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 1.55 by slacapra, Wed Nov 22 14:57:26 2006 UTC vs.
Revision 1.183 by spiga, Wed Apr 30 18:21:07 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	<	import math
5	>	from BlackWhiteListParser import BlackWhiteListParser
6		import common
7	–	import PsetManipulator
8	–
9	–	import DBSInfo
10	–	import DataDiscovery
11	–	import DataLocation
7		import Scram
8	+	from LFNBaseName import *
9
10	<	import glob, os, string, re
10	>	import os, string, glob
11
12		class Cmssw(JobType):
13		def __init__(self, cfg_params, ncjobs):
14		JobType.__init__(self, 'CMSSW')
15		common.logger.debug(3,'CMSSW::__init__')
16
17	<	# Marco.
17	>	self.argsList = []
18	>
19		self._params = {}
20		self.cfg_params = cfg_params
21	+	# init BlackWhiteListParser
22	+	self.blackWhiteListParser = BlackWhiteListParser(cfg_params)
23	+
24	+	self.MaxTarBallSize = float(self.cfg_params.get('EDG.maxtarballsize',9.5))
25
26		# number of jobs requested to be created, limit obj splitting
27		self.ncjobs = ncjobs
28
29		log = common.logger
30	<
30	>
31		self.scram = Scram.Scram(cfg_params)
31	–	scramArea = ''
32		self.additional_inbox_files = []
33		self.scriptExe = ''
34		self.executable = ''
35	+	self.executable_arch = self.scram.getArch()
36		self.tgz_name = 'default.tgz'
37	<	self.pset = ''      #scrip use case Da
37	>	self.scriptName = 'CMSSW.sh'
38	>	self.pset = ''      #scrip use case Da
39		self.datasetPath = '' #scrip use case Da
40
41		# set FJR file name
42		self.fjrFileName = 'crab_fjr.xml'
43
44		self.version = self.scram.getSWVersion()
45	<	common.taskDB.setDict('codeVersion',self.version)
46	<	self.setParam_('application', self.version)
47	<
46	<	### collect Data cards
45	>	version_array = self.version.split('_')
46	>	self.major_version = 0
47	>	self.minor_version = 0
48		try:
49	<	tmp =  cfg_params['CMSSW.datasetpath']
50	<	log.debug(6, "CMSSW::CMSSW(): datasetPath = "+tmp)
51	<	if string.lower(tmp)=='none':
52	<	self.datasetPath = None
52	<	self.selectNoInput = 1
53	<	else:
54	<	self.datasetPath = tmp
55	<	self.selectNoInput = 0
56	<	except KeyError:
57	<	msg = "Error: datasetpath not defined "
49	>	self.major_version = int(version_array[1])
50	>	self.minor_version = int(version_array[2])
51	>	except:
52	>	msg = "Cannot parse CMSSW version string: " + "_".join(version_array) + " for major and minor release number!"
53		raise CrabException(msg)
54
60	–	# ML monitoring
61	–	# split dataset path style: /PreProdR3Minbias/SIM/GEN-SIM
62	–	if not self.datasetPath:
63	–	self.setParam_('dataset', 'None')
64	–	self.setParam_('owner', 'None')
65	–	else:
66	–	datasetpath_split = self.datasetPath.split("/")
67	–	self.setParam_('dataset', datasetpath_split[1])
68	–	self.setParam_('owner', datasetpath_split[-1])
55
56	<	self.setTaskid_()
57	<	self.setParam_('taskId', self.cfg_params['taskId'])
56	>	#
57	>	# Try to block creation in case of arch/version mismatch
58	>	#
59	>
60	>	#        a = string.split(self.version, "_")
61	>	#
62	>	#        if int(a[1]) == 1 and (int(a[2]) < 5 and self.executable_arch.find('slc4') == 0):
63	>	#            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)
64	>	#            common.logger.message(msg)
65	>	#        if int(a[1]) == 1 and (int(a[2]) >= 5 and self.executable_arch.find('slc3') == 0):
66	>	#            msg = "Error: CMS does not support %s with %s architecture"%(self.version, self.executable_arch)
67	>	#            raise CrabException(msg)
68	>	#
69	>
70	>
71	>	### collect Data cards
72	>
73	>	if not cfg_params.has_key('CMSSW.datasetpath'):
74	>	msg = "Error: datasetpath not defined "
75	>	raise CrabException(msg)
76	>	tmp =  cfg_params['CMSSW.datasetpath']
77	>	log.debug(6, "CMSSW::CMSSW(): datasetPath = "+tmp)
78	>	if string.lower(tmp)=='none':
79	>	self.datasetPath = None
80	>	self.selectNoInput = 1
81	>	else:
82	>	self.datasetPath = tmp
83	>	self.selectNoInput = 0
84
85		self.dataTiers = []
86
87		## now the application
88	<	try:
89	<	self.executable = cfg_params['CMSSW.executable']
78	<	self.setParam_('exe', self.executable)
79	<	log.debug(6, "CMSSW::CMSSW(): executable = "+self.executable)
80	<	msg = "Default executable cmsRun overridden. Switch to " + self.executable
81	<	log.debug(3,msg)
82	<	except KeyError:
83	<	self.executable = 'cmsRun'
84	<	self.setParam_('exe', self.executable)
85	<	msg = "User executable not defined. Use cmsRun"
86	<	log.debug(3,msg)
87	<	pass
88	>	self.executable = cfg_params.get('CMSSW.executable','cmsRun')
89	>	log.debug(6, "CMSSW::CMSSW(): executable = "+self.executable)
90
91	<	try:
90	<	self.pset = cfg_params['CMSSW.pset']
91	<	log.debug(6, "Cmssw::Cmssw(): PSet file = "+self.pset)
92	<	if self.pset.lower() != 'none' :
93	<	if (not os.path.exists(self.pset)):
94	<	raise CrabException("User defined PSet file "+self.pset+" does not exist")
95	<	else:
96	<	self.pset = None
97	<	except KeyError:
91	>	if not cfg_params.has_key('CMSSW.pset'):
92		raise CrabException("PSet file missing. Cannot run cmsRun ")
93	+	self.pset = cfg_params['CMSSW.pset']
94	+	log.debug(6, "Cmssw::Cmssw(): PSet file = "+self.pset)
95	+	if self.pset.lower() != 'none' :
96	+	if (not os.path.exists(self.pset)):
97	+	raise CrabException("User defined PSet file "+self.pset+" does not exist")
98	+	else:
99	+	self.pset = None
100
101		# output files
102		## stuff which must be returned always via sandbox
#	Line 105 | Line 106 | class Cmssw(JobType):
106		self.output_file_sandbox.append(self.fjrFileName)
107
108		# other output files to be returned via sandbox or copied to SE
109	<	try:
110	<	self.output_file = []
111	<
112	<
113	<	tmp = cfg_params['CMSSW.output_file']
114	<	if tmp != '':
115	<	tmpOutFiles = string.split(cfg_params['CMSSW.output_file'],',')
116	<	log.debug(7, 'cmssw::cmssw(): output files '+str(tmpOutFiles))
116	<	for tmp in tmpOutFiles:
117	<	tmp=string.strip(tmp)
118	<	self.output_file.append(tmp)
119	<	pass
120	<	else:
121	<	log.message("No output file defined: only stdout/err and the CRAB Framework Job Report will be available")
109	>	self.output_file = []
110	>	tmp = cfg_params.get('CMSSW.output_file',None)
111	>	if tmp :
112	>	tmpOutFiles = string.split(tmp,',')
113	>	log.debug(7, 'cmssw::cmssw(): output files '+str(tmpOutFiles))
114	>	for tmp in tmpOutFiles:
115	>	tmp=string.strip(tmp)
116	>	self.output_file.append(tmp)
117		pass
118	<	pass
119	<	except KeyError:
120	<	log.message("No output file defined: only stdout/err and the CRAB Framework Job Report will be available")
126	<	pass
118	>	else:
119	>	log.message("No output file defined: only stdout/err and the CRAB Framework Job Report will be available\n")
120	>	pass
121
122		# script_exe file as additional file in inputSandbox
123	<	try:
124	<	self.scriptExe = cfg_params['USER.script_exe']
125	<	if self.scriptExe != '':
126	<	if not os.path.isfile(self.scriptExe):
127	<	msg ="WARNING. file "+self.scriptExe+" not found"
128	<	raise CrabException(msg)
129	<	self.additional_inbox_files.append(string.strip(self.scriptExe))
136	<	except KeyError:
137	<	self.scriptExe = ''
123	>	self.scriptExe = cfg_params.get('USER.script_exe',None)
124	>	if self.scriptExe :
125	>	if not os.path.isfile(self.scriptExe):
126	>	msg ="ERROR. file "+self.scriptExe+" not found"
127	>	raise CrabException(msg)
128	>	self.additional_inbox_files.append(string.strip(self.scriptExe))
129	>
130		#CarlosDaniele
131		if self.datasetPath == None and self.pset == None and self.scriptExe == '' :
132	<	msg ="WARNING. script_exe  not defined"
133	<	raise CrabException(msg)
132	>	msg ="Error. script_exe  not defined"
133	>	raise CrabException(msg)
134
135		## additional input files
136	<	try:
136	>	if cfg_params.has_key('USER.additional_input_files'):
137		tmpAddFiles = string.split(cfg_params['USER.additional_input_files'],',')
138		for tmp in tmpAddFiles:
139		tmp = string.strip(tmp)
140		dirname = ''
141		if not tmp[0]=="/": dirname = "."
142	<	files = glob.glob(os.path.join(dirname, tmp))
142	>	files = []
143	>	if string.find(tmp,"*")>-1:
144	>	files = glob.glob(os.path.join(dirname, tmp))
145	>	if len(files)==0:
146	>	raise CrabException("No additional input file found with this pattern: "+tmp)
147	>	else:
148	>	files.append(tmp)
149		for file in files:
150		if not os.path.exists(file):
151		raise CrabException("Additional input file not found: "+file)
152		pass
153	+	# fname = string.split(file, '/')[-1]
154	+	# storedFile = common.work_space.pathForTgz()+'share/'+fname
155	+	# shutil.copyfile(file, storedFile)
156		self.additional_inbox_files.append(string.strip(file))
157		pass
158		pass
159		common.logger.debug(5,"Additional input files: "+str(self.additional_inbox_files))
160	<	except KeyError:
160	<	pass
161	<
162	<	# files per job
163	<	try:
164	<	if (cfg_params['CMSSW.files_per_jobs']):
165	<	raise CrabException("files_per_jobs no longer supported.  Quitting.")
166	<	except KeyError:
167	<	pass
160	>	pass
161
162		## Events per job
163	<	try:
163	>	if cfg_params.has_key('CMSSW.events_per_job'):
164		self.eventsPerJob =int( cfg_params['CMSSW.events_per_job'])
165		self.selectEventsPerJob = 1
166	<	except KeyError:
166	>	else:
167		self.eventsPerJob = -1
168		self.selectEventsPerJob = 0
169	<
169	>
170		## number of jobs
171	<	try:
171	>	if cfg_params.has_key('CMSSW.number_of_jobs'):
172		self.theNumberOfJobs =int( cfg_params['CMSSW.number_of_jobs'])
173		self.selectNumberOfJobs = 1
174	<	except KeyError:
174	>	else:
175		self.theNumberOfJobs = 0
176		self.selectNumberOfJobs = 0
177
178	<	try:
178	>	if cfg_params.has_key('CMSSW.total_number_of_events'):
179		self.total_number_of_events = int(cfg_params['CMSSW.total_number_of_events'])
180		self.selectTotalNumberEvents = 1
181	<	except KeyError:
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: #CarlosDaniele
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 198 | 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")
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	>	## Old method of dealing with seeds
209	>	## FUTURE: This is for old CMSSW and old CRAB. Can throw exceptions after a couple of CRAB releases and then
210	>	## remove
211	>	self.sourceSeed = cfg_params.get('CMSSW.pythia_seed',None)
212	>	if self.sourceSeed:
213	>	print "pythia_seed is a deprecated parameter. Use preserve_seeds or increment_seeds in the future.\n","Added to increment_seeds."
214	>	self.incrementSeeds.append('sourceSeed')
215	>
216	>	self.sourceSeedVtx = cfg_params.get('CMSSW.vtx_seed',None)
217	>	if self.sourceSeedVtx:
218	>	print "vtx_seed is a deprecated parameter. Use preserve_seeds or increment_seeds in the future.\n","Added to increment_seeds."
219	>	self.incrementSeeds.append('VtxSmeared')
220	>
221	>	self.sourceSeedG4 = cfg_params.get('CMSSW.g4_seed',None)
222	>	if self.sourceSeedG4:
223	>	print "g4_seed is a deprecated parameter. Use preserve_seeds or increment_seeds in the future.\n","Added to increment_seeds."
224	>	self.incrementSeeds.append('g4SimHits')
225	>
226	>	self.sourceSeedMix = cfg_params.get('CMSSW.mix_seed',None)
227	>	if self.sourceSeedMix:
228	>	print "mix_seed is a deprecated parameter. Use preserve_seeds or increment_seeds in the future.\n","Added to increment_seeds."
229	>	self.incrementSeeds.append('mix')
230	>
231	>	self.firstRun = cfg_params.get('CMSSW.first_run',None)
232
208	–	try:
209	–	self.sourceSeedVtx = int(cfg_params['CMSSW.vtx_seed'])
210	–	except KeyError:
211	–	self.sourceSeedVtx = None
212	–	common.logger.debug(5,"No vertex seed given")
233		if self.pset != None: #CarlosDaniele
234	<	self.PsetEdit = PsetManipulator.PsetManipulator(self.pset) #Daniele Pset
234	>	import PsetManipulator as pp
235	>	PsetEdit = pp.PsetManipulator(self.pset) #Daniele Pset
236	>
237	>	# Copy/return
238	>
239	>	self.copy_data = int(cfg_params.get('USER.copy_data',0))
240	>	self.return_data = int(cfg_params.get('USER.return_data',0))
241
242		#DBSDLS-start
243	<	## Initialize the variables that are extracted from DBS/DLS and needed in other places of the code
243	>	## Initialize the variables that are extracted from DBS/DLS and needed in other places of the code
244		self.maxEvents=0  # max events available   ( --> check the requested nb. of evts in Creator.py)
245		self.DBSPaths={}  # all dbs paths requested ( --> input to the site local discovery script)
246		self.jobDestination=[]  # Site destination(s) for each job (list of lists)
#	Line 223 | Line 249 | class Cmssw(JobType):
249		blockSites = {}
250		if self.datasetPath:
251		blockSites = self.DataDiscoveryAndLocation(cfg_params)
252	<	#DBSDLS-end
252	>	#DBSDLS-end
253	>
254
228	–	self.tgzNameWithPath = self.getTarBall(self.executable)
229	–
255		## Select Splitting
256	<	if self.selectNoInput:
256	>	if self.selectNoInput:
257		if self.pset == None: #CarlosDaniele
258		self.jobSplittingForScript()
259		else:
260		self.jobSplittingNoInput()
261	<	else: self.jobSplittingByBlocks(blockSites)
261	>	else:
262	>	self.jobSplittingByBlocks(blockSites)
263
264		# modify Pset
265		if self.pset != None: #CarlosDaniele
266		try:
267	<	if (self.datasetPath): # standard job
268	<	# allow to processa a fraction of events in a file
269	<	self.PsetEdit.inputModule("INPUT")
270	<	self.PsetEdit.maxEvent("INPUTMAXEVENTS")
271	<	self.PsetEdit.skipEvent("INPUTSKIPEVENTS")
246	<	else:  # pythia like job
247	<	self.PsetEdit.maxEvent(self.eventsPerJob)
248	<	if (self.sourceSeed) :
249	<	self.PsetEdit.pythiaSeed("INPUT")
250	<	if (self.sourceSeedVtx) :
251	<	self.PsetEdit.pythiaSeedVtx("INPUTVTX")
252	<	# add FrameworkJobReport to parameter-set
253	<	self.PsetEdit.addCrabFJR(self.fjrFileName)
254	<	self.PsetEdit.psetWriter(self.configFilename())
267	>	# Add FrameworkJobReport to parameter-set, set max events.
268	>	# Reset later for data jobs by writeCFG which does all modifications
269	>	PsetEdit.addCrabFJR(self.fjrFileName) # FUTURE: Job report addition not needed by CMSSW>1.5
270	>	PsetEdit.maxEvent(self.eventsPerJob)
271	>	PsetEdit.psetWriter(self.configFilename())
272		except:
273		msg='Error while manipuliating ParameterSet: exiting...'
274		raise CrabException(msg)
275	+	self.tgzNameWithPath = self.getTarBall(self.executable)
276
277		def DataDiscoveryAndLocation(self, cfg_params):
278
279	+	import DataDiscovery
280	+	import DataLocation
281		common.logger.debug(10,"CMSSW::DataDiscoveryAndLocation()")
282
283		datasetPath=self.datasetPath
284
265	–	## TODO
266	–	dataTiersList = ""
267	–	dataTiers = dataTiersList.split(',')
268	–
285		## Contact the DBS
286	<	common.logger.message("Contacting DBS...")
286	>	common.logger.message("Contacting Data Discovery Services ...")
287		try:
288	<	self.pubdata=DataDiscovery.DataDiscovery(datasetPath, dataTiers, cfg_params)
288	>	self.pubdata=DataDiscovery.DataDiscovery(datasetPath, cfg_params)
289		self.pubdata.fetchDBSInfo()
290
291		except DataDiscovery.NotExistingDatasetError, ex :
292		msg = 'ERROR ***: failed Data Discovery in DBS : %s'%ex.getErrorMessage()
293		raise CrabException(msg)
278	–
294		except DataDiscovery.NoDataTierinProvenanceError, ex :
295		msg = 'ERROR ***: failed Data Discovery in DBS : %s'%ex.getErrorMessage()
296		raise CrabException(msg)
297		except DataDiscovery.DataDiscoveryError, ex:
298	<	msg = 'ERROR ***: failed Data Discovery in DBS  %s'%ex.getErrorMessage()
298	>	msg = 'ERROR ***: failed Data Discovery in DBS :  %s'%ex.getErrorMessage()
299		raise CrabException(msg)
300
286	–	## get list of all required data in the form of dbs paths  (dbs path = /dataset/datatier/owner)
287	–	## self.DBSPaths=self.pubdata.getDBSPaths()
288	–	common.logger.message("Required data are :"+self.datasetPath)
289	–
301		self.filesbyblock=self.pubdata.getFiles()
302		self.eventsbyblock=self.pubdata.getEventsPerBlock()
303		self.eventsbyfile=self.pubdata.getEventsPerFile()
293	–	# print str(self.filesbyblock)
294	–	# print 'self.eventsbyfile',len(self.eventsbyfile)
295	–	# print str(self.eventsbyfile)
304
305		## get max number of events
306	<	self.maxEvents=self.pubdata.getMaxEvents() ##  self.maxEvents used in Creator.py
299	<	common.logger.message("The number of available events is %s\n"%self.maxEvents)
306	>	self.maxEvents=self.pubdata.getMaxEvents() ##  self.maxEvents used in Creator.py
307
301	–	common.logger.message("Contacting DLS...")
308		## Contact the DLS and build a list of sites hosting the fileblocks
309		try:
310		dataloc=DataLocation.DataLocation(self.filesbyblock.keys(),cfg_params)
#	Line 306 | Line 312 | class Cmssw(JobType):
312		except DataLocation.DataLocationError , ex:
313		msg = 'ERROR ***: failed Data Location in DLS \n %s '%ex.getErrorMessage()
314		raise CrabException(msg)
315	<
315	>
316
317		sites = dataloc.getSites()
318		allSites = []
319		listSites = sites.values()
320	<	for list in listSites:
321	<	for oneSite in list:
320	>	for listSite in listSites:
321	>	for oneSite in listSite:
322		allSites.append(oneSite)
323		allSites = self.uniquelist(allSites)
324
325	<	common.logger.message("Sites ("+str(len(allSites))+") hosting part/all of dataset: "+str(allSites))
326	<	common.logger.debug(6, "List of Sites: "+str(allSites))
325	>	# screen output
326	>	common.logger.message("Requested dataset: " + datasetPath + " has " + str(self.maxEvents) + " events in " + str(len(self.filesbyblock.keys())) + " blocks.\n")
327	>
328		return sites
329	<
329	>
330	>	# to Be Removed  DS -- BL
331	>	#  def setArgsList(self, argsList):
332	>	#      self.argsList = argsList
333	>
334		def jobSplittingByBlocks(self, blockSites):
335		"""
336		Perform job splitting. Jobs run over an integer number of files
#	Line 369 | Line 380 | class Cmssw(JobType):
380		totalNumberOfJobs = 999999999
381		else :
382		totalNumberOfJobs = self.ncjobs
372	–
383
384		blocks = blockSites.keys()
385		blockCount = 0
#	Line 379 | Line 389 | class Cmssw(JobType):
389		jobCount = 0
390		list_of_lists = []
391
392	+	# list tracking which jobs are in which jobs belong to which block
393	+	jobsOfBlock = {}
394	+
395		# ---- Iterate over the blocks in the dataset until ---- #
396		# ---- we've met the requested total # of events    ---- #
397		while ( (eventsRemaining > 0) and (blockCount < numBlocksInDataset) and (jobCount < totalNumberOfJobs)):
398		block = blocks[blockCount]
399		blockCount += 1
400	<
400	>	if block not in jobsOfBlock.keys() :
401	>	jobsOfBlock[block] = []
402
403	<	numEventsInBlock = self.eventsbyblock[block]
404	<	common.logger.debug(5,'Events in Block File '+str(numEventsInBlock))
405	<
406	<	files = self.filesbyblock[block]
407	<	numFilesInBlock = len(files)
408	<	if (numFilesInBlock <= 0):
409	<	continue
410	<	fileCount = 0
411	<
412	<	# ---- New block => New job ---- #
413	<	parString = "\\{"
414	<	# counter for number of events in files currently worked on
415	<	filesEventCount = 0
416	<	# flag if next while loop should touch new file
417	<	newFile = 1
418	<	# job event counter
419	<	jobSkipEventCount = 0
420	<
421	<	# ---- Iterate over the files in the block until we've met the requested ---- #
422	<	# ---- total # of events or we've gone over all the files in this block  ---- #
423	<	while ( (eventsRemaining > 0) and (fileCount < numFilesInBlock) and (jobCount < totalNumberOfJobs) ):
424	<	file = files[fileCount]
425	<	if newFile :
426	<	try:
427	<	numEventsInFile = self.eventsbyfile[file]
428	<	common.logger.debug(6, "File "+str(file)+" has "+str(numEventsInFile)+" events")
429	<	# increase filesEventCount
430	<	filesEventCount += numEventsInFile
431	<	# Add file to current job
432	<	parString += '\\\"' + file + '\\\"\,'
433	<	newFile = 0
434	<	except KeyError:
435	<	common.logger.message("File "+str(file)+" has unknown number of events: skipping")
436	<
437	<
438	<	# if less events in file remain than eventsPerJobRequested
439	<	if ( filesEventCount - jobSkipEventCount < eventsPerJobRequested ) :
440	<	# if last file in block
441	<	if ( fileCount == numFilesInBlock-1 ) :
442	<	# end job using last file, use remaining events in block
403	>	if self.eventsbyblock.has_key(block) :
404	>	numEventsInBlock = self.eventsbyblock[block]
405	>	common.logger.debug(5,'Events in Block File '+str(numEventsInBlock))
406	>
407	>	files = self.filesbyblock[block]
408	>	numFilesInBlock = len(files)
409	>	if (numFilesInBlock <= 0):
410	>	continue
411	>	fileCount = 0
412	>
413	>	# ---- New block => New job ---- #
414	>	parString = ""
415	>	# counter for number of events in files currently worked on
416	>	filesEventCount = 0
417	>	# flag if next while loop should touch new file
418	>	newFile = 1
419	>	# job event counter
420	>	jobSkipEventCount = 0
421	>
422	>	# ---- Iterate over the files in the block until we've met the requested ---- #
423	>	# ---- total # of events or we've gone over all the files in this block  ---- #
424	>	while ( (eventsRemaining > 0) and (fileCount < numFilesInBlock) and (jobCount < totalNumberOfJobs) ):
425	>	file = files[fileCount]
426	>	if newFile :
427	>	try:
428	>	numEventsInFile = self.eventsbyfile[file]
429	>	common.logger.debug(6, "File "+str(file)+" has "+str(numEventsInFile)+" events")
430	>	# increase filesEventCount
431	>	filesEventCount += numEventsInFile
432	>	# Add file to current job
433	>	parString += '\\\"' + file + '\\\"\,'
434	>	newFile = 0
435	>	except KeyError:
436	>	common.logger.message("File "+str(file)+" has unknown number of events: skipping")
437	>
438	>	eventsPerJobRequested = min(eventsPerJobRequested, eventsRemaining)
439	>	# if less events in file remain than eventsPerJobRequested
440	>	if ( filesEventCount - jobSkipEventCount < eventsPerJobRequested):
441	>	# if last file in block
442	>	if ( fileCount == numFilesInBlock-1 ) :
443	>	# end job using last file, use remaining events in block
444	>	# close job and touch new file
445	>	fullString = parString[:-2]
446	>	list_of_lists.append([fullString,str(-1),str(jobSkipEventCount)])
447	>	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(filesEventCount - jobSkipEventCount)+" events (last file in block).")
448	>	self.jobDestination.append(blockSites[block])
449	>	common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
450	>	# fill jobs of block dictionary
451	>	jobsOfBlock[block].append(jobCount+1)
452	>	# reset counter
453	>	jobCount = jobCount + 1
454	>	totalEventCount = totalEventCount + filesEventCount - jobSkipEventCount
455	>	eventsRemaining = eventsRemaining - filesEventCount + jobSkipEventCount
456	>	jobSkipEventCount = 0
457	>	# reset file
458	>	parString = ""
459	>	filesEventCount = 0
460	>	newFile = 1
461	>	fileCount += 1
462	>	else :
463	>	# go to next file
464	>	newFile = 1
465	>	fileCount += 1
466	>	# if events in file equal to eventsPerJobRequested
467	>	elif ( filesEventCount - jobSkipEventCount == eventsPerJobRequested ) :
468		# close job and touch new file
469		fullString = parString[:-2]
470	<	fullString += '\\}'
471	<	list_of_lists.append([fullString,str(-1),str(jobSkipEventCount)])
433	<	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(filesEventCount - jobSkipEventCount)+" events (last file in block).")
470	>	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
471	>	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.")
472		self.jobDestination.append(blockSites[block])
473		common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
474	+	jobsOfBlock[block].append(jobCount+1)
475		# reset counter
476		jobCount = jobCount + 1
477	<	totalEventCount = totalEventCount + filesEventCount - jobSkipEventCount
478	<	eventsRemaining = eventsRemaining - filesEventCount + jobSkipEventCount
477	>	totalEventCount = totalEventCount + eventsPerJobRequested
478	>	eventsRemaining = eventsRemaining - eventsPerJobRequested
479		jobSkipEventCount = 0
480		# reset file
481	<	parString = "\\{"
481	>	parString = ""
482		filesEventCount = 0
483		newFile = 1
484		fileCount += 1
485	+
486	+	# if more events in file remain than eventsPerJobRequested
487		else :
488	<	# go to next file
489	<	newFile = 1
490	<	fileCount += 1
491	<	# if events in file equal to eventsPerJobRequested
492	<	elif ( filesEventCount - jobSkipEventCount == eventsPerJobRequested ) :
493	<	# close job and touch new file
494	<	fullString = parString[:-2]
495	<	fullString += '\\}'
496	<	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
497	<	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.")
498	<	self.jobDestination.append(blockSites[block])
499	<	common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
500	<	# reset counter
501	<	jobCount = jobCount + 1
502	<	totalEventCount = totalEventCount + eventsPerJobRequested
503	<	eventsRemaining = eventsRemaining - eventsPerJobRequested
504	<	jobSkipEventCount = 0
505	<	# reset file
506	<	parString = "\\{"
466	<	filesEventCount = 0
467	<	newFile = 1
468	<	fileCount += 1
469	<
470	<	# if more events in file remain than eventsPerJobRequested
471	<	else :
472	<	# close job but don't touch new file
473	<	fullString = parString[:-2]
474	<	fullString += '\\}'
475	<	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
476	<	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.")
477	<	self.jobDestination.append(blockSites[block])
478	<	common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
479	<	# increase counter
480	<	jobCount = jobCount + 1
481	<	totalEventCount = totalEventCount + eventsPerJobRequested
482	<	eventsRemaining = eventsRemaining - eventsPerJobRequested
483	<	# calculate skip events for last file
484	<	# use filesEventCount (contains several files), jobSkipEventCount and eventsPerJobRequest
485	<	jobSkipEventCount = eventsPerJobRequested - (filesEventCount - jobSkipEventCount - self.eventsbyfile[file])
486	<	# remove all but the last file
487	<	filesEventCount = self.eventsbyfile[file]
488	<	parString = "\\{"
489	<	parString += '\\\"' + file + '\\\"\,'
490	<	pass # END if
491	<	pass # END while (iterate over files in the block)
488	>	# close job but don't touch new file
489	>	fullString = parString[:-2]
490	>	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
491	>	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.")
492	>	self.jobDestination.append(blockSites[block])
493	>	common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
494	>	jobsOfBlock[block].append(jobCount+1)
495	>	# increase counter
496	>	jobCount = jobCount + 1
497	>	totalEventCount = totalEventCount + eventsPerJobRequested
498	>	eventsRemaining = eventsRemaining - eventsPerJobRequested
499	>	# calculate skip events for last file
500	>	# use filesEventCount (contains several files), jobSkipEventCount and eventsPerJobRequest
501	>	jobSkipEventCount = eventsPerJobRequested - (filesEventCount - jobSkipEventCount - self.eventsbyfile[file])
502	>	# remove all but the last file
503	>	filesEventCount = self.eventsbyfile[file]
504	>	parString = '\\\"' + file + '\\\"\,'
505	>	pass # END if
506	>	pass # END while (iterate over files in the block)
507		pass # END while (iterate over blocks in the dataset)
508		self.ncjobs = self.total_number_of_jobs = jobCount
509		if (eventsRemaining > 0 and jobCount < totalNumberOfJobs ):
510		common.logger.message("Could not run on all requested events because some blocks not hosted at allowed sites.")
511	<	common.logger.message("\n"+str(jobCount)+" job(s) can run on "+str(totalEventCount)+" events.\n")
512	<
511	>	common.logger.message(str(jobCount)+" job(s) can run on "+str(totalEventCount)+" events.\n")
512	>
513	>	# screen output
514	>	screenOutput = "List of jobs and available destination sites:\n\n"
515	>
516	>	# keep trace of block with no sites to print a warning at the end
517	>	noSiteBlock = []
518	>	bloskNoSite = []
519	>
520	>	blockCounter = 0
521	>	for block in blocks:
522	>	if block in jobsOfBlock.keys() :
523	>	blockCounter += 1
524	>	screenOutput += "Block %5i: jobs %20s: sites: %s\n" % (blockCounter,spanRanges(jobsOfBlock[block]),
525	>	','.join(self.blackWhiteListParser.checkWhiteList(self.blackWhiteListParser.checkBlackList(blockSites[block],block),block)))
526	>	if len(self.blackWhiteListParser.checkWhiteList(self.blackWhiteListParser.checkBlackList(blockSites[block],block),block)) == 0:
527	>	noSiteBlock.append( spanRanges(jobsOfBlock[block]) )
528	>	bloskNoSite.append( blockCounter )
529	>
530	>	common.logger.message(screenOutput)
531	>	if len(noSiteBlock) > 0 and len(bloskNoSite) > 0:
532	>	msg = 'WARNING: No sites are hosting any part of data for block:\n                '
533	>	virgola = ""
534	>	if len(bloskNoSite) > 1:
535	>	virgola = ","
536	>	for block in bloskNoSite:
537	>	msg += ' ' + str(block) + virgola
538	>	msg += '\n               Related jobs:\n                 '
539	>	virgola = ""
540	>	if len(noSiteBlock) > 1:
541	>	virgola = ","
542	>	for range_jobs in noSiteBlock:
543	>	msg += str(range_jobs) + virgola
544	>	msg += '\n               will not be submitted and this block of data can not be analyzed!\n'
545	>	if self.cfg_params.has_key('EDG.se_white_list'):
546	>	msg += 'WARNING: SE White List: '+self.cfg_params['EDG.se_white_list']+'\n'
547	>	msg += '(Hint: By whitelisting you force the job to run at this particular site(s).\n'
548	>	msg += 'Please check if the dataset is available at this site!)\n'
549	>	if self.cfg_params.has_key('EDG.ce_white_list'):
550	>	msg += 'WARNING: CE White List: '+self.cfg_params['EDG.ce_white_list']+'\n'
551	>	msg += '(Hint: By whitelisting you force the job to run at this particular site(s).\n'
552	>	msg += 'Please check if the dataset is available at this site!)\n'
553	>
554	>	common.logger.message(msg)
555	>
556		self.list_of_args = list_of_lists
557		return
558
#	Line 503 | Line 561 | class Cmssw(JobType):
561		Perform job splitting based on number of event per job
562		"""
563		common.logger.debug(5,'Splitting per events')
564	<	common.logger.message('Required '+str(self.eventsPerJob)+' events per job ')
565	<	common.logger.message('Required '+str(self.theNumberOfJobs)+' jobs in total ')
566	<	common.logger.message('Required '+str(self.total_number_of_events)+' events in total ')
564	>
565	>	if (self.selectEventsPerJob):
566	>	common.logger.message('Required '+str(self.eventsPerJob)+' events per job ')
567	>	if (self.selectNumberOfJobs):
568	>	common.logger.message('Required '+str(self.theNumberOfJobs)+' jobs in total ')
569	>	if (self.selectTotalNumberEvents):
570	>	common.logger.message('Required '+str(self.total_number_of_events)+' events in total ')
571
572		if (self.total_number_of_events < 0):
573		msg='Cannot split jobs per Events with "-1" as total number of events'
574		raise CrabException(msg)
575
576		if (self.selectEventsPerJob):
577	<	self.total_number_of_jobs = int(self.total_number_of_events/self.eventsPerJob)
577	>	if (self.selectTotalNumberEvents):
578	>	self.total_number_of_jobs = int(self.total_number_of_events/self.eventsPerJob)
579	>	elif(self.selectNumberOfJobs) :
580	>	self.total_number_of_jobs =self.theNumberOfJobs
581	>	self.total_number_of_events =int(self.theNumberOfJobs*self.eventsPerJob)
582	>
583		elif (self.selectNumberOfJobs) :
584		self.total_number_of_jobs = self.theNumberOfJobs
585		self.eventsPerJob = int(self.total_number_of_events/self.total_number_of_jobs)
#	Line 532 | Line 599 | class Cmssw(JobType):
599		self.list_of_args = []
600		for i in range(self.total_number_of_jobs):
601		## Since there is no input, any site is good
602	<	# self.jobDestination.append(["Any"])
603	<	self.jobDestination.append([""]) #must be empty to write correctly the xml
604	<	if (self.sourceSeed):
605	<	if (self.sourceSeedVtx):
606	<	## pythia + vtx random seed
607	<	self.list_of_args.append([
541	<	str(self.sourceSeed)+str(i),
542	<	str(self.sourceSeedVtx)+str(i)
543	<	])
544	<	else:
545	<	## only pythia random seed
546	<	self.list_of_args.append([(str(self.sourceSeed)+str(i))])
547	<	else:
548	<	## no random seed
549	<	self.list_of_args.append([str(i)])
550	<	#print self.list_of_args
602	>	self.jobDestination.append([""]) #must be empty to write correctly the xml
603	>	args=[]
604	>	if (self.firstRun):
605	>	## pythia first run
606	>	args.append(str(self.firstRun)+str(i))
607	>	self.list_of_args.append(args)
608
609		return
610
#	Line 576 | Line 633 | class Cmssw(JobType):
633		return
634
635		def split(self, jobParams):
636	<
580	<	common.jobDB.load()
636	>
637		#### Fabio
638		njobs = self.total_number_of_jobs
639		arglist = self.list_of_args
640		# create the empty structure
641		for i in range(njobs):
642		jobParams.append("")
643	<
643	>
644	>	listID=[]
645	>	listField=[]
646		for job in range(njobs):
647		jobParams[job] = arglist[job]
648	<	# print str(arglist[job])
649	<	# print jobParams[job]
650	<	common.jobDB.setArguments(job, jobParams[job])
651	<	common.logger.debug(5,"Job "+str(job)+" Destination: "+str(self.jobDestination[job]))
652	<	common.jobDB.setDestination(job, self.jobDestination[job])
648	>	listID.append(job+1)
649	>	job_ToSave ={}
650	>	concString = ' '
651	>	argu=''
652	>	if len(jobParams[job]):
653	>	argu +=   concString.join(jobParams[job] )
654	>	job_ToSave['arguments']= str(job+1)+' '+argu## new BL--DS
655	>	job_ToSave['dlsDestination']= self.jobDestination[job]## new BL--DS
656	>	#common._db.updateJob_(job,job_ToSave)## new BL--DS
657	>	listField.append(job_ToSave)
658	>	msg="Job "+str(job)+" Arguments:   "+str(job+1)+" "+argu+"\n"  \
659	>	+"                     Destination: "+str(self.jobDestination[job])
660	>	common.logger.debug(5,msg)
661	>	#common.logger.debug(5,"Job "+str(job)+" Destination: "+str(self.jobDestination[job]))
662	>	common._db.updateJob_(listID,listField)## new BL--DS
663	>	## Pay Attention Here....DS--BL
664	>	self.argsList = (len(jobParams[0])+1)
665
596	–	common.jobDB.save()
666		return
667	<
599	<	def getJobTypeArguments(self, nj, sched):
600	<	result = ''
601	<	for i in common.jobDB.arguments(nj):
602	<	result=result+str(i)+" "
603	<	return result
604	<
667	>
668		def numberOfJobs(self):
669		# Fabio
670		return self.total_number_of_jobs
#	Line 610 | Line 673 | class Cmssw(JobType):
673		"""
674		Return the TarBall with lib and exe
675		"""
676	<
676	>
677		# if it exist, just return it
678	<	self.tgzNameWithPath = common.work_space.shareDir()+self.tgz_name
678	>	#
679	>	# Marco. Let's start to use relative path for Boss XML files
680	>	#
681	>	self.tgzNameWithPath = common.work_space.pathForTgz()+'share/'+self.tgz_name
682		if os.path.exists(self.tgzNameWithPath):
683		return self.tgzNameWithPath
684
#	Line 626 | Line 692 | class Cmssw(JobType):
692		# First of all declare the user Scram area
693		swArea = self.scram.getSWArea_()
694		#print "swArea = ", swArea
695	<	swVersion = self.scram.getSWVersion()
696	<	#print "swVersion = ", swVersion
695	>	# swVersion = self.scram.getSWVersion()
696	>	# print "swVersion = ", swVersion
697		swReleaseTop = self.scram.getReleaseTop_()
698		#print "swReleaseTop = ", swReleaseTop
699	<
699	>
700		## check if working area is release top
701		if swReleaseTop == '' or swArea == swReleaseTop:
702	+	common.logger.debug(3,"swArea = "+swArea+" swReleaseTop ="+swReleaseTop)
703		return
704
705	<	filesToBeTarred = []
706	<	## First find the executable
707	<	if (self.executable != ''):
708	<	exeWithPath = self.scram.findFile_(executable)
709	<	#           print exeWithPath
710	<	if ( not exeWithPath ):
711	<	raise CrabException('User executable '+executable+' not found')
712	<
713	<	## then check if it's private or not
714	<	if exeWithPath.find(swReleaseTop) == -1:
715	<	# the exe is private, so we must ship
716	<	common.logger.debug(5,"Exe "+exeWithPath+" to be tarred")
717	<	path = swArea+'/'
718	<	exe = string.replace(exeWithPath, path,'')
719	<	filesToBeTarred.append(exe)
720	<	pass
721	<	else:
722	<	# the exe is from release, we'll find it on WN
723	<	pass
724	<
725	<	## Now get the libraries: only those in local working area
726	<	libDir = 'lib'
727	<	lib = swArea+'/' +libDir
728	<	common.logger.debug(5,"lib "+lib+" to be tarred")
729	<	if os.path.exists(lib):
730	<	filesToBeTarred.append(libDir)
731	<
732	<	## Now check if module dir is present
733	<	moduleDir = 'module'
734	<	if os.path.isdir(swArea+'/'+moduleDir):
735	<	filesToBeTarred.append(moduleDir)
736	<
737	<	## Now check if the Data dir is present
738	<	dataDir = 'src/Data/'
739	<	if os.path.isdir(swArea+'/'+dataDir):
740	<	filesToBeTarred.append(dataDir)
741	<
742	<	## copy ProdAgent dir to swArea
743	<	cmd = '\cp -rf ' + os.environ['CRABDIR'] + '/ProdAgentApi ' + swArea
744	<	cmd_out = runCommand(cmd)
745	<	if cmd_out != '':
746	<	common.logger.message('ProdAgentApi directory could not be copied to local CMSSW project directory.')
747	<	common.logger.message('No FrameworkJobreport parsing is possible on the WorkerNode.')
748	<
749	<	## Now check if the Data dir is present
750	<	paDir = 'ProdAgentApi'
751	<	if os.path.isdir(swArea+'/'+paDir):
752	<	filesToBeTarred.append(paDir)
753	<
754	<	## Create the tar-ball
755	<	if len(filesToBeTarred)>0:
756	<	cwd = os.getcwd()
757	<	os.chdir(swArea)
758	<	tarcmd = 'tar zcvf ' + self.tgzNameWithPath + ' '
759	<	for line in filesToBeTarred:
760	<	tarcmd = tarcmd + line + ' '
761	<	cout = runCommand(tarcmd)
762	<	if not cout:
763	<	raise CrabException('Could not create tar-ball')
764	<	os.chdir(cwd)
765	<	else:
766	<	common.logger.debug(5,"No files to be to be tarred")
767	<
768	<	return
769	<
770	<	def wsSetupEnvironment(self, nj):
705	>	import tarfile
706	>	try: # create tar ball
707	>	tar = tarfile.open(self.tgzNameWithPath, "w:gz")
708	>	## First find the executable
709	>	if (self.executable != ''):
710	>	exeWithPath = self.scram.findFile_(executable)
711	>	if ( not exeWithPath ):
712	>	raise CrabException('User executable '+executable+' not found')
713	>
714	>	## then check if it's private or not
715	>	if exeWithPath.find(swReleaseTop) == -1:
716	>	# the exe is private, so we must ship
717	>	common.logger.debug(5,"Exe "+exeWithPath+" to be tarred")
718	>	path = swArea+'/'
719	>	# distinguish case when script is in user project area or given by full path somewhere else
720	>	if exeWithPath.find(path) >= 0 :
721	>	exe = string.replace(exeWithPath, path,'')
722	>	tar.add(path+exe,exe)
723	>	else :
724	>	tar.add(exeWithPath,os.path.basename(executable))
725	>	pass
726	>	else:
727	>	# the exe is from release, we'll find it on WN
728	>	pass
729	>
730	>	## Now get the libraries: only those in local working area
731	>	libDir = 'lib'
732	>	lib = swArea+'/' +libDir
733	>	common.logger.debug(5,"lib "+lib+" to be tarred")
734	>	if os.path.exists(lib):
735	>	tar.add(lib,libDir)
736	>
737	>	## Now check if module dir is present
738	>	moduleDir = 'module'
739	>	module = swArea + '/' + moduleDir
740	>	if os.path.isdir(module):
741	>	tar.add(module,moduleDir)
742	>
743	>	## Now check if any data dir(s) is present
744	>	swAreaLen=len(swArea)
745	>	self.dataExist = False
746	>	for root, dirs, files in os.walk(swArea):
747	>	if "data" in dirs:
748	>	self.dataExist=True
749	>	common.logger.debug(5,"data "+root+"/data"+" to be tarred")
750	>	tar.add(root+"/data",root[swAreaLen:]+"/data")
751	>
752	>	### CMSSW ParameterSet
753	>	if not self.pset is None:
754	>	cfg_file = common.work_space.jobDir()+self.configFilename()
755	>	tar.add(cfg_file,self.configFilename())
756	>	common.logger.debug(5,"File added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
757	>
758	>
759	>	## Add ProdCommon dir to tar
760	>	prodcommonDir = 'ProdCommon'
761	>	prodcommonPath = os.environ['CRABDIR'] + '/' + 'ProdCommon'
762	>	if os.path.isdir(prodcommonPath):
763	>	tar.add(prodcommonPath,prodcommonDir)
764	>	common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
765	>
766	>	##### ML stuff
767	>	ML_file_list=['report.py', 'DashboardAPI.py', 'Logger.py', 'ProcInfo.py', 'apmon.py']
768	>	path=os.environ['CRABDIR'] + '/python/'
769	>	for file in ML_file_list:
770	>	tar.add(path+file,file)
771	>	common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
772	>
773	>	##### Utils
774	>	Utils_file_list=['parseCrabFjr.py','writeCfg.py', 'JobReportErrorCode.py']
775	>	for file in Utils_file_list:
776	>	tar.add(path+file,file)
777	>	common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
778	>
779	>	##### AdditionalFiles
780	>	for file in self.additional_inbox_files:
781	>	tar.add(file,string.split(file,'/')[-1])
782	>	common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
783	>
784	>	tar.close()
785	>	except :
786	>	raise CrabException('Could not create tar-ball')
787	>
788	>	## check for tarball size
789	>	tarballinfo = os.stat(self.tgzNameWithPath)
790	>	if ( tarballinfo.st_size > self.MaxTarBallSize*1024*1024 ) :
791	>	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.')
792	>
793	>	## create tar-ball with ML stuff
794	>
795	>	def wsSetupEnvironment(self, nj=0):
796		"""
797		Returns part of a job script which prepares
798		the execution environment for the job 'nj'.
799		"""
800		# Prepare JobType-independent part
801	<	txt = ''
802	<
803	<	## OLI_Daniele at this level  middleware already known
712	<
713	<	txt += 'if [ $middleware == LCG ]; then \n'
801	>	txt = '\n#Written by cms_cmssw::wsSetupEnvironment\n'
802	>	txt += 'echo ">>> setup environment"\n'
803	>	txt += 'if [ $middleware == LCG ]; then \n'
804		txt += self.wsSetupCMSLCGEnvironment_()
805		txt += 'elif [ $middleware == OSG ]; then\n'
806		txt += '    WORKING_DIR=`/bin/mktemp  -d $OSG_WN_TMP/cms_XXXXXXXXXXXX`\n'
807	<	txt += '    echo "Created working directory: $WORKING_DIR"\n'
808	<	txt += '    if [ ! -d $WORKING_DIR ] ;then\n'
809	<	txt += '        echo "SET_CMS_ENV 10016 ==> OSG $WORKING_DIR could not be created on WN `hostname`"\n'
810	<	txt += '        echo "JOB_EXIT_STATUS = 10016"\n'
721	<	txt += '        echo "JobExitCode=10016" | tee -a $RUNTIME_AREA/$repo\n'
722	<	txt += '        dumpStatus $RUNTIME_AREA/$repo\n'
723	<	txt += '        rm -f $RUNTIME_AREA/$repo \n'
724	<	txt += '        echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
725	<	txt += '        echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
726	<	txt += '        exit 1\n'
807	>	txt += '    if [ ! $? == 0 ] ;then\n'
808	>	txt += '        echo "ERROR ==> OSG $WORKING_DIR could not be created on WN `hostname`"\n'
809	>	txt += '        job_exit_code=10016\n'
810	>	txt += '        func_exit\n'
811		txt += '    fi\n'
812	+	txt += '    echo ">>> Created working directory: $WORKING_DIR"\n'
813		txt += '\n'
814		txt += '    echo "Change to working directory: $WORKING_DIR"\n'
815		txt += '    cd $WORKING_DIR\n'
816	<	txt += self.wsSetupCMSOSGEnvironment_()
816	>	txt += '    echo ">>> current directory (WORKING_DIR): $WORKING_DIR"\n'
817	>	txt += self.wsSetupCMSOSGEnvironment_()
818	>	#txt += '    echo "### Set SCRAM ARCH to ' + self.executable_arch + ' ###"\n'
819	>	#txt += '    export SCRAM_ARCH='+self.executable_arch+'\n'
820		txt += 'fi\n'
821
822		# Prepare JobType-specific part
823		scram = self.scram.commandName()
824		txt += '\n\n'
825	<	txt += 'echo "### SPECIFIC JOB SETUP ENVIRONMENT ###"\n'
825	>	txt += 'echo ">>> specific cmssw setup environment:"\n'
826	>	txt += 'echo "CMSSW_VERSION =  '+self.version+'"\n'
827		txt += scram+' project CMSSW '+self.version+'\n'
828		txt += 'status=$?\n'
829		txt += 'if [ $status != 0 ] ; then\n'
830	<	txt += '   echo "SET_EXE_ENV 10034 ==>ERROR CMSSW '+self.version+' not found on `hostname`" \n'
831	<	txt += '   echo "JOB_EXIT_STATUS = 10034"\n'
832	<	txt += '   echo "JobExitCode=10034" | tee -a $RUNTIME_AREA/$repo\n'
744	<	txt += '   dumpStatus $RUNTIME_AREA/$repo\n'
745	<	txt += '   rm -f $RUNTIME_AREA/$repo \n'
746	<	txt += '   echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
747	<	txt += '   echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
748	<	## OLI_Daniele
749	<	txt += '    if [ $middleware == OSG ]; then \n'
750	<	txt += '        echo "Remove working directory: $WORKING_DIR"\n'
751	<	txt += '        cd $RUNTIME_AREA\n'
752	<	txt += '        /bin/rm -rf $WORKING_DIR\n'
753	<	txt += '        if [ -d $WORKING_DIR ] ;then\n'
754	<	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'
755	<	txt += '            echo "JOB_EXIT_STATUS = 10018"\n'
756	<	txt += '            echo "JobExitCode=10018" | tee -a $RUNTIME_AREA/$repo\n'
757	<	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
758	<	txt += '            rm -f $RUNTIME_AREA/$repo \n'
759	<	txt += '            echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
760	<	txt += '            echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
761	<	txt += '        fi\n'
762	<	txt += '    fi \n'
763	<	txt += '   exit 1 \n'
830	>	txt += '    echo "ERROR ==> CMSSW '+self.version+' not found on `hostname`" \n'
831	>	txt += '    job_exit_code=10034\n'
832	>	txt += '    func_exit\n'
833		txt += 'fi \n'
765	–	txt += 'echo "CMSSW_VERSION =  '+self.version+'"\n'
834		txt += 'cd '+self.version+'\n'
835	<	### needed grep for bug in scramv1 ###
835	>	txt += 'SOFTWARE_DIR=`pwd`\n'
836	>	txt += 'echo ">>> current directory (SOFTWARE_DIR): $SOFTWARE_DIR" \n'
837		txt += 'eval `'+scram+' runtime -sh | grep -v SCRAMRT_LSB_JOBNAME`\n'
838	<
838	>	txt += 'if [ $? != 0 ] ; then\n'
839	>	txt += '    echo "ERROR ==> Problem with the command: "\n'
840	>	txt += '    echo "eval \`'+scram+' runtime -sh | grep -v SCRAMRT_LSB_JOBNAME \` at `hostname`"\n'
841	>	txt += '    job_exit_code=10034\n'
842	>	txt += '    func_exit\n'
843	>	txt += 'fi \n'
844		# Handle the arguments:
845		txt += "\n"
846		txt += "## number of arguments (first argument always jobnumber)\n"
847		txt += "\n"
848	<	#        txt += "narg=$#\n"
775	<	txt += "if [ $nargs -lt 2 ]\n"
848	>	txt += "if [ $nargs -lt "+str(self.argsList)+" ]\n"
849		txt += "then\n"
850	<	txt += "    echo 'SET_EXE_ENV 1 ==> ERROR Too few arguments' +$nargs+ \n"
851	<	txt += '    echo "JOB_EXIT_STATUS = 50113"\n'
852	<	txt += '    echo "JobExitCode=50113" | tee -a $RUNTIME_AREA/$repo\n'
780	<	txt += '    dumpStatus $RUNTIME_AREA/$repo\n'
781	<	txt += '    rm -f $RUNTIME_AREA/$repo \n'
782	<	txt += '    echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
783	<	txt += '    echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
784	<	## OLI_Daniele
785	<	txt += '    if [ $middleware == OSG ]; then \n'
786	<	txt += '        echo "Remove working directory: $WORKING_DIR"\n'
787	<	txt += '        cd $RUNTIME_AREA\n'
788	<	txt += '        /bin/rm -rf $WORKING_DIR\n'
789	<	txt += '        if [ -d $WORKING_DIR ] ;then\n'
790	<	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'
791	<	txt += '            echo "JOB_EXIT_STATUS = 50114"\n'
792	<	txt += '            echo "JobExitCode=50114" | tee -a $RUNTIME_AREA/$repo\n'
793	<	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
794	<	txt += '            rm -f $RUNTIME_AREA/$repo \n'
795	<	txt += '            echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
796	<	txt += '            echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
797	<	txt += '        fi\n'
798	<	txt += '    fi \n'
799	<	txt += "    exit 1\n"
850	>	txt += "    echo 'ERROR ==> Too few arguments' +$nargs+ \n"
851	>	txt += '    job_exit_code=50113\n'
852	>	txt += "    func_exit\n"
853		txt += "fi\n"
854		txt += "\n"
855
856		# Prepare job-specific part
857		job = common.job_list[nj]
858	<	if self.pset != None: #CarlosDaniele
858	>	### FEDE FOR DBS OUTPUT PUBLICATION
859	>	if (self.datasetPath):
860	>	txt += '\n'
861	>	txt += 'DatasetPath='+self.datasetPath+'\n'
862	>
863	>	datasetpath_split = self.datasetPath.split("/")
864	>
865	>	txt += 'PrimaryDataset='+datasetpath_split[1]+'\n'
866	>	txt += 'DataTier='+datasetpath_split[2]+'\n'
867	>	txt += 'ApplicationFamily=cmsRun\n'
868	>
869	>	else:
870	>	txt += 'DatasetPath=MCDataTier\n'
871	>	txt += 'PrimaryDataset=null\n'
872	>	txt += 'DataTier=null\n'
873	>	txt += 'ApplicationFamily=MCDataTier\n'
874	>	if self.pset != None:
875		pset = os.path.basename(job.configFilename())
876		txt += '\n'
877	+	txt += 'cp  $RUNTIME_AREA/'+pset+' .\n'
878		if (self.datasetPath): # standard job
879	<	#txt += 'InputFiles=$2\n'
880	<	txt += 'InputFiles=${args[1]}\n'
881	<	txt += 'MaxEvents=${args[2]}\n'
812	<	txt += 'SkipEvents=${args[3]}\n'
879	>	txt += 'InputFiles=${args[1]}; export InputFiles\n'
880	>	txt += 'MaxEvents=${args[2]}; export MaxEvents\n'
881	>	txt += 'SkipEvents=${args[3]}; export SkipEvents\n'
882		txt += 'echo "Inputfiles:<$InputFiles>"\n'
814	–	txt += 'sed "s#{\'INPUT\'}#$InputFiles#" $RUNTIME_AREA/'+pset+' > pset_tmp_1.cfg\n'
883		txt += 'echo "MaxEvents:<$MaxEvents>"\n'
816	–	txt += 'sed "s#INPUTMAXEVENTS#$MaxEvents#" pset_tmp_1.cfg > pset_tmp_2.cfg\n'
884		txt += 'echo "SkipEvents:<$SkipEvents>"\n'
818	–	txt += 'sed "s#INPUTSKIPEVENTS#$SkipEvents#" pset_tmp_2.cfg > pset.cfg\n'
885		else:  # pythia like job
886	<	if (self.sourceSeed):
887	<	#                    txt += 'Seed=$2\n'
888	<	txt += 'Seed=${args[1]}\n'
889	<	txt += 'echo "Seed: <$Seed>"\n'
890	<	txt += 'sed "s#\<INPUT\>#$Seed#" $RUNTIME_AREA/'+pset+' > tmp.cfg\n'
891	<	if (self.sourceSeedVtx):
892	<	#                        txt += 'VtxSeed=$3\n'
827	<	txt += 'VtxSeed=${args[2]}\n'
828	<	txt += 'echo "VtxSeed: <$VtxSeed>"\n'
829	<	txt += 'sed "s#INPUTVTX#$VtxSeed#" tmp.cfg > pset.cfg\n'
830	<	else:
831	<	txt += 'mv tmp.cfg pset.cfg\n'
832	<	else:
833	<	txt += '# Copy untouched pset\n'
834	<	txt += 'cp $RUNTIME_AREA/'+pset+' pset.cfg\n'
886	>	txt += 'PreserveSeeds='  + ','.join(self.preserveSeeds)  + '; export PreserveSeeds\n'
887	>	txt += 'IncrementSeeds=' + ','.join(self.incrementSeeds) + '; export IncrementSeeds\n'
888	>	txt += 'echo "PreserveSeeds: <$PreserveSeeds>"\n'
889	>	txt += 'echo "IncrementSeeds:<$IncrementSeeds>"\n'
890	>	if (self.firstRun):
891	>	txt += 'FirstRun=${args[1]}; export FirstRun\n'
892	>	txt += 'echo "FirstRun: <$FirstRun>"\n'
893
894	+	txt += 'mv -f '+pset+' pset.cfg\n'
895
837	–	if len(self.additional_inbox_files) > 0:
838	–	for file in self.additional_inbox_files:
839	–	relFile = file.split("/")[-1]
840	–	txt += 'if [ -e $RUNTIME_AREA/'+relFile+' ] ; then\n'
841	–	txt += '   cp $RUNTIME_AREA/'+relFile+' .\n'
842	–	txt += '   chmod +x '+relFile+'\n'
843	–	txt += 'fi\n'
844	–	pass
896
897	<	if self.pset != None: #CarlosDaniele
847	<	txt += 'echo "### END JOB SETUP ENVIRONMENT ###"\n\n'
848	<
897	>	if self.pset != None:
898		txt += '\n'
899		txt += 'echo "***** cat pset.cfg *********"\n'
900		txt += 'cat pset.cfg\n'
901		txt += 'echo "****** end pset.cfg ********"\n'
902		txt += '\n'
903	<	# txt += 'echo "***** cat pset1.cfg *********"\n'
904	<	# txt += 'cat pset1.cfg\n'
905	<	# txt += 'echo "****** end pset1.cfg ********"\n'
903	>	txt += 'PSETHASH=`EdmConfigHash < pset.cfg` \n'
904	>	txt += 'echo "PSETHASH = $PSETHASH" \n'
905	>	txt += '\n'
906		return txt
907
908	<	def wsBuildExe(self, nj):
908	>	def wsUntarSoftware(self, nj=0):
909		"""
910		Put in the script the commands to build an executable
911		or a library.
912		"""
913
914	<	txt = ""
914	>	txt = '\n#Written by cms_cmssw::wsUntarSoftware\n'
915
916		if os.path.isfile(self.tgzNameWithPath):
917	<	txt += 'echo "tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+'"\n'
917	>	txt += 'echo ">>> tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+' :" \n'
918		txt += 'tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+'\n'
919	+	txt += 'ls -Al \n'
920		txt += 'untar_status=$? \n'
921		txt += 'if [ $untar_status -ne 0 ]; then \n'
922	<	txt += '   echo "SET_EXE 1 ==> ERROR Untarring .tgz file failed"\n'
923	<	txt += '   echo "JOB_EXIT_STATUS = $untar_status" \n'
924	<	txt += '   echo "JobExitCode=$untar_status" | tee -a $RUNTIME_AREA/$repo\n'
875	<	txt += '   if [ $middleware == OSG ]; then \n'
876	<	txt += '       echo "Remove working directory: $WORKING_DIR"\n'
877	<	txt += '       cd $RUNTIME_AREA\n'
878	<	txt += '       /bin/rm -rf $WORKING_DIR\n'
879	<	txt += '       if [ -d $WORKING_DIR ] ;then\n'
880	<	txt += '           echo "SET_EXE 50999 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after Untarring .tgz file failed"\n'
881	<	txt += '           echo "JOB_EXIT_STATUS = 50999"\n'
882	<	txt += '           echo "JobExitCode=50999" | tee -a $RUNTIME_AREA/$repo\n'
883	<	txt += '           dumpStatus $RUNTIME_AREA/$repo\n'
884	<	txt += '           rm -f $RUNTIME_AREA/$repo \n'
885	<	txt += '           echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
886	<	txt += '           echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
887	<	txt += '       fi\n'
888	<	txt += '   fi \n'
889	<	txt += '   \n'
890	<	txt += '   exit 1 \n'
922	>	txt += '   echo "ERROR ==> Untarring .tgz file failed"\n'
923	>	txt += '   job_exit_code=$untar_status\n'
924	>	txt += '   func_exit\n'
925		txt += 'else \n'
926		txt += '   echo "Successful untar" \n'
927		txt += 'fi \n'
928		txt += '\n'
929	<	txt += 'echo "Include ProdAgentApi in PYTHONPATH"\n'
929	>	txt += 'echo ">>> Include ProdCommon in PYTHONPATH:"\n'
930		txt += 'if [ -z "$PYTHONPATH" ]; then\n'
931	<	txt += '   export PYTHONPATH=ProdAgentApi\n'
931	>	txt += '   export PYTHONPATH=$RUNTIME_AREA/ProdCommon\n'
932		txt += 'else\n'
933	<	txt += '   export PYTHONPATH=ProdAgentApi:${PYTHONPATH}\n'
933	>	txt += '   export PYTHONPATH=$RUNTIME_AREA/ProdCommon:${PYTHONPATH}\n'
934	>	txt += 'echo "PYTHONPATH=$PYTHONPATH"\n'
935		txt += 'fi\n'
936		txt += '\n'
937
938		pass
939	<
939	>
940	>	return txt
941	>
942	>	def wsBuildExe(self, nj=0):
943	>	"""
944	>	Put in the script the commands to build an executable
945	>	or a library.
946	>	"""
947	>
948	>	txt = '\n#Written by cms_cmssw::wsBuildExe\n'
949	>	txt += 'echo ">>> moving CMSSW software directories in `pwd`" \n'
950	>
951	>	txt += 'rm -r lib/ module/ \n'
952	>	txt += 'mv $RUNTIME_AREA/lib/ . \n'
953	>	txt += 'mv $RUNTIME_AREA/module/ . \n'
954	>	if self.dataExist == True: txt += 'mv $RUNTIME_AREA/src/ . \n'
955	>	if len(self.additional_inbox_files)>0:
956	>	for file in self.additional_inbox_files:
957	>	txt += 'mv $RUNTIME_AREA/'+file+' . \n'
958	>	txt += 'mv $RUNTIME_AREA/ProdCommon/ . \n'
959	>
960	>	txt += 'if [ -z "$PYTHONPATH" ]; then\n'
961	>	txt += '   export PYTHONPATH=$SOFTWARE_DIR/ProdCommon\n'
962	>	txt += 'else\n'
963	>	txt += '   export PYTHONPATH=$SOFTWARE_DIR/ProdCommon:${PYTHONPATH}\n'
964	>	txt += 'echo "PYTHONPATH=$PYTHONPATH"\n'
965	>	txt += 'fi\n'
966	>	txt += '\n'
967	>
968		return txt
969
970		def modifySteeringCards(self, nj):
971		"""
972	<	modify the card provided by the user,
972	>	modify the card provided by the user,
973		writing a new card into share dir
974		"""
975	<
975	>
976		def executableName(self):
977	<	if self.pset == None: #CarlosDaniele
977	>	if self.scriptExe: #CarlosDaniele
978		return "sh "
979		else:
980		return self.executable
981
982		def executableArgs(self):
983	<	if self.pset == None:#CarlosDaniele
983	>	# FUTURE: This function tests the CMSSW version. Can be simplified as we drop support for old versions
984	>	if self.scriptExe:#CarlosDaniele
985		return   self.scriptExe + " $NJob"
986	<	else:
987	<	return " -p pset.cfg"
986	>	else:
987	>	ex_args = ""
988	>	# FUTURE: This tests the CMSSW version. Can remove code as versions deprecated
989	>	# Framework job report
990	>	if (self.major_version >= 1 and self.minor_version >= 5) or (self.major_version >= 2):
991	>	ex_args += " -j $RUNTIME_AREA/crab_fjr_$NJob.xml"
992	>	# Type of cfg file
993	>	if self.major_version >= 2 :
994	>	ex_args += " -p pset.py"
995	>	else:
996	>	ex_args += " -p pset.cfg"
997	>	return ex_args
998
999		def inputSandbox(self, nj):
1000		"""
#	Line 932 | Line 1006 | class Cmssw(JobType):
1006		## code
1007		if os.path.isfile(self.tgzNameWithPath):
1008		inp_box.append(self.tgzNameWithPath)
1009	<	## config
1010	<	if not self.pset is None: #CarlosDaniele
937	<	inp_box.append(common.job_list[nj].configFilename())
938	<	## additional input files
939	<	#for file in self.additional_inbox_files:
940	<	#    inp_box.append(common.work_space.cwdDir()+file)
1009	>	wrapper = os.path.basename(str(common._db.queryTask('scriptName')))
1010	>	inp_box.append(common.work_space.pathForTgz() +'job/'+ wrapper)
1011		return inp_box
1012
1013		def outputSandbox(self, nj):
#	Line 948 | Line 1018 | class Cmssw(JobType):
1018
1019		## User Declared output files
1020		for out in (self.output_file+self.output_file_sandbox):
1021	<	n_out = nj + 1
1021	>	n_out = nj + 1
1022		out_box.append(self.numberFile_(out,str(n_out)))
1023		return out_box
1024
#	Line 963 | Line 1033 | class Cmssw(JobType):
1033		Returns part of a job script which renames the produced files.
1034		"""
1035
1036	<	txt = '\n'
1037	<	txt += '# directory content\n'
1036	>	txt = '\n#Written by cms_cmssw::wsRenameOutput\n'
1037	>	txt += 'echo ">>> current directory (SOFTWARE_DIR): $SOFTWARE_DIR" \n'
1038	>	txt += 'echo ">>> current directory content:"\n'
1039		txt += 'ls \n'
1040	+	txt += '\n'
1041
1042	<	for fileWithSuffix in (self.output_file+self.output_file_sandbox):
1042	>	for fileWithSuffix in (self.output_file):
1043		output_file_num = self.numberFile_(fileWithSuffix, '$NJob')
1044		txt += '\n'
1045		txt += '# check output file\n'
1046	<	txt += 'ls '+fileWithSuffix+'\n'
1047	<	txt += 'ls_result=$?\n'
1048	<	txt += 'if [ $ls_result -ne 0 ] ; then\n'
1049	<	txt += '   echo "ERROR: Problem with output file"\n'
1050	<	if common.scheduler.boss_scheduler_name == 'condor_g':
1046	>	txt += 'if [ -e ./'+fileWithSuffix+' ] ; then\n'
1047	>	if (self.copy_data == 1):  # For OSG nodes, file is in $WORKING_DIR, should not be moved to $RUNTIME_AREA
1048	>	txt += '    mv '+fileWithSuffix+' '+output_file_num+'\n'
1049	>	txt += '    ln -s `pwd`/'+output_file_num+' $RUNTIME_AREA/'+fileWithSuffix+'\n'
1050	>	else:
1051	>	txt += '    mv '+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n'
1052	>	txt += '    ln -s $RUNTIME_AREA/'+output_file_num+' $RUNTIME_AREA/'+fileWithSuffix+'\n'
1053	>	txt += 'else\n'
1054	>	txt += '    job_exit_code=60302\n'
1055	>	txt += '    echo "WARNING: Output file '+fileWithSuffix+' not found"\n'
1056	>	if common.scheduler.name().upper() == 'CONDOR_G':
1057		txt += '    if [ $middleware == OSG ]; then \n'
1058		txt += '        echo "prepare dummy output file"\n'
1059		txt += '        echo "Processing of job output failed" > $RUNTIME_AREA/'+output_file_num+'\n'
1060		txt += '    fi \n'
983	–	txt += 'else\n'
984	–	txt += '   cp '+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n'
1061		txt += 'fi\n'
1062	<
1063	<	txt += 'cd $RUNTIME_AREA\n'
1064	<	txt += 'cd $RUNTIME_AREA\n'
989	<	### OLI_DANIELE
990	<	txt += 'if [ $middleware == OSG ]; then\n'
991	<	txt += '    cd $RUNTIME_AREA\n'
992	<	txt += '    echo "Remove working directory: $WORKING_DIR"\n'
993	<	txt += '    /bin/rm -rf $WORKING_DIR\n'
994	<	txt += '    if [ -d $WORKING_DIR ] ;then\n'
995	<	txt += '        echo "SET_EXE 60999 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after cleanup of WN"\n'
996	<	txt += '        echo "JOB_EXIT_STATUS = 60999"\n'
997	<	txt += '        echo "JobExitCode=60999" | tee -a $RUNTIME_AREA/$repo\n'
998	<	txt += '        dumpStatus $RUNTIME_AREA/$repo\n'
999	<	txt += '        rm -f $RUNTIME_AREA/$repo \n'
1000	<	txt += '        echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1001	<	txt += '        echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1002	<	txt += '    fi\n'
1003	<	txt += 'fi\n'
1004	<	txt += '\n'
1005	<
1006	<	file_list = ''
1007	<	## Add to filelist only files to be possibly copied to SE
1008	<	for fileWithSuffix in self.output_file:
1009	<	output_file_num = self.numberFile_(fileWithSuffix, '$NJob')
1010	<	file_list=file_list+output_file_num+' '
1011	<	file_list=file_list[:-1]
1012	<	txt += 'file_list="'+file_list+'"\n'
1062	>	file_list = []
1063	>	for fileWithSuffix in (self.output_file):
1064	>	file_list.append(self.numberFile_(fileWithSuffix, '$NJob'))
1065
1066	+	txt += 'file_list="'+string.join(file_list,' ')+'"\n'
1067	+	txt += '\n'
1068	+	txt += 'echo ">>> current directory (SOFTWARE_DIR): $SOFTWARE_DIR" \n'
1069	+	txt += 'echo ">>> current directory content:"\n'
1070	+	txt += 'ls \n'
1071	+	txt += '\n'
1072	+	txt += 'cd $RUNTIME_AREA\n'
1073	+	txt += 'echo ">>> current directory (RUNTIME_AREA):  $RUNTIME_AREA"\n'
1074		return txt
1075
1076		def numberFile_(self, file, txt):
#	Line 1021 | Line 1081 | class Cmssw(JobType):
1081		# take away last extension
1082		name = p[0]
1083		for x in p[1:-1]:
1084	<	name=name+"."+x
1084	>	name=name+"."+x
1085		# add "_txt"
1086		if len(p)>1:
1087	<	ext = p[len(p)-1]
1088	<	result = name + '_' + txt + "." + ext
1087	>	ext = p[len(p)-1]
1088	>	result = name + '_' + txt + "." + ext
1089		else:
1090	<	result = name + '_' + txt
1091	<
1090	>	result = name + '_' + txt
1091	>
1092		return result
1093
1094	<	def getRequirements(self):
1094	>	def getRequirements(self, nj=[]):
1095		"""
1096	<	return job requirements to add to jdl files
1096	>	return job requirements to add to jdl files
1097		"""
1098		req = ''
1099		if self.version:
1100		req='Member("VO-cms-' + \
1101		self.version + \
1102		'", other.GlueHostApplicationSoftwareRunTimeEnvironment)'
1103	+	## SL add requirement for OS version only if SL4
1104	+	#reSL4 = re.compile( r'slc4' )
1105	+	if self.executable_arch: # and reSL4.search(self.executable_arch):
1106	+	req+=' && Member("VO-cms-' + \
1107	+	self.executable_arch + \
1108	+	'", other.GlueHostApplicationSoftwareRunTimeEnvironment)'
1109
1110		req = req + ' && (other.GlueHostNetworkAdapterOutboundIP)'
1111	+	if common.scheduler.name() == "glitecoll":
1112	+	req += ' && other.GlueCEStateStatus == "Production" '
1113
1114		return req
1115
1116		def configFilename(self):
1117		""" return the config filename """
1118	<	return self.name()+'.cfg'
1118	>	# FUTURE: Can remove cfg mode for CMSSW >= 2_1_x
1119	>	if (self.major_version >= 2 and self.minor_version >= 1) or (self.major_version >= 3):
1120	>	return self.name()+'.py'
1121	>	else:
1122	>	return self.name()+'.cfg'
1123
1052	–	### OLI_DANIELE
1124		def wsSetupCMSOSGEnvironment_(self):
1125		"""
1126		Returns part of a job script which is prepares
1127		the execution environment and which is common for all CMS jobs.
1128		"""
1129	<	txt = '\n'
1130	<	txt += '   echo "### SETUP CMS OSG  ENVIRONMENT ###"\n'
1131	<	txt += '   if [ -f $GRID3_APP_DIR/cmssoft/cmsset_default.sh ] ;then\n'
1132	<	txt += '      # Use $GRID3_APP_DIR/cmssoft/cmsset_default.sh to setup cms software\n'
1133	<	txt += '       source $GRID3_APP_DIR/cmssoft/cmsset_default.sh '+self.version+'\n'
1134	<	txt += '   elif [ -f $OSG_APP/cmssoft/cms/cmsset_default.sh ] ;then\n'
1129	>	txt = '\n#Written by cms_cmssw::wsSetupCMSOSGEnvironment_\n'
1130	>	txt += '    echo ">>> setup CMS OSG environment:"\n'
1131	>	txt += '    echo "set SCRAM ARCH to ' + self.executable_arch + '"\n'
1132	>	txt += '    export SCRAM_ARCH='+self.executable_arch+'\n'
1133	>	txt += '    echo "SCRAM_ARCH = $SCRAM_ARCH"\n'
1134	>	txt += '    if [ -f $OSG_APP/cmssoft/cms/cmsset_default.sh ] ;then\n'
1135		txt += '      # Use $OSG_APP/cmssoft/cms/cmsset_default.sh to setup cms software\n'
1136	<	txt += '       source $OSG_APP/cmssoft/cms/cmsset_default.sh '+self.version+'\n'
1137	<	txt += '   else\n'
1138	<	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'
1139	<	txt += '       echo "JOB_EXIT_STATUS = 10020"\n'
1140	<	txt += '       echo "JobExitCode=10020" | tee -a $RUNTIME_AREA/$repo\n'
1141	<	txt += '       dumpStatus $RUNTIME_AREA/$repo\n'
1071	<	txt += '       rm -f $RUNTIME_AREA/$repo \n'
1072	<	txt += '       echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1073	<	txt += '       echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1074	<	txt += '       exit 1\n'
1075	<	txt += '\n'
1076	<	txt += '       echo "Remove working directory: $WORKING_DIR"\n'
1077	<	txt += '       cd $RUNTIME_AREA\n'
1078	<	txt += '       /bin/rm -rf $WORKING_DIR\n'
1079	<	txt += '       if [ -d $WORKING_DIR ] ;then\n'
1080	<	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'
1081	<	txt += '            echo "JOB_EXIT_STATUS = 10017"\n'
1082	<	txt += '            echo "JobExitCode=10017" | tee -a $RUNTIME_AREA/$repo\n'
1083	<	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
1084	<	txt += '            rm -f $RUNTIME_AREA/$repo \n'
1085	<	txt += '            echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1086	<	txt += '            echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1087	<	txt += '       fi\n'
1088	<	txt += '\n'
1089	<	txt += '       exit 1\n'
1090	<	txt += '   fi\n'
1136	>	txt += '        source $OSG_APP/cmssoft/cms/cmsset_default.sh '+self.version+'\n'
1137	>	txt += '    else\n'
1138	>	txt += '        echo "ERROR ==> $OSG_APP/cmssoft/cms/cmsset_default.sh file not found"\n'
1139	>	txt += '        job_exit_code=10020\n'
1140	>	txt += '        func_exit\n'
1141	>	txt += '    fi\n'
1142		txt += '\n'
1143	<	txt += '   echo "SET_CMS_ENV 0 ==> setup cms environment ok"\n'
1144	<	txt += '   echo " END SETUP CMS OSG  ENVIRONMENT "\n'
1143	>	txt += '    echo "==> setup cms environment ok"\n'
1144	>	txt += '    echo "SCRAM_ARCH = $SCRAM_ARCH"\n'
1145
1146		return txt
1147	<
1147	>
1148		### OLI_DANIELE
1149		def wsSetupCMSLCGEnvironment_(self):
1150		"""
1151		Returns part of a job script which is prepares
1152		the execution environment and which is common for all CMS jobs.
1153		"""
1154	<	txt  = '   \n'
1155	<	txt += '   echo " ### SETUP CMS LCG  ENVIRONMENT ### "\n'
1156	<	txt += '   if [ ! $VO_CMS_SW_DIR ] ;then\n'
1157	<	txt += '       echo "SET_CMS_ENV 10031 ==> ERROR CMS software dir not found on WN `hostname`"\n'
1158	<	txt += '       echo "JOB_EXIT_STATUS = 10031" \n'
1159	<	txt += '       echo "JobExitCode=10031" | tee -a $RUNTIME_AREA/$repo\n'
1160	<	txt += '       dumpStatus $RUNTIME_AREA/$repo\n'
1161	<	txt += '       rm -f $RUNTIME_AREA/$repo \n'
1162	<	txt += '       echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1163	<	txt += '       echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1164	<	txt += '       exit 1\n'
1165	<	txt += '   else\n'
1166	<	txt += '       echo "Sourcing environment... "\n'
1167	<	txt += '       if [ ! -s $VO_CMS_SW_DIR/cmsset_default.sh ] ;then\n'
1168	<	txt += '           echo "SET_CMS_ENV 10020 ==> ERROR cmsset_default.sh file not found into dir $VO_CMS_SW_DIR"\n'
1169	<	txt += '           echo "JOB_EXIT_STATUS = 10020"\n'
1170	<	txt += '           echo "JobExitCode=10020" | tee -a $RUNTIME_AREA/$repo\n'
1171	<	txt += '           dumpStatus $RUNTIME_AREA/$repo\n'
1172	<	txt += '           rm -f $RUNTIME_AREA/$repo \n'
1173	<	txt += '           echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1174	<	txt += '           echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1175	<	txt += '           exit 1\n'
1176	<	txt += '       fi\n'
1177	<	txt += '       echo "sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
1178	<	txt += '       source $VO_CMS_SW_DIR/cmsset_default.sh\n'
1179	<	txt += '       result=$?\n'
1180	<	txt += '       if [ $result -ne 0 ]; then\n'
1181	<	txt += '           echo "SET_CMS_ENV 10032 ==> ERROR problem sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
1182	<	txt += '           echo "JOB_EXIT_STATUS = 10032"\n'
1183	<	txt += '           echo "JobExitCode=10032" | tee -a $RUNTIME_AREA/$repo\n'
1184	<	txt += '           dumpStatus $RUNTIME_AREA/$repo\n'
1185	<	txt += '           rm -f $RUNTIME_AREA/$repo \n'
1186	<	txt += '           echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1187	<	txt += '           echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1188	<	txt += '           exit 1\n'
1189	<	txt += '       fi\n'
1190	<	txt += '   fi\n'
1191	<	txt += '   \n'
1192	<	txt += '   string=`cat /etc/redhat-release`\n'
1193	<	txt += '   echo $string\n'
1194	<	txt += '   if [[ $string = *alhalla* ]]; then\n'
1195	<	txt += '       echo "SCRAM_ARCH= $SCRAM_ARCH"\n'
1196	<	txt += '   elif [[ $string = *Enterprise* ]] || [[ $string = *cientific* ]]; then\n'
1197	<	txt += '       export SCRAM_ARCH=slc3_ia32_gcc323\n'
1198	<	txt += '       echo "SCRAM_ARCH= $SCRAM_ARCH"\n'
1199	<	txt += '   else\n'
1200	<	txt += '       echo "SET_CMS_ENV 10033 ==> ERROR OS unknown, LCG environment not initialized"\n'
1201	<	txt += '       echo "JOB_EXIT_STATUS = 10033"\n'
1202	<	txt += '       echo "JobExitCode=10033" | tee -a $RUNTIME_AREA/$repo\n'
1203	<	txt += '       dumpStatus $RUNTIME_AREA/$repo\n'
1204	<	txt += '       rm -f $RUNTIME_AREA/$repo \n'
1205	<	txt += '       echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1206	<	txt += '       echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1207	<	txt += '       exit 1\n'
1208	<	txt += '   fi\n'
1209	<	txt += '   echo "SET_CMS_ENV 0 ==> setup cms environment ok"\n'
1210	<	txt += '   echo "### END SETUP CMS LCG ENVIRONMENT ###"\n'
1154	>	txt = '\n#Written by cms_cmssw::wsSetupCMSLCGEnvironment_\n'
1155	>	txt += '    echo ">>> setup CMS LCG environment:"\n'
1156	>	txt += '    echo "set SCRAM ARCH and BUILD_ARCH to ' + self.executable_arch + ' ###"\n'
1157	>	txt += '    export SCRAM_ARCH='+self.executable_arch+'\n'
1158	>	txt += '    export BUILD_ARCH='+self.executable_arch+'\n'
1159	>	txt += '    if [ ! $VO_CMS_SW_DIR ] ;then\n'
1160	>	txt += '        echo "ERROR ==> CMS software dir not found on WN `hostname`"\n'
1161	>	txt += '        job_exit_code=10031\n'
1162	>	txt += '        func_exit\n'
1163	>	txt += '    else\n'
1164	>	txt += '        echo "Sourcing environment... "\n'
1165	>	txt += '        if [ ! -s $VO_CMS_SW_DIR/cmsset_default.sh ] ;then\n'
1166	>	txt += '            echo "ERROR ==> cmsset_default.sh file not found into dir $VO_CMS_SW_DIR"\n'
1167	>	txt += '            job_exit_code=10020\n'
1168	>	txt += '            func_exit\n'
1169	>	txt += '        fi\n'
1170	>	txt += '        echo "sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
1171	>	txt += '        source $VO_CMS_SW_DIR/cmsset_default.sh\n'
1172	>	txt += '        result=$?\n'
1173	>	txt += '        if [ $result -ne 0 ]; then\n'
1174	>	txt += '            echo "ERROR ==> problem sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
1175	>	txt += '            job_exit_code=10032\n'
1176	>	txt += '            func_exit\n'
1177	>	txt += '        fi\n'
1178	>	txt += '    fi\n'
1179	>	txt += '    \n'
1180	>	txt += '    echo "==> setup cms environment ok"\n'
1181	>	return txt
1182	>
1183	>	### FEDE FOR DBS OUTPUT PUBLICATION
1184	>	def modifyReport(self, nj):
1185	>	"""
1186	>	insert the part of the script that modifies the FrameworkJob Report
1187	>	"""
1188	>
1189	>	txt = '\n#Written by cms_cmssw::modifyReport\n'
1190	>	publish_data = int(self.cfg_params.get('USER.publish_data',0))
1191	>	if (publish_data == 1):
1192	>	processedDataset = self.cfg_params['USER.publish_data_name']
1193	>	LFNBaseName = LFNBase(processedDataset)
1194	>
1195	>	txt += 'if [ $copy_exit_status -eq 0 ]; then\n'
1196	>	txt += '    FOR_LFN=%s_${PSETHASH}/\n'%(LFNBaseName)
1197	>	txt += 'else\n'
1198	>	txt += '    FOR_LFN=/copy_problems/ \n'
1199	>	txt += '    SE=""\n'
1200	>	txt += '    SE_PATH=""\n'
1201	>	txt += 'fi\n'
1202	>
1203	>	txt += 'echo ">>> Modify Job Report:" \n'
1204	>	txt += 'chmod a+x $SOFTWARE_DIR/ProdCommon/ProdCommon/FwkJobRep/ModifyJobReport.py\n'
1205	>	txt += 'ProcessedDataset='+processedDataset+'\n'
1206	>	txt += 'echo "ProcessedDataset = $ProcessedDataset"\n'
1207	>	txt += 'echo "SE = $SE"\n'
1208	>	txt += 'echo "SE_PATH = $SE_PATH"\n'
1209	>	txt += 'echo "FOR_LFN = $FOR_LFN" \n'
1210	>	txt += 'echo "CMSSW_VERSION = $CMSSW_VERSION"\n\n'
1211	>	txt += 'echo "$SOFTWARE_DIR/ProdCommon/ProdCommon/FwkJobRep/ModifyJobReport.py $RUNTIME_AREA/crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH"\n'
1212	>	txt += '$SOFTWARE_DIR/ProdCommon/ProdCommon/FwkJobRep/ModifyJobReport.py $RUNTIME_AREA/crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH\n'
1213	>	txt += 'modifyReport_result=$?\n'
1214	>	txt += 'if [ $modifyReport_result -ne 0 ]; then\n'
1215	>	txt += '    modifyReport_result=70500\n'
1216	>	txt += '    job_exit_code=$modifyReport_result\n'
1217	>	txt += '    echo "ModifyReportResult=$modifyReport_result" | tee -a $RUNTIME_AREA/$repo\n'
1218	>	txt += '    echo "WARNING: Problem with ModifyJobReport"\n'
1219	>	txt += 'else\n'
1220	>	txt += '    mv NewFrameworkJobReport.xml $RUNTIME_AREA/crab_fjr_$NJob.xml\n'
1221	>	txt += 'fi\n'
1222		return txt
1223
1224		def setParam_(self, param, value):
#	Line 1165 | Line 1227 | class Cmssw(JobType):
1227		def getParams(self):
1228		return self._params
1229
1168	–	def setTaskid_(self):
1169	–	self._taskId = self.cfg_params['taskId']
1170	–
1171	–	def getTaskid(self):
1172	–	return self._taskId
1173	–
1174	–	#######################################################################
1230		def uniquelist(self, old):
1231		"""
1232		remove duplicates from a list
#	Line 1180 | Line 1235 | class Cmssw(JobType):
1235		for e in old:
1236		nd[e]=0
1237		return nd.keys()
1238	+
1239	+	def outList(self):
1240	+	"""
1241	+	check the dimension of the output files
1242	+	"""
1243	+	txt = ''
1244	+	txt += 'echo ">>> list of expected files on output sandbox"\n'
1245	+	listOutFiles = []
1246	+	stdout = 'CMSSW_$NJob.stdout'
1247	+	stderr = 'CMSSW_$NJob.stderr'
1248	+	if (self.return_data == 1):
1249	+	for file in (self.output_file+self.output_file_sandbox):
1250	+	listOutFiles.append(self.numberFile_(file, '$NJob'))
1251	+	listOutFiles.append(stdout)
1252	+	listOutFiles.append(stderr)
1253	+	else:
1254	+	for file in (self.output_file_sandbox):
1255	+	listOutFiles.append(self.numberFile_(file, '$NJob'))
1256	+	listOutFiles.append(stdout)
1257	+	listOutFiles.append(stderr)
1258	+	txt += 'echo "output files: '+string.join(listOutFiles,' ')+'"\n'
1259	+	txt += 'filesToCheck="'+string.join(listOutFiles,' ')+'"\n'
1260	+	txt += 'export filesToCheck\n'
1261	+	return txt

Diff Legend

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