[ViewVC] Diff of: cvsroot/COMP/CRAB/python/cms

Comparing COMP/CRAB/python/cms_cmssw.py (file contents):
Revision 1.132 by ewv, Mon Oct 29 21:03:11 2007 UTC vs.
Revision 1.244 by spiga, Thu Sep 25 15:08:01 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
18	–
21		# init BlackWhiteListParser
22	<	self.blackWhiteListParser = BlackWhiteListParser(cfg_params)
22	>	self.blackWhiteListParser = SEBlackWhiteListParser(cfg_params)
23
24	<	try:
25	<	self.MaxTarBallSize = float(self.cfg_params['EDG.maxtarballsize'])
26	<	except KeyError:
27	<	self.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 35 \| Line 39 \| class Cmssw(JobType):
39		self.executable = ''
40		self.executable_arch = self.scram.getArch()
41		self.tgz_name = 'default.tgz'
38	–	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	+	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	<	#
49	<	# Try to block creation in case of arch/version mismatch
50	<	#
51	<
52	<	a = string.split(self.version, "_")
62	>	### collect Data cards
63
64	<	if int(a[1]) == 1 and (int(a[2]) < 5 and self.executable_arch.find('slc4') == 0):
65	<	msg = "Error: CMS does not support %s with %s architecture"%(self.version, self.executable_arch)
56	<	raise CrabException(msg)
57	<	if int(a[1]) == 1 and (int(a[2]) >= 5 and self.executable_arch.find('slc3') == 0):
58	<	msg = "Error: CMS does not support %s with %s architecture"%(self.version, self.executable_arch)
64	>	if not cfg_params.has_key('CMSSW.datasetpath'):
65	>	msg = "Error: datasetpath not defined "
66		raise CrabException(msg)
67
68	<	common.taskDB.setDict('codeVersion',self.version)
69	<	self.setParam_('application', self.version)
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	<	### collect Data cards
72	<
66	<	## get DBS mode
67	<	try:
68	<	self.use_dbs_1 = int(self.cfg_params['CMSSW.use_dbs_1'])
69	<	except KeyError:
70	<	self.use_dbs_1 = 0
71	>	tmp = cfg_params['CMSSW.datasetpath']
72	>	log.debug(6, "CMSSW::CMSSW(): datasetPath = "+tmp)
73
74	<	try:
73	<	tmp = cfg_params['CMSSW.datasetpath']
74	<	log.debug(6, "CMSSW::CMSSW(): datasetPath = "+tmp)
75	<	if string.lower(tmp)=='none':
76	<	self.datasetPath = None
77	<	self.selectNoInput = 1
78	<	else:
79	<	self.datasetPath = tmp
80	<	self.selectNoInput = 0
81	<	except KeyError:
74	>	if tmp =='':
75		msg = "Error: datasetpath not defined "
76		raise CrabException(msg)
77	<
78	<	# ML monitoring
79	<	# split dataset path style: /PreProdR3Minbias/SIM/GEN-SIM
87	<	if not self.datasetPath:
88	<	self.setParam_('dataset', 'None')
89	<	self.setParam_('owner', 'None')
77	>	elif string.lower(tmp)=='none':
78	>	self.datasetPath = None
79	>	self.selectNoInput = 1
80		else:
81	<	try:
82	<	datasetpath_split = self.datasetPath.split("/")
93	<	# standard style
94	<	self.setParam_('datasetFull', self.datasetPath)
95	<	if self.use_dbs_1 == 1 :
96	<	self.setParam_('dataset', datasetpath_split[1])
97	<	self.setParam_('owner', datasetpath_split[-1])
98	<	else:
99	<	self.setParam_('dataset', datasetpath_split[1])
100	<	self.setParam_('owner', datasetpath_split[2])
101	<	except:
102	<	self.setParam_('dataset', self.datasetPath)
103	<	self.setParam_('owner', self.datasetPath)
104	<
105	<	self.setTaskid_()
106	<	self.setParam_('taskId', self.cfg_params['taskId'])
81	>	self.datasetPath = tmp
82	>	self.selectNoInput = 0
83
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	<	try:
90	<	self.executable = cfg_params['CMSSW.executable']
113	<	self.setParam_('exe', self.executable)
114	<	log.debug(6, "CMSSW::CMSSW(): executable = "+self.executable)
115	<	msg = "Default executable cmsRun overridden. Switch to " + self.executable
116	<	log.debug(3,msg)
117	<	except KeyError:
118	<	self.executable = 'cmsRun'
119	<	self.setParam_('exe', self.executable)
120	<	msg = "User executable not defined. Use cmsRun"
121	<	log.debug(3,msg)
122	<	pass
89	>	self.executable = cfg_params.get('CMSSW.executable','cmsRun')
90	>	log.debug(6, "CMSSW::CMSSW(): executable = "+self.executable)
91
92	<	try:
125	<	self.pset = cfg_params['CMSSW.pset']
126	<	log.debug(6, "Cmssw::Cmssw(): PSet file = "+self.pset)
127	<	if self.pset.lower() != 'none' :
128	<	if (not os.path.exists(self.pset)):
129	<	raise CrabException("User defined PSet file "+self.pset+" does not exist")
130	<	else:
131	<	self.pset = None
132	<	except KeyError:
92	>	if not cfg_params.has_key('CMSSW.pset'):
93		raise CrabException("PSet file missing. Cannot run cmsRun ")
94	+	self.pset = cfg_params['CMSSW.pset']
95	+	log.debug(6, "Cmssw::Cmssw(): PSet file = "+self.pset)
96	+	if self.pset.lower() != 'none' :
97	+	if (not os.path.exists(self.pset)):
98	+	raise CrabException("User defined PSet file "+self.pset+" does not exist")
99	+	else:
100	+	self.pset = None
101
102		# output files
103		## stuff which must be returned always via sandbox
#	Line 140 \| 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	<	try:
111	<	self.output_file = []
112	<	tmp = cfg_params['CMSSW.output_file']
113	<	if tmp != '':
114	<	tmpOutFiles = string.split(cfg_params['CMSSW.output_file'],',')
115	<	log.debug(7, 'cmssw::cmssw(): output files '+str(tmpOutFiles))
116	<	for tmp in tmpOutFiles:
117	<	tmp=string.strip(tmp)
151	<	self.output_file.append(tmp)
152	<	pass
153	<	else:
154	<	log.message("No output file defined: only stdout/err and the CRAB Framework Job Report will be available\n")
155	<	pass
156	<	pass
157	<	except KeyError:
158	<	log.message("No output file defined: only stdout/err and the CRAB Framework Job Report will be available\n")
159	<	pass
110	>	outfileflag = False
111	>	self.output_file = []
112	>	tmp = cfg_params.get('CMSSW.output_file',None)
113	>	if tmp :
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	<	try:
121	<	self.scriptExe = cfg_params['USER.script_exe']
122	<	if self.scriptExe != '':
123	<	if not os.path.isfile(self.scriptExe):
124	<	msg ="ERROR. file "+self.scriptExe+" not found"
125	<	raise CrabException(msg)
168	<	self.additional_inbox_files.append(string.strip(self.scriptExe))
169	<	except KeyError:
170	<	self.scriptExe = ''
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))
126
172	–	#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	<	try:
135	>	if cfg_params.has_key('USER.additional_input_files'):
136		tmpAddFiles = string.split(cfg_params['USER.additional_input_files'],',')
137		for tmp in tmpAddFiles:
138		tmp = string.strip(tmp)
#	Line 192 \| Line 149 \| class Cmssw(JobType):
149		if not os.path.exists(file):
150		raise CrabException("Additional input file not found: "+file)
151		pass
195	–	# fname = string.split(file, '/')[-1]
196	–	# storedFile = common.work_space.pathForTgz()+'share/'+fname
197	–	# shutil.copyfile(file, storedFile)
152		self.additional_inbox_files.append(string.strip(file))
153		pass
154		pass
155		common.logger.debug(5,"Additional input files: "+str(self.additional_inbox_files))
156	<	except KeyError:
203	<	pass
204	<
205	<	# files per job
206	<	try:
207	<	if (cfg_params['CMSSW.files_per_jobs']):
208	<	raise CrabException("files_per_jobs no longer supported. Quitting.")
209	<	except KeyError:
210	<	pass
156	>	pass
157
158		## Events per job
159	<	try:
159	>	if cfg_params.has_key('CMSSW.events_per_job'):
160		self.eventsPerJob =int( cfg_params['CMSSW.events_per_job'])
161		self.selectEventsPerJob = 1
162	<	except KeyError:
162	>	else:
163		self.eventsPerJob = -1
164		self.selectEventsPerJob = 0
165
166		## number of jobs
167	<	try:
167	>	if cfg_params.has_key('CMSSW.number_of_jobs'):
168		self.theNumberOfJobs =int( cfg_params['CMSSW.number_of_jobs'])
169		self.selectNumberOfJobs = 1
170	<	except KeyError:
170	>	else:
171		self.theNumberOfJobs = 0
172		self.selectNumberOfJobs = 0
173
174	<	try:
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	<	except KeyError:
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 241 \| Line 191 \| class Cmssw(JobType):
191		msg = 'Must specify number_of_jobs.'
192		raise CrabException(msg)
193
194	<	## source seed for pythia
195	<	try:
196	<	self.sourceSeed = int(cfg_params['CMSSW.pythia_seed'])
197	<	except KeyError:
198	<	self.sourceSeed = None
199	<	common.logger.debug(5,"No seed given")
200	<
201	<	try:
202	<	self.sourceSeedVtx = int(cfg_params['CMSSW.vtx_seed'])
203	<	except KeyError:
204	<	self.sourceSeedVtx = None
205	<	common.logger.debug(5,"No vertex seed given")
206	<
207	<	try:
208	<	self.sourceSeedG4 = int(cfg_params['CMSSW.g4_seed'])
209	<	except KeyError:
210	<	self.sourceSeedG4 = None
211	<	common.logger.debug(5,"No g4 sim hits seed given")
194	>	## New method of dealing with seeds
195	>	self.incrementSeeds = []
196	>	self.preserveSeeds = []
197	>	if cfg_params.has_key('CMSSW.preserve_seeds'):
198	>	tmpList = cfg_params['CMSSW.preserve_seeds'].split(',')
199	>	for tmp in tmpList:
200	>	tmp.strip()
201	>	self.preserveSeeds.append(tmp)
202	>	if cfg_params.has_key('CMSSW.increment_seeds'):
203	>	tmpList = cfg_params['CMSSW.increment_seeds'].split(',')
204	>	for tmp in tmpList:
205	>	tmp.strip()
206	>	self.incrementSeeds.append(tmp)
207	>
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	>	self.sourceSeedG4 = cfg_params.get('CMSSW.g4_seed',None)
212	>	self.sourceSeedMix = cfg_params.get('CMSSW.mix_seed',None)
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	<	try:
264	<	self.sourceSeedMix = int(cfg_params['CMSSW.mix_seed'])
265	<	except KeyError:
266	<	self.sourceSeedMix = None
267	<	common.logger.debug(5,"No mix seed given")
217	>	self.firstRun = cfg_params.get('CMSSW.first_run',None)
218
219	<	try:
220	<	self.firstRun = int(cfg_params['CMSSW.first_run'])
221	<	except KeyError:
272	<	self.firstRun = None
273	<	common.logger.debug(5,"No first run given")
274	<	if self.pset != None: #CarlosDaniele
275	<	import PsetManipulator as pp
276	<	PsetEdit = pp.PsetManipulator(self.pset) #Daniele Pset
219	>	# Copy/return
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
223		#DBSDLS-start
224		## Initialize the variables that are extracted from DBS/DLS and needed in other places of the code
#	Line 287 \| Line 232 \| class Cmssw(JobType):
232		blockSites = self.DataDiscoveryAndLocation(cfg_params)
233		#DBSDLS-end
234
290	–	self.tgzNameWithPath = self.getTarBall(self.executable)
291	–
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	<	if (self.datasetPath): # standard job
250	<	# allow to processa a fraction of events in a file
251	<	PsetEdit.inputModule("INPUTFILE")
252	<	PsetEdit.maxEvent(0)
253	<	PsetEdit.skipEvent(0)
254	<	else: # pythia like job
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	<	if (self.firstRun):
257	<	PsetEdit.pythiaFirstRun(0) #First Run
258	<	if (self.sourceSeed) :
259	<	PsetEdit.pythiaSeed(0)
260	<	if (self.sourceSeedVtx) :
261	<	PsetEdit.vtxSeed(0)
262	<	if (self.sourceSeedG4) :
263	<	PsetEdit.g4Seed(0)
264	<	if (self.sourceSeedMix) :
265	<	PsetEdit.mixSeed(0)
266	<	# add FrameworkJobReport to parameter-set
267	<	PsetEdit.addCrabFJR(self.fjrFileName)
268	<	PsetEdit.psetWriter(self.configFilename())
269	<	except:
270	<	msg='Error while manipuliating ParameterSet: exiting...'
271	<	raise CrabException(msg)
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
288		import DataDiscovery
331	–	import DataDiscovery_DBS2
289		import DataLocation
290		common.logger.debug(10,"CMSSW::DataDiscoveryAndLocation()")
291
#	Line 337 \| Line 294 \| class Cmssw(JobType):
294		## Contact the DBS
295		common.logger.message("Contacting Data Discovery Services ...")
296		try:
297	<
341	<	if self.use_dbs_1 == 1 :
342	<	self.pubdata=DataDiscovery.DataDiscovery(datasetPath, cfg_params)
343	<	else :
344	<	self.pubdata=DataDiscovery_DBS2.DataDiscovery_DBS2(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 353 \| Line 306 \| class Cmssw(JobType):
306		except DataDiscovery.DataDiscoveryError, ex:
307		msg = 'ERROR ***: failed Data Discovery in DBS : %s'%ex.getErrorMessage()
308		raise CrabException(msg)
356	–	except DataDiscovery_DBS2.NotExistingDatasetError_DBS2, ex :
357	–	msg = 'ERROR ***: failed Data Discovery in DBS : %s'%ex.getErrorMessage()
358	–	raise CrabException(msg)
359	–	except DataDiscovery_DBS2.NoDataTierinProvenanceError_DBS2, ex :
360	–	msg = 'ERROR ***: failed Data Discovery in DBS : %s'%ex.getErrorMessage()
361	–	raise CrabException(msg)
362	–	except DataDiscovery_DBS2.DataDiscoveryError_DBS2, ex:
363	–	msg = 'ERROR ***: failed Data Discovery in DBS : %s'%ex.getErrorMessage()
364	–	raise CrabException(msg)
309
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 442 \| Line 387 \| class Cmssw(JobType):
387		else :
388		totalNumberOfJobs = self.ncjobs
389
445	–
390		blocks = blockSites.keys()
391		blockCount = 0
392		# Backup variable in case self.maxEvents counted events in a non-included block
#	Line 483 \| 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 497 \| 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 517 \| 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 529 \| 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 540 \| 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 549 \| 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 563 \| 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	<	parString = ""
532	<	parString += '\\\"' + 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)
536		pass # END while (iterate over blocks in the dataset)
#	Line 584 \| 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 604 \| Line 571 \| class Cmssw(JobType):
571		for range_jobs in noSiteBlock:
572		msg += str(range_jobs) + virgola
573		msg += '\n will not be submitted and this block of data can not be analyzed!\n'
574	+	if self.cfg_params.has_key('EDG.se_white_list'):
575	+	msg += 'WARNING: SE White List: '+self.cfg_params['EDG.se_white_list']+'\n'
576	+	msg += '(Hint: By whitelisting you force the job to run at this particular site(s).\n'
577	+	msg += 'Please check if the dataset is available at this site!)\n'
578	+	if self.cfg_params.has_key('EDG.ce_white_list'):
579	+	msg += 'WARNING: CE White List: '+self.cfg_params['EDG.ce_white_list']+'\n'
580	+	msg += '(Hint: By whitelisting you force the job to run at this particular site(s).\n'
581	+	msg += 'Please check if the dataset is available at this site!)\n'
582	+
583		common.logger.message(msg)
584
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 652 \| Line 817 \| class Cmssw(JobType):
817		self.list_of_args = []
818		for i in range(self.total_number_of_jobs):
819		## Since there is no input, any site is good
655	–	# self.jobDestination.append(["Any"])
820		self.jobDestination.append([""]) #must be empty to write correctly the xml
821		args=[]
822		if (self.firstRun):
823	<	## pythia first run
660	<	#self.list_of_args.append([(str(self.firstRun)+str(i))])
823	>	## pythia first run
824		args.append(str(self.firstRun)+str(i))
662	–	else:
663	–	## no first run
664	–	#self.list_of_args.append([str(i)])
665	–	args.append(str(i))
666	–	if (self.sourceSeed):
667	–	args.append(str(self.sourceSeed)+str(i))
668	–	if (self.sourceSeedVtx):
669	–	## + vtx random seed
670	–	args.append(str(self.sourceSeedVtx)+str(i))
671	–	if (self.sourceSeedG4):
672	–	## + G4 random seed
673	–	args.append(str(self.sourceSeedG4)+str(i))
674	–	if (self.sourceSeedMix):
675	–	## + Mix random seed
676	–	args.append(str(self.sourceSeedMix)+str(i))
677	–	pass
678	–	pass
825		self.list_of_args.append(args)
680	–	pass
681	–
682	–	# print self.list_of_args
826
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 700 \| 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):
703	–	## Since there is no input, any site is good
704	–	# self.jobDestination.append(["Any"])
846		self.jobDestination.append([""])
706	–	## 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
712	–	common.jobDB.load()
713	–	#### Fabio
852		njobs = self.total_number_of_jobs
853		arglist = self.list_of_args
854		# create the empty structure
855		for i in range(njobs):
856		jobParams.append("")
857
858	<	for job in range(njobs):
859	<	jobParams[job] = arglist[job]
860	<	# print str(arglist[job])
861	<	# print jobParams[job]
862	<	common.jobDB.setArguments(job, jobParams[job])
863	<	common.logger.debug(5,"Job "+str(job)+" Destination: "+str(self.jobDestination[job]))
864	<	common.jobDB.setDestination(job, self.jobDestination[job])
858	>	listID=[]
859	>	listField=[]
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[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[id])
874	>	common.logger.debug(5,msg)
875	>	common._db.updateJob_(listID,listField)
876	>	self.argsList = (len(jobParams[0])+1)
877
728	–	common.jobDB.save()
878		return
879
731	–	def getJobTypeArguments(self, nj, sched):
732	–	result = ''
733	–	for i in common.jobDB.arguments(nj):
734	–	result=result+str(i)+" "
735	–	return result
736	–
880		def numberOfJobs(self):
738	–	# 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	<
746	<	# if it exist, just return it
747	<	#
748	<	# Marco. Let's start to use relative path for Boss XML files
749	<	#
750	<	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 760 \| Line 897 \| class Cmssw(JobType):
897
898		# First of all declare the user Scram area
899		swArea = self.scram.getSWArea_()
763	–	#print "swArea = ", swArea
764	–	# swVersion = self.scram.getSWVersion()
765	–	# print "swVersion = ", swVersion
900		swReleaseTop = self.scram.getReleaseTop_()
767	–	#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 809 \| 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")
951	<
952	<	## Add ProdAgent dir to tar
953	<	paDir = 'ProdAgentApi'
954	<	pa = os.environ['CRABDIR'] + '/' + 'ProdAgentApi'
955	<	if os.path.isdir(pa):
956	<	tar.add(pa,paDir)
957	<
958	<	### FEDE FOR DBS PUBLICATION
959	<	## Add PRODCOMMON dir to tar
960	<	prodcommonDir = 'ProdCommon'
961	<	prodcommonPath = os.environ['CRABDIR'] + '/' + 'ProdCommon'
962	<	if os.path.isdir(prodcommonPath):
963	<	tar.add(prodcommonPath,prodcommonDir)
964	<	#############################
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 = './'
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.MaxTarBallSize10241024 ) :
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
843	–	self.MLtgzfile = common.work_space.pathForTgz()+'share/MLfiles.tgz'
844	–	try:
845	–	tar = tarfile.open(self.MLtgzfile, "w:gz")
846	–	path=os.environ['CRABDIR'] + '/python/'
847	–	for file in ['report.py', 'DashboardAPI.py', 'Logger.py', 'ProcInfo.py', 'apmon.py', 'parseCrabFjr.py']:
848	–	tar.add(path+file,file)
849	–	common.logger.debug(5,"Files added to "+self.MLtgzfile+" : "+str(tar.getnames()))
850	–	tar.close()
851	–	except :
852	–	raise CrabException('Could not create ML files tar-ball')
853	–
854	–	return
855	–
856	–	def additionalInputFileTgz(self):
857	–	"""
858	–	Put all additional files into a tar ball and return its name
859	–	"""
860	–	import tarfile
861	–	tarName= common.work_space.pathForTgz()+'share/'+self.additional_tgz_name
862	–	tar = tarfile.open(tarName, "w:gz")
863	–	for file in self.additional_inbox_files:
864	–	tar.add(file,string.split(file,'/')[-1])
865	–	common.logger.debug(5,"Files added to "+self.additional_tgz_name+" : "+str(tar.getnames()))
866	–	tar.close()
867	–	return tarName
1015
1016	<	def wsSetupEnvironment(self, nj):
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 = ''
1027	<
877	<	## OLI_Daniele at this level middleware already known
878	<
1026	>	txt = '\n#Written by cms_cmssw::wsSetupEnvironment\n'
1027	>	txt += 'echo ">>> setup environment"\n'
1028		txt += 'if [ $middleware == LCG ]; then \n'
880	–	txt += ' echo "### First set SCRAM ARCH and BUILD_ARCH to ' + self.executable_arch + ' ###"\n'
881	–	txt += ' export SCRAM_ARCH='+self.executable_arch+'\n'
882	–	txt += ' export BUILD_ARCH='+self.executable_arch+'\n'
1029		txt += self.wsSetupCMSLCGEnvironment_()
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'
1033	<	txt += ' echo "SET_CMS_ENV 10016 ==> OSG $WORKING_DIR could not be created on WN `hostname`"\n'
1034	<	txt += ' echo "JOB_EXIT_STATUS = 10016"\n'
1035	<	txt += ' echo "JobExitCode=10016" \| tee -a $RUNTIME_AREA/$repo\n'
890	<	txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
891	<	txt += ' rm -f $RUNTIME_AREA/$repo \n'
892	<	txt += ' echo "MonitorJobID=`echo $MonitorJobID`" \| tee -a $RUNTIME_AREA/$repo \n'
893	<	txt += ' echo "MonitorID=`echo $MonitorID`" \| tee -a $RUNTIME_AREA/$repo\n'
894	<	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'
1036		txt += ' fi\n'
1037	<	txt += ' echo "Created working directory: $WORKING_DIR"\n'
1037	>	txt += ' echo ">>> Created working directory: $WORKING_DIR"\n'
1038		txt += '\n'
1039		txt += ' echo "Change to working directory: $WORKING_DIR"\n'
1040		txt += ' cd $WORKING_DIR\n'
1041	+	txt += ' echo ">>> current directory (WORKING_DIR): $WORKING_DIR"\n'
1042		txt += self.wsSetupCMSOSGEnvironment_()
901	–	txt += ' echo "### Set SCRAM ARCH to ' + self.executable_arch + ' ###"\n'
902	–	txt += ' export SCRAM_ARCH='+self.executable_arch+'\n'
1043		txt += 'fi\n'
1044
1045		# Prepare JobType-specific part
1046		scram = self.scram.commandName()
1047		txt += '\n\n'
1048	<	txt += 'echo "### SPECIFIC JOB SETUP ENVIRONMENT ###"\n'
1048	>	txt += 'echo ">>> specific cmssw setup environment:"\n'
1049	>	txt += 'echo "CMSSW_VERSION = '+self.version+'"\n'
1050		txt += scram+' project CMSSW '+self.version+'\n'
1051		txt += 'status=$?\n'
1052		txt += 'if [ $status != 0 ] ; then\n'
1053	<	txt += ' echo "SET_EXE_ENV 10034 ==>ERROR CMSSW '+self.version+' not found on `hostname`" \n'
1054	<	txt += ' echo "JOB_EXIT_STATUS = 10034"\n'
1055	<	txt += ' echo "JobExitCode=10034" \| tee -a $RUNTIME_AREA/$repo\n'
915	<	txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
916	<	txt += ' rm -f $RUNTIME_AREA/$repo \n'
917	<	txt += ' echo "MonitorJobID=`echo $MonitorJobID`" \| tee -a $RUNTIME_AREA/$repo \n'
918	<	txt += ' echo "MonitorID=`echo $MonitorID`" \| tee -a $RUNTIME_AREA/$repo\n'
919	<	## OLI_Daniele
920	<	txt += ' if [ $middleware == OSG ]; then \n'
921	<	txt += ' echo "Remove working directory: $WORKING_DIR"\n'
922	<	txt += ' cd $RUNTIME_AREA\n'
923	<	txt += ' /bin/rm -rf $WORKING_DIR\n'
924	<	txt += ' if [ -d $WORKING_DIR ] ;then\n'
925	<	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'
926	<	txt += ' echo "JOB_EXIT_STATUS = 10018"\n'
927	<	txt += ' echo "JobExitCode=10018" \| tee -a $RUNTIME_AREA/$repo\n'
928	<	txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
929	<	txt += ' rm -f $RUNTIME_AREA/$repo \n'
930	<	txt += ' echo "MonitorJobID=`echo $MonitorJobID`" \| tee -a $RUNTIME_AREA/$repo \n'
931	<	txt += ' echo "MonitorID=`echo $MonitorID`" \| tee -a $RUNTIME_AREA/$repo\n'
932	<	txt += ' fi\n'
933	<	txt += ' fi \n'
934	<	txt += ' exit 1 \n'
1053	>	txt += ' echo "ERROR ==> CMSSW '+self.version+' not found on `hostname`" \n'
1054	>	txt += ' job_exit_code=10034\n'
1055	>	txt += ' func_exit\n'
1056		txt += 'fi \n'
936	–	txt += 'echo "CMSSW_VERSION = '+self.version+'"\n'
1057		txt += 'cd '+self.version+'\n'
938	–	########## FEDE FOR DBS2 ######################
1058		txt += 'SOFTWARE_DIR=`pwd`\n'
1059	<	txt += 'echo SOFTWARE_DIR=$SOFTWARE_DIR \n'
941	<	###############################################
942	<	### needed grep for bug in scramv1 ###
943	<	txt += scram+' runtime -sh\n'
1059	>	txt += 'echo ">>> current directory (SOFTWARE_DIR): $SOFTWARE_DIR" \n'
1060		txt += 'eval `'+scram+' runtime -sh \| grep -v SCRAMRT_LSB_JOBNAME`\n'
1061	<	txt += 'echo $PATH\n'
1062	<
1061	>	txt += 'if [ $? != 0 ] ; then\n'
1062	>	txt += ' echo "ERROR ==> Problem with the command: "\n'
1063	>	txt += ' echo "eval \`'+scram+' runtime -sh \| grep -v SCRAMRT_LSB_JOBNAME \` at `hostname`"\n'
1064	>	txt += ' job_exit_code=10034\n'
1065	>	txt += ' func_exit\n'
1066	>	txt += 'fi \n'
1067		# Handle the arguments:
1068		txt += "\n"
1069		txt += "## number of arguments (first argument always jobnumber)\n"
1070		txt += "\n"
1071	<	# txt += "narg=$#\n"
952	<	txt += "if [ $nargs -lt 2 ]\n"
1071	>	txt += "if [ $nargs -lt "+str(self.argsList)+" ]\n"
1072		txt += "then\n"
1073	<	txt += " echo 'SET_EXE_ENV 1 ==> ERROR Too few arguments' +$nargs+ \n"
1074	<	txt += ' echo "JOB_EXIT_STATUS = 50113"\n'
1075	<	txt += ' echo "JobExitCode=50113" \| tee -a $RUNTIME_AREA/$repo\n'
957	<	txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
958	<	txt += ' rm -f $RUNTIME_AREA/$repo \n'
959	<	txt += ' echo "MonitorJobID=`echo $MonitorJobID`" \| tee -a $RUNTIME_AREA/$repo \n'
960	<	txt += ' echo "MonitorID=`echo $MonitorID`" \| tee -a $RUNTIME_AREA/$repo\n'
961	<	## OLI_Daniele
962	<	txt += ' if [ $middleware == OSG ]; then \n'
963	<	txt += ' echo "Remove working directory: $WORKING_DIR"\n'
964	<	txt += ' cd $RUNTIME_AREA\n'
965	<	txt += ' /bin/rm -rf $WORKING_DIR\n'
966	<	txt += ' if [ -d $WORKING_DIR ] ;then\n'
967	<	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'
968	<	txt += ' echo "JOB_EXIT_STATUS = 50114"\n'
969	<	txt += ' echo "JobExitCode=50114" \| tee -a $RUNTIME_AREA/$repo\n'
970	<	txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
971	<	txt += ' rm -f $RUNTIME_AREA/$repo \n'
972	<	txt += ' echo "MonitorJobID=`echo $MonitorJobID`" \| tee -a $RUNTIME_AREA/$repo \n'
973	<	txt += ' echo "MonitorID=`echo $MonitorID`" \| tee -a $RUNTIME_AREA/$repo\n'
974	<	txt += ' fi\n'
975	<	txt += ' fi \n'
976	<	txt += " exit 1\n"
1073	>	txt += " echo 'ERROR ==> Too few arguments' +$nargs+ \n"
1074	>	txt += ' job_exit_code=50113\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]
982	–	### 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	<
989	<	txt += 'PrimaryDataset='+datasetpath_split[1]+'\n'
990	<	txt += 'DataTier='+datasetpath_split[2]+'\n'
991	<	#txt += 'ProcessedDataset='+datasetpath_split[3]+'\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'
998	–	#txt += 'ProcessedDataset=null\n'
1096		txt += 'ApplicationFamily=MCDataTier\n'
1097	<	if self.pset != None: #CarlosDaniele
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=$2\n'
1103	<	txt += 'InputFiles=${args[1]}\n'
1104	<	txt += 'MaxEvents=${args[2]}\n'
1105	<	txt += 'SkipEvents=${args[3]}\n'
1102	>	txt += 'InputFiles=${args[1]}; export InputFiles\n'
1103	>	if (self.useParent):
1104	>	txt += 'ParentFiles=${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:<$InputFiles>"\n'
1111	<	txt += 'sed "s#\'INPUTFILE\'#$InputFiles#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
1111	>	if (self.useParent): txt += 'echo "ParentFiles:<$ParentFiles>"\n'
1112		txt += 'echo "MaxEvents:<$MaxEvents>"\n'
1012	–	txt += 'sed "s#int32 input = 0#int32 input = $MaxEvents#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
1113		txt += 'echo "SkipEvents:<$SkipEvents>"\n'
1014	–	txt += 'sed "s#uint32 skipEvents = 0#uint32 skipEvents = $SkipEvents#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
1114		else: # pythia like job
1115	<	seedIndex=1
1115	>	txt += 'PreserveSeeds=' + ','.join(self.preserveSeeds) + '; export PreserveSeeds\n'
1116	>	txt += 'IncrementSeeds=' + ','.join(self.incrementSeeds) + '; export IncrementSeeds\n'
1117	>	txt += 'echo "PreserveSeeds: <$PreserveSeeds>"\n'
1118	>	txt += 'echo "IncrementSeeds:<$IncrementSeeds>"\n'
1119		if (self.firstRun):
1120	<	txt += 'FirstRun=${args['+str(seedIndex)+']}\n'
1120	>	txt += 'FirstRun=${args[1]}; export FirstRun\n'
1121		txt += 'echo "FirstRun: <$FirstRun>"\n'
1020	–	txt += 'sed "s#uint32 firstRun = 0#uint32 firstRun = $FirstRun#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
1021	–	seedIndex=seedIndex+1
1022	–
1023	–	if (self.sourceSeed):
1024	–	txt += 'Seed=${args['+str(seedIndex)+']}\n'
1025	–	txt += 'sed "s#uint32 sourceSeed = 0#uint32 sourceSeed = $Seed#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
1026	–	seedIndex=seedIndex+1
1027	–	## the following seeds are not always present
1028	–	if (self.sourceSeedVtx):
1029	–	txt += 'VtxSeed=${args['+str(seedIndex)+']}\n'
1030	–	txt += 'echo "VtxSeed: <$VtxSeed>"\n'
1031	–	txt += 'sed "s#uint32 VtxSmeared = 0#uint32 VtxSmeared = $VtxSeed#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
1032	–	seedIndex += 1
1033	–	if (self.sourceSeedG4):
1034	–	txt += 'G4Seed=${args['+str(seedIndex)+']}\n'
1035	–	txt += 'echo "G4Seed: <$G4Seed>"\n'
1036	–	txt += 'sed "s#uint32 g4SimHits = 0#uint32 g4SimHits = $G4Seed#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
1037	–	seedIndex += 1
1038	–	if (self.sourceSeedMix):
1039	–	txt += 'mixSeed=${args['+str(seedIndex)+']}\n'
1040	–	txt += 'echo "MixSeed: <$mixSeed>"\n'
1041	–	txt += 'sed "s#uint32 mix = 0#uint32 mix = $mixSeed#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
1042	–	seedIndex += 1
1043	–	pass
1044	–	pass
1045	–	txt += 'mv -f '+pset+' pset.cfg\n'
1122
1123	<	if len(self.additional_inbox_files) > 0:
1048	<	txt += 'if [ -e $RUNTIME_AREA/'+self.additional_tgz_name+' ] ; then\n'
1049	<	txt += ' tar xzvf $RUNTIME_AREA/'+self.additional_tgz_name+'\n'
1050	<	txt += 'fi\n'
1051	<	pass
1123	>	txt += 'mv -f ' + pset + ' ' + psetName + '\n'
1124
1053	–	if self.pset != None: #CarlosDaniele
1054	–	txt += 'echo "### END JOB SETUP ENVIRONMENT ###"\n\n'
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	<	### FEDE FOR DBS OUTPUT PUBLICATION
1134	<	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'
1064	–	##############
1139		txt += '\n'
1066	–	# txt += 'echo "*** cat pset1.cfg *******"\n'
1067	–	# txt += 'cat pset1.cfg\n'
1068	–	# txt += 'echo "**** end pset1.cfg ******"\n'
1140		return txt
1141
1142	<	def wsBuildExe(self, nj=0):
1142	>	def wsUntarSoftware(self, nj=0):
1143		"""
1144		Put in the script the commands to build an executable
1145		or a library.
1146		"""
1147
1148	<	txt = ""
1148	>	txt = '\n#Written by cms_cmssw::wsUntarSoftware\n'
1149
1150		if os.path.isfile(self.tgzNameWithPath):
1151	<	txt += 'echo "tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+'"\n'
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 "SET_EXE 1 ==> ERROR Untarring .tgz file failed"\n'
1158	<	txt += ' echo "JOB_EXIT_STATUS = $untar_status" \n'
1159	<	txt += ' echo "JobExitCode=$untar_status" \| tee -a $RUNTIME_AREA/$repo\n'
1087	<	txt += ' if [ $middleware == OSG ]; then \n'
1088	<	txt += ' echo "Remove working directory: $WORKING_DIR"\n'
1089	<	txt += ' cd $RUNTIME_AREA\n'
1090	<	txt += ' /bin/rm -rf $WORKING_DIR\n'
1091	<	txt += ' if [ -d $WORKING_DIR ] ;then\n'
1092	<	txt += ' echo "SET_EXE 50999 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after Untarring .tgz file failed"\n'
1093	<	txt += ' echo "JOB_EXIT_STATUS = 50999"\n'
1094	<	txt += ' echo "JobExitCode=50999" \| tee -a $RUNTIME_AREA/$repo\n'
1095	<	txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
1096	<	txt += ' rm -f $RUNTIME_AREA/$repo \n'
1097	<	txt += ' echo "MonitorJobID=`echo $MonitorJobID`" \| tee -a $RUNTIME_AREA/$repo \n'
1098	<	txt += ' echo "MonitorID=`echo $MonitorID`" \| tee -a $RUNTIME_AREA/$repo\n'
1099	<	txt += ' fi\n'
1100	<	txt += ' fi \n'
1101	<	txt += ' \n'
1102	<	txt += ' exit 1 \n'
1157	>	txt += ' echo "ERROR ==> Untarring .tgz file failed"\n'
1158	>	txt += ' job_exit_code=$untar_status\n'
1159	>	txt += ' func_exit\n'
1160		txt += 'else \n'
1161		txt += ' echo "Successful untar" \n'
1162		txt += 'fi \n'
1163		txt += '\n'
1164	<	txt += 'echo "Include ProdAgentApi and PRODCOMMON in PYTHONPATH"\n'
1164	>	txt += 'echo ">>> Include $RUNTIME_AREA in PYTHONPATH:"\n'
1165		txt += 'if [ -z "$PYTHONPATH" ]; then\n'
1166	<	#### FEDE FOR DBS OUTPUT PUBLICATION
1110	<	txt += ' export PYTHONPATH=$SOFTWARE_DIR/ProdAgentApi:$SOFTWARE_DIR/ProdCommon\n'
1111	<	#txt += ' export PYTHONPATH=`pwd`/ProdAgentApi:`pwd`/ProdCommon\n'
1112	<	#txt += ' export PYTHONPATH=ProdAgentApi\n'
1166	>	txt += ' export PYTHONPATH=$RUNTIME_AREA/\n'
1167		txt += 'else\n'
1168	<	txt += ' export PYTHONPATH=$SOFTWARE_DIR/ProdAgentApi:$SOFTWARE_DIR/ProdCommon:${PYTHONPATH}\n'
1115	<	#txt += ' export PYTHONPATH=`pwd`/ProdAgentApi:`pwd`/ProdCommon:${PYTHONPATH}\n'
1116	<	#txt += ' export PYTHONPATH=ProdAgentApi:${PYTHONPATH}\n'
1168	>	txt += ' export PYTHONPATH=$RUNTIME_AREA/:${PYTHONPATH}\n'
1169		txt += 'echo "PYTHONPATH=$PYTHONPATH"\n'
1118	–	###################
1170		txt += 'fi\n'
1171		txt += '\n'
1172
#	Line 1123 \| Line 1174 \| class Cmssw(JobType):
1174
1175		return txt
1176
1177	<	def modifySteeringCards(self, nj):
1177	>	def wsBuildExe(self, nj=0):
1178		"""
1179	<	modify the card provided by the user,
1180	<	writing a new card into share dir
1179	>	Put in the script the commands to build an executable
1180	>	or a library.
1181		"""
1182
1183	+	txt = '\n#Written by cms_cmssw::wsBuildExe\n'
1184	+	txt += 'echo ">>> moving CMSSW software directories in `pwd`" \n'
1185	+
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=$RUNTIME_AREA/\n'
1201	+	txt += 'else\n'
1202	+	txt += ' export PYTHONPATH=$RUNTIME_AREA/:${PYTHONPATH}\n'
1203	+	txt += 'echo "PYTHONPATH=$PYTHONPATH"\n'
1204	+	txt += 'fi\n'
1205	+	txt += '\n'
1206	+
1207	+	return txt
1208	+
1209	+
1210		def executableName(self):
1211	<	if self.scriptExe: #CarlosDaniele
1211	>	if self.scriptExe:
1212		return "sh "
1213		else:
1214		return self.executable
1215
1216		def executableArgs(self):
1217	+	# FUTURE: This function tests the CMSSW version. Can be simplified as we drop support for old versions
1218		if self.scriptExe:#CarlosDaniele
1219		return self.scriptExe + " $NJob"
1220		else:
1221	<	# if >= CMSSW_1_5_X, add -e
1222	<	version_array = self.scram.getSWVersion().split('_')
1223	<	major = 0
1224	<	minor = 0
1225	<	try:
1226	<	major = int(version_array[1])
1227	<	minor = int(version_array[2])
1228	<	except:
1150	<	msg = "Cannot parse CMSSW version string: " + "_".join(version_array) + " for major and minor release number!"
1151	<	raise CrabException(msg)
1152	<	if major >= 1 and minor >= 5 :
1153	<	return " -e -p pset.cfg"
1221	>	ex_args = ""
1222	>	# FUTURE: This tests the CMSSW version. Can remove code as versions deprecated
1223	>	# Framework job report
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	>	# Type of config file
1227	>	if self.CMSSW_major >= 2 :
1228	>	ex_args += " -p pset.py"
1229		else:
1230	<	return " -p pset.cfg"
1230	>	ex_args += " -p pset.cfg"
1231	>	return ex_args
1232
1233		def inputSandbox(self, nj):
1234		"""
1235		Returns a list of filenames to be put in JDL input sandbox.
1236		"""
1237		inp_box = []
1162	–	# # dict added to delete duplicate from input sandbox file list
1163	–	# seen = {}
1164	–	## code
1238		if os.path.isfile(self.tgzNameWithPath):
1239		inp_box.append(self.tgzNameWithPath)
1240	<	if os.path.isfile(self.MLtgzfile):
1168	<	inp_box.append(self.MLtgzfile)
1169	<	## config
1170	<	if not self.pset is None:
1171	<	inp_box.append(common.work_space.pathForTgz() + 'job/' + self.configFilename())
1172	<	## additional input files
1173	<	tgz = self.additionalInputFileTgz()
1174	<	inp_box.append(tgz)
1240	>	inp_box.append(common.work_space.jobDir() + self.scriptName)
1241		return inp_box
1242
1243		def outputSandbox(self, nj):
#	Line 1183 \| 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
1189	–	def prepareSteeringCards(self):
1190	–	"""
1191	–	Make initial modifications of the user's steering card file.
1192	–	"""
1193	–	return
1255
1256		def wsRenameOutput(self, nj):
1257		"""
1258		Returns part of a job script which renames the produced files.
1259		"""
1260
1261	<	txt = '\n'
1262	<	txt += '# directory content\n'
1263	<	txt += 'ls \n'
1264	<
1265	<	txt += 'output_exit_status=0\n'
1266	<
1206	<	for fileWithSuffix in (self.output_file_sandbox):
1207	<	output_file_num = self.numberFile_(fileWithSuffix, '$NJob')
1208	<	txt += '\n'
1209	<	txt += '# check output file\n'
1210	<	txt += 'if [ -e ./'+fileWithSuffix+' ] ; then\n'
1211	<	txt += ' mv '+fileWithSuffix+' $RUNTIME_AREA\n'
1212	<	txt += ' cp $RUNTIME_AREA/'+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n'
1213	<	txt += 'else\n'
1214	<	txt += ' exit_status=60302\n'
1215	<	txt += ' echo "ERROR: Problem with output file '+fileWithSuffix+'"\n'
1216	<	if common.scheduler.boss_scheduler_name == 'condor_g':
1217	<	txt += ' if [ $middleware == OSG ]; then \n'
1218	<	txt += ' echo "prepare dummy output file"\n'
1219	<	txt += ' echo "Processing of job output failed" > $RUNTIME_AREA/'+output_file_num+'\n'
1220	<	txt += ' fi \n'
1221	<	txt += 'fi\n'
1261	>	txt = '\n#Written by cms_cmssw::wsRenameOutput\n'
1262	>	txt += 'echo ">>> current directory (SOFTWARE_DIR): $SOFTWARE_DIR" \n'
1263	>	txt += 'echo ">>> current directory content:"\n'
1264	>	if self.debug_wrapper:
1265	>	txt += 'ls -Al\n'
1266	>	txt += '\n'
1267
1268		for fileWithSuffix in (self.output_file):
1269	<	output_file_num = self.numberFile_(fileWithSuffix, '$NJob')
1269	>	output_file_num = numberFile(fileWithSuffix, '$NJob')
1270		txt += '\n'
1271		txt += '# check output file\n'
1272		txt += 'if [ -e ./'+fileWithSuffix+' ] ; then\n'
1273	<	txt += ' mv '+fileWithSuffix+' $RUNTIME_AREA\n'
1274	<	txt += ' cp $RUNTIME_AREA/'+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n'
1273	>	if (self.copy_data == 1): # For OSG nodes, file is in $WORKING_DIR, should not be moved to $RUNTIME_AREA
1274	>	txt += ' mv '+fileWithSuffix+' '+output_file_num+'\n'
1275	>	txt += ' ln -s `pwd`/'+output_file_num+' $RUNTIME_AREA/'+fileWithSuffix+'\n'
1276	>	else:
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'
1280	<	txt += ' exit_status=60302\n'
1281	<	txt += ' echo "ERROR: Problem with output file '+fileWithSuffix+'"\n'
1282	<	txt += ' echo "JOB_EXIT_STATUS = $exit_status"\n'
1234	<	txt += ' output_exit_status=$exit_status\n'
1235	<	if common.scheduler.boss_scheduler_name == 'condor_g':
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':
1283		txt += ' if [ $middleware == OSG ]; then \n'
1284		txt += ' echo "prepare dummy output file"\n'
1285		txt += ' echo "Processing of job output failed" > $RUNTIME_AREA/'+output_file_num+'\n'
#	Line 1240 \| Line 1287 \| class Cmssw(JobType):
1287		txt += 'fi\n'
1288		file_list = []
1289		for fileWithSuffix in (self.output_file):
1290	<	file_list.append(self.numberFile_(fileWithSuffix, '$NJob'))
1290	>	file_list.append(numberFile(fileWithSuffix, '$NJob'))
1291
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	+	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
1249	–	def numberFile_(self, file, txt):
1250	–	"""
1251	–	append _'txt' before last extension of a file
1252	–	"""
1253	–	p = string.split(file,".")
1254	–	# take away last extension
1255	–	name = p[0]
1256	–	for x in p[1:-1]:
1257	–	name=name+"."+x
1258	–	# add "_txt"
1259	–	if len(p)>1:
1260	–	ext = p[len(p)-1]
1261	–	result = name + '_' + txt + "." + ext
1262	–	else:
1263	–	result = name + '_' + txt
1264	–
1265	–	return result
1266	–
1303		def getRequirements(self, nj=[]):
1304		"""
1305		return job requirements to add to jdl files
#	Line 1273 \| 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
1277	<	#reSL4 = re.compile( r'slc4' )
1278	<	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" ) 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
1291	–	### OLI_DANIELE
1331		def wsSetupCMSOSGEnvironment_(self):
1332		"""
1333		Returns part of a job script which is prepares
1334		the execution environment and which is common for all CMS jobs.
1335		"""
1336	<	txt = '\n'
1337	<	txt += ' echo "### SETUP CMS OSG ENVIRONMENT ###"\n'
1338	<	txt += ' if [ -f $GRID3_APP_DIR/cmssoft/cmsset_default.sh ] ;then\n'
1339	<	txt += ' # Use $GRID3_APP_DIR/cmssoft/cmsset_default.sh to setup cms software\n'
1340	<	txt += ' export SCRAM_ARCH='+self.executable_arch+'\n'
1341	<	txt += ' source $GRID3_APP_DIR/cmssoft/cmsset_default.sh '+self.version+'\n'
1303	<	txt += ' elif [ -f $OSG_APP/cmssoft/cms/cmsset_default.sh ] ;then\n'
1336	>	txt = '\n#Written by cms_cmssw::wsSetupCMSOSGEnvironment_\n'
1337	>	txt += ' echo ">>> setup CMS OSG environment:"\n'
1338	>	txt += ' echo "set SCRAM ARCH to ' + self.executable_arch + '"\n'
1339	>	txt += ' export SCRAM_ARCH='+self.executable_arch+'\n'
1340	>	txt += ' echo "SCRAM_ARCH = $SCRAM_ARCH"\n'
1341	>	txt += ' if [ -f $OSG_APP/cmssoft/cms/cmsset_default.sh ] ;then\n'
1342		txt += ' # Use $OSG_APP/cmssoft/cms/cmsset_default.sh to setup cms software\n'
1343	<	txt += ' export SCRAM_ARCH='+self.executable_arch+'\n'
1344	<	txt += ' source $OSG_APP/cmssoft/cms/cmsset_default.sh '+self.version+'\n'
1345	<	txt += ' else\n'
1346	<	txt += ' echo "SET_CMS_ENV 10020 ==> ERROR $GRID3_APP_DIR/cmssoft/cmsset_default.sh and $OSG_APP/cmssoft/cms/cmsset_default.sh file not found"\n'
1347	<	txt += ' echo "JOB_EXIT_STATUS = 10020"\n'
1348	<	txt += ' echo "JobExitCode=10020" \| tee -a $RUNTIME_AREA/$repo\n'
1311	<	txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
1312	<	txt += ' rm -f $RUNTIME_AREA/$repo \n'
1313	<	txt += ' echo "MonitorJobID=`echo $MonitorJobID`" \| tee -a $RUNTIME_AREA/$repo \n'
1314	<	txt += ' echo "MonitorID=`echo $MonitorID`" \| tee -a $RUNTIME_AREA/$repo\n'
1315	<	txt += ' exit 1\n'
1316	<	txt += '\n'
1317	<	txt += ' echo "Remove working directory: $WORKING_DIR"\n'
1318	<	txt += ' cd $RUNTIME_AREA\n'
1319	<	txt += ' /bin/rm -rf $WORKING_DIR\n'
1320	<	txt += ' if [ -d $WORKING_DIR ] ;then\n'
1321	<	txt += ' echo "SET_CMS_ENV 10017 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after $GRID3_APP_DIR/cmssoft/cmsset_default.sh and $OSG_APP/cmssoft/cms/cmsset_default.sh file not found"\n'
1322	<	txt += ' echo "JOB_EXIT_STATUS = 10017"\n'
1323	<	txt += ' echo "JobExitCode=10017" \| tee -a $RUNTIME_AREA/$repo\n'
1324	<	txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
1325	<	txt += ' rm -f $RUNTIME_AREA/$repo \n'
1326	<	txt += ' echo "MonitorJobID=`echo $MonitorJobID`" \| tee -a $RUNTIME_AREA/$repo \n'
1327	<	txt += ' echo "MonitorID=`echo $MonitorID`" \| tee -a $RUNTIME_AREA/$repo\n'
1328	<	txt += ' fi\n'
1329	<	txt += '\n'
1330	<	txt += ' exit 1\n'
1331	<	txt += ' fi\n'
1343	>	txt += ' source $OSG_APP/cmssoft/cms/cmsset_default.sh '+self.version+'\n'
1344	>	txt += ' else\n'
1345	>	txt += ' echo "ERROR ==> $OSG_APP/cmssoft/cms/cmsset_default.sh file not found"\n'
1346	>	txt += ' job_exit_code=10020\n'
1347	>	txt += ' func_exit\n'
1348	>	txt += ' fi\n'
1349		txt += '\n'
1350	<	txt += ' echo "SET_CMS_ENV 0 ==> setup cms environment ok"\n'
1351	<	txt += ' echo " END SETUP CMS OSG ENVIRONMENT "\n'
1350	>	txt += ' echo "==> setup cms environment ok"\n'
1351	>	txt += ' echo "SCRAM_ARCH = $SCRAM_ARCH"\n'
1352
1353		return txt
1354
1338	–	### OLI_DANIELE
1355		def wsSetupCMSLCGEnvironment_(self):
1356		"""
1357		Returns part of a job script which is prepares
1358		the execution environment and which is common for all CMS jobs.
1359		"""
1360	<	txt = ' \n'
1361	<	txt += ' echo " ### SETUP CMS LCG ENVIRONMENT ### "\n'
1362	<	txt += ' if [ ! $VO_CMS_SW_DIR ] ;then\n'
1363	<	txt += ' echo "SET_CMS_ENV 10031 ==> ERROR CMS software dir not found on WN `hostname`"\n'
1364	<	txt += ' echo "JOB_EXIT_STATUS = 10031" \n'
1365	<	txt += ' echo "JobExitCode=10031" \| tee -a $RUNTIME_AREA/$repo\n'
1366	<	txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
1367	<	txt += ' rm -f $RUNTIME_AREA/$repo \n'
1368	<	txt += ' echo "MonitorJobID=`echo $MonitorJobID`" \| tee -a $RUNTIME_AREA/$repo \n'
1369	<	txt += ' echo "MonitorID=`echo $MonitorID`" \| tee -a $RUNTIME_AREA/$repo\n'
1370	<	txt += ' exit 1\n'
1371	<	txt += ' else\n'
1372	<	txt += ' echo "Sourcing environment... "\n'
1373	<	txt += ' if [ ! -s $VO_CMS_SW_DIR/cmsset_default.sh ] ;then\n'
1374	<	txt += ' echo "SET_CMS_ENV 10020 ==> ERROR cmsset_default.sh file not found into dir $VO_CMS_SW_DIR"\n'
1375	<	txt += ' echo "JOB_EXIT_STATUS = 10020"\n'
1376	<	txt += ' echo "JobExitCode=10020" \| tee -a $RUNTIME_AREA/$repo\n'
1377	<	txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
1378	<	txt += ' rm -f $RUNTIME_AREA/$repo \n'
1379	<	txt += ' echo "MonitorJobID=`echo $MonitorJobID`" \| tee -a $RUNTIME_AREA/$repo \n'
1380	<	txt += ' echo "MonitorID=`echo $MonitorID`" \| tee -a $RUNTIME_AREA/$repo\n'
1381	<	txt += ' exit 1\n'
1382	<	txt += ' fi\n'
1383	<	txt += ' echo "sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
1384	<	txt += ' source $VO_CMS_SW_DIR/cmsset_default.sh\n'
1385	<	txt += ' result=$?\n'
1386	<	txt += ' if [ $result -ne 0 ]; then\n'
1371	<	txt += ' echo "SET_CMS_ENV 10032 ==> ERROR problem sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
1372	<	txt += ' echo "JOB_EXIT_STATUS = 10032"\n'
1373	<	txt += ' echo "JobExitCode=10032" \| tee -a $RUNTIME_AREA/$repo\n'
1374	<	txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
1375	<	txt += ' rm -f $RUNTIME_AREA/$repo \n'
1376	<	txt += ' echo "MonitorJobID=`echo $MonitorJobID`" \| tee -a $RUNTIME_AREA/$repo \n'
1377	<	txt += ' echo "MonitorID=`echo $MonitorID`" \| tee -a $RUNTIME_AREA/$repo\n'
1378	<	txt += ' exit 1\n'
1379	<	txt += ' fi\n'
1380	<	txt += ' fi\n'
1381	<	txt += ' \n'
1382	<	txt += ' echo "SET_CMS_ENV 0 ==> setup cms environment ok"\n'
1383	<	txt += ' echo "### END SETUP CMS LCG ENVIRONMENT ###"\n'
1360	>	txt = '\n#Written by cms_cmssw::wsSetupCMSLCGEnvironment_\n'
1361	>	txt += ' echo ">>> setup CMS LCG environment:"\n'
1362	>	txt += ' echo "set SCRAM ARCH and BUILD_ARCH to ' + self.executable_arch + ' ###"\n'
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'
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'
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'
1375	>	txt += ' fi\n'
1376	>	txt += ' echo "sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
1377	>	txt += ' source $VO_CMS_SW_DIR/cmsset_default.sh\n'
1378	>	txt += ' result=$?\n'
1379	>	txt += ' if [ $result -ne 0 ]; then\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'
1383	>	txt += ' fi\n'
1384	>	txt += ' fi\n'
1385	>	txt += ' \n'
1386	>	txt += ' echo "==> setup cms environment ok"\n'
1387		return txt
1388
1389	<	### FEDE FOR DBS OUTPUT PUBLICATION
1387	<	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 = ''
1393	<	try:
1394	<	publish_data = int(self.cfg_params['USER.publish_data'])
1395	<	except KeyError:
1396	<	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):
1398	–	txt += 'echo "Modify Job Report" \n'
1399	–	#txt += 'chmod a+x $RUNTIME_AREA/'+self.version+'/ProdAgentApi/FwkJobRep/ModifyJobReport.py\n'
1400	–	################ FEDE FOR DBS2 #############################################
1401	–	txt += 'chmod a+x $SOFTWARE_DIR/ProdAgentApi/FwkJobRep/ModifyJobReport.py\n'
1402	–	#############################################################################
1403	–	#try:
1404	–	# publish_data = int(self.cfg_params['USER.publish_data'])
1405	–	#except KeyError:
1406	–	# publish_data = 0
1396
1397	<	txt += 'if [ -z "$SE" ]; then\n'
1398	<	txt += ' SE="" \n'
1399	<	txt += 'fi \n'
1411	<	txt += 'if [ -z "$SE_PATH" ]; then\n'
1412	<	txt += ' SE_PATH="" \n'
1413	<	txt += 'fi \n'
1414	<	txt += 'echo "SE = $SE"\n'
1415	<	txt += 'echo "SE_PATH = $SE_PATH"\n'
1416	<
1417	<	#if (publish_data == 1):
1418	<	#processedDataset = self.cfg_params['USER.processed_datasetname']
1419	<	processedDataset = self.cfg_params['USER.publish_data_name']
1420	<	txt += 'ProcessedDataset='+processedDataset+'\n'
1421	<	#### LFN=/store/user/<user>/processedDataset_PSETHASH
1422	<	txt += 'if [ "$SE_PATH" == "" ]; then\n'
1423	<	#### FEDE: added slash in LFN ##############
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 += 'else \n'
1402	<	txt += ' tmp=`echo $SE_PATH \| awk -F \'store\' \'{print$2}\'` \n'
1403	<	##### FEDE TO BE CHANGED, BECAUSE STORE IS HARDCODED!!!! ########
1404	<	txt += ' FOR_LFN=/store$tmp \n'
1405	<	txt += 'fi \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 = $CMSSW_VERSION"\n\n'
1413	<	#txt += 'echo "$RUNTIME_AREA/'+self.version+'/ProdAgentApi/FwkJobRep/ModifyJobReport.py crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH"\n'
1414	<	txt += 'echo "$SOFTWARE_DIR/ProdAgentApi/FwkJobRep/ModifyJobReport.py crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH"\n'
1415	<	txt += '$SOFTWARE_DIR/ProdAgentApi/FwkJobRep/ModifyJobReport.py crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH\n'
1416	<	#txt += '$RUNTIME_AREA/'+self.version+'/ProdAgentApi/FwkJobRep/ModifyJobReport.py crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH\n'
1417	<
1413	>	args = '$RUNTIME_AREA/crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier ' \
1414	>	'$User -$ProcessedDataset-$PSETHASH $ApplicationFamily '+ \
1415	>	' $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH"\n'
1416	>	txt += 'echo "$RUNTIME_AREA/ProdCommon/FwkJobRep/ModifyJobReport.py '+str(args)
1417	>	txt += '$RUNTIME_AREA/ProdCommon/FwkJobRep/ModifyJobReport.py '+str(args)
1418		txt += 'modifyReport_result=$?\n'
1439	–	txt += 'echo modifyReport_result = $modifyReport_result\n'
1419		txt += 'if [ $modifyReport_result -ne 0 ]; then\n'
1420	<	txt += ' exit_status=1\n'
1421	<	txt += ' echo "ERROR: Problem with ModifyJobReport"\n'
1420	>	txt += ' modifyReport_result=70500\n'
1421	>	txt += ' job_exit_code=$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 crab_fjr_$NJob.xml\n'
1425	>	txt += ' mv NewFrameworkJobReport.xml $RUNTIME_AREA/crab_fjr_$NJob.xml\n'
1426		txt += 'fi\n'
1446	–	else:
1447	–	txt += 'echo "no data publication required"\n'
1448	–	#txt += 'ProcessedDataset=no_data_to_publish \n'
1449	–	#### FEDE: added slash in LFN ##############
1450	–	#txt += 'FOR_LFN=/local/ \n'
1451	–	#txt += 'echo "ProcessedDataset = $ProcessedDataset"\n'
1452	–	#txt += 'echo "FOR_LFN = $FOR_LFN" \n'
1427		return txt
1428
1429	<	def cleanEnv(self):
1430	<	### OLI_DANIELE
1431	<	txt = ''
1432	<	txt += 'if [ $middleware == OSG ]; then\n'
1433	<	txt += ' cd $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 "SET_EXE 60999 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after cleanup of WN"\n'
1438	<	txt += ' echo "JOB_EXIT_STATUS = 60999"\n'
1439	<	txt += ' echo "JobExitCode=60999" \| tee -a $RUNTIME_AREA/$repo\n'
1440	<	txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
1441	<	txt += ' rm -f $RUNTIME_AREA/$repo \n'
1442	<	txt += ' echo "MonitorJobID=`echo $MonitorJobID`" \| tee -a $RUNTIME_AREA/$repo \n'
1443	<	txt += ' echo "MonitorID=`echo $MonitorID`" \| tee -a $RUNTIME_AREA/$repo\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 1478 \| Line 1499 \| class Cmssw(JobType):
1499		def getParams(self):
1500		return self._params
1501
1481	–	def setTaskid_(self):
1482	–	self._taskId = self.cfg_params['taskId']
1483	–
1484	–	def getTaskid(self):
1485	–	return self._taskId
1486	–
1502		def uniquelist(self, old):
1503		"""
1504		remove duplicates from a list
#	Line 1493 \| Line 1508 \| class Cmssw(JobType):
1508		nd[e]=0
1509		return nd.keys()
1510
1511	<
1497	<	def checkOut(self, limit):
1511	>	def outList(self):
1512		"""
1513		check the dimension of the output files
1514		"""
1515	<	txt = 'echo "*****************************************"\n'
1516	<	txt += 'echo " Starting output sandbox limit check "\n'
1503	<	txt += 'echo "*****************************************"\n'
1504	<	allOutFiles = ""
1515	>	txt = ''
1516	>	txt += 'echo ">>> list of expected files on output sandbox"\n'
1517		listOutFiles = []
1518	<	for fileOut in (self.output_file+self.output_file_sandbox):
1519	<	if fileOut.find('crab_fjr') == -1:
1520	<	allOutFiles = allOutFiles + " " + self.numberFile_(fileOut, '$NJob')
1521	<	listOutFiles.append(self.numberFile_(fileOut, '$NJob'))
1522	<	txt += 'echo "OUTPUT files: '+str(allOutFiles)+'";\n'
1523	<	txt += 'ls -gGhrta;\n'
1524	<	txt += 'sum=0;\n'
1525	<	txt += 'for file in '+str(allOutFiles)+' ; do\n'
1526	<	txt += ' if [ -e $file ]; then\n'
1527	<	txt += ' tt=`ls -gGrta $file \| awk \'{ print $3 }\'`\n'
1528	<	txt += ' sum=`expr $sum + $tt`\n'
1529	<	txt += ' else\n'
1530	<	txt += ' echo "WARNING: output file $file not found!"\n'
1531	<	txt += ' fi\n'
1532	<	txt += 'done\n'
1521	<	txt += 'echo "Total Output dimension: $sum";\n'
1522	<	txt += 'limit='+str(limit)+';\n'
1523	<	txt += 'echo "OUTPUT FILES LIMIT SET TO: $limit";\n'
1524	<	txt += 'if [ $limit -lt $sum ]; then\n'
1525	<	txt += ' echo "WARNING: output files have to big size - something will be lost;"\n'
1526	<	txt += ' echo " checking the output file sizes..."\n'
1527	<	"""
1528	<	txt += ' dim=0;\n'
1529	<	txt += ' exclude=0;\n'
1530	<	txt += ' for files in '+str(allOutFiles)+' ; do\n'
1531	<	txt += ' sumTemp=0;\n'
1532	<	txt += ' for file2 in '+str(allOutFiles)+' ; do\n'
1533	<	txt += ' if [ $file != $file2 ]; then\n'
1534	<	txt += ' tt=`ls -gGrta $file2 \| awk \'{ print $3 }\';`\n'
1535	<	txt += ' sumTemp=`expr $sumTemp + $tt`;\n'
1536	<	txt += ' fi\n'
1537	<	txt += ' done\n'
1538	<	txt += ' if [ $sumTemp -lt $limit ]; then\n'
1539	<	txt += ' if [ $dim -lt $sumTemp ]; then\n'
1540	<	txt += ' dim=$sumTemp;\n'
1541	<	txt += ' exclude=$file;\n'
1542	<	txt += ' fi\n'
1543	<	txt += ' fi\n'
1544	<	txt += ' done\n'
1545	<	txt += ' echo "Dimension calculated: $dim"; echo "File to exclude: $exclude";\n'
1546	<	"""
1547	<	txt += ' tot=0;\n'
1548	<	txt += ' for file2 in '+str(allOutFiles)+' ; do\n'
1549	<	txt += ' tt=`ls -gGrta $file2 \| awk \'{ print $3 }\';`\n'
1550	<	txt += ' tot=`expr $tot + $tt`;\n'
1551	<	txt += ' if [ $limit -lt $tot ]; then\n'
1552	<	txt += ' tot=`expr $tot - $tt`;\n'
1553	<	txt += ' fileLast=$file;\n'
1554	<	txt += ' break;\n'
1555	<	txt += ' fi\n'
1556	<	txt += ' done\n'
1557	<	txt += ' echo "Dimension calculated: $tot"; echo "First file to exclude: $file";\n'
1558	<	txt += ' flag=0;\n'
1559	<	txt += ' for filess in '+str(allOutFiles)+' ; do\n'
1560	<	txt += ' if [ $fileLast = $filess ]; then\n'
1561	<	txt += ' flag=1;\n'
1562	<	txt += ' fi\n'
1563	<	txt += ' if [ $flag -eq 1 ]; then\n'
1564	<	txt += ' rm -f $filess;\n'
1565	<	txt += ' fi\n'
1566	<	txt += ' done\n'
1567	<	txt += ' ls -agGhrt;\n'
1568	<	txt += ' echo "WARNING: output files are too big in dimension: can not put in the output_sandbox.";\n'
1569	<	txt += ' echo "JOB_EXIT_STATUS = 70000";\n'
1570	<	txt += ' exit_status=70000;\n'
1571	<	txt += 'else'
1572	<	txt += ' echo "Total Output dimension $sum is fine.";\n'
1573	<	txt += 'fi\n'
1574	<	txt += 'echo "*****************************************"\n'
1575	<	txt += 'echo "* Ending output sandbox limit check *"\n'
1576	<	txt += 'echo "*****************************************"\n'
1518	>	stdout = 'CMSSW_$NJob.stdout'
1519	>	stderr = 'CMSSW_$NJob.stderr'
1520	>	if (self.return_data == 1):
1521	>	for file in (self.output_file+self.output_file_sandbox):
1522	>	listOutFiles.append(numberFile(file, '$NJob'))
1523	>	listOutFiles.append(stdout)
1524	>	listOutFiles.append(stderr)
1525	>	else:
1526	>	for file in (self.output_file_sandbox):
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

Diff Legend

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

Comparing COMP/CRAB/python/cms_cmssw.py (file contents): Revision 1.132 by ewv, Mon Oct 29 21:03:11 2007 UTC vs. Revision 1.244 by spiga, Thu Sep 25 15:08:01 2008 UTC

Diff Legend

Comparing COMP/CRAB/python/cms_cmssw.py (file contents):
Revision 1.132 by ewv, Mon Oct 29 21:03:11 2007 UTC vs.
Revision 1.244 by spiga, Thu Sep 25 15:08:01 2008 UTC