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 1.137 by slacapra, Fri Nov 16 11:09:31 2007 UTC vs.
Revision 1.235 by spiga, Fri Aug 29 15:06:41 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
18	–
22		# init BlackWhiteListParser
23	<	self.blackWhiteListParser = BlackWhiteListParser(cfg_params)
23	>	self.blackWhiteListParser = SEBlackWhiteListParser(cfg_params)
24
25	<	try:
26	<	self.MaxTarBallSize = float(self.cfg_params['EDG.maxtarballsize'])
27	<	except KeyError:
28	<	self.MaxTarBallSize = 9.5
25	>	### Temporary patch to automatically skip the ISB size check:
26	>	server=self.cfg_params.get('CRAB.server_name',None)
27	>	size = 9.5
28	>	if server: size = 99999
29	>	### D.S.
30	>	self.MaxTarBallSize = float(self.cfg_params.get('EDG.maxtarballsize',size))
31
32		# number of jobs requested to be created, limit obj splitting
33		self.ncjobs = ncjobs
#	Line 35 | Line 40 | class Cmssw(JobType):
40		self.executable = ''
41		self.executable_arch = self.scram.getArch()
42		self.tgz_name = 'default.tgz'
38	–	self.additional_tgz_name = 'additional.tgz'
43		self.scriptName = 'CMSSW.sh'
44	<	self.pset = ''      #scrip use case Da
45	<	self.datasetPath = '' #scrip use case Da
44	>	self.pset = ''
45	>	self.datasetPath = ''
46
47		# set FJR file name
48		self.fjrFileName = 'crab_fjr.xml'
49
50		self.version = self.scram.getSWVersion()
51	<
52	<	#
53	<	# Try to block creation in case of arch/version mismatch
54	<	#
55	<
56	<	a = string.split(self.version, "_")
57	<
58	<	if int(a[1]) == 1 and (int(a[2]) < 5 and self.executable_arch.find('slc4') == 0):
59	<	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)
60	<	common.logger.message(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)
51	>	version_array = self.version.split('_')
52	>	self.CMSSW_major = 0
53	>	self.CMSSW_minor = 0
54	>	self.CMSSW_patch = 0
55	>	try:
56	>	self.CMSSW_major = int(version_array[1])
57	>	self.CMSSW_minor = int(version_array[2])
58	>	self.CMSSW_patch = int(version_array[3])
59	>	except:
60	>	msg = "Cannot parse CMSSW version string: " + self.version + " for major and minor release number!"
61		raise CrabException(msg)
62
61	–	common.taskDB.setDict('codeVersion',self.version)
62	–	self.setParam_('application', self.version)
63	–
63		### collect Data cards
64
65	<	try:
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	<	except KeyError:
65	>	if not cfg_params.has_key('CMSSW.datasetpath'):
66		msg = "Error: datasetpath not defined "
67		raise CrabException(msg)
68
69	<	# ML monitoring
70	<	# split dataset path style: /PreProdR3Minbias/SIM/GEN-SIM
81	<	if not self.datasetPath:
82	<	self.setParam_('dataset', 'None')
83	<	self.setParam_('owner', 'None')
84	<	else:
85	<	try:
86	<	datasetpath_split = self.datasetPath.split("/")
87	<	# standard style
88	<	self.setParam_('datasetFull', self.datasetPath)
89	<	self.setParam_('dataset', datasetpath_split[1])
90	<	self.setParam_('owner', datasetpath_split[2])
91	<	except:
92	<	self.setParam_('dataset', self.datasetPath)
93	<	self.setParam_('owner', self.datasetPath)
69	>	### Temporary: added to remove input file control in the case of PU
70	>	self.dataset_pu = cfg_params.get('CMSSW.dataset_pu', None)
71
72	<	self.setTaskid_()
73	<	self.setParam_('taskId', self.cfg_params['taskId'])
72	>	tmp =  cfg_params['CMSSW.datasetpath']
73	>	log.debug(6, "CMSSW::CMSSW(): datasetPath = "+tmp)
74	>	if string.lower(tmp)=='none':
75	>	self.datasetPath = None
76	>	self.selectNoInput = 1
77	>	else:
78	>	self.datasetPath = tmp
79	>	self.selectNoInput = 0
80
81		self.dataTiers = []
82	<
82	>	self.debugWrap = ''
83	>	self.debug_wrapper = cfg_params.get('USER.debug_wrapper',False)
84	>	if self.debug_wrapper: self.debugWrap='--debug'
85		## now the application
86	<	try:
87	<	self.executable = cfg_params['CMSSW.executable']
103	<	self.setParam_('exe', self.executable)
104	<	log.debug(6, "CMSSW::CMSSW(): executable = "+self.executable)
105	<	msg = "Default executable cmsRun overridden. Switch to " + self.executable
106	<	log.debug(3,msg)
107	<	except KeyError:
108	<	self.executable = 'cmsRun'
109	<	self.setParam_('exe', self.executable)
110	<	msg = "User executable not defined. Use cmsRun"
111	<	log.debug(3,msg)
112	<	pass
86	>	self.executable = cfg_params.get('CMSSW.executable','cmsRun')
87	>	log.debug(6, "CMSSW::CMSSW(): executable = "+self.executable)
88
89	<	try:
115	<	self.pset = cfg_params['CMSSW.pset']
116	<	log.debug(6, "Cmssw::Cmssw(): PSet file = "+self.pset)
117	<	if self.pset.lower() != 'none' :
118	<	if (not os.path.exists(self.pset)):
119	<	raise CrabException("User defined PSet file "+self.pset+" does not exist")
120	<	else:
121	<	self.pset = None
122	<	except KeyError:
89	>	if not cfg_params.has_key('CMSSW.pset'):
90		raise CrabException("PSet file missing. Cannot run cmsRun ")
91	+	self.pset = cfg_params['CMSSW.pset']
92	+	log.debug(6, "Cmssw::Cmssw(): PSet file = "+self.pset)
93	+	if self.pset.lower() != 'none' :
94	+	if (not os.path.exists(self.pset)):
95	+	raise CrabException("User defined PSet file "+self.pset+" does not exist")
96	+	else:
97	+	self.pset = None
98
99		# output files
100		## stuff which must be returned always via sandbox
#	Line 130 | Line 104 | class Cmssw(JobType):
104		self.output_file_sandbox.append(self.fjrFileName)
105
106		# other output files to be returned via sandbox or copied to SE
107	<	try:
108	<	self.output_file = []
109	<	tmp = cfg_params['CMSSW.output_file']
110	<	if tmp != '':
111	<	tmpOutFiles = string.split(cfg_params['CMSSW.output_file'],',')
112	<	log.debug(7, 'cmssw::cmssw(): output files '+str(tmpOutFiles))
113	<	for tmp in tmpOutFiles:
114	<	tmp=string.strip(tmp)
141	<	self.output_file.append(tmp)
142	<	pass
143	<	else:
144	<	log.message("No output file defined: only stdout/err and the CRAB Framework Job Report will be available\n")
145	<	pass
146	<	pass
147	<	except KeyError:
148	<	log.message("No output file defined: only stdout/err and the CRAB Framework Job Report will be available\n")
149	<	pass
107	>	outfileflag = False
108	>	self.output_file = []
109	>	tmp = cfg_params.get('CMSSW.output_file',None)
110	>	if tmp :
111	>	self.output_file = [x.strip() for x in tmp.split(',')]
112	>	outfileflag = True #output found
113	>	#else:
114	>	#    log.message("No output file defined: only stdout/err and the CRAB Framework Job Report will be available\n")
115
116		# script_exe file as additional file in inputSandbox
117	<	try:
118	<	self.scriptExe = cfg_params['USER.script_exe']
119	<	if self.scriptExe != '':
120	<	if not os.path.isfile(self.scriptExe):
121	<	msg ="ERROR. file "+self.scriptExe+" not found"
122	<	raise CrabException(msg)
158	<	self.additional_inbox_files.append(string.strip(self.scriptExe))
159	<	except KeyError:
160	<	self.scriptExe = ''
117	>	self.scriptExe = cfg_params.get('USER.script_exe',None)
118	>	if self.scriptExe :
119	>	if not os.path.isfile(self.scriptExe):
120	>	msg ="ERROR. file "+self.scriptExe+" not found"
121	>	raise CrabException(msg)
122	>	self.additional_inbox_files.append(string.strip(self.scriptExe))
123
162	–	#CarlosDaniele
124		if self.datasetPath == None and self.pset == None and self.scriptExe == '' :
125	<	msg ="Error. script_exe  not defined"
126	<	raise CrabException(msg)
125	>	msg ="Error. script_exe  not defined"
126	>	raise CrabException(msg)
127	>
128	>	# use parent files...
129	>	self.useParent = self.cfg_params.get('CMSSW.use_parent',False)
130
131		## additional input files
132	<	try:
132	>	if cfg_params.has_key('USER.additional_input_files'):
133		tmpAddFiles = string.split(cfg_params['USER.additional_input_files'],',')
134		for tmp in tmpAddFiles:
135		tmp = string.strip(tmp)
#	Line 182 | Line 146 | class Cmssw(JobType):
146		if not os.path.exists(file):
147		raise CrabException("Additional input file not found: "+file)
148		pass
185	–	# fname = string.split(file, '/')[-1]
186	–	# storedFile = common.work_space.pathForTgz()+'share/'+fname
187	–	# shutil.copyfile(file, storedFile)
149		self.additional_inbox_files.append(string.strip(file))
150		pass
151		pass
152		common.logger.debug(5,"Additional input files: "+str(self.additional_inbox_files))
153	<	except KeyError:
193	<	pass
194	<
195	<	# files per job
196	<	try:
197	<	if (cfg_params['CMSSW.files_per_jobs']):
198	<	raise CrabException("files_per_jobs no longer supported.  Quitting.")
199	<	except KeyError:
200	<	pass
153	>	pass
154
155		## Events per job
156	<	try:
156	>	if cfg_params.has_key('CMSSW.events_per_job'):
157		self.eventsPerJob =int( cfg_params['CMSSW.events_per_job'])
158		self.selectEventsPerJob = 1
159	<	except KeyError:
159	>	else:
160		self.eventsPerJob = -1
161		self.selectEventsPerJob = 0
162
163		## number of jobs
164	<	try:
164	>	if cfg_params.has_key('CMSSW.number_of_jobs'):
165		self.theNumberOfJobs =int( cfg_params['CMSSW.number_of_jobs'])
166		self.selectNumberOfJobs = 1
167	<	except KeyError:
167	>	else:
168		self.theNumberOfJobs = 0
169		self.selectNumberOfJobs = 0
170
171	<	try:
171	>	if cfg_params.has_key('CMSSW.total_number_of_events'):
172		self.total_number_of_events = int(cfg_params['CMSSW.total_number_of_events'])
173		self.selectTotalNumberEvents = 1
174	<	except KeyError:
174	>	if self.selectNumberOfJobs  == 1:
175	>	if (self.total_number_of_events != -1) and int(self.total_number_of_events) < int(self.theNumberOfJobs):
176	>	msg = 'Must specify at least one event per job. total_number_of_events > number_of_jobs '
177	>	raise CrabException(msg)
178	>	else:
179		self.total_number_of_events = 0
180		self.selectTotalNumberEvents = 0
181
182	<	if self.pset != None: #CarlosDaniele
182	>	if self.pset != None:
183		if ( (self.selectTotalNumberEvents + self.selectEventsPerJob + self.selectNumberOfJobs) != 2 ):
184		msg = 'Must define exactly two of total_number_of_events, events_per_job, or number_of_jobs.'
185		raise CrabException(msg)
#	Line 231 | Line 188 | class Cmssw(JobType):
188		msg = 'Must specify  number_of_jobs.'
189		raise CrabException(msg)
190
191	<	## source seed for pythia
192	<	try:
193	<	self.sourceSeed = int(cfg_params['CMSSW.pythia_seed'])
194	<	except KeyError:
195	<	self.sourceSeed = None
196	<	common.logger.debug(5,"No seed given")
197	<
198	<	try:
199	<	self.sourceSeedVtx = int(cfg_params['CMSSW.vtx_seed'])
200	<	except KeyError:
201	<	self.sourceSeedVtx = None
202	<	common.logger.debug(5,"No vertex seed given")
191	>	## New method of dealing with seeds
192	>	self.incrementSeeds = []
193	>	self.preserveSeeds = []
194	>	if cfg_params.has_key('CMSSW.preserve_seeds'):
195	>	tmpList = cfg_params['CMSSW.preserve_seeds'].split(',')
196	>	for tmp in tmpList:
197	>	tmp.strip()
198	>	self.preserveSeeds.append(tmp)
199	>	if cfg_params.has_key('CMSSW.increment_seeds'):
200	>	tmpList = cfg_params['CMSSW.increment_seeds'].split(',')
201	>	for tmp in tmpList:
202	>	tmp.strip()
203	>	self.incrementSeeds.append(tmp)
204	>
205	>	## FUTURE: Can remove in CRAB 2.4.0
206	>	self.sourceSeed    = cfg_params.get('CMSSW.pythia_seed',None)
207	>	self.sourceSeedVtx = cfg_params.get('CMSSW.vtx_seed',None)
208	>	self.sourceSeedG4  = cfg_params.get('CMSSW.g4_seed',None)
209	>	self.sourceSeedMix = cfg_params.get('CMSSW.mix_seed',None)
210	>	if self.sourceSeed or self.sourceSeedVtx or self.sourceSeedG4 or self.sourceSeedMix:
211	>	msg = 'pythia_seed, vtx_seed, g4_seed, and mix_seed are no longer valid settings. You must use increment_seeds or preserve_seeds'
212	>	raise CrabException(msg)
213
214	<	try:
248	<	self.sourceSeedG4 = int(cfg_params['CMSSW.g4_seed'])
249	<	except KeyError:
250	<	self.sourceSeedG4 = None
251	<	common.logger.debug(5,"No g4 sim hits seed given")
214	>	self.firstRun = cfg_params.get('CMSSW.first_run',None)
215
216	<	try:
217	<	self.sourceSeedMix = int(cfg_params['CMSSW.mix_seed'])
218	<	except KeyError:
256	<	self.sourceSeedMix = None
257	<	common.logger.debug(5,"No mix seed given")
258	<
259	<	try:
260	<	self.firstRun = int(cfg_params['CMSSW.first_run'])
261	<	except KeyError:
262	<	self.firstRun = None
263	<	common.logger.debug(5,"No first run given")
264	<	if self.pset != None: #CarlosDaniele
265	<	import PsetManipulator as pp
266	<	PsetEdit = pp.PsetManipulator(self.pset) #Daniele Pset
216	>	# Copy/return
217	>	self.copy_data = int(cfg_params.get('USER.copy_data',0))
218	>	self.return_data = int(cfg_params.get('USER.return_data',0))
219
220		#DBSDLS-start
221		## Initialize the variables that are extracted from DBS/DLS and needed in other places of the code
#	Line 277 | Line 229 | class Cmssw(JobType):
229		blockSites = self.DataDiscoveryAndLocation(cfg_params)
230		#DBSDLS-end
231
280	–	self.tgzNameWithPath = self.getTarBall(self.executable)
281	–
232		## Select Splitting
233		if self.selectNoInput:
234	<	if self.pset == None: #CarlosDaniele
234	>	if self.pset == None:
235		self.jobSplittingForScript()
236		else:
237		self.jobSplittingNoInput()
238		else:
239		self.jobSplittingByBlocks(blockSites)
240
241	<	# modify Pset
242	<	if self.pset != None: #CarlosDaniele
243	<	try:
244	<	if (self.datasetPath): # standard job
245	<	# allow to processa a fraction of events in a file
246	<	PsetEdit.inputModule("INPUTFILE")
247	<	PsetEdit.maxEvent(0)
248	<	PsetEdit.skipEvent(0)
249	<	else:  # pythia like job
241	>	# modify Pset only the first time
242	>	if isNew:
243	>	if self.pset != None:
244	>	import PsetManipulator as pp
245	>	PsetEdit = pp.PsetManipulator(self.pset)
246	>	try:
247	>	# Add FrameworkJobReport to parameter-set, set max events.
248	>	# Reset later for data jobs by writeCFG which does all modifications
249	>	PsetEdit.addCrabFJR(self.fjrFileName) # FUTURE: Job report addition not needed by CMSSW>1.5
250		PsetEdit.maxEvent(self.eventsPerJob)
251	<	if (self.firstRun):
252	<	PsetEdit.pythiaFirstRun(0)  #First Run
253	<	if (self.sourceSeed) :
254	<	PsetEdit.pythiaSeed(0)
255	<	if (self.sourceSeedVtx) :
256	<	PsetEdit.vtxSeed(0)
257	<	if (self.sourceSeedG4) :
258	<	PsetEdit.g4Seed(0)
259	<	if (self.sourceSeedMix) :
260	<	PsetEdit.mixSeed(0)
261	<	# add FrameworkJobReport to parameter-set
262	<	PsetEdit.addCrabFJR(self.fjrFileName)
263	<	PsetEdit.psetWriter(self.configFilename())
264	<	except:
265	<	msg='Error while manipuliating ParameterSet: exiting...'
266	<	raise CrabException(msg)
251	>	PsetEdit.psetWriter(self.configFilename())
252	>	## If present, add TFileService to output files
253	>	if not int(cfg_params.get('CMSSW.skip_TFileService_output',0)):
254	>	tfsOutput = PsetEdit.getTFileService()
255	>	if tfsOutput:
256	>	if tfsOutput in self.output_file:
257	>	common.logger.debug(5,"Output from TFileService "+tfsOutput+" already in output files")
258	>	else:
259	>	outfileflag = True #output found
260	>	self.output_file.append(tfsOutput)
261	>	common.logger.message("Adding "+tfsOutput+" to output files (from TFileService)")
262	>	pass
263	>	pass
264	>	## If present and requested, add PoolOutputModule to output files
265	>	if int(cfg_params.get('CMSSW.get_edm_output',0)):
266	>	edmOutput = PsetEdit.getPoolOutputModule()
267	>	if edmOutput:
268	>	if edmOutput in self.output_file:
269	>	common.logger.debug(5,"Output from PoolOutputModule "+edmOutput+" already in output files")
270	>	else:
271	>	self.output_file.append(edmOutput)
272	>	common.logger.message("Adding "+edmOutput+" to output files (from PoolOutputModule)")
273	>	pass
274	>	pass
275	>	except CrabException:
276	>	msg='Error while manipulating ParameterSet: exiting...'
277	>	raise CrabException(msg)
278	>	## Prepare inputSandbox TarBall (only the first time)
279	>	self.tgzNameWithPath = self.getTarBall(self.executable)
280
281		def DataDiscoveryAndLocation(self, cfg_params):
282
#	Line 326 | Line 289 | class Cmssw(JobType):
289		## Contact the DBS
290		common.logger.message("Contacting Data Discovery Services ...")
291		try:
292	<
330	<	self.pubdata=DataDiscovery.DataDiscovery(datasetPath, cfg_params)
292	>	self.pubdata=DataDiscovery.DataDiscovery(datasetPath, cfg_params,self.skip_blocks)
293		self.pubdata.fetchDBSInfo()
294
295		except DataDiscovery.NotExistingDatasetError, ex :
#	Line 343 | Line 305 | class Cmssw(JobType):
305		self.filesbyblock=self.pubdata.getFiles()
306		self.eventsbyblock=self.pubdata.getEventsPerBlock()
307		self.eventsbyfile=self.pubdata.getEventsPerFile()
308	+	self.parentFiles=self.pubdata.getParent()
309
310		## get max number of events
311	<	self.maxEvents=self.pubdata.getMaxEvents() ##  self.maxEvents used in Creator.py
311	>	self.maxEvents=self.pubdata.getMaxEvents()
312
313		## Contact the DLS and build a list of sites hosting the fileblocks
314		try:
#	Line 419 | Line 382 | class Cmssw(JobType):
382		else :
383		totalNumberOfJobs = self.ncjobs
384
422	–
385		blocks = blockSites.keys()
386		blockCount = 0
387		# Backup variable in case self.maxEvents counted events in a non-included block
#	Line 460 | Line 422 | class Cmssw(JobType):
422
423		# ---- Iterate over the files in the block until we've met the requested ---- #
424		# ---- total # of events or we've gone over all the files in this block  ---- #
425	+	pString=''
426		while ( (eventsRemaining > 0) and (fileCount < numFilesInBlock) and (jobCount < totalNumberOfJobs) ):
427		file = files[fileCount]
428	+	if self.useParent:
429	+	parent = self.parentFiles[file]
430	+	for f in parent :
431	+	pString += '\\\"' + f + '\\\"\,'
432	+	common.logger.debug(6, "File "+str(file)+" has the following parents: "+str(parent))
433	+	common.logger.write("File "+str(file)+" has the following parents: "+str(parent))
434		if newFile :
435		try:
436		numEventsInFile = self.eventsbyfile[file]
#	Line 474 | Line 443 | class Cmssw(JobType):
443		except KeyError:
444		common.logger.message("File "+str(file)+" has unknown number of events: skipping")
445
446	<
446	>	eventsPerJobRequested = min(eventsPerJobRequested, eventsRemaining)
447		# if less events in file remain than eventsPerJobRequested
448	<	if ( filesEventCount - jobSkipEventCount < eventsPerJobRequested ) :
448	>	if ( filesEventCount - jobSkipEventCount < eventsPerJobRequested):
449		# if last file in block
450		if ( fileCount == numFilesInBlock-1 ) :
451		# end job using last file, use remaining events in block
452		# close job and touch new file
453		fullString = parString[:-2]
454	<	list_of_lists.append([fullString,str(-1),str(jobSkipEventCount)])
454	>	if self.useParent:
455	>	fullParentString = pString[:-2]
456	>	list_of_lists.append([fullString,fullParentString,str(-1),str(jobSkipEventCount)])
457	>	else:
458	>	list_of_lists.append([fullString,str(-1),str(jobSkipEventCount)])
459		common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(filesEventCount - jobSkipEventCount)+" events (last file in block).")
460		self.jobDestination.append(blockSites[block])
461		common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
#	Line 494 | Line 467 | class Cmssw(JobType):
467		eventsRemaining = eventsRemaining - filesEventCount + jobSkipEventCount
468		jobSkipEventCount = 0
469		# reset file
470	+	pString = ""
471		parString = ""
472		filesEventCount = 0
473		newFile = 1
#	Line 506 | Line 480 | class Cmssw(JobType):
480		elif ( filesEventCount - jobSkipEventCount == eventsPerJobRequested ) :
481		# close job and touch new file
482		fullString = parString[:-2]
483	<	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
483	>	if self.useParent:
484	>	fullParentString = pString[:-2]
485	>	list_of_lists.append([fullString,fullParentString,str(eventsPerJobRequested),str(jobSkipEventCount)])
486	>	else:
487	>	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
488		common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.")
489		self.jobDestination.append(blockSites[block])
490		common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
#	Line 517 | Line 495 | class Cmssw(JobType):
495		eventsRemaining = eventsRemaining - eventsPerJobRequested
496		jobSkipEventCount = 0
497		# reset file
498	+	pString = ""
499		parString = ""
500		filesEventCount = 0
501		newFile = 1
#	Line 526 | Line 505 | class Cmssw(JobType):
505		else :
506		# close job but don't touch new file
507		fullString = parString[:-2]
508	<	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
508	>	if self.useParent:
509	>	fullParentString = pString[:-2]
510	>	list_of_lists.append([fullString,fullParentString,str(eventsPerJobRequested),str(jobSkipEventCount)])
511	>	else:
512	>	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
513		common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.")
514		self.jobDestination.append(blockSites[block])
515		common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
#	Line 540 | Line 523 | class Cmssw(JobType):
523		jobSkipEventCount = eventsPerJobRequested - (filesEventCount - jobSkipEventCount - self.eventsbyfile[file])
524		# remove all but the last file
525		filesEventCount = self.eventsbyfile[file]
526	<	parString = ""
527	<	parString += '\\\"' + file + '\\\"\,'
526	>	if self.useParent:
527	>	for f in parent : pString += '\\\"' + f + '\\\"\,'
528	>	parString = '\\\"' + file + '\\\"\,'
529		pass # END if
530		pass # END while (iterate over files in the block)
531		pass # END while (iterate over blocks in the dataset)
#	Line 561 | Line 545 | class Cmssw(JobType):
545		for block in blocks:
546		if block in jobsOfBlock.keys() :
547		blockCounter += 1
548	<	screenOutput += "Block %5i: jobs %20s: sites: %s\n" % (blockCounter,spanRanges(jobsOfBlock[block]),','.join(self.blackWhiteListParser.checkWhiteList(self.blackWhiteListParser.checkBlackList(blockSites[block],block),block)))
548	>	screenOutput += "Block %5i: jobs %20s: sites: %s\n" % (blockCounter,spanRanges(jobsOfBlock[block]),
549	>	','.join(self.blackWhiteListParser.checkWhiteList(self.blackWhiteListParser.checkBlackList(blockSites[block],block),block)))
550		if len(self.blackWhiteListParser.checkWhiteList(self.blackWhiteListParser.checkBlackList(blockSites[block],block),block)) == 0:
551		noSiteBlock.append( spanRanges(jobsOfBlock[block]) )
552		bloskNoSite.append( blockCounter )
#	Line 581 | Line 566 | class Cmssw(JobType):
566		for range_jobs in noSiteBlock:
567		msg += str(range_jobs) + virgola
568		msg += '\n               will not be submitted and this block of data can not be analyzed!\n'
569	+	if self.cfg_params.has_key('EDG.se_white_list'):
570	+	msg += 'WARNING: SE White List: '+self.cfg_params['EDG.se_white_list']+'\n'
571	+	msg += '(Hint: By whitelisting you force the job to run at this particular site(s).\n'
572	+	msg += 'Please check if the dataset is available at this site!)\n'
573	+	if self.cfg_params.has_key('EDG.ce_white_list'):
574	+	msg += 'WARNING: CE White List: '+self.cfg_params['EDG.ce_white_list']+'\n'
575	+	msg += '(Hint: By whitelisting you force the job to run at this particular site(s).\n'
576	+	msg += 'Please check if the dataset is available at this site!)\n'
577	+
578		common.logger.message(msg)
579
580		self.list_of_args = list_of_lists
#	Line 629 | Line 623 | class Cmssw(JobType):
623		self.list_of_args = []
624		for i in range(self.total_number_of_jobs):
625		## Since there is no input, any site is good
632	–	# self.jobDestination.append(["Any"])
626		self.jobDestination.append([""]) #must be empty to write correctly the xml
627		args=[]
628		if (self.firstRun):
629	<	## pythia first run
637	<	#self.list_of_args.append([(str(self.firstRun)+str(i))])
629	>	## pythia first run
630		args.append(str(self.firstRun)+str(i))
639	–	else:
640	–	## no first run
641	–	#self.list_of_args.append([str(i)])
642	–	args.append(str(i))
643	–	if (self.sourceSeed):
644	–	args.append(str(self.sourceSeed)+str(i))
645	–	if (self.sourceSeedVtx):
646	–	## + vtx random seed
647	–	args.append(str(self.sourceSeedVtx)+str(i))
648	–	if (self.sourceSeedG4):
649	–	## + G4 random seed
650	–	args.append(str(self.sourceSeedG4)+str(i))
651	–	if (self.sourceSeedMix):
652	–	## + Mix random seed
653	–	args.append(str(self.sourceSeedMix)+str(i))
654	–	pass
655	–	pass
631		self.list_of_args.append(args)
657	–	pass
658	–
659	–	# print self.list_of_args
632
633		return
634
635
636	<	def jobSplittingForScript(self):#CarlosDaniele
636	>	def jobSplittingForScript(self):
637		"""
638		Perform job splitting based on number of job
639		"""
#	Line 677 | Line 649 | class Cmssw(JobType):
649		# argument is seed number.$i
650		self.list_of_args = []
651		for i in range(self.total_number_of_jobs):
680	–	## Since there is no input, any site is good
681	–	# self.jobDestination.append(["Any"])
652		self.jobDestination.append([""])
683	–	## no random seed
653		self.list_of_args.append([str(i)])
654		return
655
656	<	def split(self, jobParams):
656	>	def split(self, jobParams,firstJobID):
657
689	–	common.jobDB.load()
690	–	#### Fabio
658		njobs = self.total_number_of_jobs
659		arglist = self.list_of_args
660		# create the empty structure
661		for i in range(njobs):
662		jobParams.append("")
663
664	<	for job in range(njobs):
665	<	jobParams[job] = arglist[job]
666	<	# print str(arglist[job])
667	<	# print jobParams[job]
668	<	common.jobDB.setArguments(job, jobParams[job])
669	<	common.logger.debug(5,"Job "+str(job)+" Destination: "+str(self.jobDestination[job]))
670	<	common.jobDB.setDestination(job, self.jobDestination[job])
664	>	listID=[]
665	>	listField=[]
666	>	for id in range(njobs):
667	>	job = id + int(firstJobID)
668	>	jobParams[id] = arglist[id]
669	>	listID.append(job+1)
670	>	job_ToSave ={}
671	>	concString = ' '
672	>	argu=''
673	>	if len(jobParams[id]):
674	>	argu +=   concString.join(jobParams[id] )
675	>	job_ToSave['arguments']= str(job+1)+' '+argu
676	>	job_ToSave['dlsDestination']= self.jobDestination[id]
677	>	listField.append(job_ToSave)
678	>	msg="Job "+str(job)+" Arguments:   "+str(job+1)+" "+argu+"\n"  \
679	>	+"                     Destination: "+str(self.jobDestination[id])
680	>	common.logger.debug(5,msg)
681	>	common._db.updateJob_(listID,listField)
682	>	self.argsList = (len(jobParams[0])+1)
683
705	–	common.jobDB.save()
684		return
685
708	–	def getJobTypeArguments(self, nj, sched):
709	–	result = ''
710	–	for i in common.jobDB.arguments(nj):
711	–	result=result+str(i)+" "
712	–	return result
713	–
686		def numberOfJobs(self):
715	–	# Fabio
687		return self.total_number_of_jobs
688
689		def getTarBall(self, exe):
690		"""
691		Return the TarBall with lib and exe
692		"""
722	–
723	–	# if it exist, just return it
724	–	#
725	–	# Marco. Let's start to use relative path for Boss XML files
726	–	#
693		self.tgzNameWithPath = common.work_space.pathForTgz()+'share/'+self.tgz_name
694		if os.path.exists(self.tgzNameWithPath):
695		return self.tgzNameWithPath
#	Line 737 | Line 703 | class Cmssw(JobType):
703
704		# First of all declare the user Scram area
705		swArea = self.scram.getSWArea_()
740	–	#print "swArea = ", swArea
741	–	# swVersion = self.scram.getSWVersion()
742	–	# print "swVersion = ", swVersion
706		swReleaseTop = self.scram.getReleaseTop_()
744	–	#print "swReleaseTop = ", swReleaseTop
707
708		## check if working area is release top
709		if swReleaseTop == '' or swArea == swReleaseTop:
710	+	common.logger.debug(3,"swArea = "+swArea+" swReleaseTop ="+swReleaseTop)
711		return
712
713		import tarfile
#	Line 786 | Line 749 | class Cmssw(JobType):
749		tar.add(module,moduleDir)
750
751		## Now check if any data dir(s) is present
752	<	swAreaLen=len(swArea)
753	<	for root, dirs, files in os.walk(swArea):
754	<	if "data" in dirs:
755	<	common.logger.debug(5,"data "+root+"/data"+" to be tarred")
756	<	tar.add(root+"/data",root[swAreaLen:]+"/data")
757	<
758	<	## Add ProdAgent dir to tar
759	<	paDir = 'ProdAgentApi'
760	<	pa = os.environ['CRABDIR'] + '/' + 'ProdAgentApi'
761	<	if os.path.isdir(pa):
762	<	tar.add(pa,paDir)
763	<
764	<	### FEDE FOR DBS PUBLICATION
765	<	## Add PRODCOMMON dir to tar
766	<	prodcommonDir = 'ProdCommon'
767	<	prodcommonPath = os.environ['CRABDIR'] + '/' + 'ProdCommon'
768	<	if os.path.isdir(prodcommonPath):
769	<	tar.add(prodcommonPath,prodcommonDir)
770	<	#############################
752	>	self.dataExist = False
753	>	todo_list = [(i, i) for i in  os.listdir(swArea+"/src")]
754	>	while len(todo_list):
755	>	entry, name = todo_list.pop()
756	>	if name.startswith('crab_0_') or  name.startswith('.') or name == 'CVS':
757	>	continue
758	>	if os.path.isdir(swArea+"/src/"+entry):
759	>	entryPath = entry + '/'
760	>	todo_list += [(entryPath + i, i) for i in  os.listdir(swArea+"/src/"+entry)]
761	>	if name == 'data':
762	>	self.dataExist=True
763	>	common.logger.debug(5,"data "+entry+" to be tarred")
764	>	tar.add(swArea+"/src/"+entry,"src/"+entry)
765	>	pass
766	>	pass
767	>
768	>	### CMSSW ParameterSet
769	>	if not self.pset is None:
770	>	cfg_file = common.work_space.jobDir()+self.configFilename()
771	>	tar.add(cfg_file,self.configFilename())
772	>	common.logger.debug(5,"File added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
773	>
774
775	+	## Add ProdCommon dir to tar
776	+	prodcommonDir = './'
777	+	prodcommonPath = os.environ['CRABDIR'] + '/' + 'external/'
778	+	neededStuff = ['ProdCommon/__init__.py','ProdCommon/FwkJobRep', 'ProdCommon/CMSConfigTools','ProdCommon/Core','ProdCommon/MCPayloads', 'IMProv']
779	+	for file in neededStuff:
780	+	tar.add(prodcommonPath+file,prodcommonDir+file)
781		common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
782	+
783	+	##### ML stuff
784	+	ML_file_list=['report.py', 'DashboardAPI.py', 'Logger.py', 'ProcInfo.py', 'apmon.py']
785	+	path=os.environ['CRABDIR'] + '/python/'
786	+	for file in ML_file_list:
787	+	tar.add(path+file,file)
788	+	common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
789	+
790	+	##### Utils
791	+	Utils_file_list=['parseCrabFjr.py','writeCfg.py', 'fillCrabFjr.py']
792	+	for file in Utils_file_list:
793	+	tar.add(path+file,file)
794	+	common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
795	+
796	+	##### AdditionalFiles
797	+	for file in self.additional_inbox_files:
798	+	tar.add(file,string.split(file,'/')[-1])
799	+	common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
800	+
801		tar.close()
802	<	except :
803	<	raise CrabException('Could not create tar-ball')
802	>	except IOError:
803	>	raise CrabException('Could not create tar-ball '+self.tgzNameWithPath)
804	>	except tarfile.TarError:
805	>	raise CrabException('Could not create tar-ball '+self.tgzNameWithPath)
806
807		## check for tarball size
808		tarballinfo = os.stat(self.tgzNameWithPath)
#	Line 817 | Line 810 | class Cmssw(JobType):
810		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.')
811
812		## create tar-ball with ML stuff
820	–	self.MLtgzfile =  common.work_space.pathForTgz()+'share/MLfiles.tgz'
821	–	try:
822	–	tar = tarfile.open(self.MLtgzfile, "w:gz")
823	–	path=os.environ['CRABDIR'] + '/python/'
824	–	for file in ['report.py', 'DashboardAPI.py', 'Logger.py', 'ProcInfo.py', 'apmon.py', 'parseCrabFjr.py']:
825	–	tar.add(path+file,file)
826	–	common.logger.debug(5,"Files added to "+self.MLtgzfile+" : "+str(tar.getnames()))
827	–	tar.close()
828	–	except :
829	–	raise CrabException('Could not create ML files tar-ball')
813
814	<	return
832	<
833	<	def additionalInputFileTgz(self):
834	<	"""
835	<	Put all additional files into a tar ball and return its name
836	<	"""
837	<	import tarfile
838	<	tarName=  common.work_space.pathForTgz()+'share/'+self.additional_tgz_name
839	<	tar = tarfile.open(tarName, "w:gz")
840	<	for file in self.additional_inbox_files:
841	<	tar.add(file,string.split(file,'/')[-1])
842	<	common.logger.debug(5,"Files added to "+self.additional_tgz_name+" : "+str(tar.getnames()))
843	<	tar.close()
844	<	return tarName
845	<
846	<	def wsSetupEnvironment(self, nj):
814	>	def wsSetupEnvironment(self, nj=0):
815		"""
816		Returns part of a job script which prepares
817		the execution environment for the job 'nj'.
818		"""
819	+	if (self.CMSSW_major >= 2 and self.CMSSW_minor >= 1) or (self.CMSSW_major >= 3):
820	+	psetName = 'pset.py'
821	+	else:
822	+	psetName = 'pset.cfg'
823		# Prepare JobType-independent part
824	<	txt = ''
824	>	txt = '\n#Written by cms_cmssw::wsSetupEnvironment\n'
825		txt += 'echo ">>> setup environment"\n'
826		txt += 'if [ $middleware == LCG ]; then \n'
827		txt += self.wsSetupCMSLCGEnvironment_()
828		txt += 'elif [ $middleware == OSG ]; then\n'
829		txt += '    WORKING_DIR=`/bin/mktemp  -d $OSG_WN_TMP/cms_XXXXXXXXXXXX`\n'
830		txt += '    if [ ! $? == 0 ] ;then\n'
831	<	txt += '        echo "SET_CMS_ENV 10016 ==> OSG $WORKING_DIR could not be created on WN `hostname`"\n'
832	<	txt += '        echo "JOB_EXIT_STATUS = 10016"\n'
833	<	txt += '        echo "JobExitCode=10016" | tee -a $RUNTIME_AREA/$repo\n'
862	<	txt += '        dumpStatus $RUNTIME_AREA/$repo\n'
863	<	txt += '        exit 1\n'
831	>	txt += '        echo "ERROR ==> OSG $WORKING_DIR could not be created on WN `hostname`"\n'
832	>	txt += '        job_exit_code=10016\n'
833	>	txt += '        func_exit\n'
834		txt += '    fi\n'
835		txt += '    echo ">>> Created working directory: $WORKING_DIR"\n'
836		txt += '\n'
#	Line 868 | Line 838 | class Cmssw(JobType):
838		txt += '    cd $WORKING_DIR\n'
839		txt += '    echo ">>> current directory (WORKING_DIR): $WORKING_DIR"\n'
840		txt += self.wsSetupCMSOSGEnvironment_()
871	–	#txt += '    echo "### Set SCRAM ARCH to ' + self.executable_arch + ' ###"\n'
872	–	#txt += '    export SCRAM_ARCH='+self.executable_arch+'\n'
841		txt += 'fi\n'
842
843		# Prepare JobType-specific part
#	Line 880 | Line 848 | class Cmssw(JobType):
848		txt += scram+' project CMSSW '+self.version+'\n'
849		txt += 'status=$?\n'
850		txt += 'if [ $status != 0 ] ; then\n'
851	<	txt += '    echo "SET_EXE_ENV 10034 ==>ERROR CMSSW '+self.version+' not found on `hostname`" \n'
852	<	txt += '    echo "JOB_EXIT_STATUS = 10034"\n'
853	<	txt += '    echo "JobExitCode=10034" | tee -a $RUNTIME_AREA/$repo\n'
886	<	txt += '    dumpStatus $RUNTIME_AREA/$repo\n'
887	<	txt += '    if [ $middleware == OSG ]; then \n'
888	<	txt += '        cd $RUNTIME_AREA\n'
889	<	txt += '        echo ">>> current directory (RUNTIME_AREA): $RUNTIME_AREA"\n'
890	<	txt += '        echo ">>> Remove working directory: $WORKING_DIR"\n'
891	<	txt += '        /bin/rm -rf $WORKING_DIR\n'
892	<	txt += '        if [ -d $WORKING_DIR ] ;then\n'
893	<	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'
894	<	txt += '            echo "JOB_EXIT_STATUS = 10018"\n'
895	<	txt += '            echo "JobExitCode=10018" | tee -a $RUNTIME_AREA/$repo\n'
896	<	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
897	<	txt += '        fi\n'
898	<	txt += '    fi \n'
899	<	txt += '    exit 1 \n'
851	>	txt += '    echo "ERROR ==> CMSSW '+self.version+' not found on `hostname`" \n'
852	>	txt += '    job_exit_code=10034\n'
853	>	txt += '    func_exit\n'
854		txt += 'fi \n'
855		txt += 'cd '+self.version+'\n'
902	–	########## FEDE FOR DBS2 ######################
856		txt += 'SOFTWARE_DIR=`pwd`\n'
857		txt += 'echo ">>> current directory (SOFTWARE_DIR): $SOFTWARE_DIR" \n'
905	–	###############################################
906	–	### needed grep for bug in scramv1 ###
907	–	txt += scram+' runtime -sh\n'
858		txt += 'eval `'+scram+' runtime -sh | grep -v SCRAMRT_LSB_JOBNAME`\n'
859	<	txt += 'echo $PATH\n'
860	<
859	>	txt += 'if [ $? != 0 ] ; then\n'
860	>	txt += '    echo "ERROR ==> Problem with the command: "\n'
861	>	txt += '    echo "eval \`'+scram+' runtime -sh | grep -v SCRAMRT_LSB_JOBNAME \` at `hostname`"\n'
862	>	txt += '    job_exit_code=10034\n'
863	>	txt += '    func_exit\n'
864	>	txt += 'fi \n'
865		# Handle the arguments:
866		txt += "\n"
867		txt += "## number of arguments (first argument always jobnumber)\n"
868		txt += "\n"
869	<	txt += "if [ $nargs -lt 2 ]\n"
869	>	txt += "if [ $nargs -lt "+str(self.argsList)+" ]\n"
870		txt += "then\n"
871	<	txt += "    echo 'SET_EXE_ENV 1 ==> ERROR Too few arguments' +$nargs+ \n"
872	<	txt += '    echo "JOB_EXIT_STATUS = 50113"\n'
873	<	txt += '    echo "JobExitCode=50113" | tee -a $RUNTIME_AREA/$repo\n'
920	<	txt += '    dumpStatus $RUNTIME_AREA/$repo\n'
921	<	txt += '    if [ $middleware == OSG ]; then \n'
922	<	txt += '        cd $RUNTIME_AREA\n'
923	<	txt += '        echo ">>> current directory (RUNTIME_AREA): $RUNTIME_AREA"\n'
924	<	txt += '        echo ">>> Remove working directory: $WORKING_DIR"\n'
925	<	txt += '        /bin/rm -rf $WORKING_DIR\n'
926	<	txt += '        if [ -d $WORKING_DIR ] ;then\n'
927	<	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'
928	<	txt += '            echo "JOB_EXIT_STATUS = 50114"\n'
929	<	txt += '            echo "JobExitCode=50114" | tee -a $RUNTIME_AREA/$repo\n'
930	<	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
931	<	txt += '        fi\n'
932	<	txt += '    fi \n'
933	<	txt += "    exit 1\n"
871	>	txt += "    echo 'ERROR ==> Too few arguments' +$nargs+ \n"
872	>	txt += '    job_exit_code=50113\n'
873	>	txt += "    func_exit\n"
874		txt += "fi\n"
875		txt += "\n"
876
877		# Prepare job-specific part
878		job = common.job_list[nj]
939	–	### FEDE FOR DBS OUTPUT PUBLICATION
879		if (self.datasetPath):
880		txt += '\n'
881		txt += 'DatasetPath='+self.datasetPath+'\n'
882
883		datasetpath_split = self.datasetPath.split("/")
884	<
884	>	### FEDE FOR NEW LFN ###
885	>	self.primaryDataset = datasetpath_split[1]
886	>	########################
887		txt += 'PrimaryDataset='+datasetpath_split[1]+'\n'
888		txt += 'DataTier='+datasetpath_split[2]+'\n'
889		txt += 'ApplicationFamily=cmsRun\n'
890
891		else:
892		txt += 'DatasetPath=MCDataTier\n'
893	+	### FEDE FOR NEW LFN ###
894	+	self.primaryDataset = 'null'
895	+	########################
896		txt += 'PrimaryDataset=null\n'
897		txt += 'DataTier=null\n'
898		txt += 'ApplicationFamily=MCDataTier\n'
899	<	if self.pset != None: #CarlosDaniele
899	>	if self.pset != None:
900		pset = os.path.basename(job.configFilename())
901		txt += '\n'
902		txt += 'cp  $RUNTIME_AREA/'+pset+' .\n'
903		if (self.datasetPath): # standard job
904	<	txt += 'InputFiles=${args[1]}\n'
905	<	txt += 'MaxEvents=${args[2]}\n'
906	<	txt += 'SkipEvents=${args[3]}\n'
904	>	txt += 'InputFiles=${args[1]}; export InputFiles\n'
905	>	if (self.useParent):
906	>	txt += 'ParentFiles=${args[2]}; export ParentFiles\n'
907	>	txt += 'MaxEvents=${args[3]}; export MaxEvents\n'
908	>	txt += 'SkipEvents=${args[4]}; export SkipEvents\n'
909	>	else:
910	>	txt += 'MaxEvents=${args[2]}; export MaxEvents\n'
911	>	txt += 'SkipEvents=${args[3]}; export SkipEvents\n'
912		txt += 'echo "Inputfiles:<$InputFiles>"\n'
913	<	txt += 'sed "s#\'INPUTFILE\'#$InputFiles#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
913	>	if (self.useParent): txt += 'echo "ParentFiles:<$ParentFiles>"\n'
914		txt += 'echo "MaxEvents:<$MaxEvents>"\n'
966	–	txt += 'sed "s#int32 input = 0#int32 input = $MaxEvents#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
915		txt += 'echo "SkipEvents:<$SkipEvents>"\n'
968	–	txt += 'sed "s#uint32 skipEvents = 0#uint32 skipEvents = $SkipEvents#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
916		else:  # pythia like job
917	<	seedIndex=1
917	>	txt += 'PreserveSeeds='  + ','.join(self.preserveSeeds)  + '; export PreserveSeeds\n'
918	>	txt += 'IncrementSeeds=' + ','.join(self.incrementSeeds) + '; export IncrementSeeds\n'
919	>	txt += 'echo "PreserveSeeds: <$PreserveSeeds>"\n'
920	>	txt += 'echo "IncrementSeeds:<$IncrementSeeds>"\n'
921		if (self.firstRun):
922	<	txt += 'FirstRun=${args['+str(seedIndex)+']}\n'
922	>	txt += 'FirstRun=${args[1]}; export FirstRun\n'
923		txt += 'echo "FirstRun: <$FirstRun>"\n'
974	–	txt += 'sed "s#uint32 firstRun = 0#uint32 firstRun = $FirstRun#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
975	–	seedIndex=seedIndex+1
924
925	<	if (self.sourceSeed):
978	<	txt += 'Seed=${args['+str(seedIndex)+']}\n'
979	<	txt += 'sed "s#uint32 sourceSeed = 0#uint32 sourceSeed = $Seed#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
980	<	seedIndex=seedIndex+1
981	<	## the following seeds are not always present
982	<	if (self.sourceSeedVtx):
983	<	txt += 'VtxSeed=${args['+str(seedIndex)+']}\n'
984	<	txt += 'echo "VtxSeed: <$VtxSeed>"\n'
985	<	txt += 'sed "s#uint32 VtxSmeared = 0#uint32 VtxSmeared = $VtxSeed#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
986	<	seedIndex += 1
987	<	if (self.sourceSeedG4):
988	<	txt += 'G4Seed=${args['+str(seedIndex)+']}\n'
989	<	txt += 'echo "G4Seed: <$G4Seed>"\n'
990	<	txt += 'sed "s#uint32 g4SimHits = 0#uint32 g4SimHits = $G4Seed#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
991	<	seedIndex += 1
992	<	if (self.sourceSeedMix):
993	<	txt += 'mixSeed=${args['+str(seedIndex)+']}\n'
994	<	txt += 'echo "MixSeed: <$mixSeed>"\n'
995	<	txt += 'sed "s#uint32 mix = 0#uint32 mix = $mixSeed#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
996	<	seedIndex += 1
997	<	pass
998	<	pass
999	<	txt += 'mv -f '+pset+' pset.cfg\n'
925	>	txt += 'mv -f ' + pset + ' ' + psetName + '\n'
926
1001	–	if len(self.additional_inbox_files) > 0:
1002	–	txt += 'if [ -e $RUNTIME_AREA/'+self.additional_tgz_name+' ] ; then\n'
1003	–	txt += '  tar xzvf $RUNTIME_AREA/'+self.additional_tgz_name+'\n'
1004	–	txt += 'fi\n'
1005	–	pass
927
928	<	if self.pset != None: #CarlosDaniele
928	>	if self.pset != None:
929	>	# FUTURE: Can simply for 2_1_x and higher
930		txt += '\n'
931	<	txt += 'echo "***** cat pset.cfg *********"\n'
932	<	txt += 'cat pset.cfg\n'
933	<	txt += 'echo "****** end pset.cfg ********"\n'
934	<	txt += '\n'
935	<	### FEDE FOR DBS OUTPUT PUBLICATION
936	<	txt += 'PSETHASH=`EdmConfigHash < pset.cfg` \n'
931	>	if self.debug_wrapper==True:
932	>	txt += 'echo "***** cat ' + psetName + ' *********"\n'
933	>	txt += 'cat ' + psetName + '\n'
934	>	txt += 'echo "****** end ' + psetName + ' ********"\n'
935	>	txt += '\n'
936	>	if (self.CMSSW_major >= 2 and self.CMSSW_minor >= 1) or (self.CMSSW_major >= 3):
937	>	txt += 'PSETHASH=`edmConfigHash ' + psetName + '` \n'
938	>	else:
939	>	txt += 'PSETHASH=`edmConfigHash < ' + psetName + '` \n'
940		txt += 'echo "PSETHASH = $PSETHASH" \n'
1016	–	##############
941		txt += '\n'
942		return txt
943
944	<	def wsBuildExe(self, nj=0):
944	>	def wsUntarSoftware(self, nj=0):
945		"""
946		Put in the script the commands to build an executable
947		or a library.
948		"""
949
950	<	txt = ""
950	>	txt = '\n#Written by cms_cmssw::wsUntarSoftware\n'
951
952		if os.path.isfile(self.tgzNameWithPath):
953		txt += 'echo ">>> tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+' :" \n'
954		txt += 'tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+'\n'
955	+	if  self.debug_wrapper:
956	+	txt += 'ls -Al \n'
957		txt += 'untar_status=$? \n'
958		txt += 'if [ $untar_status -ne 0 ]; then \n'
959	<	txt += '   echo "SET_EXE 1 ==> ERROR Untarring .tgz file failed"\n'
960	<	txt += '   echo "JOB_EXIT_STATUS = $untar_status" \n'
961	<	txt += '   echo "JobExitCode=$untar_status" | tee -a $RUNTIME_AREA/$repo\n'
1036	<	txt += '   if [ $middleware == OSG ]; then \n'
1037	<	txt += '       cd $RUNTIME_AREA\n'
1038	<	txt += '        echo ">>> current directory (RUNTIME_AREA): $RUNTIME_AREA"\n'
1039	<	txt += '        echo ">>> Remove working directory: $WORKING_DIR"\n'
1040	<	txt += '       /bin/rm -rf $WORKING_DIR\n'
1041	<	txt += '       if [ -d $WORKING_DIR ] ;then\n'
1042	<	txt += '           echo "SET_EXE 50999 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after Untarring .tgz file failed"\n'
1043	<	txt += '           echo "JOB_EXIT_STATUS = 50999"\n'
1044	<	txt += '           echo "JobExitCode=50999" | tee -a $RUNTIME_AREA/$repo\n'
1045	<	txt += '           dumpStatus $RUNTIME_AREA/$repo\n'
1046	<	txt += '       fi\n'
1047	<	txt += '   fi \n'
1048	<	txt += '   \n'
1049	<	txt += '   exit 1 \n'
959	>	txt += '   echo "ERROR ==> Untarring .tgz file failed"\n'
960	>	txt += '   job_exit_code=$untar_status\n'
961	>	txt += '   func_exit\n'
962		txt += 'else \n'
963		txt += '   echo "Successful untar" \n'
964		txt += 'fi \n'
965		txt += '\n'
966	<	txt += 'echo ">>> Include ProdAgentApi and PRODCOMMON in PYTHONPATH:"\n'
966	>	txt += 'echo ">>> Include $RUNTIME_AREA in PYTHONPATH:"\n'
967		txt += 'if [ -z "$PYTHONPATH" ]; then\n'
968	<	#### FEDE FOR DBS OUTPUT PUBLICATION
1057	<	txt += '   export PYTHONPATH=$SOFTWARE_DIR/ProdAgentApi:$SOFTWARE_DIR/ProdCommon\n'
968	>	txt += '   export PYTHONPATH=$RUNTIME_AREA/\n'
969		txt += 'else\n'
970	<	txt += '   export PYTHONPATH=$SOFTWARE_DIR/ProdAgentApi:$SOFTWARE_DIR/ProdCommon:${PYTHONPATH}\n'
970	>	txt += '   export PYTHONPATH=$RUNTIME_AREA/:${PYTHONPATH}\n'
971		txt += 'echo "PYTHONPATH=$PYTHONPATH"\n'
1061	–	###################
972		txt += 'fi\n'
973		txt += '\n'
974
#	Line 1066 | Line 976 | class Cmssw(JobType):
976
977		return txt
978
979	<	def modifySteeringCards(self, nj):
979	>	def wsBuildExe(self, nj=0):
980		"""
981	<	modify the card provided by the user,
982	<	writing a new card into share dir
981	>	Put in the script the commands to build an executable
982	>	or a library.
983		"""
984
985	+	txt = '\n#Written by cms_cmssw::wsBuildExe\n'
986	+	txt += 'echo ">>> moving CMSSW software directories in `pwd`" \n'
987	+
988	+	txt += 'rm -r lib/ module/ \n'
989	+	txt += 'mv $RUNTIME_AREA/lib/ . \n'
990	+	txt += 'mv $RUNTIME_AREA/module/ . \n'
991	+	if self.dataExist == True:
992	+	txt += 'rm -r src/ \n'
993	+	txt += 'mv $RUNTIME_AREA/src/ . \n'
994	+	if len(self.additional_inbox_files)>0:
995	+	for file in self.additional_inbox_files:
996	+	txt += 'mv $RUNTIME_AREA/'+os.path.basename(file)+' . \n'
997	+	# txt += 'mv $RUNTIME_AREA/ProdCommon/ . \n'
998	+	# txt += 'mv $RUNTIME_AREA/IMProv/ . \n'
999	+
1000	+	txt += 'echo ">>> Include $RUNTIME_AREA in PYTHONPATH:"\n'
1001	+	txt += 'if [ -z "$PYTHONPATH" ]; then\n'
1002	+	txt += '   export PYTHONPATH=$RUNTIME_AREA/\n'
1003	+	txt += 'else\n'
1004	+	txt += '   export PYTHONPATH=$RUNTIME_AREA/:${PYTHONPATH}\n'
1005	+	txt += 'echo "PYTHONPATH=$PYTHONPATH"\n'
1006	+	txt += 'fi\n'
1007	+	txt += '\n'
1008	+
1009	+	return txt
1010	+
1011	+
1012		def executableName(self):
1013	<	if self.scriptExe: #CarlosDaniele
1013	>	if self.scriptExe:
1014		return "sh "
1015		else:
1016		return self.executable
1017
1018		def executableArgs(self):
1019	+	# FUTURE: This function tests the CMSSW version. Can be simplified as we drop support for old versions
1020		if self.scriptExe:#CarlosDaniele
1021		return   self.scriptExe + " $NJob"
1022		else:
1023	<	# if >= CMSSW_1_5_X, add -e
1024	<	version_array = self.scram.getSWVersion().split('_')
1025	<	major = 0
1026	<	minor = 0
1027	<	try:
1028	<	major = int(version_array[1])
1029	<	minor = int(version_array[2])
1030	<	except:
1093	<	msg = "Cannot parse CMSSW version string: " + "_".join(version_array) + " for major and minor release number!"
1094	<	raise CrabException(msg)
1095	<	if major >= 1 and minor >= 5 :
1096	<	return " -e -p pset.cfg"
1023	>	ex_args = ""
1024	>	# FUTURE: This tests the CMSSW version. Can remove code as versions deprecated
1025	>	# Framework job report
1026	>	if (self.CMSSW_major >= 1 and self.CMSSW_minor >= 5) or (self.CMSSW_major >= 2):
1027	>	ex_args += " -j $RUNTIME_AREA/crab_fjr_$NJob.xml"
1028	>	# Type of config file
1029	>	if self.CMSSW_major >= 2 :
1030	>	ex_args += " -p pset.py"
1031		else:
1032	<	return " -p pset.cfg"
1032	>	ex_args += " -p pset.cfg"
1033	>	return ex_args
1034
1035		def inputSandbox(self, nj):
1036		"""
1037		Returns a list of filenames to be put in JDL input sandbox.
1038		"""
1039		inp_box = []
1105	–	# # dict added to delete duplicate from input sandbox file list
1106	–	# seen = {}
1107	–	## code
1040		if os.path.isfile(self.tgzNameWithPath):
1041		inp_box.append(self.tgzNameWithPath)
1042	<	if os.path.isfile(self.MLtgzfile):
1043	<	inp_box.append(self.MLtgzfile)
1112	<	## config
1113	<	if not self.pset is None:
1114	<	inp_box.append(common.work_space.pathForTgz() + 'job/' + self.configFilename())
1115	<	## additional input files
1116	<	tgz = self.additionalInputFileTgz()
1117	<	inp_box.append(tgz)
1042	>	wrapper = os.path.basename(str(common._db.queryTask('scriptName')))
1043	>	inp_box.append(common.work_space.pathForTgz() +'job/'+ wrapper)
1044		return inp_box
1045
1046		def outputSandbox(self, nj):
#	Line 1126 | Line 1052 | class Cmssw(JobType):
1052		## User Declared output files
1053		for out in (self.output_file+self.output_file_sandbox):
1054		n_out = nj + 1
1055	<	out_box.append(self.numberFile_(out,str(n_out)))
1055	>	out_box.append(numberFile(out,str(n_out)))
1056		return out_box
1057
1132	–	def prepareSteeringCards(self):
1133	–	"""
1134	–	Make initial modifications of the user's steering card file.
1135	–	"""
1136	–	return
1058
1059		def wsRenameOutput(self, nj):
1060		"""
1061		Returns part of a job script which renames the produced files.
1062		"""
1063
1064	<	txt = '\n'
1065	<	txt += 'echo" >>> directory content:"\n'
1066	<	txt += 'ls \n'
1067	<	txt = '\n'
1068	<
1069	<	txt += 'output_exit_status=0\n'
1149	<
1150	<	for fileWithSuffix in (self.output_file_sandbox):
1151	<	output_file_num = self.numberFile_(fileWithSuffix, '$NJob')
1152	<	txt += '\n'
1153	<	txt += '# check output file\n'
1154	<	txt += 'if [ -e ./'+fileWithSuffix+' ] ; then\n'
1155	<	txt += '    mv '+fileWithSuffix+' $RUNTIME_AREA\n'
1156	<	txt += '    cp $RUNTIME_AREA/'+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n'
1157	<	txt += 'else\n'
1158	<	txt += '    exit_status=60302\n'
1159	<	txt += '    echo "ERROR: Problem with output file '+fileWithSuffix+'"\n'
1160	<	if common.scheduler.boss_scheduler_name == 'condor_g':
1161	<	txt += '    if [ $middleware == OSG ]; then \n'
1162	<	txt += '        echo "prepare dummy output file"\n'
1163	<	txt += '        echo "Processing of job output failed" > $RUNTIME_AREA/'+output_file_num+'\n'
1164	<	txt += '    fi \n'
1165	<	txt += 'fi\n'
1064	>	txt = '\n#Written by cms_cmssw::wsRenameOutput\n'
1065	>	txt += 'echo ">>> current directory (SOFTWARE_DIR): $SOFTWARE_DIR" \n'
1066	>	txt += 'echo ">>> current directory content:"\n'
1067	>	if self.debug_wrapper:
1068	>	txt += 'ls -Al\n'
1069	>	txt += '\n'
1070
1071		for fileWithSuffix in (self.output_file):
1072	<	output_file_num = self.numberFile_(fileWithSuffix, '$NJob')
1072	>	output_file_num = numberFile(fileWithSuffix, '$NJob')
1073		txt += '\n'
1074		txt += '# check output file\n'
1075		txt += 'if [ -e ./'+fileWithSuffix+' ] ; then\n'
1076	<	txt += '    mv '+fileWithSuffix+' $RUNTIME_AREA\n'
1077	<	txt += '    cp $RUNTIME_AREA/'+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n'
1076	>	if (self.copy_data == 1):  # For OSG nodes, file is in $WORKING_DIR, should not be moved to $RUNTIME_AREA
1077	>	txt += '    mv '+fileWithSuffix+' '+output_file_num+'\n'
1078	>	txt += '    ln -s `pwd`/'+output_file_num+' $RUNTIME_AREA/'+fileWithSuffix+'\n'
1079	>	else:
1080	>	txt += '    mv '+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n'
1081	>	txt += '    ln -s $RUNTIME_AREA/'+output_file_num+' $RUNTIME_AREA/'+fileWithSuffix+'\n'
1082		txt += 'else\n'
1083	<	txt += '    exit_status=60302\n'
1084	<	txt += '    echo "ERROR: Problem with output file '+fileWithSuffix+'"\n'
1085	<	txt += '    echo "JOB_EXIT_STATUS = $exit_status"\n'
1178	<	txt += '    output_exit_status=$exit_status\n'
1179	<	if common.scheduler.boss_scheduler_name == 'condor_g':
1083	>	txt += '    job_exit_code=60302\n'
1084	>	txt += '    echo "WARNING: Output file '+fileWithSuffix+' not found"\n'
1085	>	if common.scheduler.name().upper() == 'CONDOR_G':
1086		txt += '    if [ $middleware == OSG ]; then \n'
1087		txt += '        echo "prepare dummy output file"\n'
1088		txt += '        echo "Processing of job output failed" > $RUNTIME_AREA/'+output_file_num+'\n'
#	Line 1184 | Line 1090 | class Cmssw(JobType):
1090		txt += 'fi\n'
1091		file_list = []
1092		for fileWithSuffix in (self.output_file):
1093	<	file_list.append(self.numberFile_(fileWithSuffix, '$NJob'))
1093	>	file_list.append(numberFile(fileWithSuffix, '$NJob'))
1094
1095		txt += 'file_list="'+string.join(file_list,' ')+'"\n'
1096	+	txt += '\n'
1097	+	txt += 'echo ">>> current directory (SOFTWARE_DIR): $SOFTWARE_DIR" \n'
1098	+	txt += 'echo ">>> current directory content:"\n'
1099	+	if self.debug_wrapper:
1100	+	txt += 'ls -Al\n'
1101	+	txt += '\n'
1102		txt += 'cd $RUNTIME_AREA\n'
1103		txt += 'echo ">>> current directory (RUNTIME_AREA):  $RUNTIME_AREA"\n'
1104		return txt
1105
1194	–	def numberFile_(self, file, txt):
1195	–	"""
1196	–	append _'txt' before last extension of a file
1197	–	"""
1198	–	p = string.split(file,".")
1199	–	# take away last extension
1200	–	name = p[0]
1201	–	for x in p[1:-1]:
1202	–	name=name+"."+x
1203	–	# add "_txt"
1204	–	if len(p)>1:
1205	–	ext = p[len(p)-1]
1206	–	result = name + '_' + txt + "." + ext
1207	–	else:
1208	–	result = name + '_' + txt
1209	–
1210	–	return result
1211	–
1106		def getRequirements(self, nj=[]):
1107		"""
1108		return job requirements to add to jdl files
#	Line 1218 | Line 1112 | class Cmssw(JobType):
1112		req='Member("VO-cms-' + \
1113		self.version + \
1114		'", other.GlueHostApplicationSoftwareRunTimeEnvironment)'
1115	<	## SL add requirement for OS version only if SL4
1222	<	#reSL4 = re.compile( r'slc4' )
1223	<	if self.executable_arch: # and reSL4.search(self.executable_arch):
1115	>	if self.executable_arch:
1116		req+=' && Member("VO-cms-' + \
1117		self.executable_arch + \
1118		'", other.GlueHostApplicationSoftwareRunTimeEnvironment)'
1119
1120		req = req + ' && (other.GlueHostNetworkAdapterOutboundIP)'
1121	+	if ( common.scheduler.name() == "glitecoll" ) or ( common.scheduler.name() == "glite"):
1122	+	req += ' && other.GlueCEStateStatus == "Production" '
1123
1124		return req
1125
1126		def configFilename(self):
1127		""" return the config filename """
1128	<	return self.name()+'.cfg'
1128	>	# FUTURE: Can remove cfg mode for CMSSW >= 2_1_x
1129	>	if (self.CMSSW_major >= 2 and self.CMSSW_minor >= 1) or (self.CMSSW_major >= 3):
1130	>	return self.name()+'.py'
1131	>	else:
1132	>	return self.name()+'.cfg'
1133
1134		def wsSetupCMSOSGEnvironment_(self):
1135		"""
1136		Returns part of a job script which is prepares
1137		the execution environment and which is common for all CMS jobs.
1138		"""
1139	<	txt = '    echo ">>> setup CMS OSG environment:"\n'
1139	>	txt = '\n#Written by cms_cmssw::wsSetupCMSOSGEnvironment_\n'
1140	>	txt += '    echo ">>> setup CMS OSG environment:"\n'
1141		txt += '    echo "set SCRAM ARCH to ' + self.executable_arch + '"\n'
1142		txt += '    export SCRAM_ARCH='+self.executable_arch+'\n'
1143		txt += '    echo "SCRAM_ARCH = $SCRAM_ARCH"\n'
#	Line 1246 | Line 1145 | class Cmssw(JobType):
1145		txt += '      # Use $OSG_APP/cmssoft/cms/cmsset_default.sh to setup cms software\n'
1146		txt += '        source $OSG_APP/cmssoft/cms/cmsset_default.sh '+self.version+'\n'
1147		txt += '    else\n'
1148	<	txt += '        echo "SET_CMS_ENV 10020 ==> ERROR $OSG_APP/cmssoft/cms/cmsset_default.sh file not found"\n'
1149	<	txt += '        echo "JOB_EXIT_STATUS = 10020"\n'
1150	<	txt += '        echo "JobExitCode=10020" | tee -a $RUNTIME_AREA/$repo\n'
1252	<	txt += '        dumpStatus $RUNTIME_AREA/$repo\n'
1253	<	txt += '\n'
1254	<	txt += '        cd $RUNTIME_AREA\n'
1255	<	txt += '        echo ">>> current directory (RUNTIME_AREA): $RUNTIME_AREA"\n'
1256	<	txt += '        echo ">>> Remove working directory: $WORKING_DIR"\n'
1257	<	txt += '        /bin/rm -rf $WORKING_DIR\n'
1258	<	txt += '        if [ -d $WORKING_DIR ] ;then\n'
1259	<	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'
1260	<	txt += '            echo "JOB_EXIT_STATUS = 10017"\n'
1261	<	txt += '            echo "JobExitCode=10017" | tee -a $RUNTIME_AREA/$repo\n'
1262	<	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
1263	<	txt += '        fi\n'
1264	<	txt += '\n'
1265	<	txt += '        exit 1\n'
1148	>	txt += '        echo "ERROR ==> $OSG_APP/cmssoft/cms/cmsset_default.sh file not found"\n'
1149	>	txt += '        job_exit_code=10020\n'
1150	>	txt += '        func_exit\n'
1151		txt += '    fi\n'
1152		txt += '\n'
1153	<	txt += '    echo "SET_CMS_ENV 0 ==> setup cms environment ok"\n'
1153	>	txt += '    echo "==> setup cms environment ok"\n'
1154		txt += '    echo "SCRAM_ARCH = $SCRAM_ARCH"\n'
1155
1156		return txt
1157
1273	–	### OLI_DANIELE
1158		def wsSetupCMSLCGEnvironment_(self):
1159		"""
1160		Returns part of a job script which is prepares
1161		the execution environment and which is common for all CMS jobs.
1162		"""
1163	<	txt = '    echo ">>> setup CMS LCG environment:"\n'
1163	>	txt = '\n#Written by cms_cmssw::wsSetupCMSLCGEnvironment_\n'
1164	>	txt += '    echo ">>> setup CMS LCG environment:"\n'
1165		txt += '    echo "set SCRAM ARCH and BUILD_ARCH to ' + self.executable_arch + ' ###"\n'
1166		txt += '    export SCRAM_ARCH='+self.executable_arch+'\n'
1167		txt += '    export BUILD_ARCH='+self.executable_arch+'\n'
1168		txt += '    if [ ! $VO_CMS_SW_DIR ] ;then\n'
1169	<	txt += '        echo "SET_CMS_ENV 10031 ==> ERROR CMS software dir not found on WN `hostname`"\n'
1170	<	txt += '        echo "JOB_EXIT_STATUS = 10031" \n'
1171	<	txt += '        echo "JobExitCode=10031" | tee -a $RUNTIME_AREA/$repo\n'
1287	<	txt += '        dumpStatus $RUNTIME_AREA/$repo\n'
1288	<	txt += '        exit 1\n'
1169	>	txt += '        echo "ERROR ==> CMS software dir not found on WN `hostname`"\n'
1170	>	txt += '        job_exit_code=10031\n'
1171	>	txt += '        func_exit\n'
1172		txt += '    else\n'
1173		txt += '        echo "Sourcing environment... "\n'
1174		txt += '        if [ ! -s $VO_CMS_SW_DIR/cmsset_default.sh ] ;then\n'
1175	<	txt += '            echo "SET_CMS_ENV 10020 ==> ERROR cmsset_default.sh file not found into dir $VO_CMS_SW_DIR"\n'
1176	<	txt += '            echo "JOB_EXIT_STATUS = 10020"\n'
1177	<	txt += '            echo "JobExitCode=10020" | tee -a $RUNTIME_AREA/$repo\n'
1295	<	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
1296	<	txt += '            exit 1\n'
1175	>	txt += '            echo "ERROR ==> cmsset_default.sh file not found into dir $VO_CMS_SW_DIR"\n'
1176	>	txt += '            job_exit_code=10020\n'
1177	>	txt += '            func_exit\n'
1178		txt += '        fi\n'
1179		txt += '        echo "sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
1180		txt += '        source $VO_CMS_SW_DIR/cmsset_default.sh\n'
1181		txt += '        result=$?\n'
1182		txt += '        if [ $result -ne 0 ]; then\n'
1183	<	txt += '            echo "SET_CMS_ENV 10032 ==> ERROR problem sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
1184	<	txt += '            echo "JOB_EXIT_STATUS = 10032"\n'
1185	<	txt += '            echo "JobExitCode=10032" | tee -a $RUNTIME_AREA/$repo\n'
1305	<	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
1306	<	txt += '            exit 1\n'
1183	>	txt += '            echo "ERROR ==> problem sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
1184	>	txt += '            job_exit_code=10032\n'
1185	>	txt += '            func_exit\n'
1186		txt += '        fi\n'
1187		txt += '    fi\n'
1188		txt += '    \n'
1189	<	txt += '    echo "SET_CMS_ENV 0 ==> setup cms environment ok"\n'
1189	>	txt += '    echo "==> setup cms environment ok"\n'
1190		return txt
1191
1313	–	### FEDE FOR DBS OUTPUT PUBLICATION
1192		def modifyReport(self, nj):
1193		"""
1194		insert the part of the script that modifies the FrameworkJob Report
1195		"""
1196	<
1197	<	txt = ''
1320	<	try:
1321	<	publish_data = int(self.cfg_params['USER.publish_data'])
1322	<	except KeyError:
1323	<	publish_data = 0
1196	>	txt = '\n#Written by cms_cmssw::modifyReport\n'
1197	>	publish_data = int(self.cfg_params.get('USER.publish_data',0))
1198		if (publish_data == 1):
1325	–	txt += 'echo ">>> Modify Job Report:" \n'
1326	–	################ FEDE FOR DBS2 #############################################
1327	–	txt += 'chmod a+x $SOFTWARE_DIR/ProdAgentApi/FwkJobRep/ModifyJobReport.py\n'
1328	–	#############################################################################
1329	–
1330	–	txt += 'if [ -z "$SE" ]; then\n'
1331	–	txt += '    SE="" \n'
1332	–	txt += 'fi \n'
1333	–	txt += 'if [ -z "$SE_PATH" ]; then\n'
1334	–	txt += '    SE_PATH="" \n'
1335	–	txt += 'fi \n'
1336	–	txt += 'echo "SE = $SE"\n'
1337	–	txt += 'echo "SE_PATH = $SE_PATH"\n'
1338	–
1199		processedDataset = self.cfg_params['USER.publish_data_name']
1200	<	txt += 'ProcessedDataset='+processedDataset+'\n'
1201	<	#### LFN=/store/user/<user>/processedDataset_PSETHASH
1202	<	txt += 'if [ "$SE_PATH" == "" ]; then\n'
1203	<	#### FEDE: added slash in LFN ##############
1200	>	if (self.primaryDataset == 'null'):
1201	>	self.primaryDataset = processedDataset
1202	>	if (common.scheduler.name().upper() == "CAF" or common.scheduler.name().upper() == "LSF"):
1203	>	### FEDE FOR NEW LFN ###
1204	>	LFNBaseName = LFNBase(self.primaryDataset, processedDataset, LocalUser=True)
1205	>	self.user = getUserName(LocalUser=True)
1206	>	########################
1207	>	else :
1208	>	### FEDE FOR NEW LFN ###
1209	>	LFNBaseName = LFNBase(self.primaryDataset, processedDataset)
1210	>	self.user = getUserName()
1211	>	########################
1212	>
1213	>	txt += 'if [ $copy_exit_status -eq 0 ]; then\n'
1214	>	### FEDE FOR NEW LFN ###
1215	>	#txt += '    FOR_LFN=%s_${PSETHASH}/\n'%(LFNBaseName)
1216	>	txt += '    FOR_LFN=%s/${PSETHASH}/\n'%(LFNBaseName)
1217	>	########################
1218	>	txt += 'else\n'
1219		txt += '    FOR_LFN=/copy_problems/ \n'
1220	<	txt += 'else \n'
1221	<	txt += '    tmp=`echo $SE_PATH | awk -F \'store\' \'{print$2}\'` \n'
1222	<	#####  FEDE TO BE CHANGED, BECAUSE STORE IS HARDCODED!!!! ########
1223	<	txt += '    FOR_LFN=/store$tmp \n'
1224	<	txt += 'fi \n'
1220	>	txt += '    SE=""\n'
1221	>	txt += '    SE_PATH=""\n'
1222	>	txt += 'fi\n'
1223	>
1224	>	txt += 'echo ">>> Modify Job Report:" \n'
1225	>	txt += 'chmod a+x $RUNTIME_AREA/ProdCommon/FwkJobRep/ModifyJobReport.py\n'
1226	>	txt += 'ProcessedDataset='+processedDataset+'\n'
1227		txt += 'echo "ProcessedDataset = $ProcessedDataset"\n'
1228	+	txt += 'echo "SE = $SE"\n'
1229	+	txt += 'echo "SE_PATH = $SE_PATH"\n'
1230		txt += 'echo "FOR_LFN = $FOR_LFN" \n'
1231		txt += 'echo "CMSSW_VERSION = $CMSSW_VERSION"\n\n'
1232	<	#txt += 'echo "$RUNTIME_AREA/'+self.version+'/ProdAgentApi/FwkJobRep/ModifyJobReport.py crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH"\n'
1233	<	txt += 'echo "$SOFTWARE_DIR/ProdAgentApi/FwkJobRep/ModifyJobReport.py crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH"\n'
1234	<	txt += '$SOFTWARE_DIR/ProdAgentApi/FwkJobRep/ModifyJobReport.py crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH\n'
1235	<	#txt += '$RUNTIME_AREA/'+self.version+'/ProdAgentApi/FwkJobRep/ModifyJobReport.py crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH\n'
1357	<
1232	>	### FEDE FOR NEW LFN ###
1233	>	txt += 'echo "$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 $SE_PATH"\n'
1234	>	txt += '$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 $SE_PATH\n'
1235	>	########################
1236		txt += 'modifyReport_result=$?\n'
1359	–	txt += 'echo modifyReport_result = $modifyReport_result\n'
1237		txt += 'if [ $modifyReport_result -ne 0 ]; then\n'
1238	<	txt += '    exit_status=1\n'
1239	<	txt += '    echo "ERROR: Problem with ModifyJobReport"\n'
1238	>	txt += '    modifyReport_result=70500\n'
1239	>	txt += '    job_exit_code=$modifyReport_result\n'
1240	>	txt += '    echo "ModifyReportResult=$modifyReport_result" | tee -a $RUNTIME_AREA/$repo\n'
1241	>	txt += '    echo "WARNING: Problem with ModifyJobReport"\n'
1242		txt += 'else\n'
1243	<	txt += '    mv NewFrameworkJobReport.xml crab_fjr_$NJob.xml\n'
1243	>	txt += '    mv NewFrameworkJobReport.xml $RUNTIME_AREA/crab_fjr_$NJob.xml\n'
1244		txt += 'fi\n'
1366	–	else:
1367	–	txt += 'echo "no data publication required"\n'
1245		return txt
1246
1247	<	def cleanEnv(self):
1248	<	txt = ''
1249	<	txt += 'if [ $middleware == OSG ]; then\n'
1250	<	txt += '    cd $RUNTIME_AREA\n'
1251	<	txt += '    echo ">>> current directory (RUNTIME_AREA): $RUNTIME_AREA"\n'
1252	<	txt += '    echo ">>> Remove working directory: $WORKING_DIR"\n'
1253	<	txt += '    /bin/rm -rf $WORKING_DIR\n'
1254	<	txt += '    if [ -d $WORKING_DIR ] ;then\n'
1255	<	txt += '        echo "SET_EXE 60999 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after cleanup of WN"\n'
1256	<	txt += '        echo "JOB_EXIT_STATUS = 60999"\n'
1257	<	txt += '        echo "JobExitCode=60999" | tee -a $RUNTIME_AREA/$repo\n'
1258	<	txt += '        dumpStatus $RUNTIME_AREA/$repo\n'
1247	>	def wsParseFJR(self):
1248	>	"""
1249	>	Parse the FrameworkJobReport to obtain useful infos
1250	>	"""
1251	>	txt = '\n#Written by cms_cmssw::wsParseFJR\n'
1252	>	txt += 'echo ">>> Parse FrameworkJobReport crab_fjr.xml"\n'
1253	>	txt += 'if [ -s $RUNTIME_AREA/crab_fjr_$NJob.xml ]; then\n'
1254	>	txt += '    if [ -s $RUNTIME_AREA/parseCrabFjr.py ]; then\n'
1255	>	txt += '        cmd_out=`python $RUNTIME_AREA/parseCrabFjr.py --input $RUNTIME_AREA/crab_fjr_$NJob.xml --dashboard $MonitorID,$MonitorJobID '+self.debugWrap+'`\n'
1256	>	if self.debug_wrapper :
1257	>	txt += '        echo "Result of parsing the FrameworkJobReport crab_fjr.xml: $cmd_out"\n'
1258	>	txt += '        executable_exit_status=`python $RUNTIME_AREA/parseCrabFjr.py --input $RUNTIME_AREA/crab_fjr_$NJob.xml --exitcode`\n'
1259	>	txt += '        if [ $executable_exit_status -eq 50115 ];then\n'
1260	>	txt += '            echo ">>> crab_fjr.xml contents: "\n'
1261	>	txt += '            cat $RUNTIME_AREA/crab_fjr_$NJob.xml\n'
1262	>	txt += '            echo "Wrong FrameworkJobReport --> does not contain useful info. ExitStatus: $executable_exit_status"\n'
1263	>	txt += '        elif [ $executable_exit_status -eq -999 ];then\n'
1264	>	txt += '            echo "ExitStatus from FrameworkJobReport not available. not available. Using exit code of executable from command line."\n'
1265	>	txt += '        else\n'
1266	>	txt += '            echo "Extracted ExitStatus from FrameworkJobReport parsing output: $executable_exit_status"\n'
1267	>	txt += '        fi\n'
1268	>	txt += '    else\n'
1269	>	txt += '        echo "CRAB python script to parse CRAB FrameworkJobReport crab_fjr.xml is not available, using exit code of executable from command line."\n'
1270		txt += '    fi\n'
1271	+	#### Patch to check input data reading for CMSSW16x Hopefully we-ll remove it asap
1272	+
1273	+	txt += '    if [ $executable_exit_status -eq 0 ];then\n'
1274	+	txt += '      echo ">>> Executable succeded  $executable_exit_status"\n'
1275	+	if (self.datasetPath and not (self.dataset_pu or self.useParent)) :
1276	+	# VERIFY PROCESSED DATA
1277	+	txt += '      echo ">>> Verify list of processed files:"\n'
1278	+	txt += '      echo $InputFiles |tr -d \'\\\\\' |tr \',\' \'\\n\'|tr -d \'"\' > input-files.txt\n'
1279	+	txt += '      python $RUNTIME_AREA/parseCrabFjr.py --input $RUNTIME_AREA/crab_fjr_$NJob.xml --lfn > processed-files.txt\n'
1280	+	txt += '      cat input-files.txt  | sort | uniq > tmp.txt\n'
1281	+	txt += '      mv tmp.txt input-files.txt\n'
1282	+	txt += '      echo "cat input-files.txt"\n'
1283	+	txt += '      echo "----------------------"\n'
1284	+	txt += '      cat input-files.txt\n'
1285	+	txt += '      cat processed-files.txt | sort | uniq > tmp.txt\n'
1286	+	txt += '      mv tmp.txt processed-files.txt\n'
1287	+	txt += '      echo "----------------------"\n'
1288	+	txt += '      echo "cat processed-files.txt"\n'
1289	+	txt += '      echo "----------------------"\n'
1290	+	txt += '      cat processed-files.txt\n'
1291	+	txt += '      echo "----------------------"\n'
1292	+	txt += '      diff -q input-files.txt processed-files.txt\n'
1293	+	txt += '      fileverify_status=$?\n'
1294	+	txt += '      if [ $fileverify_status -ne 0 ]; then\n'
1295	+	txt += '         executable_exit_status=30001\n'
1296	+	txt += '         echo "ERROR ==> not all input files processed"\n'
1297	+	txt += '         echo "      ==> list of processed files from crab_fjr.xml differs from list in pset.cfg"\n'
1298	+	txt += '         echo "      ==> diff input-files.txt processed-files.txt"\n'
1299	+	txt += '      fi\n'
1300	+	txt += '    elif [ $executable_exit_status -ne 0 ] || [ $executable_exit_status -ne 50015 ] || [ $executable_exit_status -ne 50017 ];then\n'
1301	+	txt += '      echo ">>> Executable failed  $executable_exit_status"\n'
1302	+	txt += '      func_exit\n'
1303	+	txt += '    fi\n'
1304	+	txt += '\n'
1305	+	txt += 'else\n'
1306	+	txt += '    echo "CRAB FrameworkJobReport crab_fjr.xml is not available, using exit code of executable from command line."\n'
1307		txt += 'fi\n'
1308		txt += '\n'
1309	+	txt += 'echo "ExeExitCode=$executable_exit_status" | tee -a $RUNTIME_AREA/$repo\n'
1310	+	txt += 'echo "EXECUTABLE_EXIT_STATUS = $executable_exit_status"\n'
1311	+	txt += 'job_exit_code=$executable_exit_status\n'
1312	+
1313		return txt
1314
1315		def setParam_(self, param, value):
#	Line 1390 | Line 1318 | class Cmssw(JobType):
1318		def getParams(self):
1319		return self._params
1320
1393	–	def setTaskid_(self):
1394	–	self._taskId = self.cfg_params['taskId']
1395	–
1396	–	def getTaskid(self):
1397	–	return self._taskId
1398	–
1321		def uniquelist(self, old):
1322		"""
1323		remove duplicates from a list
#	Line 1405 | Line 1327 | class Cmssw(JobType):
1327		nd[e]=0
1328		return nd.keys()
1329
1330	<
1409	<	def checkOut(self, limit):
1330	>	def outList(self):
1331		"""
1332		check the dimension of the output files
1333		"""
1334	<	txt += 'echo ">>> Starting output sandbox limit check :"\n'
1335	<	allOutFiles = ""
1334	>	txt = ''
1335	>	txt += 'echo ">>> list of expected files on output sandbox"\n'
1336		listOutFiles = []
1337	<	for fileOut in (self.output_file+self.output_file_sandbox):
1338	<	if fileOut.find('crab_fjr') == -1:
1339	<	allOutFiles = allOutFiles + " " + self.numberFile_(fileOut, '$NJob')
1340	<	listOutFiles.append(self.numberFile_(fileOut, '$NJob'))
1341	<	txt += 'echo "OUTPUT files: '+str(allOutFiles)+'";\n'
1342	<	txt += 'ls -gGhrta;\n'
1343	<	txt += 'sum=0;\n'
1344	<	txt += 'for file in '+str(allOutFiles)+' ; do\n'
1345	<	txt += '    if [ -e $file ]; then\n'
1346	<	txt += '        tt=`ls -gGrta $file | awk \'{ print $3 }\'`\n'
1347	<	txt += '        sum=`expr $sum + $tt`\n'
1348	<	txt += '    else\n'
1349	<	txt += '        echo "WARNING: output file $file not found!"\n'
1350	<	txt += '    fi\n'
1351	<	txt += 'done\n'
1431	<	txt += 'echo "Total Output dimension: $sum";\n'
1432	<	txt += 'limit='+str(limit)+';\n'
1433	<	txt += 'echo "OUTPUT FILES LIMIT SET TO: $limit";\n'
1434	<	txt += 'if [ $limit -lt $sum ]; then\n'
1435	<	txt += '    echo "WARNING: output files have to big size - something will be lost;"\n'
1436	<	txt += '    echo "         checking the output file sizes..."\n'
1437	<	"""
1438	<	txt += '    dim=0;\n'
1439	<	txt += '    exclude=0;\n'
1440	<	txt += '    for files in '+str(allOutFiles)+' ; do\n'
1441	<	txt += '        sumTemp=0;\n'
1442	<	txt += '        for file2 in '+str(allOutFiles)+' ; do\n'
1443	<	txt += '            if [ $file != $file2 ]; then\n'
1444	<	txt += '                tt=`ls -gGrta $file2 | awk \'{ print $3 }\';`\n'
1445	<	txt += '                sumTemp=`expr $sumTemp + $tt`;\n'
1446	<	txt += '            fi\n'
1447	<	txt += '        done\n'
1448	<	txt += '        if [ $sumTemp -lt $limit ]; then\n'
1449	<	txt += '            if [ $dim -lt $sumTemp ]; then\n'
1450	<	txt += '                dim=$sumTemp;\n'
1451	<	txt += '                exclude=$file;\n'
1452	<	txt += '            fi\n'
1453	<	txt += '        fi\n'
1454	<	txt += '    done\n'
1455	<	txt += '    echo "Dimension calculated: $dim"; echo "File to exclude: $exclude";\n'
1456	<	"""
1457	<	txt += '    tot=0;\n'
1458	<	txt += '    for file2 in '+str(allOutFiles)+' ; do\n'
1459	<	txt += '        tt=`ls -gGrta $file2 | awk \'{ print $3 }\';`\n'
1460	<	txt += '        tot=`expr $tot + $tt`;\n'
1461	<	txt += '        if [ $limit -lt $tot ]; then\n'
1462	<	txt += '            tot=`expr $tot - $tt`;\n'
1463	<	txt += '            fileLast=$file;\n'
1464	<	txt += '            break;\n'
1465	<	txt += '        fi\n'
1466	<	txt += '    done\n'
1467	<	txt += '    echo "Dimension calculated: $tot"; echo "First file to exclude: $file";\n'
1468	<	txt += '    flag=0;\n'
1469	<	txt += '    for filess in '+str(allOutFiles)+' ; do\n'
1470	<	txt += '        if [ $fileLast = $filess ]; then\n'
1471	<	txt += '            flag=1;\n'
1472	<	txt += '        fi\n'
1473	<	txt += '        if [ $flag -eq 1 ]; then\n'
1474	<	txt += '            rm -f $filess;\n'
1475	<	txt += '        fi\n'
1476	<	txt += '    done\n'
1477	<	txt += '    ls -agGhrt;\n'
1478	<	txt += '    echo "WARNING: output files are too big in dimension: can not put in the output_sandbox.";\n'
1479	<	txt += '    echo "JOB_EXIT_STATUS = 70000";\n'
1480	<	txt += '    exit_status=70000;\n'
1481	<	txt += 'else'
1482	<	txt += '    echo "Total Output dimension $sum is fine.";\n'
1483	<	txt += 'fi\n'
1484	<	txt += 'echo "Ending output sandbox limit check"\n'
1337	>	stdout = 'CMSSW_$NJob.stdout'
1338	>	stderr = 'CMSSW_$NJob.stderr'
1339	>	if (self.return_data == 1):
1340	>	for file in (self.output_file+self.output_file_sandbox):
1341	>	listOutFiles.append(numberFile(file, '$NJob'))
1342	>	listOutFiles.append(stdout)
1343	>	listOutFiles.append(stderr)
1344	>	else:
1345	>	for file in (self.output_file_sandbox):
1346	>	listOutFiles.append(numberFile(file, '$NJob'))
1347	>	listOutFiles.append(stdout)
1348	>	listOutFiles.append(stderr)
1349	>	txt += 'echo "output files: '+string.join(listOutFiles,' ')+'"\n'
1350	>	txt += 'filesToCheck="'+string.join(listOutFiles,' ')+'"\n'
1351	>	txt += 'export filesToCheck\n'
1352		return txt

Diff Legend

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