt += '            exit 1\n'
1170	>	txt += '            echo "ERROR ==> cmsset_default.sh file not found into dir $VO_CMS_SW_DIR"\n'
1171	>	txt += '            job_exit_code=10020\n'
1172	>	txt += '            func_exit\n'
1173		txt += '        fi\n'
1174		txt += '        echo "sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
1175		txt += '        source $VO_CMS_SW_DIR/cmsset_default.sh\n'
1176		txt += '        result=$?\n'
1177		txt += '        if [ $result -ne 0 ]; then\n'
1178	<	txt += '            echo "SET_CMS_ENV 10032 ==> ERROR problem sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
1179	<	txt += '            echo "JOB_EXIT_STATUS = 10032"\n'
1180	<	txt += '            echo "JobExitCode=10032" | tee -a $RUNTIME_AREA/$repo\n'
1304	<	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
1305	<	txt += '            exit 1\n'
1178	>	txt += '            echo "ERROR ==> problem sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
1179	>	txt += '            job_exit_code=10032\n'
1180	>	txt += '            func_exit\n'
1181		txt += '        fi\n'
1182		txt += '    fi\n'
1183		txt += '    \n'
1184	<	txt += '    echo "SET_CMS_ENV 0 ==> setup cms environment ok"\n'
1184	>	txt += '    echo "==> setup cms environment ok"\n'
1185		return txt
1186
1312	–	### FEDE FOR DBS OUTPUT PUBLICATION
1187		def modifyReport(self, nj):
1188		"""
1189		insert the part of the script that modifies the FrameworkJob Report
1190		"""
1191	<
1192	<	txt = ''
1319	<	try:
1320	<	publish_data = int(self.cfg_params['USER.publish_data'])
1321	<	except KeyError:
1322	<	publish_data = 0
1191	>	txt = '\n#Written by cms_cmssw::modifyReport\n'
1192	>	publish_data = int(self.cfg_params.get('USER.publish_data',0))
1193		if (publish_data == 1):
1324	–	txt += 'echo ">>> Modify Job Report:" \n'
1325	–	################ FEDE FOR DBS2 #############################################
1326	–	txt += 'chmod a+x $SOFTWARE_DIR/ProdAgentApi/FwkJobRep/ModifyJobReport.py\n'
1327	–	#############################################################################
1328	–
1329	–	txt += 'if [ -z "$SE" ]; then\n'
1330	–	txt += '    SE="" \n'
1331	–	txt += 'fi \n'
1332	–	txt += 'if [ -z "$SE_PATH" ]; then\n'
1333	–	txt += '    SE_PATH="" \n'
1334	–	txt += 'fi \n'
1335	–	txt += 'echo "SE = $SE"\n'
1336	–	txt += 'echo "SE_PATH = $SE_PATH"\n'
1337	–
1194		processedDataset = self.cfg_params['USER.publish_data_name']
1195	<	txt += 'ProcessedDataset='+processedDataset+'\n'
1196	<	#### LFN=/store/user/<user>/processedDataset_PSETHASH
1197	<	txt += 'if [ "$SE_PATH" == "" ]; then\n'
1198	<	#### FEDE: added slash in LFN ##############
1195	>	if (self.primaryDataset == 'null'):
1196	>	self.primaryDataset = processedDataset
1197	>	if (common.scheduler.name().upper() == "CAF" or common.scheduler.name().upper() == "LSF"):
1198	>	### FEDE FOR NEW LFN ###
1199	>	LFNBaseName = LFNBase(self.primaryDataset, processedDataset, LocalUser=True)
1200	>	self.user = getUserName(LocalUser=True)
1201	>	########################
1202	>	else :
1203	>	### FEDE FOR NEW LFN ###
1204	>	LFNBaseName = LFNBase(self.primaryDataset, processedDataset)
1205	>	self.user = getUserName()
1206	>	########################
1207	>
1208	>	txt += 'if [ $copy_exit_status -eq 0 ]; then\n'
1209	>	### FEDE FOR NEW LFN ###
1210	>	#txt += '    FOR_LFN=%s_${PSETHASH}/\n'%(LFNBaseName)
1211	>	txt += '    FOR_LFN=%s/${PSETHASH}/\n'%(LFNBaseName)
1212	>	########################
1213	>	txt += 'else\n'
1214		txt += '    FOR_LFN=/copy_problems/ \n'
1215	<	txt += 'else \n'
1216	<	txt += '    tmp=`echo $SE_PATH | awk -F \'store\' \'{print$2}\'` \n'
1217	<	#####  FEDE TO BE CHANGED, BECAUSE STORE IS HARDCODED!!!! ########
1218	<	txt += '    FOR_LFN=/store$tmp \n'
1219	<	txt += 'fi \n'
1215	>	txt += '    SE=""\n'
1216	>	txt += '    SE_PATH=""\n'
1217	>	txt += 'fi\n'
1218	>
1219	>	txt += 'echo ">>> Modify Job Report:" \n'
1220	>	txt += 'chmod a+x $RUNTIME_AREA/ProdCommon/FwkJobRep/ModifyJobReport.py\n'
1221	>	txt += 'ProcessedDataset='+processedDataset+'\n'
1222		txt += 'echo "ProcessedDataset = $ProcessedDataset"\n'
1223	+	txt += 'echo "SE = $SE"\n'
1224	+	txt += 'echo "SE_PATH = $SE_PATH"\n'
1225		txt += 'echo "FOR_LFN = $FOR_LFN" \n'
1226		txt += 'echo "CMSSW_VERSION = $CMSSW_VERSION"\n\n'
1227	<	#txt += 'echo "$RUNTIME_AREA/'+self.version+'/ProdAgentApi/FwkJobRep/ModifyJobReport.py crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH"\n'
1228	<	txt += 'echo "$SOFTWARE_DIR/ProdAgentApi/FwkJobRep/ModifyJobReport.py crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH"\n'
1229	<	txt += '$SOFTWARE_DIR/ProdAgentApi/FwkJobRep/ModifyJobReport.py crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH\n'
1230	<	#txt += '$RUNTIME_AREA/'+self.version+'/ProdAgentApi/FwkJobRep/ModifyJobReport.py crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH\n'
1356	<
1227	>	### FEDE FOR NEW LFN ###
1228	>	txt += 'echo "$RUNTIME_AREA/ProdCommon/FwkJobRep/ModifyJobReport.py $RUNTIME_AREA/crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier ' + self.user + '-$ProcessedDataset-$PSETHASH $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH"\n'
1229	>	txt += '$RUNTIME_AREA/ProdCommon/FwkJobRep/ModifyJobReport.py $RUNTIME_AREA/crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier ' + self.user + '-$ProcessedDataset-$PSETHASH $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH\n'
1230	>	########################
1231		txt += 'modifyReport_result=$?\n'
1358	–	txt += 'echo modifyReport_result = $modifyReport_result\n'
1232		txt += 'if [ $modifyReport_result -ne 0 ]; then\n'
1233	<	txt += '    exit_status=1\n'
1234	<	txt += '    echo "ERROR: Problem with ModifyJobReport"\n'
1233	>	txt += '    modifyReport_result=70500\n'
1234	>	txt += '    job_exit_code=$modifyReport_result\n'
1235	>	txt += '    echo "ModifyReportResult=$modifyReport_result" | tee -a $RUNTIME_AREA/$repo\n'
1236	>	txt += '    echo "WARNING: Problem with ModifyJobReport"\n'
1237		txt += 'else\n'
1238	<	txt += '    mv NewFrameworkJobReport.xml crab_fjr_$NJob.xml\n'
1238	>	txt += '    mv NewFrameworkJobReport.xml $RUNTIME_AREA/crab_fjr_$NJob.xml\n'
1239		txt += 'fi\n'
1365	–	else:
1366	–	txt += 'echo "no data publication required"\n'
1240		return txt
1241
1242	<	def cleanEnv(self):
1243	<	txt = ''
1244	<	txt += 'if [ $middleware == OSG ]; then\n'
1245	<	txt += '    cd $RUNTIME_AREA\n'
1246	<	txt += '    echo ">>> current directory (RUNTIME_AREA): $RUNTIME_AREA"\n'
1247	<	txt += '    echo ">>> Remove working directory: $WORKING_DIR"\n'
1248	<	txt += '    /bin/rm -rf $WORKING_DIR\n'
1249	<	txt += '    if [ -d $WORKING_DIR ] ;then\n'
1250	<	txt += '        echo "SET_EXE 60999 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after cleanup of WN"\n'
1251	<	txt += '        echo "JOB_EXIT_STATUS = 60999"\n'
1252	<	txt += '        echo "JobExitCode=60999" | tee -a $RUNTIME_AREA/$repo\n'
1253	<	txt += '        dumpStatus $RUNTIME_AREA/$repo\n'
1242	>	def wsParseFJR(self):
1243	>	"""
1244	>	Parse the FrameworkJobReport to obtain useful infos
1245	>	"""
1246	>	txt = '\n#Written by cms_cmssw::wsParseFJR\n'
1247	>	txt += 'echo ">>> Parse FrameworkJobReport crab_fjr.xml"\n'
1248	>	txt += 'if [ -s $RUNTIME_AREA/crab_fjr_$NJob.xml ]; then\n'
1249	>	txt += '    if [ -s $RUNTIME_AREA/parseCrabFjr.py ]; then\n'
1250	>	txt += '        cmd_out=`python $RUNTIME_AREA/parseCrabFjr.py --input $RUNTIME_AREA/crab_fjr_$NJob.xml --dashboard $MonitorID,$MonitorJobID '+self.debugWrap+'`\n'
1251	>	if self.debug_wrapper :
1252	>	txt += '        echo "Result of parsing the FrameworkJobReport crab_fjr.xml: $cmd_out"\n'
1253	>	txt += '        executable_exit_status=`python $RUNTIME_AREA/parseCrabFjr.py --input $RUNTIME_AREA/crab_fjr_$NJob.xml --exitcode`\n'
1254	>	txt += '        if [ $executable_exit_status -eq 50115 ];then\n'
1255	>	txt += '            echo ">>> crab_fjr.xml contents: "\n'
1256	>	txt += '            cat $RUNTIME_AREA/crab_fjr_$NJob.xml\n'
1257	>	txt += '            echo "Wrong FrameworkJobReport --> does not contain useful info. ExitStatus: $executable_exit_status"\n'
1258	>	txt += '        elif [ $executable_exit_status -eq -999 ];then\n'
1259	>	txt += '            echo "ExitStatus from FrameworkJobReport not available. not available. Using exit code of executable from command line."\n'
1260	>	txt += '        else\n'
1261	>	txt += '            echo "Extracted ExitStatus from FrameworkJobReport parsing output: $executable_exit_status"\n'
1262	>	txt += '        fi\n'
1263	>	txt += '    else\n'
1264	>	txt += '        echo "CRAB python script to parse CRAB FrameworkJobReport crab_fjr.xml is not available, using exit code of executable from command line."\n'
1265		txt += '    fi\n'
1266	+	#### Patch to check input data reading for CMSSW16x Hopefully we-ll remove it asap
1267	+
1268	+	if (self.datasetPath and not (self.dataset_pu or self.useParent) :
1269	+	# VERIFY PROCESSED DATA
1270	+	txt += '    if [ $executable_exit_status -eq 0 ];then\n'
1271	+	txt += '      echo ">>> Verify list of processed files:"\n'
1272	+	txt += '      echo $InputFiles |tr -d \'\\\\\' |tr \',\' \'\\n\'|tr -d \'"\' > input-files.txt\n'
1273	+	txt += '      python $RUNTIME_AREA/parseCrabFjr.py --input $RUNTIME_AREA/crab_fjr_$NJob.xml --lfn > processed-files.txt\n'
1274	+	txt += '      cat input-files.txt  | sort | uniq > tmp.txt\n'
1275	+	txt += '      mv tmp.txt input-files.txt\n'
1276	+	txt += '      echo "cat input-files.txt"\n'
1277	+	txt += '      echo "----------------------"\n'
1278	+	txt += '      cat input-files.txt\n'
1279	+	txt += '      cat processed-files.txt | sort | uniq > tmp.txt\n'
1280	+	txt += '      mv tmp.txt processed-files.txt\n'
1281	+	txt += '      echo "----------------------"\n'
1282	+	txt += '      echo "cat processed-files.txt"\n'
1283	+	txt += '      echo "----------------------"\n'
1284	+	txt += '      cat processed-files.txt\n'
1285	+	txt += '      echo "----------------------"\n'
1286	+	txt += '      diff -q input-files.txt processed-files.txt\n'
1287	+	txt += '      fileverify_status=$?\n'
1288	+	txt += '      if [ $fileverify_status -ne 0 ]; then\n'
1289	+	txt += '         executable_exit_status=30001\n'
1290	+	txt += '         echo "ERROR ==> not all input files processed"\n'
1291	+	txt += '         echo "      ==> list of processed files from crab_fjr.xml differs from list in pset.cfg"\n'
1292	+	txt += '         echo "      ==> diff input-files.txt processed-files.txt"\n'
1293	+	txt += '      fi\n'
1294	+	txt += '    fi\n'
1295	+	txt += '\n'
1296	+	txt += 'else\n'
1297	+	txt += '    echo "CRAB FrameworkJobReport crab_fjr.xml is not available, using exit code of executable from command line."\n'
1298		txt += 'fi\n'
1299		txt += '\n'
1300	+	txt += 'echo "ExeExitCode=$executable_exit_status" | tee -a $RUNTIME_AREA/$repo\n'
1301	+	txt += 'echo "EXECUTABLE_EXIT_STATUS = $executable_exit_status"\n'
1302	+	txt += 'job_exit_code=$executable_exit_status\n'
1303	+
1304		return txt
1305
1306		def setParam_(self, param, value):
#	Line 1389 | Line 1309 | class Cmssw(JobType):
1309		def getParams(self):
1310		return self._params
1311
1392	–	def setTaskid_(self):
1393	–	self._taskId = self.cfg_params['taskId']
1394	–
1395	–	def getTaskid(self):
1396	–	return self._taskId
1397	–
1312		def uniquelist(self, old):
1313		"""
1314		remove duplicates from a list
#	Line 1404 | Line 1318 | class Cmssw(JobType):
1318		nd[e]=0
1319		return nd.keys()
1320
1321	<
1408	<	def checkOut(self, limit):
1321	>	def outList(self):
1322		"""
1323		check the dimension of the output files
1324		"""
1325	<	txt = 'echo ">>> Starting output sandbox limit check :"\n'
1326	<	allOutFiles = ""
1325	>	txt = ''
1326	>	txt += 'echo ">>> list of expected files on output sandbox"\n'
1327		listOutFiles = []
1328	<	txt += 'stdoutFile=`ls | grep *stdout` \n'
1329	<	txt += 'stderrFile=`ls | grep *stderr` \n'
1330	<	for fileOut in (self.output_file+self.output_file_sandbox):
1331	<	if fileOut.find('crab_fjr') == -1:
1332	<	allOutFiles = allOutFiles + " " + self.numberFile_(fileOut, '$NJob') + " $stdoutFile $stderrFile"
1333	<	listOutFiles.append(self.numberFile_(fileOut, '$NJob'))
1334	<	txt += 'echo "OUTPUT files: '+str(allOutFiles)+'";\n'
1335	<	txt += 'ls -gGhrta;\n'
1336	<	txt += 'sum=0;\n'
1337	<	txt += 'for file in '+str(allOutFiles)+' ; do\n'
1338	<	txt += '    if [ -e $file ]; then\n'
1339	<	txt += '        tt=`ls -gGrta $file | awk \'{ print $3 }\'`\n'
1340	<	txt += '        sum=`expr $sum + $tt`\n'
1341	<	txt += '    else\n'
1342	<	txt += '        echo "WARNING: output file $file not found!"\n'
1430	<	txt += '    fi\n'
1431	<	txt += 'done\n'
1432	<	txt += 'echo "Total Output dimension: $sum";\n'
1433	<	txt += 'limit='+str(limit)+';\n'
1434	<	txt += 'echo "OUTPUT FILES LIMIT SET TO: $limit";\n'
1435	<	txt += 'if [ $limit -lt $sum ]; then\n'
1436	<	txt += '    echo "WARNING: output files have to big size - something will be lost;"\n'
1437	<	txt += '    echo "         checking the output file sizes..."\n'
1438	<	txt += '    tot=0;\n'
1439	<	txt += '    for filefile in '+str(allOutFiles)+' ; do\n'
1440	<	txt += '        dimFile=`ls -gGrta $filefile | awk \'{ print $3 }\';`\n'
1441	<	txt += '        tot=`expr $tot + $tt`;\n'
1442	<	txt += '        if [ $limit -lt $dimFile ]; then\n'
1443	<	txt += '            echo "deleting file: $filefile";\n'
1444	<	txt += '            rm -f $filefile\n'
1445	<	txt += '        elif [ $limit -lt $tot ]; then\n'
1446	<	txt += '            echo "deleting file: $filefile";\n'
1447	<	txt += '            rm -f $filefile\n'
1448	<	txt += '        else\n'
1449	<	txt += '            echo "saving file: $filefile"\n'
1450	<	txt += '        fi\n'
1451	<	txt += '    done\n'
1452	<
1453	<	txt += '    ls -agGhrt;\n'
1454	<	txt += '    echo "WARNING: output files are too big in dimension: can not put in the output_sandbox.";\n'
1455	<	txt += '    echo "JOB_EXIT_STATUS = 70000";\n'
1456	<	txt += '    exit_status=70000;\n'
1457	<	txt += 'else'
1458	<	txt += '    echo "Total Output dimension $sum is fine.";\n'
1459	<	txt += 'fi\n'
1460	<	txt += 'echo "Ending output sandbox limit check"\n'
1328	>	stdout = 'CMSSW_$NJob.stdout'
1329	>	stderr = 'CMSSW_$NJob.stderr'
1330	>	if (self.return_data == 1):
1331	>	for file in (self.output_file+self.output_file_sandbox):
1332	>	listOutFiles.append(numberFile(file, '$NJob'))
1333	>	listOutFiles.append(stdout)
1334	>	listOutFiles.append(stderr)
1335	>	else:
1336	>	for file in (self.output_file_sandbox):
1337	>	listOutFiles.append(numberFile(file, '$NJob'))
1338	>	listOutFiles.append(stdout)
1339	>	listOutFiles.append(stderr)
1340	>	txt += 'echo "output files: '+string.join(listOutFiles,' ')+'"\n'
1341	>	txt += 'filesToCheck="'+string.join(listOutFiles,' ')+'"\n'
1342	>	txt += 'export filesToCheck\n'
1343		return txt

Diff Legend

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