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 1.75 by gutsche, Sun Apr 8 23:50:33 2007 UTC vs.
Revision 1.195 by slacapra, Wed May 28 15:42:01 2008 UTC

#	Line 2 | Line 2 | from JobType import JobType
2		from crab_logger import Logger
3		from crab_exceptions import *
4		from crab_util import *
5	+	from BlackWhiteListParser import BlackWhiteListParser
6		import common
6	–	import PsetManipulator
7	–	import DataDiscovery
8	–	import DataDiscovery_DBS2
9	–	import DataLocation
7		import Scram
8	+	from LFNBaseName import *
9
10	<	import os, string, re, shutil, glob
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	<	try:
24	<	self.MaxTarBallSize = float(self.cfg_params['EDG.maxtarballsize'])
25	<	except KeyError:
26	<	self.MaxTarBallSize = 100.0
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)
32		self.additional_inbox_files = []
33		self.scriptExe = ''
#	Line 37 | Line 35 | class Cmssw(JobType):
35		self.executable_arch = self.scram.getArch()
36		self.tgz_name = 'default.tgz'
37		self.scriptName = 'CMSSW.sh'
38	<	self.pset = ''      #scrip use case Da
39	<	self.datasetPath = '' #scrip use case Da
38	>	self.pset = ''
39	>	self.datasetPath = ''
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)
45	>	version_array = self.version.split('_')
46	>	self.CMSSW_major = 0
47	>	self.CMSSW_minor = 0
48	>	self.CMSSW_patch = 0
49	>	try:
50	>	self.CMSSW_major = int(version_array[1])
51	>	self.CMSSW_minor = int(version_array[2])
52	>	self.CMSSW_patch = int(version_array[3])
53	>	except:
54	>	msg = "Cannot parse CMSSW version string: " + self.version + " for major and minor release number!"
55	>	raise CrabException(msg)
56
57		### collect Data cards
58
59	<	## get DBS mode
60	<	try:
54	<	self.use_dbs_2 = int(self.cfg_params['CMSSW.use_dbs_2'])
55	<	except KeyError:
56	<	self.use_dbs_2 = 0
57	<
58	<	try:
59	<	tmp =  cfg_params['CMSSW.datasetpath']
60	<	log.debug(6, "CMSSW::CMSSW(): datasetPath = "+tmp)
61	<	if string.lower(tmp)=='none':
62	<	self.datasetPath = None
63	<	self.selectNoInput = 1
64	<	else:
65	<	self.datasetPath = tmp
66	<	self.selectNoInput = 0
67	<	except KeyError:
68	<	msg = "Error: datasetpath not defined "
59	>	if not cfg_params.has_key('CMSSW.datasetpath'):
60	>	msg = "Error: datasetpath not defined "
61		raise CrabException(msg)
62	<
63	<	# ML monitoring
64	<	# split dataset path style: /PreProdR3Minbias/SIM/GEN-SIM
65	<	if not self.datasetPath:
66	<	self.setParam_('dataset', 'None')
75	<	self.setParam_('owner', 'None')
62	>	tmp =  cfg_params['CMSSW.datasetpath']
63	>	log.debug(6, "CMSSW::CMSSW(): datasetPath = "+tmp)
64	>	if string.lower(tmp)=='none':
65	>	self.datasetPath = None
66	>	self.selectNoInput = 1
67		else:
68	<	datasetpath_split = self.datasetPath.split("/")
69	<	if self.use_dbs_2 == 1 :
79	<	self.setParam_('dataset', datasetpath_split[1])
80	<	self.setParam_('owner', datasetpath_split[2])
81	<	else :
82	<	self.setParam_('dataset', datasetpath_split[1])
83	<	self.setParam_('owner', datasetpath_split[-1])
84	<
85	<	self.setTaskid_()
86	<	self.setParam_('taskId', self.cfg_params['taskId'])
68	>	self.datasetPath = tmp
69	>	self.selectNoInput = 0
70
71		self.dataTiers = []
72
73	+	self.debug_pset = cfg_params.get('USER.debug_pset',False)
74	+
75		## now the application
76	<	try:
77	<	self.executable = cfg_params['CMSSW.executable']
93	<	self.setParam_('exe', self.executable)
94	<	log.debug(6, "CMSSW::CMSSW(): executable = "+self.executable)
95	<	msg = "Default executable cmsRun overridden. Switch to " + self.executable
96	<	log.debug(3,msg)
97	<	except KeyError:
98	<	self.executable = 'cmsRun'
99	<	self.setParam_('exe', self.executable)
100	<	msg = "User executable not defined. Use cmsRun"
101	<	log.debug(3,msg)
102	<	pass
76	>	self.executable = cfg_params.get('CMSSW.executable','cmsRun')
77	>	log.debug(6, "CMSSW::CMSSW(): executable = "+self.executable)
78
79	<	try:
105	<	self.pset = cfg_params['CMSSW.pset']
106	<	log.debug(6, "Cmssw::Cmssw(): PSet file = "+self.pset)
107	<	if self.pset.lower() != 'none' :
108	<	if (not os.path.exists(self.pset)):
109	<	raise CrabException("User defined PSet file "+self.pset+" does not exist")
110	<	else:
111	<	self.pset = None
112	<	except KeyError:
79	>	if not cfg_params.has_key('CMSSW.pset'):
80		raise CrabException("PSet file missing. Cannot run cmsRun ")
81	+	self.pset = cfg_params['CMSSW.pset']
82	+	log.debug(6, "Cmssw::Cmssw(): PSet file = "+self.pset)
83	+	if self.pset.lower() != 'none' :
84	+	if (not os.path.exists(self.pset)):
85	+	raise CrabException("User defined PSet file "+self.pset+" does not exist")
86	+	else:
87	+	self.pset = None
88
89		# output files
90		## stuff which must be returned always via sandbox
#	Line 120 | Line 94 | class Cmssw(JobType):
94		self.output_file_sandbox.append(self.fjrFileName)
95
96		# other output files to be returned via sandbox or copied to SE
97	<	try:
98	<	self.output_file = []
99	<	tmp = cfg_params['CMSSW.output_file']
100	<	if tmp != '':
101	<	tmpOutFiles = string.split(cfg_params['CMSSW.output_file'],',')
102	<	log.debug(7, 'cmssw::cmssw(): output files '+str(tmpOutFiles))
103	<	for tmp in tmpOutFiles:
104	<	tmp=string.strip(tmp)
131	<	self.output_file.append(tmp)
132	<	pass
133	<	else:
134	<	log.message("No output file defined: only stdout/err and the CRAB Framework Job Report will be available")
97	>	self.output_file = []
98	>	tmp = cfg_params.get('CMSSW.output_file',None)
99	>	if tmp :
100	>	tmpOutFiles = string.split(tmp,',')
101	>	log.debug(7, 'cmssw::cmssw(): output files '+str(tmpOutFiles))
102	>	for tmp in tmpOutFiles:
103	>	tmp=string.strip(tmp)
104	>	self.output_file.append(tmp)
105		pass
106	<	pass
107	<	except KeyError:
108	<	log.message("No output file defined: only stdout/err and the CRAB Framework Job Report will be available")
139	<	pass
106	>	else:
107	>	log.message("No output file defined: only stdout/err and the CRAB Framework Job Report will be available\n")
108	>	pass
109
110		# script_exe file as additional file in inputSandbox
111	<	try:
112	<	self.scriptExe = cfg_params['USER.script_exe']
113	<	if self.scriptExe != '':
114	<	if not os.path.isfile(self.scriptExe):
115	<	msg ="ERROR. file "+self.scriptExe+" not found"
116	<	raise CrabException(msg)
148	<	self.additional_inbox_files.append(string.strip(self.scriptExe))
149	<	except KeyError:
150	<	self.scriptExe = ''
111	>	self.scriptExe = cfg_params.get('USER.script_exe',None)
112	>	if self.scriptExe :
113	>	if not os.path.isfile(self.scriptExe):
114	>	msg ="ERROR. file "+self.scriptExe+" not found"
115	>	raise CrabException(msg)
116	>	self.additional_inbox_files.append(string.strip(self.scriptExe))
117
152	–	#CarlosDaniele
118		if self.datasetPath == None and self.pset == None and self.scriptExe == '' :
119	<	msg ="Error. script_exe  not defined"
120	<	raise CrabException(msg)
119	>	msg ="Error. script_exe  not defined"
120	>	raise CrabException(msg)
121
122		## additional input files
123	<	try:
123	>	if cfg_params.has_key('USER.additional_input_files'):
124		tmpAddFiles = string.split(cfg_params['USER.additional_input_files'],',')
125		for tmp in tmpAddFiles:
126		tmp = string.strip(tmp)
127		dirname = ''
128		if not tmp[0]=="/": dirname = "."
129	<	files = glob.glob(os.path.join(dirname, tmp))
129	>	files = []
130	>	if string.find(tmp,"*")>-1:
131	>	files = glob.glob(os.path.join(dirname, tmp))
132	>	if len(files)==0:
133	>	raise CrabException("No additional input file found with this pattern: "+tmp)
134	>	else:
135	>	files.append(tmp)
136		for file in files:
137		if not os.path.exists(file):
138		raise CrabException("Additional input file not found: "+file)
139		pass
140	<	storedFile = common.work_space.shareDir()+file
170	<	shutil.copyfile(file, storedFile)
171	<	self.additional_inbox_files.append(string.strip(storedFile))
140	>	self.additional_inbox_files.append(string.strip(file))
141		pass
142		pass
143		common.logger.debug(5,"Additional input files: "+str(self.additional_inbox_files))
144	<	except KeyError:
176	<	pass
177	<
178	<	# files per job
179	<	try:
180	<	if (cfg_params['CMSSW.files_per_jobs']):
181	<	raise CrabException("files_per_jobs no longer supported.  Quitting.")
182	<	except KeyError:
183	<	pass
144	>	pass
145
146		## Events per job
147	<	try:
147	>	if cfg_params.has_key('CMSSW.events_per_job'):
148		self.eventsPerJob =int( cfg_params['CMSSW.events_per_job'])
149		self.selectEventsPerJob = 1
150	<	except KeyError:
150	>	else:
151		self.eventsPerJob = -1
152		self.selectEventsPerJob = 0
153	<
153	>
154		## number of jobs
155	<	try:
155	>	if cfg_params.has_key('CMSSW.number_of_jobs'):
156		self.theNumberOfJobs =int( cfg_params['CMSSW.number_of_jobs'])
157		self.selectNumberOfJobs = 1
158	<	except KeyError:
158	>	else:
159		self.theNumberOfJobs = 0
160		self.selectNumberOfJobs = 0
161
162	<	try:
162	>	if cfg_params.has_key('CMSSW.total_number_of_events'):
163		self.total_number_of_events = int(cfg_params['CMSSW.total_number_of_events'])
164		self.selectTotalNumberEvents = 1
165	<	except KeyError:
165	>	if self.selectNumberOfJobs  == 1:
166	>	if int(self.total_number_of_events) < int(self.theNumberOfJobs):
167	>	msg = 'Must specify at least one event per job. total_number_of_events > number_of_jobs '
168	>	raise CrabException(msg)
169	>	else:
170		self.total_number_of_events = 0
171		self.selectTotalNumberEvents = 0
172
173	<	if self.pset != None: #CarlosDaniele
173	>	if self.pset != None:
174		if ( (self.selectTotalNumberEvents + self.selectEventsPerJob + self.selectNumberOfJobs) != 2 ):
175		msg = 'Must define exactly two of total_number_of_events, events_per_job, or number_of_jobs.'
176		raise CrabException(msg)
#	Line 214 | Line 179 | class Cmssw(JobType):
179		msg = 'Must specify  number_of_jobs.'
180		raise CrabException(msg)
181
182	<	## source seed for pythia
183	<	try:
184	<	self.sourceSeed = int(cfg_params['CMSSW.pythia_seed'])
185	<	except KeyError:
186	<	self.sourceSeed = None
187	<	common.logger.debug(5,"No seed given")
182	>	## New method of dealing with seeds
183	>	self.incrementSeeds = []
184	>	self.preserveSeeds = []
185	>	if cfg_params.has_key('CMSSW.preserve_seeds'):
186	>	tmpList = cfg_params['CMSSW.preserve_seeds'].split(',')
187	>	for tmp in tmpList:
188	>	tmp.strip()
189	>	self.preserveSeeds.append(tmp)
190	>	if cfg_params.has_key('CMSSW.increment_seeds'):
191	>	tmpList = cfg_params['CMSSW.increment_seeds'].split(',')
192	>	for tmp in tmpList:
193	>	tmp.strip()
194	>	self.incrementSeeds.append(tmp)
195	>
196	>	## Old method of dealing with seeds
197	>	## FUTURE: This is for old CMSSW and old CRAB. Can throw exceptions after a couple of CRAB releases and then
198	>	## remove
199	>	self.sourceSeed = cfg_params.get('CMSSW.pythia_seed',None)
200	>	if self.sourceSeed:
201	>	print "pythia_seed is a deprecated parameter. Use preserve_seeds or increment_seeds in the future.\n","Added to increment_seeds."
202	>	self.incrementSeeds.append('sourceSeed')
203	>	self.incrementSeeds.append('theSource')
204	>
205	>	self.sourceSeedVtx = cfg_params.get('CMSSW.vtx_seed',None)
206	>	if self.sourceSeedVtx:
207	>	print "vtx_seed is a deprecated parameter. Use preserve_seeds or increment_seeds in the future.\n","Added to increment_seeds."
208	>	self.incrementSeeds.append('VtxSmeared')
209	>
210	>	self.sourceSeedG4 = cfg_params.get('CMSSW.g4_seed',None)
211	>	if self.sourceSeedG4:
212	>	print "g4_seed is a deprecated parameter. Use preserve_seeds or increment_seeds in the future.\n","Added to increment_seeds."
213	>	self.incrementSeeds.append('g4SimHits')
214	>
215	>	self.sourceSeedMix = cfg_params.get('CMSSW.mix_seed',None)
216	>	if self.sourceSeedMix:
217	>	print "mix_seed is a deprecated parameter. Use preserve_seeds or increment_seeds in the future.\n","Added to increment_seeds."
218	>	self.incrementSeeds.append('mix')
219	>
220	>	self.firstRun = cfg_params.get('CMSSW.first_run',None)
221
224	–	try:
225	–	self.sourceSeedVtx = int(cfg_params['CMSSW.vtx_seed'])
226	–	except KeyError:
227	–	self.sourceSeedVtx = None
228	–	common.logger.debug(5,"No vertex seed given")
229	–	try:
230	–	self.firstRun = int(cfg_params['CMSSW.first_run'])
231	–	except KeyError:
232	–	self.firstRun = None
233	–	common.logger.debug(5,"No first run given")
222		if self.pset != None: #CarlosDaniele
223	<	self.PsetEdit = PsetManipulator.PsetManipulator(self.pset) #Daniele Pset
223	>	import PsetManipulator as pp
224	>	PsetEdit = pp.PsetManipulator(self.pset) #Daniele Pset
225	>
226	>	# Copy/return
227	>
228	>	self.copy_data = int(cfg_params.get('USER.copy_data',0))
229	>	self.return_data = int(cfg_params.get('USER.return_data',0))
230
231		#DBSDLS-start
232	<	## Initialize the variables that are extracted from DBS/DLS and needed in other places of the code
232	>	## Initialize the variables that are extracted from DBS/DLS and needed in other places of the code
233		self.maxEvents=0  # max events available   ( --> check the requested nb. of evts in Creator.py)
234		self.DBSPaths={}  # all dbs paths requested ( --> input to the site local discovery script)
235		self.jobDestination=[]  # Site destination(s) for each job (list of lists)
#	Line 244 | Line 238 | class Cmssw(JobType):
238		blockSites = {}
239		if self.datasetPath:
240		blockSites = self.DataDiscoveryAndLocation(cfg_params)
241	<	#DBSDLS-end
241	>	#DBSDLS-end
242	>
243
249	–	self.tgzNameWithPath = self.getTarBall(self.executable)
250	–
244		## Select Splitting
245	<	if self.selectNoInput:
246	<	if self.pset == None: #CarlosDaniele
245	>	if self.selectNoInput:
246	>	if self.pset == None:
247		self.jobSplittingForScript()
248		else:
249		self.jobSplittingNoInput()
250	<	else:
250	>	else:
251		self.jobSplittingByBlocks(blockSites)
252
253		# modify Pset
254	<	if self.pset != None: #CarlosDaniele
254	>	if self.pset != None:
255		try:
256	<	if (self.datasetPath): # standard job
257	<	# allow to processa a fraction of events in a file
258	<	self.PsetEdit.inputModule("INPUT")
259	<	self.PsetEdit.maxEvent("INPUTMAXEVENTS")
260	<	self.PsetEdit.skipEvent("INPUTSKIPEVENTS")
268	<	else:  # pythia like job
269	<	self.PsetEdit.maxEvent(self.eventsPerJob)
270	<	if (self.firstRun):
271	<	self.PsetEdit.pythiaFirstRun("INPUTFIRSTRUN")  #First Run
272	<	if (self.sourceSeed) :
273	<	self.PsetEdit.pythiaSeed("INPUT")
274	<	if (self.sourceSeedVtx) :
275	<	self.PsetEdit.pythiaSeedVtx("INPUTVTX")
276	<	# add FrameworkJobReport to parameter-set
277	<	self.PsetEdit.addCrabFJR(self.fjrFileName)
278	<	self.PsetEdit.psetWriter(self.configFilename())
256	>	# Add FrameworkJobReport to parameter-set, set max events.
257	>	# Reset later for data jobs by writeCFG which does all modifications
258	>	PsetEdit.addCrabFJR(self.fjrFileName) # FUTURE: Job report addition not needed by CMSSW>1.5
259	>	PsetEdit.maxEvent(self.eventsPerJob)
260	>	PsetEdit.psetWriter(self.configFilename())
261		except:
262	<	msg='Error while manipuliating ParameterSet: exiting...'
262	>	msg='Error while manipulating ParameterSet: exiting...'
263		raise CrabException(msg)
264	+	self.tgzNameWithPath = self.getTarBall(self.executable)
265
266		def DataDiscoveryAndLocation(self, cfg_params):
267
268	+	import DataDiscovery
269	+	import DataLocation
270		common.logger.debug(10,"CMSSW::DataDiscoveryAndLocation()")
271
272		datasetPath=self.datasetPath
273
274		## Contact the DBS
275	<	common.logger.message("Contacting DBS...")
275	>	common.logger.message("Contacting Data Discovery Services ...")
276		try:
277	<
293	<	if self.use_dbs_2 == 1 :
294	<	self.pubdata=DataDiscovery_DBS2.DataDiscovery_DBS2(datasetPath, cfg_params)
295	<	else :
296	<	self.pubdata=DataDiscovery.DataDiscovery(datasetPath, cfg_params)
277	>	self.pubdata=DataDiscovery.DataDiscovery(datasetPath, cfg_params)
278		self.pubdata.fetchDBSInfo()
279
280		except DataDiscovery.NotExistingDatasetError, ex :
#	Line 305 | Line 286 | class Cmssw(JobType):
286		except DataDiscovery.DataDiscoveryError, ex:
287		msg = 'ERROR ***: failed Data Discovery in DBS :  %s'%ex.getErrorMessage()
288		raise CrabException(msg)
308	–	except DataDiscovery_DBS2.NotExistingDatasetError_DBS2, ex :
309	–	msg = 'ERROR ***: failed Data Discovery in DBS : %s'%ex.getErrorMessage()
310	–	raise CrabException(msg)
311	–	except DataDiscovery_DBS2.NoDataTierinProvenanceError_DBS2, ex :
312	–	msg = 'ERROR ***: failed Data Discovery in DBS : %s'%ex.getErrorMessage()
313	–	raise CrabException(msg)
314	–	except DataDiscovery_DBS2.DataDiscoveryError_DBS2, ex:
315	–	msg = 'ERROR ***: failed Data Discovery in DBS :  %s'%ex.getErrorMessage()
316	–	raise CrabException(msg)
317	–
318	–	## get list of all required data in the form of dbs paths  (dbs path = /dataset/datatier/owner)
319	–	common.logger.message("Required data are :"+self.datasetPath)
289
290		self.filesbyblock=self.pubdata.getFiles()
291		self.eventsbyblock=self.pubdata.getEventsPerBlock()
292		self.eventsbyfile=self.pubdata.getEventsPerFile()
293
294		## get max number of events
295	<	self.maxEvents=self.pubdata.getMaxEvents() ##  self.maxEvents used in Creator.py
327	<	common.logger.message("The number of available events is %s\n"%self.maxEvents)
295	>	self.maxEvents=self.pubdata.getMaxEvents()
296
329	–	common.logger.message("Contacting DLS...")
297		## Contact the DLS and build a list of sites hosting the fileblocks
298		try:
299		dataloc=DataLocation.DataLocation(self.filesbyblock.keys(),cfg_params)
#	Line 334 | Line 301 | class Cmssw(JobType):
301		except DataLocation.DataLocationError , ex:
302		msg = 'ERROR ***: failed Data Location in DLS \n %s '%ex.getErrorMessage()
303		raise CrabException(msg)
304	<
304	>
305
306		sites = dataloc.getSites()
307		allSites = []
#	Line 344 | Line 311 | class Cmssw(JobType):
311		allSites.append(oneSite)
312		allSites = self.uniquelist(allSites)
313
314	<	common.logger.message("Sites ("+str(len(allSites))+") hosting part/all of dataset: "+str(allSites))
315	<	common.logger.debug(6, "List of Sites: "+str(allSites))
314	>	# screen output
315	>	common.logger.message("Requested dataset: " + datasetPath + " has " + str(self.maxEvents) + " events in " + str(len(self.filesbyblock.keys())) + " blocks.\n")
316	>
317		return sites
318	<
318	>
319		def jobSplittingByBlocks(self, blockSites):
320		"""
321		Perform job splitting. Jobs run over an integer number of files
#	Line 397 | Line 365 | class Cmssw(JobType):
365		totalNumberOfJobs = 999999999
366		else :
367		totalNumberOfJobs = self.ncjobs
400	–
368
369		blocks = blockSites.keys()
370		blockCount = 0
#	Line 407 | Line 374 | class Cmssw(JobType):
374		jobCount = 0
375		list_of_lists = []
376
377	+	# list tracking which jobs are in which jobs belong to which block
378	+	jobsOfBlock = {}
379	+
380		# ---- Iterate over the blocks in the dataset until ---- #
381		# ---- we've met the requested total # of events    ---- #
382		while ( (eventsRemaining > 0) and (blockCount < numBlocksInDataset) and (jobCount < totalNumberOfJobs)):
383		block = blocks[blockCount]
384		blockCount += 1
385	<
385	>	if block not in jobsOfBlock.keys() :
386	>	jobsOfBlock[block] = []
387	>
388		if self.eventsbyblock.has_key(block) :
389		numEventsInBlock = self.eventsbyblock[block]
390		common.logger.debug(5,'Events in Block File '+str(numEventsInBlock))
391	<
391	>
392		files = self.filesbyblock[block]
393		numFilesInBlock = len(files)
394		if (numFilesInBlock <= 0):
#	Line 424 | Line 396 | class Cmssw(JobType):
396		fileCount = 0
397
398		# ---- New block => New job ---- #
399	<	parString = "\\{"
399	>	parString = ""
400		# counter for number of events in files currently worked on
401		filesEventCount = 0
402		# flag if next while loop should touch new file
403		newFile = 1
404		# job event counter
405		jobSkipEventCount = 0
406	<
406	>
407		# ---- Iterate over the files in the block until we've met the requested ---- #
408		# ---- total # of events or we've gone over all the files in this block  ---- #
409		while ( (eventsRemaining > 0) and (fileCount < numFilesInBlock) and (jobCount < totalNumberOfJobs) ):
#	Line 447 | Line 419 | class Cmssw(JobType):
419		newFile = 0
420		except KeyError:
421		common.logger.message("File "+str(file)+" has unknown number of events: skipping")
450	–
422
423	+	eventsPerJobRequested = min(eventsPerJobRequested, eventsRemaining)
424		# if less events in file remain than eventsPerJobRequested
425	<	if ( filesEventCount - jobSkipEventCount < eventsPerJobRequested ) :
425	>	if ( filesEventCount - jobSkipEventCount < eventsPerJobRequested):
426		# if last file in block
427		if ( fileCount == numFilesInBlock-1 ) :
428		# end job using last file, use remaining events in block
429		# close job and touch new file
430		fullString = parString[:-2]
459	–	fullString += '\\}'
431		list_of_lists.append([fullString,str(-1),str(jobSkipEventCount)])
432		common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(filesEventCount - jobSkipEventCount)+" events (last file in block).")
433		self.jobDestination.append(blockSites[block])
434		common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
435	+	# fill jobs of block dictionary
436	+	jobsOfBlock[block].append(jobCount+1)
437		# reset counter
438		jobCount = jobCount + 1
439		totalEventCount = totalEventCount + filesEventCount - jobSkipEventCount
440		eventsRemaining = eventsRemaining - filesEventCount + jobSkipEventCount
441		jobSkipEventCount = 0
442		# reset file
443	<	parString = "\\{"
443	>	parString = ""
444		filesEventCount = 0
445		newFile = 1
446		fileCount += 1
#	Line 479 | Line 452 | class Cmssw(JobType):
452		elif ( filesEventCount - jobSkipEventCount == eventsPerJobRequested ) :
453		# close job and touch new file
454		fullString = parString[:-2]
482	–	fullString += '\\}'
455		list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
456		common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.")
457		self.jobDestination.append(blockSites[block])
458		common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
459	+	jobsOfBlock[block].append(jobCount+1)
460		# reset counter
461		jobCount = jobCount + 1
462		totalEventCount = totalEventCount + eventsPerJobRequested
463		eventsRemaining = eventsRemaining - eventsPerJobRequested
464		jobSkipEventCount = 0
465		# reset file
466	<	parString = "\\{"
466	>	parString = ""
467		filesEventCount = 0
468		newFile = 1
469		fileCount += 1
470	<
470	>
471		# if more events in file remain than eventsPerJobRequested
472		else :
473		# close job but don't touch new file
474		fullString = parString[:-2]
502	–	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	+	jobsOfBlock[block].append(jobCount+1)
480		# increase counter
481		jobCount = jobCount + 1
482		totalEventCount = totalEventCount + eventsPerJobRequested
#	Line 513 | Line 486 | class Cmssw(JobType):
486		jobSkipEventCount = eventsPerJobRequested - (filesEventCount - jobSkipEventCount - self.eventsbyfile[file])
487		# remove all but the last file
488		filesEventCount = self.eventsbyfile[file]
489	<	parString = "\\{"
517	<	parString += '\\\"' + file + '\\\"\,'
489	>	parString = '\\\"' + file + '\\\"\,'
490		pass # END if
491		pass # END while (iterate over files in the block)
492		pass # END while (iterate over blocks in the dataset)
493		self.ncjobs = self.total_number_of_jobs = jobCount
494		if (eventsRemaining > 0 and jobCount < totalNumberOfJobs ):
495		common.logger.message("Could not run on all requested events because some blocks not hosted at allowed sites.")
496	<	common.logger.message("\n"+str(jobCount)+" job(s) can run on "+str(totalEventCount)+" events.\n")
497	<
496	>	common.logger.message(str(jobCount)+" job(s) can run on "+str(totalEventCount)+" events.\n")
497	>
498	>	# screen output
499	>	screenOutput = "List of jobs and available destination sites:\n\n"
500	>
501	>	# keep trace of block with no sites to print a warning at the end
502	>	noSiteBlock = []
503	>	bloskNoSite = []
504	>
505	>	blockCounter = 0
506	>	for block in blocks:
507	>	if block in jobsOfBlock.keys() :
508	>	blockCounter += 1
509	>	screenOutput += "Block %5i: jobs %20s: sites: %s\n" % (blockCounter,spanRanges(jobsOfBlock[block]),
510	>	','.join(self.blackWhiteListParser.checkWhiteList(self.blackWhiteListParser.checkBlackList(blockSites[block],block),block)))
511	>	if len(self.blackWhiteListParser.checkWhiteList(self.blackWhiteListParser.checkBlackList(blockSites[block],block),block)) == 0:
512	>	noSiteBlock.append( spanRanges(jobsOfBlock[block]) )
513	>	bloskNoSite.append( blockCounter )
514	>
515	>	common.logger.message(screenOutput)
516	>	if len(noSiteBlock) > 0 and len(bloskNoSite) > 0:
517	>	msg = 'WARNING: No sites are hosting any part of data for block:\n                '
518	>	virgola = ""
519	>	if len(bloskNoSite) > 1:
520	>	virgola = ","
521	>	for block in bloskNoSite:
522	>	msg += ' ' + str(block) + virgola
523	>	msg += '\n               Related jobs:\n                 '
524	>	virgola = ""
525	>	if len(noSiteBlock) > 1:
526	>	virgola = ","
527	>	for range_jobs in noSiteBlock:
528	>	msg += str(range_jobs) + virgola
529	>	msg += '\n               will not be submitted and this block of data can not be analyzed!\n'
530	>	if self.cfg_params.has_key('EDG.se_white_list'):
531	>	msg += 'WARNING: SE White List: '+self.cfg_params['EDG.se_white_list']+'\n'
532	>	msg += '(Hint: By whitelisting you force the job to run at this particular site(s).\n'
533	>	msg += 'Please check if the dataset is available at this site!)\n'
534	>	if self.cfg_params.has_key('EDG.ce_white_list'):
535	>	msg += 'WARNING: CE White List: '+self.cfg_params['EDG.ce_white_list']+'\n'
536	>	msg += '(Hint: By whitelisting you force the job to run at this particular site(s).\n'
537	>	msg += 'Please check if the dataset is available at this site!)\n'
538	>
539	>	common.logger.message(msg)
540	>
541		self.list_of_args = list_of_lists
542		return
543
#	Line 531 | Line 546 | class Cmssw(JobType):
546		Perform job splitting based on number of event per job
547		"""
548		common.logger.debug(5,'Splitting per events')
549	<	common.logger.message('Required '+str(self.eventsPerJob)+' events per job ')
550	<	common.logger.message('Required '+str(self.theNumberOfJobs)+' jobs in total ')
551	<	common.logger.message('Required '+str(self.total_number_of_events)+' events in total ')
549	>
550	>	if (self.selectEventsPerJob):
551	>	common.logger.message('Required '+str(self.eventsPerJob)+' events per job ')
552	>	if (self.selectNumberOfJobs):
553	>	common.logger.message('Required '+str(self.theNumberOfJobs)+' jobs in total ')
554	>	if (self.selectTotalNumberEvents):
555	>	common.logger.message('Required '+str(self.total_number_of_events)+' events in total ')
556
557		if (self.total_number_of_events < 0):
558		msg='Cannot split jobs per Events with "-1" as total number of events'
#	Line 542 | Line 561 | class Cmssw(JobType):
561		if (self.selectEventsPerJob):
562		if (self.selectTotalNumberEvents):
563		self.total_number_of_jobs = int(self.total_number_of_events/self.eventsPerJob)
564	<	elif(self.selectNumberOfJobs) :
564	>	elif(self.selectNumberOfJobs) :
565		self.total_number_of_jobs =self.theNumberOfJobs
566	<	self.total_number_of_events =int(self.theNumberOfJobs*self.eventsPerJob)
566	>	self.total_number_of_events =int(self.theNumberOfJobs*self.eventsPerJob)
567
568		elif (self.selectNumberOfJobs) :
569		self.total_number_of_jobs = self.theNumberOfJobs
570		self.eventsPerJob = int(self.total_number_of_events/self.total_number_of_jobs)
571	<
571	>
572		common.logger.debug(5,'N jobs  '+str(self.total_number_of_jobs))
573
574		# is there any remainder?
#	Line 565 | Line 584 | class Cmssw(JobType):
584		self.list_of_args = []
585		for i in range(self.total_number_of_jobs):
586		## Since there is no input, any site is good
587	<	# self.jobDestination.append(["Any"])
588	<	self.jobDestination.append([""]) #must be empty to write correctly the xml
570	<	args=''
587	>	self.jobDestination.append([""]) #must be empty to write correctly the xml
588	>	args=[]
589		if (self.firstRun):
590	<	## pythia first run
591	<	#self.list_of_args.append([(str(self.firstRun)+str(i))])
592	<	args=args+(str(self.firstRun)+str(i))
575	<	else:
576	<	## no first run
577	<	#self.list_of_args.append([str(i)])
578	<	args=args+str(i)
579	<	if (self.sourceSeed):
580	<	if (self.sourceSeedVtx):
581	<	## pythia + vtx random seed
582	<	#self.list_of_args.append([
583	<	#                          str(self.sourceSeed)+str(i),
584	<	#                          str(self.sourceSeedVtx)+str(i)
585	<	#                          ])
586	<	args=args+str(',')+str(self.sourceSeed)+str(i)+str(',')+str(self.sourceSeedVtx)+str(i)
587	<	else:
588	<	## only pythia random seed
589	<	#self.list_of_args.append([(str(self.sourceSeed)+str(i))])
590	<	args=args +str(',')+str(self.sourceSeed)+str(i)
591	<	else:
592	<	## no random seed
593	<	if str(args)=='': args=args+(str(self.firstRun)+str(i))
594	<	arguments=args.split(',')
595	<	if len(arguments)==3:self.list_of_args.append([str(arguments[0]),str(arguments[1]),str(arguments[2])])
596	<	elif len(arguments)==2:self.list_of_args.append([str(arguments[0]),str(arguments[1])])
597	<	else :self.list_of_args.append([str(arguments[0])])
598	<
599	<	#   print self.list_of_args
590	>	## pythia first run
591	>	args.append(str(self.firstRun)+str(i))
592	>	self.list_of_args.append(args)
593
594		return
595
596
597	<	def jobSplittingForScript(self):#CarlosDaniele
597	>	def jobSplittingForScript(self):
598		"""
599		Perform job splitting based on number of job
600		"""
#	Line 617 | Line 610 | class Cmssw(JobType):
610		# argument is seed number.$i
611		self.list_of_args = []
612		for i in range(self.total_number_of_jobs):
620	–	## Since there is no input, any site is good
621	–	# self.jobDestination.append(["Any"])
613		self.jobDestination.append([""])
623	–	## no random seed
614		self.list_of_args.append([str(i)])
615		return
616
617		def split(self, jobParams):
618	<
629	<	common.jobDB.load()
630	<	#### Fabio
618	>
619		njobs = self.total_number_of_jobs
620		arglist = self.list_of_args
621		# create the empty structure
622		for i in range(njobs):
623		jobParams.append("")
624	<
624	>
625	>	listID=[]
626	>	listField=[]
627		for job in range(njobs):
628		jobParams[job] = arglist[job]
629	<	# print str(arglist[job])
630	<	# print jobParams[job]
631	<	common.jobDB.setArguments(job, jobParams[job])
632	<	common.logger.debug(5,"Job "+str(job)+" Destination: "+str(self.jobDestination[job]))
633	<	common.jobDB.setDestination(job, self.jobDestination[job])
629	>	listID.append(job+1)
630	>	job_ToSave ={}
631	>	concString = ' '
632	>	argu=''
633	>	if len(jobParams[job]):
634	>	argu +=   concString.join(jobParams[job] )
635	>	job_ToSave['arguments']= str(job+1)+' '+argu
636	>	job_ToSave['dlsDestination']= self.jobDestination[job]
637	>	listField.append(job_ToSave)
638	>	msg="Job "+str(job)+" Arguments:   "+str(job+1)+" "+argu+"\n"  \
639	>	+"                     Destination: "+str(self.jobDestination[job])
640	>	common.logger.debug(5,msg)
641	>	common._db.updateJob_(listID,listField)
642	>	self.argsList = (len(jobParams[0])+1)
643
645	–	common.jobDB.save()
644		return
645	<
648	<	def getJobTypeArguments(self, nj, sched):
649	<	result = ''
650	<	for i in common.jobDB.arguments(nj):
651	<	result=result+str(i)+" "
652	<	return result
653	<
645	>
646		def numberOfJobs(self):
655	–	# Fabio
647		return self.total_number_of_jobs
648
649		def getTarBall(self, exe):
650		"""
651		Return the TarBall with lib and exe
652		"""
662	–
663	–	# if it exist, just return it
664	–	#
665	–	# Marco. Let's start to use relative path for Boss XML files
666	–	#
653		self.tgzNameWithPath = common.work_space.pathForTgz()+'share/'+self.tgz_name
654		if os.path.exists(self.tgzNameWithPath):
655		return self.tgzNameWithPath
#	Line 677 | Line 663 | class Cmssw(JobType):
663
664		# First of all declare the user Scram area
665		swArea = self.scram.getSWArea_()
680	–	#print "swArea = ", swArea
681	–	# swVersion = self.scram.getSWVersion()
682	–	# print "swVersion = ", swVersion
666		swReleaseTop = self.scram.getReleaseTop_()
667	<	#print "swReleaseTop = ", swReleaseTop
685	<
667	>
668		## check if working area is release top
669		if swReleaseTop == '' or swArea == swReleaseTop:
670	+	common.logger.debug(3,"swArea = "+swArea+" swReleaseTop ="+swReleaseTop)
671		return
672
673		import tarfile
674		try: # create tar ball
675		tar = tarfile.open(self.tgzNameWithPath, "w:gz")
676		## First find the executable
677	<	if (executable != ''):
677	>	if (self.executable != ''):
678		exeWithPath = self.scram.findFile_(executable)
679		if ( not exeWithPath ):
680		raise CrabException('User executable '+executable+' not found')
681	<
681	>
682		## then check if it's private or not
683		if exeWithPath.find(swReleaseTop) == -1:
684		# the exe is private, so we must ship
#	Line 704 | Line 687 | class Cmssw(JobType):
687		# distinguish case when script is in user project area or given by full path somewhere else
688		if exeWithPath.find(path) >= 0 :
689		exe = string.replace(exeWithPath, path,'')
690	<	tar.add(path+exe,os.path.basename(executable))
690	>	tar.add(path+exe,exe)
691		else :
692		tar.add(exeWithPath,os.path.basename(executable))
693		pass
694		else:
695		# the exe is from release, we'll find it on WN
696		pass
697	<
697	>
698		## Now get the libraries: only those in local working area
699		libDir = 'lib'
700		lib = swArea+'/' +libDir
701		common.logger.debug(5,"lib "+lib+" to be tarred")
702		if os.path.exists(lib):
703		tar.add(lib,libDir)
704	<
704	>
705		## Now check if module dir is present
706		moduleDir = 'module'
707		module = swArea + '/' + moduleDir
#	Line 727 | Line 710 | class Cmssw(JobType):
710
711		## Now check if any data dir(s) is present
712		swAreaLen=len(swArea)
713	+	self.dataExist = False
714		for root, dirs, files in os.walk(swArea):
715		if "data" in dirs:
716	+	self.dataExist=True
717		common.logger.debug(5,"data "+root+"/data"+" to be tarred")
718		tar.add(root+"/data",root[swAreaLen:]+"/data")
719
720	<	## Add ProdAgent dir to tar
721	<	paDir = 'ProdAgentApi'
722	<	pa = os.environ['CRABDIR'] + '/' + 'ProdAgentApi'
723	<	if os.path.isdir(pa):
724	<	tar.add(pa,paDir)
725	<
720	>	### CMSSW ParameterSet
721	>	if not self.pset is None:
722	>	cfg_file = common.work_space.jobDir()+self.configFilename()
723	>	tar.add(cfg_file,self.configFilename())
724	>	common.logger.debug(5,"File added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
725	>
726	>
727	>	## Add ProdCommon dir to tar
728	>	prodcommonDir = 'ProdCommon'
729	>	prodcommonPath = os.environ['CRABDIR'] + '/' + 'ProdCommon'
730	>	if os.path.isdir(prodcommonPath):
731	>	tar.add(prodcommonPath,prodcommonDir)
732	>	common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
733	>
734	>	##### ML stuff
735	>	ML_file_list=['report.py', 'DashboardAPI.py', 'Logger.py', 'ProcInfo.py', 'apmon.py']
736	>	path=os.environ['CRABDIR'] + '/python/'
737	>	for file in ML_file_list:
738	>	tar.add(path+file,file)
739	>	common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
740	>
741	>	##### Utils
742	>	Utils_file_list=['parseCrabFjr.py','writeCfg.py', 'JobReportErrorCode.py']
743	>	for file in Utils_file_list:
744	>	tar.add(path+file,file)
745	>	common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
746	>
747	>	##### AdditionalFiles
748	>	for file in self.additional_inbox_files:
749	>	tar.add(file,string.split(file,'/')[-1])
750		common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
751	+
752		tar.close()
753		except :
754		raise CrabException('Could not create tar-ball')
#	Line 749 | Line 759 | class Cmssw(JobType):
759		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.')
760
761		## create tar-ball with ML stuff
762	<	self.MLtgzfile =  common.work_space.pathForTgz()+'share/MLfiles.tgz'
763	<	try:
754	<	tar = tarfile.open(self.MLtgzfile, "w:gz")
755	<	path=os.environ['CRABDIR'] + '/python/'
756	<	for file in ['report.py', 'DashboardAPI.py', 'Logger.py', 'ProcInfo.py', 'apmon.py', 'parseCrabFjr.py']:
757	<	tar.add(path+file,file)
758	<	common.logger.debug(5,"Files added to "+self.MLtgzfile+" : "+str(tar.getnames()))
759	<	tar.close()
760	<	except :
761	<	raise CrabException('Could not create ML files tar-ball')
762	<
763	<	return
764	<
765	<	def wsSetupEnvironment(self, nj):
762	>
763	>	def wsSetupEnvironment(self, nj=0):
764		"""
765		Returns part of a job script which prepares
766		the execution environment for the job 'nj'.
767		"""
768	+	if (self.CMSSW_major >= 2 and self.CMSSW_minor >= 1) or (self.CMSSW_major >= 3):
769	+	psetName = 'pset.py'
770	+	else:
771	+	psetName = 'pset.cfg'
772		# Prepare JobType-independent part
773	<	txt = ''
774	<
775	<	## OLI_Daniele at this level  middleware already known
774	<
775	<	txt += 'if [ $middleware == LCG ]; then \n'
773	>	txt = '\n#Written by cms_cmssw::wsSetupEnvironment\n'
774	>	txt += 'echo ">>> setup environment"\n'
775	>	txt += 'if [ $middleware == LCG ]; then \n'
776		txt += self.wsSetupCMSLCGEnvironment_()
777		txt += 'elif [ $middleware == OSG ]; then\n'
778		txt += '    WORKING_DIR=`/bin/mktemp  -d $OSG_WN_TMP/cms_XXXXXXXXXXXX`\n'
779	<	txt += '    echo "Created working directory: $WORKING_DIR"\n'
780	<	txt += '    if [ ! -d $WORKING_DIR ] ;then\n'
781	<	txt += '        echo "SET_CMS_ENV 10016 ==> OSG $WORKING_DIR could not be created on WN `hostname`"\n'
782	<	txt += '        echo "JOB_EXIT_STATUS = 10016"\n'
783	<	txt += '        echo "JobExitCode=10016" | tee -a $RUNTIME_AREA/$repo\n'
784	<	txt += '        dumpStatus $RUNTIME_AREA/$repo\n'
785	<	txt += '        rm -f $RUNTIME_AREA/$repo \n'
786	<	txt += '        echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
787	<	txt += '        echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
788	<	txt += '        exit 1\n'
779	>	txt += '    if [ ! $? == 0 ] ;then\n'
780	>	txt += '        echo "ERROR ==> OSG $WORKING_DIR could not be created on WN `hostname`"\n'
781	>	txt += '        job_exit_code=10016\n'
782	>	txt += '        func_exit\n'
783		txt += '    fi\n'
784	+	txt += '    echo ">>> Created working directory: $WORKING_DIR"\n'
785		txt += '\n'
786		txt += '    echo "Change to working directory: $WORKING_DIR"\n'
787		txt += '    cd $WORKING_DIR\n'
788	<	txt += self.wsSetupCMSOSGEnvironment_()
788	>	txt += '    echo ">>> current directory (WORKING_DIR): $WORKING_DIR"\n'
789	>	txt += self.wsSetupCMSOSGEnvironment_()
790		txt += 'fi\n'
791
792		# Prepare JobType-specific part
793		scram = self.scram.commandName()
794		txt += '\n\n'
795	<	txt += 'echo "### SPECIFIC JOB SETUP ENVIRONMENT ###"\n'
795	>	txt += 'echo ">>> specific cmssw setup environment:"\n'
796	>	txt += 'echo "CMSSW_VERSION =  '+self.version+'"\n'
797		txt += scram+' project CMSSW '+self.version+'\n'
798		txt += 'status=$?\n'
799		txt += 'if [ $status != 0 ] ; then\n'
800	<	txt += '   echo "SET_EXE_ENV 10034 ==>ERROR CMSSW '+self.version+' not found on `hostname`" \n'
801	<	txt += '   echo "JOB_EXIT_STATUS = 10034"\n'
802	<	txt += '   echo "JobExitCode=10034" | tee -a $RUNTIME_AREA/$repo\n'
806	<	txt += '   dumpStatus $RUNTIME_AREA/$repo\n'
807	<	txt += '   rm -f $RUNTIME_AREA/$repo \n'
808	<	txt += '   echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
809	<	txt += '   echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
810	<	## OLI_Daniele
811	<	txt += '    if [ $middleware == OSG ]; then \n'
812	<	txt += '        echo "Remove working directory: $WORKING_DIR"\n'
813	<	txt += '        cd $RUNTIME_AREA\n'
814	<	txt += '        /bin/rm -rf $WORKING_DIR\n'
815	<	txt += '        if [ -d $WORKING_DIR ] ;then\n'
816	<	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'
817	<	txt += '            echo "JOB_EXIT_STATUS = 10018"\n'
818	<	txt += '            echo "JobExitCode=10018" | tee -a $RUNTIME_AREA/$repo\n'
819	<	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
820	<	txt += '            rm -f $RUNTIME_AREA/$repo \n'
821	<	txt += '            echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
822	<	txt += '            echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
823	<	txt += '        fi\n'
824	<	txt += '    fi \n'
825	<	txt += '   exit 1 \n'
800	>	txt += '    echo "ERROR ==> CMSSW '+self.version+' not found on `hostname`" \n'
801	>	txt += '    job_exit_code=10034\n'
802	>	txt += '    func_exit\n'
803		txt += 'fi \n'
827	–	txt += 'echo "CMSSW_VERSION =  '+self.version+'"\n'
828	–	txt += 'export SCRAM_ARCH='+self.executable_arch+'\n'
804		txt += 'cd '+self.version+'\n'
805	<	### needed grep for bug in scramv1 ###
806	<	txt += scram+' runtime -sh\n'
805	>	txt += 'SOFTWARE_DIR=`pwd`\n'
806	>	txt += 'echo ">>> current directory (SOFTWARE_DIR): $SOFTWARE_DIR" \n'
807		txt += 'eval `'+scram+' runtime -sh | grep -v SCRAMRT_LSB_JOBNAME`\n'
808	<	txt += 'echo $PATH\n'
809	<
808	>	txt += 'if [ $? != 0 ] ; then\n'
809	>	txt += '    echo "ERROR ==> Problem with the command: "\n'
810	>	txt += '    echo "eval \`'+scram+' runtime -sh | grep -v SCRAMRT_LSB_JOBNAME \` at `hostname`"\n'
811	>	txt += '    job_exit_code=10034\n'
812	>	txt += '    func_exit\n'
813	>	txt += 'fi \n'
814		# Handle the arguments:
815		txt += "\n"
816		txt += "## number of arguments (first argument always jobnumber)\n"
817		txt += "\n"
818	<	#        txt += "narg=$#\n"
840	<	txt += "if [ $nargs -lt 2 ]\n"
818	>	txt += "if [ $nargs -lt "+str(self.argsList)+" ]\n"
819		txt += "then\n"
820	<	txt += "    echo 'SET_EXE_ENV 1 ==> ERROR Too few arguments' +$nargs+ \n"
821	<	txt += '    echo "JOB_EXIT_STATUS = 50113"\n'
822	<	txt += '    echo "JobExitCode=50113" | tee -a $RUNTIME_AREA/$repo\n'
845	<	txt += '    dumpStatus $RUNTIME_AREA/$repo\n'
846	<	txt += '    rm -f $RUNTIME_AREA/$repo \n'
847	<	txt += '    echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
848	<	txt += '    echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
849	<	## OLI_Daniele
850	<	txt += '    if [ $middleware == OSG ]; then \n'
851	<	txt += '        echo "Remove working directory: $WORKING_DIR"\n'
852	<	txt += '        cd $RUNTIME_AREA\n'
853	<	txt += '        /bin/rm -rf $WORKING_DIR\n'
854	<	txt += '        if [ -d $WORKING_DIR ] ;then\n'
855	<	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'
856	<	txt += '            echo "JOB_EXIT_STATUS = 50114"\n'
857	<	txt += '            echo "JobExitCode=50114" | tee -a $RUNTIME_AREA/$repo\n'
858	<	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
859	<	txt += '            rm -f $RUNTIME_AREA/$repo \n'
860	<	txt += '            echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
861	<	txt += '            echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
862	<	txt += '        fi\n'
863	<	txt += '    fi \n'
864	<	txt += "    exit 1\n"
820	>	txt += "    echo 'ERROR ==> Too few arguments' +$nargs+ \n"
821	>	txt += '    job_exit_code=50113\n'
822	>	txt += "    func_exit\n"
823		txt += "fi\n"
824		txt += "\n"
825
826		# Prepare job-specific part
827		job = common.job_list[nj]
828	<	if self.pset != None: #CarlosDaniele
828	>	if (self.datasetPath):
829	>	txt += '\n'
830	>	txt += 'DatasetPath='+self.datasetPath+'\n'
831	>
832	>	datasetpath_split = self.datasetPath.split("/")
833	>
834	>	txt += 'PrimaryDataset='+datasetpath_split[1]+'\n'
835	>	txt += 'DataTier='+datasetpath_split[2]+'\n'
836	>	txt += 'ApplicationFamily=cmsRun\n'
837	>
838	>	else:
839	>	txt += 'DatasetPath=MCDataTier\n'
840	>	txt += 'PrimaryDataset=null\n'
841	>	txt += 'DataTier=null\n'
842	>	txt += 'ApplicationFamily=MCDataTier\n'
843	>	if self.pset != None:
844		pset = os.path.basename(job.configFilename())
845		txt += '\n'
846	+	txt += 'cp  $RUNTIME_AREA/'+pset+' .\n'
847		if (self.datasetPath): # standard job
848	<	#txt += 'InputFiles=$2\n'
849	<	txt += 'InputFiles=${args[1]}\n'
850	<	txt += 'MaxEvents=${args[2]}\n'
877	<	txt += 'SkipEvents=${args[3]}\n'
848	>	txt += 'InputFiles=${args[1]}; export InputFiles\n'
849	>	txt += 'MaxEvents=${args[2]}; export MaxEvents\n'
850	>	txt += 'SkipEvents=${args[3]}; export SkipEvents\n'
851		txt += 'echo "Inputfiles:<$InputFiles>"\n'
879	–	txt += 'sed "s#{\'INPUT\'}#$InputFiles#" $RUNTIME_AREA/'+pset+' > pset_tmp_1.cfg\n'
852		txt += 'echo "MaxEvents:<$MaxEvents>"\n'
881	–	txt += 'sed "s#INPUTMAXEVENTS#$MaxEvents#" pset_tmp_1.cfg > pset_tmp_2.cfg\n'
853		txt += 'echo "SkipEvents:<$SkipEvents>"\n'
883	–	txt += 'sed "s#INPUTSKIPEVENTS#$SkipEvents#" pset_tmp_2.cfg > pset.cfg\n'
854		else:  # pythia like job
855	<	if (self.sourceSeed):
856	<	txt += 'FirstRun=${args[1]}\n'
855	>	txt += 'PreserveSeeds='  + ','.join(self.preserveSeeds)  + '; export PreserveSeeds\n'
856	>	txt += 'IncrementSeeds=' + ','.join(self.incrementSeeds) + '; export IncrementSeeds\n'
857	>	txt += 'echo "PreserveSeeds: <$PreserveSeeds>"\n'
858	>	txt += 'echo "IncrementSeeds:<$IncrementSeeds>"\n'
859	>	if (self.firstRun):
860	>	txt += 'FirstRun=${args[1]}; export FirstRun\n'
861		txt += 'echo "FirstRun: <$FirstRun>"\n'
888	–	txt += 'sed "s#\<INPUTFIRSTRUN\>#$FirstRun#" $RUNTIME_AREA/'+pset+' > tmp_1.cfg\n'
889	–	else:
890	–	txt += '# Copy untouched pset\n'
891	–	txt += 'cp $RUNTIME_AREA/'+pset+' tmp_1.cfg\n'
892	–	if (self.sourceSeed):
893	–	#                    txt += 'Seed=$2\n'
894	–	txt += 'Seed=${args[2]}\n'
895	–	txt += 'echo "Seed: <$Seed>"\n'
896	–	txt += 'sed "s#\<INPUT\>#$Seed#" tmp_1.cfg > tmp_2.cfg\n'
897	–	if (self.sourceSeedVtx):
898	–	#                        txt += 'VtxSeed=$3\n'
899	–	txt += 'VtxSeed=${args[3]}\n'
900	–	txt += 'echo "VtxSeed: <$VtxSeed>"\n'
901	–	txt += 'sed "s#INPUTVTX#$VtxSeed#" tmp_2.cfg > pset.cfg\n'
902	–	else:
903	–	txt += 'mv tmp_2.cfg pset.cfg\n'
904	–	else:
905	–	txt += 'mv tmp_1.cfg pset.cfg\n'
906	–	# txt += '# Copy untouched pset\n'
907	–	# txt += 'cp $RUNTIME_AREA/'+pset+' pset.cfg\n'
862
863	+	txt += 'mv -f ' + pset + ' ' + psetName + '\n'
864
910	–	if len(self.additional_inbox_files) > 0:
911	–	for file in self.additional_inbox_files:
912	–	relFile = file.split("/")[-1]
913	–	txt += 'if [ -e $RUNTIME_AREA/'+relFile+' ] ; then\n'
914	–	txt += '   cp $RUNTIME_AREA/'+relFile+' .\n'
915	–	txt += '   chmod +x '+relFile+'\n'
916	–	txt += 'fi\n'
917	–	pass
865
866	<	if self.pset != None: #CarlosDaniele
867	<	txt += 'echo "### END JOB SETUP ENVIRONMENT ###"\n\n'
921	<
866	>	if self.pset != None:
867	>	# FUTURE: Can simply for 2_1_x and higher
868		txt += '\n'
869	<	txt += 'echo "***** cat pset.cfg *********"\n'
870	<	txt += 'cat pset.cfg\n'
871	<	txt += 'echo "****** end pset.cfg ********"\n'
869	>	if self.debug_pset==True:
870	>	txt += 'echo "***** cat ' + psetName + ' *********"\n'
871	>	txt += 'cat ' + psetName + '\n'
872	>	txt += 'echo "****** end ' + psetName + ' ********"\n'
873	>	txt += '\n'
874	>	txt += 'PSETHASH=`edmConfigHash < ' + psetName + '` \n'
875	>	txt += 'echo "PSETHASH = $PSETHASH" \n'
876		txt += '\n'
927	–	# txt += 'echo "***** cat pset1.cfg *********"\n'
928	–	# txt += 'cat pset1.cfg\n'
929	–	# txt += 'echo "****** end pset1.cfg ********"\n'
877		return txt
878
879	<	def wsBuildExe(self, nj=0):
879	>	def wsUntarSoftware(self, nj=0):
880		"""
881		Put in the script the commands to build an executable
882		or a library.
883		"""
884
885	<	txt = ""
885	>	txt = '\n#Written by cms_cmssw::wsUntarSoftware\n'
886
887		if os.path.isfile(self.tgzNameWithPath):
888	<	txt += 'echo "tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+'"\n'
888	>	txt += 'echo ">>> tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+' :" \n'
889		txt += 'tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+'\n'
890	+	txt += 'ls -Al \n'
891		txt += 'untar_status=$? \n'
892		txt += 'if [ $untar_status -ne 0 ]; then \n'
893	<	txt += '   echo "SET_EXE 1 ==> ERROR Untarring .tgz file failed"\n'
894	<	txt += '   echo "JOB_EXIT_STATUS = $untar_status" \n'
895	<	txt += '   echo "JobExitCode=$untar_status" | tee -a $RUNTIME_AREA/$repo\n'
948	<	txt += '   if [ $middleware == OSG ]; then \n'
949	<	txt += '       echo "Remove working directory: $WORKING_DIR"\n'
950	<	txt += '       cd $RUNTIME_AREA\n'
951	<	txt += '       /bin/rm -rf $WORKING_DIR\n'
952	<	txt += '       if [ -d $WORKING_DIR ] ;then\n'
953	<	txt += '           echo "SET_EXE 50999 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after Untarring .tgz file failed"\n'
954	<	txt += '           echo "JOB_EXIT_STATUS = 50999"\n'
955	<	txt += '           echo "JobExitCode=50999" | tee -a $RUNTIME_AREA/$repo\n'
956	<	txt += '           dumpStatus $RUNTIME_AREA/$repo\n'
957	<	txt += '           rm -f $RUNTIME_AREA/$repo \n'
958	<	txt += '           echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
959	<	txt += '           echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
960	<	txt += '       fi\n'
961	<	txt += '   fi \n'
962	<	txt += '   \n'
963	<	txt += '   exit 1 \n'
893	>	txt += '   echo "ERROR ==> Untarring .tgz file failed"\n'
894	>	txt += '   job_exit_code=$untar_status\n'
895	>	txt += '   func_exit\n'
896		txt += 'else \n'
897		txt += '   echo "Successful untar" \n'
898		txt += 'fi \n'
899		txt += '\n'
900	<	txt += 'echo "Include ProdAgentApi in PYTHONPATH"\n'
900	>	txt += 'echo ">>> Include ProdCommon in PYTHONPATH:"\n'
901		txt += 'if [ -z "$PYTHONPATH" ]; then\n'
902	<	txt += '   export PYTHONPATH=ProdAgentApi\n'
902	>	txt += '   export PYTHONPATH=$RUNTIME_AREA/ProdCommon\n'
903		txt += 'else\n'
904	<	txt += '   export PYTHONPATH=ProdAgentApi:${PYTHONPATH}\n'
904	>	txt += '   export PYTHONPATH=$RUNTIME_AREA/ProdCommon:${PYTHONPATH}\n'
905	>	txt += 'echo "PYTHONPATH=$PYTHONPATH"\n'
906		txt += 'fi\n'
907		txt += '\n'
908
909		pass
910	<
910	>
911	>	return txt
912	>
913	>	def wsBuildExe(self, nj=0):
914	>	"""
915	>	Put in the script the commands to build an executable
916	>	or a library.
917	>	"""
918	>
919	>	txt = '\n#Written by cms_cmssw::wsBuildExe\n'
920	>	txt += 'echo ">>> moving CMSSW software directories in `pwd`" \n'
921	>
922	>	txt += 'rm -r lib/ module/ \n'
923	>	txt += 'mv $RUNTIME_AREA/lib/ . \n'
924	>	txt += 'mv $RUNTIME_AREA/module/ . \n'
925	>	if self.dataExist == True:
926	>	txt += 'rm -r src/ \n'
927	>	txt += 'mv $RUNTIME_AREA/src/ . \n'
928	>	if len(self.additional_inbox_files)>0:
929	>	for file in self.additional_inbox_files:
930	>	txt += 'mv $RUNTIME_AREA/'+os.path.basename(file)+' . \n'
931	>	txt += 'mv $RUNTIME_AREA/ProdCommon/ . \n'
932	>
933	>	txt += 'if [ -z "$PYTHONPATH" ]; then\n'
934	>	txt += '   export PYTHONPATH=$SOFTWARE_DIR/ProdCommon\n'
935	>	txt += 'else\n'
936	>	txt += '   export PYTHONPATH=$SOFTWARE_DIR/ProdCommon:${PYTHONPATH}\n'
937	>	txt += 'echo "PYTHONPATH=$PYTHONPATH"\n'
938	>	txt += 'fi\n'
939	>	txt += '\n'
940	>
941		return txt
942
943		def modifySteeringCards(self, nj):
944		"""
945	<	modify the card provided by the user,
945	>	modify the card provided by the user,
946		writing a new card into share dir
947		"""
948	<
948	>
949		def executableName(self):
950	<	if self.scriptExe: #CarlosDaniele
950	>	if self.scriptExe:
951		return "sh "
952		else:
953		return self.executable
954
955		def executableArgs(self):
956	+	# FUTURE: This function tests the CMSSW version. Can be simplified as we drop support for old versions
957		if self.scriptExe:#CarlosDaniele
958		return   self.scriptExe + " $NJob"
959	<	else:
960	<	return " -p pset.cfg"
959	>	else:
960	>	ex_args = ""
961	>	# FUTURE: This tests the CMSSW version. Can remove code as versions deprecated
962	>	# Framework job report
963	>	if (self.CMSSW_major >= 1 and self.CMSSW_minor >= 5) or (self.CMSSW_major >= 2):
964	>	ex_args += " -j $RUNTIME_AREA/crab_fjr_$NJob.xml"
965	>	# Type of config file
966	>	if self.CMSSW_major >= 2 :
967	>	ex_args += " -p pset.py"
968	>	else:
969	>	ex_args += " -p pset.cfg"
970	>	return ex_args
971
972		def inputSandbox(self, nj):
973		"""
974		Returns a list of filenames to be put in JDL input sandbox.
975		"""
976		inp_box = []
1003	–	# # dict added to delete duplicate from input sandbox file list
1004	–	# seen = {}
1005	–	## code
977		if os.path.isfile(self.tgzNameWithPath):
978		inp_box.append(self.tgzNameWithPath)
979	<	if os.path.isfile(self.MLtgzfile):
980	<	inp_box.append(self.MLtgzfile)
1010	<	## config
1011	<	if not self.pset is None:
1012	<	inp_box.append(common.work_space.pathForTgz() + 'job/' + self.configFilename())
1013	<	## additional input files
1014	<	for file in self.additional_inbox_files:
1015	<	inp_box.append(file)
979	>	wrapper = os.path.basename(str(common._db.queryTask('scriptName')))
980	>	inp_box.append(common.work_space.pathForTgz() +'job/'+ wrapper)
981		return inp_box
982
983		def outputSandbox(self, nj):
#	Line 1023 | Line 988 | class Cmssw(JobType):
988
989		## User Declared output files
990		for out in (self.output_file+self.output_file_sandbox):
991	<	n_out = nj + 1
991	>	n_out = nj + 1
992		out_box.append(self.numberFile_(out,str(n_out)))
993		return out_box
994
#	Line 1038 | Line 1003 | class Cmssw(JobType):
1003		Returns part of a job script which renames the produced files.
1004		"""
1005
1006	<	txt = '\n'
1007	<	txt += '# directory content\n'
1006	>	txt = '\n#Written by cms_cmssw::wsRenameOutput\n'
1007	>	txt += 'echo ">>> current directory (SOFTWARE_DIR): $SOFTWARE_DIR" \n'
1008	>	txt += 'echo ">>> current directory content:"\n'
1009		txt += 'ls \n'
1010	+	txt += '\n'
1011
1012	<	for fileWithSuffix in (self.output_file+self.output_file_sandbox):
1012	>	for fileWithSuffix in (self.output_file):
1013		output_file_num = self.numberFile_(fileWithSuffix, '$NJob')
1014		txt += '\n'
1015		txt += '# check output file\n'
1016	<	txt += 'ls '+fileWithSuffix+'\n'
1017	<	txt += 'ls_result=$?\n'
1018	<	txt += 'if [ $ls_result -ne 0 ] ; then\n'
1019	<	txt += '   echo "ERROR: Problem with output file"\n'
1020	<	if common.scheduler.boss_scheduler_name == 'condor_g':
1016	>	txt += 'if [ -e ./'+fileWithSuffix+' ] ; then\n'
1017	>	if (self.copy_data == 1):  # For OSG nodes, file is in $WORKING_DIR, should not be moved to $RUNTIME_AREA
1018	>	txt += '    mv '+fileWithSuffix+' '+output_file_num+'\n'
1019	>	txt += '    ln -s `pwd`/'+output_file_num+' $RUNTIME_AREA/'+fileWithSuffix+'\n'
1020	>	else:
1021	>	txt += '    mv '+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n'
1022	>	txt += '    ln -s $RUNTIME_AREA/'+output_file_num+' $RUNTIME_AREA/'+fileWithSuffix+'\n'
1023	>	txt += 'else\n'
1024	>	txt += '    job_exit_code=60302\n'
1025	>	txt += '    echo "WARNING: Output file '+fileWithSuffix+' not found"\n'
1026	>	if common.scheduler.name().upper() == 'CONDOR_G':
1027		txt += '    if [ $middleware == OSG ]; then \n'
1028		txt += '        echo "prepare dummy output file"\n'
1029		txt += '        echo "Processing of job output failed" > $RUNTIME_AREA/'+output_file_num+'\n'
1030		txt += '    fi \n'
1058	–	txt += 'else\n'
1059	–	txt += '   cp '+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n'
1031		txt += 'fi\n'
1032	<
1033	<	txt += 'cd $RUNTIME_AREA\n'
1034	<	txt += 'cd $RUNTIME_AREA\n'
1064	<	### OLI_DANIELE
1065	<	txt += 'if [ $middleware == OSG ]; then\n'
1066	<	txt += '    cd $RUNTIME_AREA\n'
1067	<	txt += '    echo "Remove working directory: $WORKING_DIR"\n'
1068	<	txt += '    /bin/rm -rf $WORKING_DIR\n'
1069	<	txt += '    if [ -d $WORKING_DIR ] ;then\n'
1070	<	txt += '        echo "SET_EXE 60999 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after cleanup of WN"\n'
1071	<	txt += '        echo "JOB_EXIT_STATUS = 60999"\n'
1072	<	txt += '        echo "JobExitCode=60999" | tee -a $RUNTIME_AREA/$repo\n'
1073	<	txt += '        dumpStatus $RUNTIME_AREA/$repo\n'
1074	<	txt += '        rm -f $RUNTIME_AREA/$repo \n'
1075	<	txt += '        echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1076	<	txt += '        echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1077	<	txt += '    fi\n'
1078	<	txt += 'fi\n'
1079	<	txt += '\n'
1080	<
1081	<	file_list = ''
1082	<	## Add to filelist only files to be possibly copied to SE
1083	<	for fileWithSuffix in self.output_file:
1084	<	output_file_num = self.numberFile_(fileWithSuffix, '$NJob')
1085	<	file_list=file_list+output_file_num+' '
1086	<	file_list=file_list[:-1]
1087	<	txt += 'file_list="'+file_list+'"\n'
1032	>	file_list = []
1033	>	for fileWithSuffix in (self.output_file):
1034	>	file_list.append(self.numberFile_(fileWithSuffix, '$NJob'))
1035
1036	+	txt += 'file_list="'+string.join(file_list,' ')+'"\n'
1037	+	txt += '\n'
1038	+	txt += 'echo ">>> current directory (SOFTWARE_DIR): $SOFTWARE_DIR" \n'
1039	+	txt += 'echo ">>> current directory content:"\n'
1040	+	txt += 'ls \n'
1041	+	txt += '\n'
1042	+	txt += 'cd $RUNTIME_AREA\n'
1043	+	txt += 'echo ">>> current directory (RUNTIME_AREA):  $RUNTIME_AREA"\n'
1044		return txt
1045
1046		def numberFile_(self, file, txt):
#	Line 1096 | Line 1051 | class Cmssw(JobType):
1051		# take away last extension
1052		name = p[0]
1053		for x in p[1:-1]:
1054	<	name=name+"."+x
1054	>	name=name+"."+x
1055		# add "_txt"
1056		if len(p)>1:
1057	<	ext = p[len(p)-1]
1058	<	result = name + '_' + txt + "." + ext
1057	>	ext = p[len(p)-1]
1058	>	result = name + '_' + txt + "." + ext
1059		else:
1060	<	result = name + '_' + txt
1061	<
1060	>	result = name + '_' + txt
1061	>
1062		return result
1063
1064		def getRequirements(self, nj=[]):
1065		"""
1066	<	return job requirements to add to jdl files
1066	>	return job requirements to add to jdl files
1067		"""
1068		req = ''
1069		if self.version:
1070		req='Member("VO-cms-' + \
1071		self.version + \
1072		'", other.GlueHostApplicationSoftwareRunTimeEnvironment)'
1073	+	if self.executable_arch:
1074	+	req+=' && Member("VO-cms-' + \
1075	+	self.executable_arch + \
1076	+	'", other.GlueHostApplicationSoftwareRunTimeEnvironment)'
1077
1078		req = req + ' && (other.GlueHostNetworkAdapterOutboundIP)'
1079	+	if common.scheduler.name() == "glitecoll":
1080	+	req += ' && other.GlueCEStateStatus == "Production" '
1081
1082		return req
1083
1084		def configFilename(self):
1085		""" return the config filename """
1086	<	return self.name()+'.cfg'
1086	>	# FUTURE: Can remove cfg mode for CMSSW >= 2_1_x
1087	>	if (self.CMSSW_major >= 2 and self.CMSSW_minor >= 1) or (self.CMSSW_major >= 3):
1088	>	return self.name()+'.py'
1089	>	else:
1090	>	return self.name()+'.cfg'
1091
1127	–	### OLI_DANIELE
1092		def wsSetupCMSOSGEnvironment_(self):
1093		"""
1094		Returns part of a job script which is prepares
1095		the execution environment and which is common for all CMS jobs.
1096		"""
1097	<	txt = '\n'
1098	<	txt += '   echo "### SETUP CMS OSG  ENVIRONMENT ###"\n'
1099	<	txt += '   if [ -f $GRID3_APP_DIR/cmssoft/cmsset_default.sh ] ;then\n'
1100	<	txt += '      # Use $GRID3_APP_DIR/cmssoft/cmsset_default.sh to setup cms software\n'
1101	<	txt += '       source $GRID3_APP_DIR/cmssoft/cmsset_default.sh '+self.version+'\n'
1102	<	txt += '   elif [ -f $OSG_APP/cmssoft/cms/cmsset_default.sh ] ;then\n'
1097	>	txt = '\n#Written by cms_cmssw::wsSetupCMSOSGEnvironment_\n'
1098	>	txt += '    echo ">>> setup CMS OSG environment:"\n'
1099	>	txt += '    echo "set SCRAM ARCH to ' + self.executable_arch + '"\n'
1100	>	txt += '    export SCRAM_ARCH='+self.executable_arch+'\n'
1101	>	txt += '    echo "SCRAM_ARCH = $SCRAM_ARCH"\n'
1102	>	txt += '    if [ -f $OSG_APP/cmssoft/cms/cmsset_default.sh ] ;then\n'
1103		txt += '      # Use $OSG_APP/cmssoft/cms/cmsset_default.sh to setup cms software\n'
1104	<	txt += '       source $OSG_APP/cmssoft/cms/cmsset_default.sh '+self.version+'\n'
1105	<	txt += '   else\n'
1106	<	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'
1107	<	txt += '       echo "JOB_EXIT_STATUS = 10020"\n'
1108	<	txt += '       echo "JobExitCode=10020" | tee -a $RUNTIME_AREA/$repo\n'
1109	<	txt += '       dumpStatus $RUNTIME_AREA/$repo\n'
1146	<	txt += '       rm -f $RUNTIME_AREA/$repo \n'
1147	<	txt += '       echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1148	<	txt += '       echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1149	<	txt += '       exit 1\n'
1150	<	txt += '\n'
1151	<	txt += '       echo "Remove working directory: $WORKING_DIR"\n'
1152	<	txt += '       cd $RUNTIME_AREA\n'
1153	<	txt += '       /bin/rm -rf $WORKING_DIR\n'
1154	<	txt += '       if [ -d $WORKING_DIR ] ;then\n'
1155	<	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'
1156	<	txt += '            echo "JOB_EXIT_STATUS = 10017"\n'
1157	<	txt += '            echo "JobExitCode=10017" | tee -a $RUNTIME_AREA/$repo\n'
1158	<	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
1159	<	txt += '            rm -f $RUNTIME_AREA/$repo \n'
1160	<	txt += '            echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1161	<	txt += '            echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1162	<	txt += '       fi\n'
1163	<	txt += '\n'
1164	<	txt += '       exit 1\n'
1165	<	txt += '   fi\n'
1104	>	txt += '        source $OSG_APP/cmssoft/cms/cmsset_default.sh '+self.version+'\n'
1105	>	txt += '    else\n'
1106	>	txt += '        echo "ERROR ==> $OSG_APP/cmssoft/cms/cmsset_default.sh file not found"\n'
1107	>	txt += '        job_exit_code=10020\n'
1108	>	txt += '        func_exit\n'
1109	>	txt += '    fi\n'
1110		txt += '\n'
1111	<	txt += '   echo "SET_CMS_ENV 0 ==> setup cms environment ok"\n'
1112	<	txt += '   echo " END SETUP CMS OSG  ENVIRONMENT "\n'
1111	>	txt += '    echo "==> setup cms environment ok"\n'
1112	>	txt += '    echo "SCRAM_ARCH = $SCRAM_ARCH"\n'
1113
1114		return txt
1115	<
1172	<	### OLI_DANIELE
1115	>
1116		def wsSetupCMSLCGEnvironment_(self):
1117		"""
1118		Returns part of a job script which is prepares
1119		the execution environment and which is common for all CMS jobs.
1120		"""
1121	<	txt  = '   \n'
1122	<	txt += '   echo " ### SETUP CMS LCG  ENVIRONMENT ### "\n'
1123	<	txt += '   if [ ! $VO_CMS_SW_DIR ] ;then\n'
1124	<	txt += '       echo "SET_CMS_ENV 10031 ==> ERROR CMS software dir not found on WN `hostname`"\n'
1125	<	txt += '       echo "JOB_EXIT_STATUS = 10031" \n'
1126	<	txt += '       echo "JobExitCode=10031" | tee -a $RUNTIME_AREA/$repo\n'
1127	<	txt += '       dumpStatus $RUNTIME_AREA/$repo\n'
1128	<	txt += '       rm -f $RUNTIME_AREA/$repo \n'
1129	<	txt += '       echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1130	<	txt += '       echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1131	<	txt += '       exit 1\n'
1132	<	txt += '   else\n'
1133	<	txt += '       echo "Sourcing environment... "\n'
1134	<	txt += '       if [ ! -s $VO_CMS_SW_DIR/cmsset_default.sh ] ;then\n'
1135	<	txt += '           echo "SET_CMS_ENV 10020 ==> ERROR cmsset_default.sh file not found into dir $VO_CMS_SW_DIR"\n'
1136	<	txt += '           echo "JOB_EXIT_STATUS = 10020"\n'
1137	<	txt += '           echo "JobExitCode=10020" | tee -a $RUNTIME_AREA/$repo\n'
1138	<	txt += '           dumpStatus $RUNTIME_AREA/$repo\n'
1139	<	txt += '           rm -f $RUNTIME_AREA/$repo \n'
1140	<	txt += '           echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1141	<	txt += '           echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1142	<	txt += '           exit 1\n'
1143	<	txt += '       fi\n'
1144	<	txt += '       echo "sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
1145	<	txt += '       source $VO_CMS_SW_DIR/cmsset_default.sh\n'
1146	<	txt += '       result=$?\n'
1147	<	txt += '       if [ $result -ne 0 ]; then\n'
1148	<	txt += '           echo "SET_CMS_ENV 10032 ==> ERROR problem sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
1149	<	txt += '           echo "JOB_EXIT_STATUS = 10032"\n'
1150	<	txt += '           echo "JobExitCode=10032" | tee -a $RUNTIME_AREA/$repo\n'
1151	<	txt += '           dumpStatus $RUNTIME_AREA/$repo\n'
1152	<	txt += '           rm -f $RUNTIME_AREA/$repo \n'
1153	<	txt += '           echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1154	<	txt += '           echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1155	<	txt += '           exit 1\n'
1156	<	txt += '       fi\n'
1157	<	txt += '   fi\n'
1158	<	txt += '   \n'
1159	<	txt += '   echo "SET_CMS_ENV 0 ==> setup cms environment ok"\n'
1160	<	txt += '   echo "### END SETUP CMS LCG ENVIRONMENT ###"\n'
1121	>	txt = '\n#Written by cms_cmssw::wsSetupCMSLCGEnvironment_\n'
1122	>	txt += '    echo ">>> setup CMS LCG environment:"\n'
1123	>	txt += '    echo "set SCRAM ARCH and BUILD_ARCH to ' + self.executable_arch + ' ###"\n'
1124	>	txt += '    export SCRAM_ARCH='+self.executable_arch+'\n'
1125	>	txt += '    export BUILD_ARCH='+self.executable_arch+'\n'
1126	>	txt += '    if [ ! $VO_CMS_SW_DIR ] ;then\n'
1127	>	txt += '        echo "ERROR ==> CMS software dir not found on WN `hostname`"\n'
1128	>	txt += '        job_exit_code=10031\n'
1129	>	txt += '        func_exit\n'
1130	>	txt += '    else\n'
1131	>	txt += '        echo "Sourcing environment... "\n'
1132	>	txt += '        if [ ! -s $VO_CMS_SW_DIR/cmsset_default.sh ] ;then\n'
1133	>	txt += '            echo "ERROR ==> cmsset_default.sh file not found into dir $VO_CMS_SW_DIR"\n'
1134	>	txt += '            job_exit_code=10020\n'
1135	>	txt += '            func_exit\n'
1136	>	txt += '        fi\n'
1137	>	txt += '        echo "sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
1138	>	txt += '        source $VO_CMS_SW_DIR/cmsset_default.sh\n'
1139	>	txt += '        result=$?\n'
1140	>	txt += '        if [ $result -ne 0 ]; then\n'
1141	>	txt += '            echo "ERROR ==> problem sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
1142	>	txt += '            job_exit_code=10032\n'
1143	>	txt += '            func_exit\n'
1144	>	txt += '        fi\n'
1145	>	txt += '    fi\n'
1146	>	txt += '    \n'
1147	>	txt += '    echo "==> setup cms environment ok"\n'
1148	>	return txt
1149	>
1150	>	def modifyReport(self, nj):
1151	>	"""
1152	>	insert the part of the script that modifies the FrameworkJob Report
1153	>	"""
1154	>	txt = '\n#Written by cms_cmssw::modifyReport\n'
1155	>	publish_data = int(self.cfg_params.get('USER.publish_data',0))
1156	>	if (publish_data == 1):
1157	>	processedDataset = self.cfg_params['USER.publish_data_name']
1158	>	LFNBaseName = LFNBase(processedDataset)
1159	>
1160	>	txt += 'if [ $copy_exit_status -eq 0 ]; then\n'
1161	>	txt += '    FOR_LFN=%s_${PSETHASH}/\n'%(LFNBaseName)
1162	>	txt += 'else\n'
1163	>	txt += '    FOR_LFN=/copy_problems/ \n'
1164	>	txt += '    SE=""\n'
1165	>	txt += '    SE_PATH=""\n'
1166	>	txt += 'fi\n'
1167	>
1168	>	txt += 'echo ">>> Modify Job Report:" \n'
1169	>	txt += 'chmod a+x $SOFTWARE_DIR/ProdCommon/ProdCommon/FwkJobRep/ModifyJobReport.py\n'
1170	>	txt += 'ProcessedDataset='+processedDataset+'\n'
1171	>	txt += 'echo "ProcessedDataset = $ProcessedDataset"\n'
1172	>	txt += 'echo "SE = $SE"\n'
1173	>	txt += 'echo "SE_PATH = $SE_PATH"\n'
1174	>	txt += 'echo "FOR_LFN = $FOR_LFN" \n'
1175	>	txt += 'echo "CMSSW_VERSION = $CMSSW_VERSION"\n\n'
1176	>	txt += 'echo "$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 $SE_PATH"\n'
1177	>	txt += '$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 $SE_PATH\n'
1178	>	txt += 'modifyReport_result=$?\n'
1179	>	txt += 'if [ $modifyReport_result -ne 0 ]; then\n'
1180	>	txt += '    modifyReport_result=70500\n'
1181	>	txt += '    job_exit_code=$modifyReport_result\n'
1182	>	txt += '    echo "ModifyReportResult=$modifyReport_result" | tee -a $RUNTIME_AREA/$repo\n'
1183	>	txt += '    echo "WARNING: Problem with ModifyJobReport"\n'
1184	>	txt += 'else\n'
1185	>	txt += '    mv NewFrameworkJobReport.xml $RUNTIME_AREA/crab_fjr_$NJob.xml\n'
1186	>	txt += 'fi\n'
1187	>	return txt
1188	>
1189	>	def wsParseFJR(self):
1190	>	"""
1191	>	Parse the FrameworkJobReport to obtain useful infos
1192	>	"""
1193	>	txt = '\n#Written by cms_cmssw::wsParseFJR\n'
1194	>	txt += 'echo ">>> Parse FrameworkJobReport crab_fjr.xml"\n'
1195	>	txt += 'if [ -s $RUNTIME_AREA/crab_fjr_$NJob.xml ]; then\n'
1196	>	txt += '    if [ -s $RUNTIME_AREA/parseCrabFjr.py ]; then\n'
1197	>	txt += '        cmd_out=`python $RUNTIME_AREA/parseCrabFjr.py --input $RUNTIME_AREA/crab_fjr_$NJob.xml --MonitorID $MonitorID --MonitorJobID $MonitorJobID`\n'
1198	>	txt += '        echo "Result of parsing the FrameworkJobReport crab_fjr.xml: $cmd_out"\n'
1199	>	txt += '        tmp_executable_exit_status=`echo $cmd_out | awk -F\; \'{print $1}\' | awk -F \' \' \'{print $NF}\'`\n'
1200	>	txt += '        if [ -n $tmp_executable_exit_status ];then\n'
1201	>	txt += '            executable_exit_status=$tmp_executable_exit_status\n'
1202	>	txt += '        fi\n'
1203	>	txt += '        if [ $executable_exit_status -eq 50115 ];then\n'
1204	>	txt += '            echo ">>> crab_fjr.xml contents: "\n'
1205	>	txt += '            cat $RUNTIME_AREA/crab_fjr_NJob.xml\n'
1206	>	txt += '            echo "Wrong FrameworkJobReport --> does not contain useful info. ExitStatus: $executable_exit_status"\n'
1207	>	txt += '        else\n'
1208	>	txt += '            echo "Extracted ExitStatus from FrameworkJobReport parsing output: $executable_exit_status"\n'
1209	>	txt += '        fi\n'
1210	>	txt += '    else\n'
1211	>	txt += '        echo "CRAB python script to parse CRAB FrameworkJobReport crab_fjr.xml is not available, using exit code of executable from command line."\n'
1212	>	txt += '    fi\n'
1213	>	#### Patch to check input data reading for CMSSW16x Hopefully we-ll remove it asap
1214	>
1215	>	if self.datasetPath:
1216	>	# VERIFY PROCESSED DATA
1217	>	txt += '    if [ $executable_exit_status -eq 0 ];then\n'
1218	>	txt += '      echo ">>> Verify list of processed files:"\n'
1219	>	txt += '      echo $InputFiles |tr -d "\\\\" |tr "," "\\n"|tr -d "\\"" > input-files.txt\n'
1220	>	txt += '      grep LFN $RUNTIME_AREA/crab_fjr_$NJob.xml |cut -d">" -f2|cut -d"<" -f1|grep "/" > processed-files.txt\n'
1221	>	txt += '      cat input-files.txt  | sort | uniq > tmp.txt\n'
1222	>	txt += '      mv tmp.txt input-files.txt\n'
1223	>	txt += '      echo "cat input-files.txt"\n'
1224	>	txt += '      echo "----------------------"\n'
1225	>	txt += '      cat input-files.txt\n'
1226	>	txt += '      cat processed-files.txt | sort | uniq > tmp.txt\n'
1227	>	txt += '      mv tmp.txt processed-files.txt\n'
1228	>	txt += '      echo "----------------------"\n'
1229	>	txt += '      echo "cat processed-files.txt"\n'
1230	>	txt += '      echo "----------------------"\n'
1231	>	txt += '      cat processed-files.txt\n'
1232	>	txt += '      echo "----------------------"\n'
1233	>	txt += '      diff -q input-files.txt processed-files.txt\n'
1234	>	txt += '      fileverify_status=$?\n'
1235	>	txt += '      if [ $fileverify_status -ne 0 ]; then\n'
1236	>	txt += '         executable_exit_status=30001\n'
1237	>	txt += '         echo "ERROR ==> not all input files processed"\n'
1238	>	txt += '         echo "      ==> list of processed files from crab_fjr.xml differs from list in pset.cfg"\n'
1239	>	txt += '         echo "      ==> diff input-files.txt processed-files.txt"\n'
1240	>	txt += '      fi\n'
1241	>	txt += '    fi\n'
1242	>	txt += '\n'
1243	>	txt += 'else\n'
1244	>	txt += '    echo "CRAB FrameworkJobReport crab_fjr.xml is not available, using exit code of executable from command line."\n'
1245	>	txt += 'fi\n'
1246	>	txt += '\n'
1247	>	txt += 'echo "ExeExitCode=$executable_exit_status" | tee -a $RUNTIME_AREA/$repo\n'
1248	>	txt += 'echo "EXECUTABLE_EXIT_STATUS = $executable_exit_status"\n'
1249	>	txt += 'job_exit_code=$executable_exit_status\n'
1250	>
1251		return txt
1252
1253		def setParam_(self, param, value):
#	Line 1223 | Line 1256 | class Cmssw(JobType):
1256		def getParams(self):
1257		return self._params
1258
1226	–	def setTaskid_(self):
1227	–	self._taskId = self.cfg_params['taskId']
1228	–
1229	–	def getTaskid(self):
1230	–	return self._taskId
1231	–
1232	–	#######################################################################
1259		def uniquelist(self, old):
1260		"""
1261		remove duplicates from a list
#	Line 1238 | Line 1264 | class Cmssw(JobType):
1264		for e in old:
1265		nd[e]=0
1266		return nd.keys()
1267	+
1268	+	def outList(self):
1269	+	"""
1270	+	check the dimension of the output files
1271	+	"""
1272	+	txt = ''
1273	+	txt += 'echo ">>> list of expected files on output sandbox"\n'
1274	+	listOutFiles = []
1275	+	stdout = 'CMSSW_$NJob.stdout'
1276	+	stderr = 'CMSSW_$NJob.stderr'
1277	+	if (self.return_data == 1):
1278	+	for file in (self.output_file+self.output_file_sandbox):
1279	+	listOutFiles.append(self.numberFile_(file, '$NJob'))
1280	+	listOutFiles.append(stdout)
1281	+	listOutFiles.append(stderr)
1282	+	else:
1283	+	for file in (self.output_file_sandbox):
1284	+	listOutFiles.append(self.numberFile_(file, '$NJob'))
1285	+	listOutFiles.append(stdout)
1286	+	listOutFiles.append(stderr)
1287	+	txt += 'echo "output files: '+string.join(listOutFiles,' ')+'"\n'
1288	+	txt += 'filesToCheck="'+string.join(listOutFiles,' ')+'"\n'
1289	+	txt += 'export filesToCheck\n'
1290	+	return txt

Diff Legend

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