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 1.7 by gutsche, Tue Jun 13 20:43:00 2006 UTC vs.
Revision 1.178 by spiga, Sun Apr 20 09:34:40 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	–
8	–	import DBSInfo_EDM
9	–	import DataDiscovery_EDM
10	–	import DataLocation_EDM
7		import Scram
8	+	from LFNBaseName import *
9
10	<	import os, string, re
10	>	import os, string, glob
11
12		class Cmssw(JobType):
13	<	def __init__(self, cfg_params):
13	>	def __init__(self, cfg_params, ncjobs):
14		JobType.__init__(self, 'CMSSW')
15		common.logger.debug(3,'CMSSW::__init__')
16
17	<	self.analisys_common_info = {}
18	<	# 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)
28	–	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.additional_tgz_name = 'additional.tgz'
38	+	self.scriptName = 'CMSSW.sh'
39	+	self.pset = ''      #scrip use case Da
40	+	self.datasetPath = '' #scrip use case Da
41
42	+	# set FJR file name
43	+	self.fjrFileName = 'crab_fjr.xml'
44
45		self.version = self.scram.getSWVersion()
36	–	self.setParam_('application', self.version)
37	–	common.analisys_common_info['sw_version'] = self.version
38	–	### FEDE
39	–	common.analisys_common_info['copy_input_data'] = 0
40	–	common.analisys_common_info['events_management'] = 1
46
47	<	### collect Data cards
48	<	try:
49	<	#   self.owner = cfg_params['CMSSW.owner']
50	<	#   log.debug(6, "CMSSW::CMSSW(): owner = "+self.owner)
51	<	#   self.dataset = cfg_params['CMSSW.dataset']
52	<	self.datasetPath = cfg_params['CMSSW.datasetpath']
53	<	log.debug(6, "CMSSW::CMSSW(): datasetPath = "+self.datasetPath)
54	<	except KeyError:
55	<	#    msg = "Error: owner and/or dataset not defined "
56	<	msg = "Error: datasetpath not defined "
57	<	raise CrabException(msg)
58	<
59	<	# ML monitoring
55	<	# split dataset path style: /PreProdR3Minbias/SIM/GEN-SIM
56	<	datasetpath_split = self.datasetPath.split("/")
57	<	self.setParam_('dataset', datasetpath_split[1])
58	<	self.setParam_('owner', datasetpath_split[-1])
47	>	#
48	>	# Try to block creation in case of arch/version mismatch
49	>	#
50	>
51	>	#        a = string.split(self.version, "_")
52	>	#
53	>	#        if int(a[1]) == 1 and (int(a[2]) < 5 and self.executable_arch.find('slc4') == 0):
54	>	#            msg = "Warning: You are using %s version of CMSSW  with %s architecture. \n--> Did you compile your libraries with SLC3? Otherwise you can find some problems running on SLC4 Grid nodes.\n"%(self.version, self.executable_arch)
55	>	#            common.logger.message(msg)
56	>	#        if int(a[1]) == 1 and (int(a[2]) >= 5 and self.executable_arch.find('slc3') == 0):
57	>	#            msg = "Error: CMS does not support %s with %s architecture"%(self.version, self.executable_arch)
58	>	#            raise CrabException(msg)
59	>	#
60
61
62	+	### collect Data cards
63
64	+	if not cfg_params.has_key('CMSSW.datasetpath'):
65	+	msg = "Error: datasetpath not defined "
66	+	raise CrabException(msg)
67	+	tmp =  cfg_params['CMSSW.datasetpath']
68	+	log.debug(6, "CMSSW::CMSSW(): datasetPath = "+tmp)
69	+	if string.lower(tmp)=='none':
70	+	self.datasetPath = None
71	+	self.selectNoInput = 1
72	+	else:
73	+	self.datasetPath = tmp
74	+	self.selectNoInput = 0
75
76		self.dataTiers = []
64	–	#       try:
65	–	#           tmpDataTiers = string.split(cfg_params['CMSSW.data_tier'],',')
66	–	#           for tmp in tmpDataTiers:
67	–	#               tmp=string.strip(tmp)
68	–	#               self.dataTiers.append(tmp)
69	–	#               pass
70	–	#           pass
71	–	#       except KeyError:
72	–	#           pass
73	–	#       log.debug(6, "Cmssw::Cmssw(): dataTiers = "+str(self.dataTiers))
77
78		## now the application
79	<	try:
80	<	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
79	>	self.executable = cfg_params.get('CMSSW.executable','cmsRun')
80	>	log.debug(6, "CMSSW::CMSSW(): executable = "+self.executable)
81
82	<	try:
83	<	self.pset = cfg_params['CMSSW.pset']
84	<	log.debug(6, "Cmssw::Cmssw(): PSet file = "+self.pset)
82	>	if not cfg_params.has_key('CMSSW.pset'):
83	>	raise CrabException("PSet file missing. Cannot run cmsRun ")
84	>	self.pset = cfg_params['CMSSW.pset']
85	>	log.debug(6, "Cmssw::Cmssw(): PSet file = "+self.pset)
86	>	if self.pset.lower() != 'none' :
87		if (not os.path.exists(self.pset)):
88		raise CrabException("User defined PSet file "+self.pset+" does not exist")
89	<	except KeyError:
90	<	raise CrabException("PSet file missing. Cannot run cmsRun ")
89	>	else:
90	>	self.pset = None
91
92		# output files
93	<	try:
94	<	self.output_file = []
93	>	## stuff which must be returned always via sandbox
94	>	self.output_file_sandbox = []
95
96	<	tmp = cfg_params['CMSSW.output_file']
97	<	if tmp != '':
98	<	tmpOutFiles = string.split(cfg_params['CMSSW.output_file'],',')
99	<	log.debug(7, 'cmssw::cmssw(): output files '+str(tmpOutFiles))
100	<	for tmp in tmpOutFiles:
101	<	tmp=string.strip(tmp)
102	<	self.output_file.append(tmp)
103	<	pass
104	<	else:
105	<	log.message("No output file defined: only stdout/err will be available")
96	>	# add fjr report by default via sandbox
97	>	self.output_file_sandbox.append(self.fjrFileName)
98	>
99	>	# other output files to be returned via sandbox or copied to SE
100	>	self.output_file = []
101	>	tmp = cfg_params.get('CMSSW.output_file',None)
102	>	if tmp :
103	>	tmpOutFiles = string.split(tmp,',')
104	>	log.debug(7, 'cmssw::cmssw(): output files '+str(tmpOutFiles))
105	>	for tmp in tmpOutFiles:
106	>	tmp=string.strip(tmp)
107	>	self.output_file.append(tmp)
108		pass
109	<	pass
110	<	except KeyError:
111	<	log.message("No output file defined: only stdout/err will be available")
115	<	pass
109	>	else:
110	>	log.message("No output file defined: only stdout/err and the CRAB Framework Job Report will be available\n")
111	>	pass
112
113		# script_exe file as additional file in inputSandbox
114	<	try:
115	<	self.scriptExe = cfg_params['USER.script_exe']
116	<	self.additional_inbox_files.append(self.scriptExe)
117	<	except KeyError:
118	<	pass
119	<	if self.scriptExe != '':
120	<	if os.path.isfile(self.scriptExe):
121	<	pass
122	<	else:
123	<	log.message("WARNING. file "+self.scriptExe+" not found")
124	<	sys.exit()
125	<
114	>	self.scriptExe = cfg_params.get('USER.script_exe',None)
115	>	if self.scriptExe :
116	>	if not os.path.isfile(self.scriptExe):
117	>	msg ="ERROR. file "+self.scriptExe+" not found"
118	>	raise CrabException(msg)
119	>	self.additional_inbox_files.append(string.strip(self.scriptExe))
120	>
121	>	#CarlosDaniele
122	>	if self.datasetPath == None and self.pset == None and self.scriptExe == '' :
123	>	msg ="Error. script_exe  not defined"
124	>	raise CrabException(msg)
125	>
126		## additional input files
127	<	try:
128	<	tmpAddFiles = string.split(cfg_params['CMSSW.additional_input_files'],',')
127	>	if cfg_params.has_key('USER.additional_input_files'):
128	>	tmpAddFiles = string.split(cfg_params['USER.additional_input_files'],',')
129		for tmp in tmpAddFiles:
130	<	if not os.path.exists(tmp):
131	<	raise CrabException("Additional input file not found: "+tmp)
132	<	tmp=string.strip(tmp)
133	<	self.additional_inbox_files.append(tmp)
130	>	tmp = string.strip(tmp)
131	>	dirname = ''
132	>	if not tmp[0]=="/": dirname = "."
133	>	files = []
134	>	if string.find(tmp,"*")>-1:
135	>	files = glob.glob(os.path.join(dirname, tmp))
136	>	if len(files)==0:
137	>	raise CrabException("No additional input file found with this pattern: "+tmp)
138	>	else:
139	>	files.append(tmp)
140	>	for file in files:
141	>	if not os.path.exists(file):
142	>	raise CrabException("Additional input file not found: "+file)
143	>	pass
144	>	# fname = string.split(file, '/')[-1]
145	>	# storedFile = common.work_space.pathForTgz()+'share/'+fname
146	>	# shutil.copyfile(file, storedFile)
147	>	self.additional_inbox_files.append(string.strip(file))
148		pass
149		pass
150	<	except KeyError:
151	<	pass
142	<
143	<	try:
144	<	self.filesPerJob = int(cfg_params['CMSSW.files_per_jobs']) #Daniele
145	<	except KeyError:
146	<	self.filesPerJob = 1
147	<
148	<	## Max event   will be total_number_of_events ???  Daniele
149	<	try:
150	<	self.maxEv = cfg_params['CMSSW.event_per_job']
151	<	except KeyError:
152	<	self.maxEv = "-1"
153	<	##
154	<	try:
155	<	self.total_number_of_events = int(cfg_params['CMSSW.total_number_of_events'])
156	<	except KeyError:
157	<	msg = 'Must define total_number_of_events'
158	<	raise CrabException(msg)
159	<
160	<	CEBlackList = []
161	<	try:
162	<	tmpBad = string.split(cfg_params['EDG.ce_black_list'],',')
163	<	for tmp in tmpBad:
164	<	tmp=string.strip(tmp)
165	<	CEBlackList.append(tmp)
166	<	except KeyError:
167	<	pass
150	>	common.logger.debug(5,"Additional input files: "+str(self.additional_inbox_files))
151	>	pass
152
153	<	self.reCEBlackList=[]
154	<	for bad in CEBlackList:
155	<	self.reCEBlackList.append(re.compile( bad ))
153	>	## Events per job
154	>	if cfg_params.has_key('CMSSW.events_per_job'):
155	>	self.eventsPerJob =int( cfg_params['CMSSW.events_per_job'])
156	>	self.selectEventsPerJob = 1
157	>	else:
158	>	self.eventsPerJob = -1
159	>	self.selectEventsPerJob = 0
160
161	<	common.logger.debug(5,'CEBlackList: '+str(CEBlackList))
161	>	## number of jobs
162	>	if cfg_params.has_key('CMSSW.number_of_jobs'):
163	>	self.theNumberOfJobs =int( cfg_params['CMSSW.number_of_jobs'])
164	>	self.selectNumberOfJobs = 1
165	>	else:
166	>	self.theNumberOfJobs = 0
167	>	self.selectNumberOfJobs = 0
168
169	<	CEWhiteList = []
170	<	try:
171	<	tmpGood = string.split(cfg_params['EDG.ce_white_list'],',')
172	<	for tmp in tmpGood:
173	<	tmp=string.strip(tmp)
174	<	CEWhiteList.append(tmp)
181	<	except KeyError:
182	<	pass
169	>	if cfg_params.has_key('CMSSW.total_number_of_events'):
170	>	self.total_number_of_events = int(cfg_params['CMSSW.total_number_of_events'])
171	>	self.selectTotalNumberEvents = 1
172	>	else:
173	>	self.total_number_of_events = 0
174	>	self.selectTotalNumberEvents = 0
175
176	<	#print 'CEWhiteList: ',CEWhiteList
177	<	self.reCEWhiteList=[]
178	<	for Good in CEWhiteList:
179	<	self.reCEWhiteList.append(re.compile( Good ))
176	>	if self.pset != None: #CarlosDaniele
177	>	if ( (self.selectTotalNumberEvents + self.selectEventsPerJob + self.selectNumberOfJobs) != 2 ):
178	>	msg = 'Must define exactly two of total_number_of_events, events_per_job, or number_of_jobs.'
179	>	raise CrabException(msg)
180	>	else:
181	>	if (self.selectNumberOfJobs == 0):
182	>	msg = 'Must specify  number_of_jobs.'
183	>	raise CrabException(msg)
184	>
185	>	## New method of dealing with seeds
186	>	self.incrementSeeds = []
187	>	self.preserveSeeds = []
188	>	if cfg_params.has_key('CMSSW.preserve_seeds'):
189	>	tmpList = cfg_params['CMSSW.preserve_seeds'].split(',')
190	>	for tmp in tmpList:
191	>	tmp.strip()
192	>	self.preserveSeeds.append(tmp)
193	>	if cfg_params.has_key('CMSSW.increment_seeds'):
194	>	tmpList = cfg_params['CMSSW.increment_seeds'].split(',')
195	>	for tmp in tmpList:
196	>	tmp.strip()
197	>	self.incrementSeeds.append(tmp)
198	>
199	>	## Old method of dealing with seeds
200	>	## FUTURE: This is for old CMSSW and old CRAB. Can throw exceptions after a couple of CRAB releases and then
201	>	## remove
202	>	self.sourceSeed = cfg_params.get('CMSSW.pythia_seed',None)
203	>	if self.sourceSeed:
204	>	print "pythia_seed is a deprecated parameter. Use preserve_seeds or increment_seeds in the future.\n","Added to increment_seeds."
205	>	self.incrementSeeds.append('sourceSeed')
206	>
207	>	self.sourceSeedVtx = cfg_params.get('CMSSW.vtx_seed',None)
208	>	if self.sourceSeedVtx:
209	>	print "vtx_seed is a deprecated parameter. Use preserve_seeds or increment_seeds in the future.\n","Added to increment_seeds."
210	>	self.incrementSeeds.append('VtxSmeared')
211	>
212	>	self.sourceSeedG4 = cfg_params.get('CMSSW.g4_seed',None)
213	>	if self.sourceSeedG4:
214	>	print "g4_seed is a deprecated parameter. Use preserve_seeds or increment_seeds in the future.\n","Added to increment_seeds."
215	>	self.incrementSeeds.append('g4SimHits')
216	>
217	>	self.sourceSeedMix = cfg_params.get('CMSSW.mix_seed',None)
218	>	if self.sourceSeedMix:
219	>	print "mix_seed is a deprecated parameter. Use preserve_seeds or increment_seeds in the future.\n","Added to increment_seeds."
220	>	self.incrementSeeds.append('mix')
221	>
222	>	self.firstRun = cfg_params.get('CMSSW.first_run',None)
223	>
224	>	if self.pset != None: #CarlosDaniele
225	>	import PsetManipulator as pp
226	>	PsetEdit = pp.PsetManipulator(self.pset) #Daniele Pset
227
228	<	common.logger.debug(5,'CEWhiteList: '+str(CEWhiteList))
228	>	# Copy/return
229
230	<	self.PsetEdit = PsetManipulator.PsetManipulator(self.pset) #Daniele Pset
230	>	self.copy_data = int(cfg_params.get('USER.copy_data',0))
231	>	self.return_data = int(cfg_params.get('USER.return_data',0))
232
233		#DBSDLS-start
234	<	## Initialize the variables that are extracted from DBS/DLS and needed in other places of the code
234	>	## Initialize the variables that are extracted from DBS/DLS and needed in other places of the code
235		self.maxEvents=0  # max events available   ( --> check the requested nb. of evts in Creator.py)
236		self.DBSPaths={}  # all dbs paths requested ( --> input to the site local discovery script)
237	+	self.jobDestination=[]  # Site destination(s) for each job (list of lists)
238		## Perform the data location and discovery (based on DBS/DLS)
239	<	self.DataDiscoveryAndLocation(cfg_params)
240	<	#DBSDLS-end
239	>	## SL: Don't if NONE is specified as input (pythia use case)
240	>	blockSites = {}
241	>	if self.datasetPath:
242	>	blockSites = self.DataDiscoveryAndLocation(cfg_params)
243	>	#DBSDLS-end
244
245		self.tgzNameWithPath = self.getTarBall(self.executable)
246
247	<	self.jobSplitting()  #Daniele job Splitting
248	<	self.PsetEdit.maxEvent(self.maxEv) #Daniele
249	<	self.PsetEdit.inputModule("INPUT") #Daniele
250	<	self.PsetEdit.psetWriter(self.configFilename())
251	<
247	>	## Select Splitting
248	>	if self.selectNoInput:
249	>	if self.pset == None: #CarlosDaniele
250	>	self.jobSplittingForScript()
251	>	else:
252	>	self.jobSplittingNoInput()
253	>	else:
254	>	self.jobSplittingByBlocks(blockSites)
255
256	+	# modify Pset
257	+	if self.pset != None: #CarlosDaniele
258	+	try:
259	+	# Add FrameworkJobReport to parameter-set, set max events.
260	+	# Reset later for data jobs by writeCFG which does all modifications
261	+	PsetEdit.addCrabFJR(self.fjrFileName)
262	+	PsetEdit.maxEvent(self.eventsPerJob)
263	+	PsetEdit.psetWriter(self.configFilename())
264	+	except:
265	+	msg='Error while manipuliating ParameterSet: exiting...'
266	+	raise CrabException(msg)
267
268		def DataDiscoveryAndLocation(self, cfg_params):
269
270	+	import DataDiscovery
271	+	import DataLocation
272		common.logger.debug(10,"CMSSW::DataDiscoveryAndLocation()")
273
214	–	#datasetPath = "/"+self.owner+"/"+self.dataTiers[0]+"/"+self.dataset
215	–
274		datasetPath=self.datasetPath
275
218	–	## TODO
219	–	dataTiersList = ""
220	–	dataTiers = dataTiersList.split(',')
221	–
276		## Contact the DBS
277	+	common.logger.message("Contacting Data Discovery Services ...")
278		try:
279	<	self.pubdata=DataDiscovery_EDM.DataDiscovery_EDM(datasetPath, dataTiers, cfg_params)
279	>	self.pubdata=DataDiscovery.DataDiscovery(datasetPath, cfg_params)
280		self.pubdata.fetchDBSInfo()
281
282	<	except DataDiscovery_EDM.NotExistingDatasetError, ex :
282	>	except DataDiscovery.NotExistingDatasetError, ex :
283		msg = 'ERROR ***: failed Data Discovery in DBS : %s'%ex.getErrorMessage()
284		raise CrabException(msg)
285	<
231	<	except DataDiscovery_EDM.NoDataTierinProvenanceError, ex :
285	>	except DataDiscovery.NoDataTierinProvenanceError, ex :
286		msg = 'ERROR ***: failed Data Discovery in DBS : %s'%ex.getErrorMessage()
287		raise CrabException(msg)
288	<	except DataDiscovery_EDM.DataDiscoveryError, ex:
289	<	msg = 'ERROR ***: failed Data Discovery in DBS  %s'%ex.getErrorMessage()
288	>	except DataDiscovery.DataDiscoveryError, ex:
289	>	msg = 'ERROR ***: failed Data Discovery in DBS :  %s'%ex.getErrorMessage()
290		raise CrabException(msg)
291
292	<	## get list of all required data in the form of dbs paths  (dbs path = /dataset/datatier/owner)
293	<	## self.DBSPaths=self.pubdata.getDBSPaths()
294	<	common.logger.message("Required data are :"+self.datasetPath)
241	<
242	<	filesbyblock=self.pubdata.getFiles()
243	<	self.AllInputFiles=filesbyblock.values()
244	<	self.files = self.AllInputFiles
245	<
246	<	## TEMP
247	<	#    self.filesTmp = filesbyblock.values()
248	<	#    self.files = []
249	<	#    locPath='rfio:cmsbose2.bo.infn.it:/flatfiles/SE00/cms/fanfani/ProdTest/'
250	<	#    locPath=''
251	<	#    tmp = []
252	<	#    for file in self.filesTmp[0]:
253	<	#        tmp.append(locPath+file)
254	<	#    self.files.append(tmp)
255	<	## END TEMP
292	>	self.filesbyblock=self.pubdata.getFiles()
293	>	self.eventsbyblock=self.pubdata.getEventsPerBlock()
294	>	self.eventsbyfile=self.pubdata.getEventsPerFile()
295
296		## get max number of events
297	<	#common.logger.debug(10,"number of events for primary fileblocks %i"%self.pubdata.getMaxEvents())
259	<	self.maxEvents=self.pubdata.getMaxEvents() ##  self.maxEvents used in Creator.py
260	<	common.logger.message("\nThe number of available events is %s"%self.maxEvents)
297	>	self.maxEvents=self.pubdata.getMaxEvents() ##  self.maxEvents used in Creator.py
298
299		## Contact the DLS and build a list of sites hosting the fileblocks
300		try:
301	<	dataloc=DataLocation_EDM.DataLocation_EDM(filesbyblock.keys(),cfg_params)
301	>	dataloc=DataLocation.DataLocation(self.filesbyblock.keys(),cfg_params)
302		dataloc.fetchDLSInfo()
303	<	except DataLocation_EDM.DataLocationError , ex:
303	>	except DataLocation.DataLocationError , ex:
304		msg = 'ERROR ***: failed Data Location in DLS \n %s '%ex.getErrorMessage()
305		raise CrabException(msg)
269	–
270	–	allsites=dataloc.getSites()
271	–	common.logger.debug(5,"sites are %s"%allsites)
272	–	sites=self.checkBlackList(allsites)
273	–	common.logger.debug(5,"sites are (after black list) %s"%sites)
274	–	sites=self.checkWhiteList(sites)
275	–	common.logger.debug(5,"sites are (after white list) %s"%sites)
306
277	–	if len(sites)==0:
278	–	msg = 'No sites hosting all the needed data! Exiting... '
279	–	raise CrabException(msg)
307
308	<	common.logger.message("List of Sites hosting the data : "+str(sites))
309	<	common.logger.debug(6, "List of Sites: "+str(sites))
310	<	common.analisys_common_info['sites']=sites    ## used in SchedulerEdg.py in createSchScript
311	<	self.setParam_('TargetCE', ','.join(sites))
308	>	sites = dataloc.getSites()
309	>	allSites = []
310	>	listSites = sites.values()
311	>	for listSite in listSites:
312	>	for oneSite in listSite:
313	>	allSites.append(oneSite)
314	>	allSites = self.uniquelist(allSites)
315	>
316	>	# screen output
317	>	common.logger.message("Requested dataset: " + datasetPath + " has " + str(self.maxEvents) + " events in " + str(len(self.filesbyblock.keys())) + " blocks.\n")
318	>
319	>	return sites
320	>
321	>	# to Be Removed  DS -- BL
322	>	#  def setArgsList(self, argsList):
323	>	#      self.argsList = argsList
324	>
325	>	def jobSplittingByBlocks(self, blockSites):
326	>	"""
327	>	Perform job splitting. Jobs run over an integer number of files
328	>	and no more than one block.
329	>	ARGUMENT: blockSites: dictionary with blocks as keys and list of host sites as values
330	>	REQUIRES: self.selectTotalNumberEvents, self.selectEventsPerJob, self.selectNumberofJobs,
331	>	self.total_number_of_events, self.eventsPerJob, self.theNumberOfJobs,
332	>	self.maxEvents, self.filesbyblock
333	>	SETS: self.jobDestination - Site destination(s) for each job (a list of lists)
334	>	self.total_number_of_jobs - Total # of jobs
335	>	self.list_of_args - File(s) job will run on (a list of lists)
336	>	"""
337	>
338	>	# ---- Handle the possible job splitting configurations ---- #
339	>	if (self.selectTotalNumberEvents):
340	>	totalEventsRequested = self.total_number_of_events
341	>	if (self.selectEventsPerJob):
342	>	eventsPerJobRequested = self.eventsPerJob
343	>	if (self.selectNumberOfJobs):
344	>	totalEventsRequested = self.theNumberOfJobs * self.eventsPerJob
345	>
346	>	# If user requested all the events in the dataset
347	>	if (totalEventsRequested == -1):
348	>	eventsRemaining=self.maxEvents
349	>	# If user requested more events than are in the dataset
350	>	elif (totalEventsRequested > self.maxEvents):
351	>	eventsRemaining = self.maxEvents
352	>	common.logger.message("Requested "+str(self.total_number_of_events)+ " events, but only "+str(self.maxEvents)+" events are available.")
353	>	# If user requested less events than are in the dataset
354	>	else:
355	>	eventsRemaining = totalEventsRequested
356	>
357	>	# If user requested more events per job than are in the dataset
358	>	if (self.selectEventsPerJob and eventsPerJobRequested > self.maxEvents):
359	>	eventsPerJobRequested = self.maxEvents
360	>
361	>	# For user info at end
362	>	totalEventCount = 0
363	>
364	>	if (self.selectTotalNumberEvents and self.selectNumberOfJobs):
365	>	eventsPerJobRequested = int(eventsRemaining/self.theNumberOfJobs)
366	>
367	>	if (self.selectNumberOfJobs):
368	>	common.logger.message("May not create the exact number_of_jobs requested.")
369	>
370	>	if ( self.ncjobs == 'all' ) :
371	>	totalNumberOfJobs = 999999999
372	>	else :
373	>	totalNumberOfJobs = self.ncjobs
374	>
375	>	blocks = blockSites.keys()
376	>	blockCount = 0
377	>	# Backup variable in case self.maxEvents counted events in a non-included block
378	>	numBlocksInDataset = len(blocks)
379	>
380	>	jobCount = 0
381	>	list_of_lists = []
382	>
383	>	# list tracking which jobs are in which jobs belong to which block
384	>	jobsOfBlock = {}
385	>
386	>	# ---- Iterate over the blocks in the dataset until ---- #
387	>	# ---- we've met the requested total # of events    ---- #
388	>	while ( (eventsRemaining > 0) and (blockCount < numBlocksInDataset) and (jobCount < totalNumberOfJobs)):
389	>	block = blocks[blockCount]
390	>	blockCount += 1
391	>	if block not in jobsOfBlock.keys() :
392	>	jobsOfBlock[block] = []
393	>
394	>	if self.eventsbyblock.has_key(block) :
395	>	numEventsInBlock = self.eventsbyblock[block]
396	>	common.logger.debug(5,'Events in Block File '+str(numEventsInBlock))
397	>
398	>	files = self.filesbyblock[block]
399	>	numFilesInBlock = len(files)
400	>	if (numFilesInBlock <= 0):
401	>	continue
402	>	fileCount = 0
403	>
404	>	# ---- New block => New job ---- #
405	>	parString = ""
406	>	# counter for number of events in files currently worked on
407	>	filesEventCount = 0
408	>	# flag if next while loop should touch new file
409	>	newFile = 1
410	>	# job event counter
411	>	jobSkipEventCount = 0
412	>
413	>	# ---- Iterate over the files in the block until we've met the requested ---- #
414	>	# ---- total # of events or we've gone over all the files in this block  ---- #
415	>	while ( (eventsRemaining > 0) and (fileCount < numFilesInBlock) and (jobCount < totalNumberOfJobs) ):
416	>	file = files[fileCount]
417	>	if newFile :
418	>	try:
419	>	numEventsInFile = self.eventsbyfile[file]
420	>	common.logger.debug(6, "File "+str(file)+" has "+str(numEventsInFile)+" events")
421	>	# increase filesEventCount
422	>	filesEventCount += numEventsInFile
423	>	# Add file to current job
424	>	parString += '\\\"' + file + '\\\"\,'
425	>	newFile = 0
426	>	except KeyError:
427	>	common.logger.message("File "+str(file)+" has unknown number of events: skipping")
428	>
429	>	eventsPerJobRequested = min(eventsPerJobRequested, eventsRemaining)
430	>	# if less events in file remain than eventsPerJobRequested
431	>	if ( filesEventCount - jobSkipEventCount < eventsPerJobRequested):
432	>	# if last file in block
433	>	if ( fileCount == numFilesInBlock-1 ) :
434	>	# end job using last file, use remaining events in block
435	>	# close job and touch new file
436	>	fullString = parString[:-2]
437	>	list_of_lists.append([fullString,str(-1),str(jobSkipEventCount)])
438	>	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(filesEventCount - jobSkipEventCount)+" events (last file in block).")
439	>	self.jobDestination.append(blockSites[block])
440	>	common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
441	>	# fill jobs of block dictionary
442	>	jobsOfBlock[block].append(jobCount+1)
443	>	# reset counter
444	>	jobCount = jobCount + 1
445	>	totalEventCount = totalEventCount + filesEventCount - jobSkipEventCount
446	>	eventsRemaining = eventsRemaining - filesEventCount + jobSkipEventCount
447	>	jobSkipEventCount = 0
448	>	# reset file
449	>	parString = ""
450	>	filesEventCount = 0
451	>	newFile = 1
452	>	fileCount += 1
453	>	else :
454	>	# go to next file
455	>	newFile = 1
456	>	fileCount += 1
457	>	# if events in file equal to eventsPerJobRequested
458	>	elif ( filesEventCount - jobSkipEventCount == eventsPerJobRequested ) :
459	>	# close job and touch new file
460	>	fullString = parString[:-2]
461	>	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
462	>	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.")
463	>	self.jobDestination.append(blockSites[block])
464	>	common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
465	>	jobsOfBlock[block].append(jobCount+1)
466	>	# reset counter
467	>	jobCount = jobCount + 1
468	>	totalEventCount = totalEventCount + eventsPerJobRequested
469	>	eventsRemaining = eventsRemaining - eventsPerJobRequested
470	>	jobSkipEventCount = 0
471	>	# reset file
472	>	parString = ""
473	>	filesEventCount = 0
474	>	newFile = 1
475	>	fileCount += 1
476	>
477	>	# if more events in file remain than eventsPerJobRequested
478	>	else :
479	>	# close job but don't touch new file
480	>	fullString = parString[:-2]
481	>	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
482	>	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.")
483	>	self.jobDestination.append(blockSites[block])
484	>	common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
485	>	jobsOfBlock[block].append(jobCount+1)
486	>	# increase counter
487	>	jobCount = jobCount + 1
488	>	totalEventCount = totalEventCount + eventsPerJobRequested
489	>	eventsRemaining = eventsRemaining - eventsPerJobRequested
490	>	# calculate skip events for last file
491	>	# use filesEventCount (contains several files), jobSkipEventCount and eventsPerJobRequest
492	>	jobSkipEventCount = eventsPerJobRequested - (filesEventCount - jobSkipEventCount - self.eventsbyfile[file])
493	>	# remove all but the last file
494	>	filesEventCount = self.eventsbyfile[file]
495	>	parString = '\\\"' + file + '\\\"\,'
496	>	pass # END if
497	>	pass # END while (iterate over files in the block)
498	>	pass # END while (iterate over blocks in the dataset)
499	>	self.ncjobs = self.total_number_of_jobs = jobCount
500	>	if (eventsRemaining > 0 and jobCount < totalNumberOfJobs ):
501	>	common.logger.message("Could not run on all requested events because some blocks not hosted at allowed sites.")
502	>	common.logger.message(str(jobCount)+" job(s) can run on "+str(totalEventCount)+" events.\n")
503	>
504	>	# screen output
505	>	screenOutput = "List of jobs and available destination sites:\n\n"
506	>
507	>	# keep trace of block with no sites to print a warning at the end
508	>	noSiteBlock = []
509	>	bloskNoSite = []
510	>
511	>	blockCounter = 0
512	>	for block in blocks:
513	>	if block in jobsOfBlock.keys() :
514	>	blockCounter += 1
515	>	screenOutput += "Block %5i: jobs %20s: sites: %s\n" % (blockCounter,spanRanges(jobsOfBlock[block]),
516	>	','.join(self.blackWhiteListParser.checkWhiteList(self.blackWhiteListParser.checkBlackList(blockSites[block],block),block)))
517	>	if len(self.blackWhiteListParser.checkWhiteList(self.blackWhiteListParser.checkBlackList(blockSites[block],block),block)) == 0:
518	>	noSiteBlock.append( spanRanges(jobsOfBlock[block]) )
519	>	bloskNoSite.append( blockCounter )
520	>
521	>	common.logger.message(screenOutput)
522	>	if len(noSiteBlock) > 0 and len(bloskNoSite) > 0:
523	>	msg = 'WARNING: No sites are hosting any part of data for block:\n                '
524	>	virgola = ""
525	>	if len(bloskNoSite) > 1:
526	>	virgola = ","
527	>	for block in bloskNoSite:
528	>	msg += ' ' + str(block) + virgola
529	>	msg += '\n               Related jobs:\n                 '
530	>	virgola = ""
531	>	if len(noSiteBlock) > 1:
532	>	virgola = ","
533	>	for range_jobs in noSiteBlock:
534	>	msg += str(range_jobs) + virgola
535	>	msg += '\n               will not be submitted and this block of data can not be analyzed!\n'
536	>	if self.cfg_params.has_key('EDG.se_white_list'):
537	>	msg += 'WARNING: SE White List: '+self.cfg_params['EDG.se_white_list']+'\n'
538	>	msg += '(Hint: By whitelisting you force the job to run at this particular site(s).\n'
539	>	msg += 'Please check if the dataset is available at this site!)\n'
540	>	if self.cfg_params.has_key('EDG.ce_white_list'):
541	>	msg += 'WARNING: CE White List: '+self.cfg_params['EDG.ce_white_list']+'\n'
542	>	msg += '(Hint: By whitelisting you force the job to run at this particular site(s).\n'
543	>	msg += 'Please check if the dataset is available at this site!)\n'
544	>
545	>	common.logger.message(msg)
546	>
547	>	self.list_of_args = list_of_lists
548		return
549	<
550	<	def jobSplitting(self):
549	>
550	>	def jobSplittingNoInput(self):
551		"""
552	<	first implemntation for job splitting
553	<	"""
554	<	#  print 'eventi totali '+str(self.maxEvents)
292	<	#  print 'eventi totali richiesti dallo user '+str(self.total_number_of_events)
293	<	#print 'files per job '+str(self.filesPerJob)
294	<	common.logger.message('Required '+str(self.filesPerJob)+' files per job ')
295	<	common.logger.message('Required '+str(self.total_number_of_events)+' events in total ')
296	<
297	<	## TODO: SL need to have (from DBS) a detailed list of how many events per each file
298	<	n_tot_files = (len(self.files[0]))
299	<	## SL: this is wrong if the files have different number of events
300	<	evPerFile = int(self.maxEvents)/n_tot_files
301	<
302	<	common.logger.debug(5,'Events per File '+str(evPerFile))
303	<
304	<	## if asked to process all events, do it
305	<	if self.total_number_of_events == -1:
306	<	self.total_number_of_events=self.maxEvents
307	<	self.total_number_of_jobs = int(n_tot_files)*1/int(self.filesPerJob)
308	<	common.logger.message(str(self.total_number_of_jobs)+' jobs will be created for all available events '+str(self.total_number_of_events)+' events')
309	<
310	<	else:
311	<	self.total_number_of_files = int(self.total_number_of_events/evPerFile)
312	<	## SL: if ask for less event than what is computed to be available on a
313	<	##     file, process the first file anyhow.
314	<	if self.total_number_of_files == 0:
315	<	self.total_number_of_files = self.total_number_of_files + 1
552	>	Perform job splitting based on number of event per job
553	>	"""
554	>	common.logger.debug(5,'Splitting per events')
555
556	<	common.logger.debug(5,'N files  '+str(self.total_number_of_files))
556	>	if (self.selectEventsPerJob):
557	>	common.logger.message('Required '+str(self.eventsPerJob)+' events per job ')
558	>	if (self.selectNumberOfJobs):
559	>	common.logger.message('Required '+str(self.theNumberOfJobs)+' jobs in total ')
560	>	if (self.selectTotalNumberEvents):
561	>	common.logger.message('Required '+str(self.total_number_of_events)+' events in total ')
562
563	<	check = 0
564	<
565	<	## Compute the number of jobs
322	<	#self.total_number_of_jobs = int(n_tot_files)*1/int(self.filesPerJob)
323	<	self.total_number_of_jobs = int(self.total_number_of_files/self.filesPerJob)
324	<	common.logger.debug(5,'N jobs  '+str(self.total_number_of_jobs))
563	>	if (self.total_number_of_events < 0):
564	>	msg='Cannot split jobs per Events with "-1" as total number of events'
565	>	raise CrabException(msg)
566
567	<	## is there any remainder?
568	<	check = int(self.total_number_of_files) - (int(self.total_number_of_jobs)*self.filesPerJob)
567	>	if (self.selectEventsPerJob):
568	>	if (self.selectTotalNumberEvents):
569	>	self.total_number_of_jobs = int(self.total_number_of_events/self.eventsPerJob)
570	>	elif(self.selectNumberOfJobs) :
571	>	self.total_number_of_jobs =self.theNumberOfJobs
572	>	self.total_number_of_events =int(self.theNumberOfJobs*self.eventsPerJob)
573	>
574	>	elif (self.selectNumberOfJobs) :
575	>	self.total_number_of_jobs = self.theNumberOfJobs
576	>	self.eventsPerJob = int(self.total_number_of_events/self.total_number_of_jobs)
577	>
578	>	common.logger.debug(5,'N jobs  '+str(self.total_number_of_jobs))
579	>
580	>	# is there any remainder?
581	>	check = int(self.total_number_of_events) - (int(self.total_number_of_jobs)*self.eventsPerJob)
582	>
583	>	common.logger.debug(5,'Check  '+str(check))
584	>
585	>	common.logger.message(str(self.total_number_of_jobs)+' jobs can be created, each for '+str(self.eventsPerJob)+' for a total of '+str(self.total_number_of_jobs*self.eventsPerJob)+' events')
586	>	if check > 0:
587	>	common.logger.message('Warning: asked '+str(self.total_number_of_events)+' but can do only '+str(int(self.total_number_of_jobs)*self.eventsPerJob))
588	>
589	>	# argument is seed number.$i
590	>	self.list_of_args = []
591	>	for i in range(self.total_number_of_jobs):
592	>	## Since there is no input, any site is good
593	>	self.jobDestination.append([""]) #must be empty to write correctly the xml
594	>	args=[]
595	>	if (self.firstRun):
596	>	## pythia first run
597	>	args.append(str(self.firstRun)+str(i))
598	>	self.list_of_args.append(args)
599
600	<	common.logger.debug(5,'Check  '+str(check))
600	>	return
601
331	–	if check > 0:
332	–	self.total_number_of_jobs =  self.total_number_of_jobs + 1
333	–	common.logger.message('Warning: last job will be created with '+str(check)+' files')
602
603	<	common.logger.message(str(self.total_number_of_jobs)+' jobs will be created for a total of '+str((self.total_number_of_jobs-1)*self.filesPerJob*evPerFile + check*evPerFile)+' events')
604	<	pass
603	>	def jobSplittingForScript(self):#CarlosDaniele
604	>	"""
605	>	Perform job splitting based on number of job
606	>	"""
607	>	common.logger.debug(5,'Splitting per job')
608	>	common.logger.message('Required '+str(self.theNumberOfJobs)+' jobs in total ')
609
610	<	list_of_lists = []
611	<	for i in xrange(0, int(n_tot_files), self.filesPerJob):
612	<	list_of_lists.append(self.files[0][i: i+self.filesPerJob])
610	>	self.total_number_of_jobs = self.theNumberOfJobs
611	>
612	>	common.logger.debug(5,'N jobs  '+str(self.total_number_of_jobs))
613
614	<	self.list_of_files = list_of_lists
615	<
614	>	common.logger.message(str(self.total_number_of_jobs)+' jobs can be created')
615	>
616	>	# argument is seed number.$i
617	>	self.list_of_args = []
618	>	for i in range(self.total_number_of_jobs):
619	>	## Since there is no input, any site is good
620	>	# self.jobDestination.append(["Any"])
621	>	self.jobDestination.append([""])
622	>	## no random seed
623	>	self.list_of_args.append([str(i)])
624		return
625
626		def split(self, jobParams):
627	<
348	<	common.jobDB.load()
627	>
628		#### Fabio
629		njobs = self.total_number_of_jobs
630	<	filelist = self.list_of_files
630	>	arglist = self.list_of_args
631		# create the empty structure
632		for i in range(njobs):
633		jobParams.append("")
634	<
634	>
635	>	listID=[]
636	>	listField=[]
637		for job in range(njobs):
638	<	jobParams[job] = filelist[job]
639	<	common.jobDB.setArguments(job, jobParams[job])
638	>	jobParams[job] = arglist[job]
639	>	listID.append(job+1)
640	>	job_ToSave ={}
641	>	concString = ' '
642	>	argu=''
643	>	if len(jobParams[job]):
644	>	argu +=   concString.join(jobParams[job] )
645	>	job_ToSave['arguments']= str(job+1)+' '+argu## new BL--DS
646	>	job_ToSave['dlsDestination']= self.jobDestination[job]## new BL--DS
647	>	#common._db.updateJob_(job,job_ToSave)## new BL--DS
648	>	listField.append(job_ToSave)
649	>	msg="Job "+str(job)+" Arguments:   "+str(job+1)+" "+argu+"\n"  \
650	>	+"                     Destination: "+str(self.jobDestination[job])
651	>	common.logger.debug(5,msg)
652	>	#common.logger.debug(5,"Job "+str(job)+" Destination: "+str(self.jobDestination[job]))
653	>	common._db.updateJob_(listID,listField)## new BL--DS
654	>	## Pay Attention Here....DS--BL
655	>	self.argsList = (len(jobParams[1])+1)
656
360	–	common.jobDB.save()
657		return
658	<
363	<	def getJobTypeArguments(self, nj, sched):
364	<	params = common.jobDB.arguments(nj)
365	<	#print params
366	<	parString = "\\{"
367	<
368	<	for i in range(len(params) - 1):
369	<	parString += '\\\"' + params[i] + '\\\"\,'
370	<
371	<	parString += '\\\"' + params[len(params) - 1] + '\\\"\\}'
372	<	return parString
373	<
658	>
659		def numberOfJobs(self):
660		# Fabio
376	–
661		return self.total_number_of_jobs
378	–
379	–
380	–
381	–	def checkBlackList(self, allSites):
382	–	if len(self.reCEBlackList)==0: return allSites
383	–	sites = []
384	–	for site in allSites:
385	–	common.logger.debug(10,'Site '+site)
386	–	good=1
387	–	for re in self.reCEBlackList:
388	–	if re.search(site):
389	–	common.logger.message('CE in black list, skipping site '+site)
390	–	good=0
391	–	pass
392	–	if good: sites.append(site)
393	–	if len(sites) == 0:
394	–	common.logger.debug(3,"No sites found after BlackList")
395	–	return sites
396	–
397	–	def checkWhiteList(self, allSites):
398	–
399	–	if len(self.reCEWhiteList)==0: return allSites
400	–	sites = []
401	–	for site in allSites:
402	–	good=0
403	–	for re in self.reCEWhiteList:
404	–	if re.search(site):
405	–	common.logger.debug(5,'CE in white list, adding site '+site)
406	–	good=1
407	–	if not good: continue
408	–	sites.append(site)
409	–	if len(sites) == 0:
410	–	common.logger.message("No sites found after WhiteList\n")
411	–	else:
412	–	common.logger.debug(5,"Selected sites via WhiteList are "+str(sites)+"\n")
413	–	return sites
662
663		def getTarBall(self, exe):
664		"""
665		Return the TarBall with lib and exe
666		"""
667	<
667	>
668		# if it exist, just return it
669	<	self.tgzNameWithPath = common.work_space.shareDir()+self.tgz_name
669	>	#
670	>	# Marco. Let's start to use relative path for Boss XML files
671	>	#
672	>	self.tgzNameWithPath = common.work_space.pathForTgz()+'share/'+self.tgz_name
673		if os.path.exists(self.tgzNameWithPath):
674		return self.tgzNameWithPath
675
#	Line 432 | Line 683 | class Cmssw(JobType):
683		# First of all declare the user Scram area
684		swArea = self.scram.getSWArea_()
685		#print "swArea = ", swArea
686	<	swVersion = self.scram.getSWVersion()
687	<	#print "swVersion = ", swVersion
686	>	# swVersion = self.scram.getSWVersion()
687	>	# print "swVersion = ", swVersion
688		swReleaseTop = self.scram.getReleaseTop_()
689		#print "swReleaseTop = ", swReleaseTop
690	<
690	>
691		## check if working area is release top
692		if swReleaseTop == '' or swArea == swReleaseTop:
693	+	common.logger.debug(3,"swArea = "+swArea+" swReleaseTop ="+swReleaseTop)
694		return
695
696	<	filesToBeTarred = []
697	<	## First find the executable
698	<	if (self.executable != ''):
699	<	exeWithPath = self.scram.findFile_(executable)
700	<	#           print exeWithPath
701	<	if ( not exeWithPath ):
702	<	raise CrabException('User executable '+executable+' not found')
703	<
704	<	## then check if it's private or not
705	<	if exeWithPath.find(swReleaseTop) == -1:
706	<	# the exe is private, so we must ship
707	<	common.logger.debug(5,"Exe "+exeWithPath+" to be tarred")
708	<	path = swArea+'/'
709	<	exe = string.replace(exeWithPath, path,'')
710	<	filesToBeTarred.append(exe)
711	<	pass
712	<	else:
713	<	# the exe is from release, we'll find it on WN
714	<	pass
715	<
716	<	## Now get the libraries: only those in local working area
717	<	libDir = 'lib'
718	<	lib = swArea+'/' +libDir
719	<	common.logger.debug(5,"lib "+lib+" to be tarred")
720	<	if os.path.exists(lib):
721	<	filesToBeTarred.append(libDir)
722	<
723	<	## Now check if module dir is present
724	<	moduleDir = 'module'
725	<	if os.path.isdir(swArea+'/'+moduleDir):
726	<	filesToBeTarred.append(moduleDir)
727	<
728	<	## Now check if the Data dir is present
729	<	dataDir = 'src/Data/'
730	<	if os.path.isdir(swArea+'/'+dataDir):
731	<	filesToBeTarred.append(dataDir)
732	<
733	<	## Create the tar-ball
734	<	if len(filesToBeTarred)>0:
735	<	cwd = os.getcwd()
736	<	os.chdir(swArea)
737	<	tarcmd = 'tar zcvf ' + self.tgzNameWithPath + ' '
738	<	for line in filesToBeTarred:
739	<	tarcmd = tarcmd + line + ' '
740	<	cout = runCommand(tarcmd)
741	<	if not cout:
742	<	raise CrabException('Could not create tar-ball')
743	<	os.chdir(cwd)
744	<	else:
745	<	common.logger.debug(5,"No files to be to be tarred")
746	<
696	>	import tarfile
697	>	try: # create tar ball
698	>	tar = tarfile.open(self.tgzNameWithPath, "w:gz")
699	>	## First find the executable
700	>	if (self.executable != ''):
701	>	exeWithPath = self.scram.findFile_(executable)
702	>	if ( not exeWithPath ):
703	>	raise CrabException('User executable '+executable+' not found')
704	>
705	>	## then check if it's private or not
706	>	if exeWithPath.find(swReleaseTop) == -1:
707	>	# the exe is private, so we must ship
708	>	common.logger.debug(5,"Exe "+exeWithPath+" to be tarred")
709	>	path = swArea+'/'
710	>	# distinguish case when script is in user project area or given by full path somewhere else
711	>	if exeWithPath.find(path) >= 0 :
712	>	exe = string.replace(exeWithPath, path,'')
713	>	tar.add(path+exe,exe)
714	>	else :
715	>	tar.add(exeWithPath,os.path.basename(executable))
716	>	pass
717	>	else:
718	>	# the exe is from release, we'll find it on WN
719	>	pass
720	>
721	>	## Now get the libraries: only those in local working area
722	>	libDir = 'lib'
723	>	lib = swArea+'/' +libDir
724	>	common.logger.debug(5,"lib "+lib+" to be tarred")
725	>	if os.path.exists(lib):
726	>	tar.add(lib,libDir)
727	>
728	>	## Now check if module dir is present
729	>	moduleDir = 'module'
730	>	module = swArea + '/' + moduleDir
731	>	if os.path.isdir(module):
732	>	tar.add(module,moduleDir)
733	>
734	>	## Now check if any data dir(s) is present
735	>	swAreaLen=len(swArea)
736	>	for root, dirs, files in os.walk(swArea):
737	>	if "data" in dirs:
738	>	common.logger.debug(5,"data "+root+"/data"+" to be tarred")
739	>	tar.add(root+"/data",root[swAreaLen:]+"/data")
740	>
741	>
742	>	## Add ProdCommon dir to tar
743	>	prodcommonDir = 'ProdCommon'
744	>	prodcommonPath = os.environ['CRABDIR'] + '/' + 'ProdCommon'
745	>	if os.path.isdir(prodcommonPath):
746	>	tar.add(prodcommonPath,prodcommonDir)
747	>
748	>	common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
749	>	tar.close()
750	>	except :
751	>	raise CrabException('Could not create tar-ball')
752	>
753	>	## check for tarball size
754	>	tarballinfo = os.stat(self.tgzNameWithPath)
755	>	if ( tarballinfo.st_size > self.MaxTarBallSize*1024*1024 ) :
756	>	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.')
757	>
758	>	## create tar-ball with ML stuff
759	>	self.MLtgzfile =  common.work_space.pathForTgz()+'share/MLfiles.tgz'
760	>	try:
761	>	tar = tarfile.open(self.MLtgzfile, "w:gz")
762	>	path=os.environ['CRABDIR'] + '/python/'
763	>	for file in ['report.py', 'DashboardAPI.py', 'Logger.py', 'ProcInfo.py', 'apmon.py', 'parseCrabFjr.py','writeCfg.py', 'JobReportErrorCode.py']:
764	>	tar.add(path+file,file)
765	>	common.logger.debug(5,"Files added to "+self.MLtgzfile+" : "+str(tar.getnames()))
766	>	tar.close()
767	>	except :
768	>	raise CrabException('Could not create ML files tar-ball')
769	>
770		return
771	<
772	<	def wsSetupEnvironment(self, nj):
771	>
772	>	def additionalInputFileTgz(self):
773	>	"""
774	>	Put all additional files into a tar ball and return its name
775	>	"""
776	>	import tarfile
777	>	tarName=  common.work_space.pathForTgz()+'share/'+self.additional_tgz_name
778	>	tar = tarfile.open(tarName, "w:gz")
779	>	for file in self.additional_inbox_files:
780	>	tar.add(file,string.split(file,'/')[-1])
781	>	common.logger.debug(5,"Files added to "+self.additional_tgz_name+" : "+str(tar.getnames()))
782	>	tar.close()
783	>	return tarName
784	>
785	>	def wsSetupEnvironment(self, nj=0):
786		"""
787		Returns part of a job script which prepares
788		the execution environment for the job 'nj'.
789		"""
790		# Prepare JobType-independent part
791	<	txt = ''
792	<
793	<	## OLI_Daniele at this level  middleware already known
506	<
507	<	txt += 'if [ $middleware == LCG ]; then \n'
791	>	txt = '\n#Written by cms_cmssw::wsSetupEnvironment\n'
792	>	txt += 'echo ">>> setup environment"\n'
793	>	txt += 'if [ $middleware == LCG ]; then \n'
794		txt += self.wsSetupCMSLCGEnvironment_()
795		txt += 'elif [ $middleware == OSG ]; then\n'
796	<	txt += '    time=`date -u +"%s"`\n'
797	<	txt += '    WORKING_DIR=$OSG_WN_TMP/cms_$time\n'
798	<	txt += '    echo "Creating working directory: $WORKING_DIR"\n'
799	<	txt += '    /bin/mkdir -p $WORKING_DIR\n'
800	<	txt += '    if [ ! -d $WORKING_DIR ] ;then\n'
515	<	txt += '        echo "SET_CMS_ENV 10016 ==> OSG $WORKING_DIR could not be created on WN `hostname`"\n'
516	<	txt += '        echo "JOB_EXIT_STATUS = 10016"\n'
517	<	txt += '        echo "JobExitCode=10016" | tee -a $RUNTIME_AREA/$repo\n'
518	<	txt += '        dumpStatus $RUNTIME_AREA/$repo\n'
519	<	txt += '        exit 1\n'
796	>	txt += '    WORKING_DIR=`/bin/mktemp  -d $OSG_WN_TMP/cms_XXXXXXXXXXXX`\n'
797	>	txt += '    if [ ! $? == 0 ] ;then\n'
798	>	txt += '        echo "ERROR ==> OSG $WORKING_DIR could not be created on WN `hostname`"\n'
799	>	txt += '        job_exit_code=10016\n'
800	>	txt += '        func_exit\n'
801		txt += '    fi\n'
802	+	txt += '    echo ">>> Created working directory: $WORKING_DIR"\n'
803		txt += '\n'
804		txt += '    echo "Change to working directory: $WORKING_DIR"\n'
805		txt += '    cd $WORKING_DIR\n'
806	<	txt += self.wsSetupCMSOSGEnvironment_()
806	>	txt += '    echo ">>> current directory (WORKING_DIR): $WORKING_DIR"\n'
807	>	txt += self.wsSetupCMSOSGEnvironment_()
808	>	#txt += '    echo "### Set SCRAM ARCH to ' + self.executable_arch + ' ###"\n'
809	>	#txt += '    export SCRAM_ARCH='+self.executable_arch+'\n'
810		txt += 'fi\n'
811
812		# Prepare JobType-specific part
813		scram = self.scram.commandName()
814		txt += '\n\n'
815	<	txt += 'echo "### SPECIFIC JOB SETUP ENVIRONMENT ###"\n'
815	>	txt += 'echo ">>> specific cmssw setup environment:"\n'
816	>	txt += 'echo "CMSSW_VERSION =  '+self.version+'"\n'
817		txt += scram+' project CMSSW '+self.version+'\n'
818		txt += 'status=$?\n'
819		txt += 'if [ $status != 0 ] ; then\n'
820	<	txt += '   echo "SET_EXE_ENV 10034 ==>ERROR CMSSW '+self.version+' not found on `hostname`" \n'
821	<	txt += '   echo "JOB_EXIT_STATUS = 10034"\n'
822	<	txt += '   echo "JobExitCode=10034" | tee -a $RUNTIME_AREA/$repo\n'
537	<	txt += '   dumpStatus $RUNTIME_AREA/$repo\n'
538	<	## OLI_Daniele
539	<	txt += '    if [ $middleware == OSG ]; then \n'
540	<	txt += '        echo "Remove working directory: $WORKING_DIR"\n'
541	<	txt += '        cd $RUNTIME_AREA\n'
542	<	txt += '        /bin/rm -rf $WORKING_DIR\n'
543	<	txt += '        if [ -d $WORKING_DIR ] ;then\n'
544	<	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'
545	<	txt += '            echo "JOB_EXIT_STATUS = 10018"\n'
546	<	txt += '            echo "JobExitCode=10018" | tee -a $RUNTIME_AREA/$repo\n'
547	<	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
548	<	txt += '        fi\n'
549	<	txt += '    fi \n'
550	<	txt += '   exit 1 \n'
820	>	txt += '    echo "ERROR ==> CMSSW '+self.version+' not found on `hostname`" \n'
821	>	txt += '    job_exit_code=10034\n'
822	>	txt += '    func_exit\n'
823		txt += 'fi \n'
552	–	txt += 'echo "CMSSW_VERSION =  '+self.version+'"\n'
824		txt += 'cd '+self.version+'\n'
825	+	########## FEDE FOR DBS2 ######################
826	+	txt += 'SOFTWARE_DIR=`pwd`\n'
827	+	txt += 'echo ">>> current directory (SOFTWARE_DIR): $SOFTWARE_DIR" \n'
828	+	###############################################
829		### needed grep for bug in scramv1 ###
830		txt += 'eval `'+scram+' runtime -sh | grep -v SCRAMRT_LSB_JOBNAME`\n'
556	–
831		# Handle the arguments:
832		txt += "\n"
833		txt += "## number of arguments (first argument always jobnumber)\n"
834		txt += "\n"
835	<	txt += "narg=$#\n"
562	<	txt += "if [ $narg -lt 2 ]\n"
835	>	txt += "if [ $nargs -lt "+str(self.argsList)+" ]\n"
836		txt += "then\n"
837	<	txt += "    echo 'SET_EXE_ENV 1 ==> ERROR Too few arguments' +$narg+ \n"
838	<	txt += '    echo "JOB_EXIT_STATUS = 50113"\n'
839	<	txt += '    echo "JobExitCode=50113" | tee -a $RUNTIME_AREA/$repo\n'
567	<	txt += '    dumpStatus $RUNTIME_AREA/$repo\n'
568	<	## OLI_Daniele
569	<	txt += '    if [ $middleware == OSG ]; then \n'
570	<	txt += '        echo "Remove working directory: $WORKING_DIR"\n'
571	<	txt += '        cd $RUNTIME_AREA\n'
572	<	txt += '        /bin/rm -rf $WORKING_DIR\n'
573	<	txt += '        if [ -d $WORKING_DIR ] ;then\n'
574	<	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'
575	<	txt += '            echo "JOB_EXIT_STATUS = 50114"\n'
576	<	txt += '            echo "JobExitCode=50114" | tee -a $RUNTIME_AREA/$repo\n'
577	<	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
578	<	txt += '        fi\n'
579	<	txt += '    fi \n'
580	<	txt += "    exit 1\n"
837	>	txt += "    echo 'ERROR ==> Too few arguments' +$nargs+ \n"
838	>	txt += '    job_exit_code=50113\n'
839	>	txt += "    func_exit\n"
840		txt += "fi\n"
841		txt += "\n"
842
843		# Prepare job-specific part
844		job = common.job_list[nj]
845	<	pset = os.path.basename(job.configFilename())
846	<	txt += '\n'
847	<	txt += 'InputFiles=$2\n'
848	<	txt += 'echo "<$InputFiles>"\n'
590	<	#txt += 'echo sed "s#{\'INPUT\'}#$InputFiles#" $RUNTIME_AREA/'+pset+' \n'
591	<	txt += 'sed "s#{\'INPUT\'}#$InputFiles#" $RUNTIME_AREA/'+pset+' > pset.cfg\n'
592	<	#txt += 'sed "s#{\'INPUT\'}#${InputFiles}#" $RUNTIME_AREA/'+pset+' > pset1.cfg\n'
845	>	### FEDE FOR DBS OUTPUT PUBLICATION
846	>	if (self.datasetPath):
847	>	txt += '\n'
848	>	txt += 'DatasetPath='+self.datasetPath+'\n'
849
850	<	if len(self.additional_inbox_files) > 0:
595	<	for file in self.additional_inbox_files:
596	<	txt += 'if [ -e $RUNTIME_AREA/'+file+' ] ; then\n'
597	<	txt += '   cp $RUNTIME_AREA/'+file+' .\n'
598	<	txt += '   chmod +x '+file+'\n'
599	<	txt += 'fi\n'
600	<	pass
850	>	datasetpath_split = self.datasetPath.split("/")
851
852	<	txt += 'echo "### END JOB SETUP ENVIRONMENT ###"\n\n'
852	>	txt += 'PrimaryDataset='+datasetpath_split[1]+'\n'
853	>	txt += 'DataTier='+datasetpath_split[2]+'\n'
854	>	txt += 'ApplicationFamily=cmsRun\n'
855
856	<	txt += '\n'
857	<	txt += 'echo "***** cat pset.cfg *********"\n'
858	<	txt += 'cat pset.cfg\n'
859	<	txt += 'echo "****** end pset.cfg ********"\n'
860	<	txt += '\n'
861	<	# txt += 'echo "***** cat pset1.cfg *********"\n'
862	<	# txt += 'cat pset1.cfg\n'
863	<	# txt += 'echo "****** end pset1.cfg ********"\n'
856	>	else:
857	>	txt += 'DatasetPath=MCDataTier\n'
858	>	txt += 'PrimaryDataset=null\n'
859	>	txt += 'DataTier=null\n'
860	>	txt += 'ApplicationFamily=MCDataTier\n'
861	>	if self.pset != None:
862	>	pset = os.path.basename(job.configFilename())
863	>	txt += '\n'
864	>	txt += 'cp  $RUNTIME_AREA/'+pset+' .\n'
865	>	if (self.datasetPath): # standard job
866	>	txt += 'InputFiles=${args[1]}; export InputFiles\n'
867	>	txt += 'MaxEvents=${args[2]}; export MaxEvents\n'
868	>	txt += 'SkipEvents=${args[3]}; export SkipEvents\n'
869	>	txt += 'echo "Inputfiles:<$InputFiles>"\n'
870	>	txt += 'echo "MaxEvents:<$MaxEvents>"\n'
871	>	txt += 'echo "SkipEvents:<$SkipEvents>"\n'
872	>	else:  # pythia like job
873	>	txt += 'PreserveSeeds='  + ','.join(self.preserveSeeds)  + '; export PreserveSeeds\n'
874	>	txt += 'IncrementSeeds=' + ','.join(self.incrementSeeds) + '; export IncrementSeeds\n'
875	>	txt += 'echo "PreserveSeeds: <$PreserveSeeds>"\n'
876	>	txt += 'echo "IncrementSeeds:<$IncrementSeeds>"\n'
877	>	if (self.firstRun):
878	>	txt += 'FirstRun=${args[1]}; export FirstRun\n'
879	>	txt += 'echo "FirstRun: <$FirstRun>"\n'
880	>
881	>	txt += 'mv -f '+pset+' pset.cfg\n'
882	>
883	>	if len(self.additional_inbox_files) > 0:
884	>	txt += 'if [ -e $RUNTIME_AREA/'+self.additional_tgz_name+' ] ; then\n'
885	>	txt += '  tar xzvf $RUNTIME_AREA/'+self.additional_tgz_name+'\n'
886	>	txt += 'fi\n'
887	>	pass
888	>
889	>	if self.pset != None:
890	>	txt += '\n'
891	>	txt += 'echo "***** cat pset.cfg *********"\n'
892	>	txt += 'cat pset.cfg\n'
893	>	txt += 'echo "****** end pset.cfg ********"\n'
894	>	txt += '\n'
895	>	txt += 'PSETHASH=`EdmConfigHash < pset.cfg` \n'
896	>	txt += 'echo "PSETHASH = $PSETHASH" \n'
897	>	txt += '\n'
898		return txt
899
900	<	def wsBuildExe(self, nj):
900	>	def wsUntarSoftware(self, nj=0):
901		"""
902		Put in the script the commands to build an executable
903		or a library.
904		"""
905
906	<	txt = ""
906	>	txt = '\n#Written by cms_cmssw::wsUntarSoftware\n'
907
908		if os.path.isfile(self.tgzNameWithPath):
909	<	txt += 'echo "tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+'"\n'
909	>	txt += 'echo ">>> tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+' :" \n'
910		txt += 'tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+'\n'
911		txt += 'untar_status=$? \n'
912		txt += 'if [ $untar_status -ne 0 ]; then \n'
913	<	txt += '   echo "SET_EXE 1 ==> ERROR Untarring .tgz file failed"\n'
914	<	txt += '   echo "JOB_EXIT_STATUS = $untar_status" \n'
915	<	txt += '   echo "JobExitCode=$untar_status" | tee -a $RUNTIME_AREA/$repo\n'
630	<	txt += '   if [ $middleware == OSG ]; then \n'
631	<	txt += '       echo "Remove working directory: $WORKING_DIR"\n'
632	<	txt += '       cd $RUNTIME_AREA\n'
633	<	txt += '       /bin/rm -rf $WORKING_DIR\n'
634	<	txt += '       if [ -d $WORKING_DIR ] ;then\n'
635	<	txt += '        echo "SET_EXE 50999 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after Untarring .tgz file failed"\n'
636	<	txt += '        echo "JOB_EXIT_STATUS = 50999"\n'
637	<	txt += '        echo "JobExitCode=50999" | tee -a $RUNTIME_AREA/$repo\n'
638	<	txt += '        dumpStatus $RUNTIME_AREA/$repo\n'
639	<	txt += '       fi\n'
640	<	txt += '   fi \n'
641	<	txt += '   \n'
642	<	txt += '   exit 1 \n'
913	>	txt += '   echo "ERROR ==> Untarring .tgz file failed"\n'
914	>	txt += '   job_exit_code=$untar_status\n'
915	>	txt += '   func_exit\n'
916		txt += 'else \n'
917		txt += '   echo "Successful untar" \n'
918		txt += 'fi \n'
919	+	txt += '\n'
920	+	txt += 'echo ">>> Include ProdCommon in PYTHONPATH:"\n'
921	+	txt += 'if [ -z "$PYTHONPATH" ]; then\n'
922	+	txt += '   export PYTHONPATH=$RUNTIME_AREA/ProdCommon\n'
923	+	txt += 'else\n'
924	+	txt += '   export PYTHONPATH=$RUNTIME_AREA/ProdCommon:${PYTHONPATH}\n'
925	+	txt += 'echo "PYTHONPATH=$PYTHONPATH"\n'
926	+	txt += 'fi\n'
927	+	txt += '\n'
928	+
929		pass
930	<
930	>
931	>	return txt
932	>
933	>	def wsBuildExe(self, nj=0):
934	>	"""
935	>	Put in the script the commands to build an executable
936	>	or a library.
937	>	"""
938	>
939	>	txt = '\n#Written by cms_cmssw::wsBuildExe\n'
940	>	txt += 'echo ">>> moving CMSSW software directories in `pwd`" \n'
941	>
942	>	txt += 'rm -r lib/ module/ \n'
943	>	txt += 'mv $RUNTIME_AREA/lib/ . \n'
944	>	txt += 'mv $RUNTIME_AREA/module/ . \n'
945	>	txt += 'mv $RUNTIME_AREA/ProdCommon/ . \n'
946	>
947	>	txt += 'if [ -z "$PYTHONPATH" ]; then\n'
948	>	txt += '   export PYTHONPATH=$SOFTWARE_DIR/ProdCommon\n'
949	>	txt += 'else\n'
950	>	txt += '   export PYTHONPATH=$SOFTWARE_DIR/ProdCommon:${PYTHONPATH}\n'
951	>	txt += 'echo "PYTHONPATH=$PYTHONPATH"\n'
952	>	txt += 'fi\n'
953	>	txt += '\n'
954	>
955		return txt
956
957		def modifySteeringCards(self, nj):
958		"""
959	<	modify the card provided by the user,
959	>	modify the card provided by the user,
960		writing a new card into share dir
961		"""
962	<
962	>
963		def executableName(self):
964	<	return self.executable
964	>	if self.scriptExe: #CarlosDaniele
965	>	return "sh "
966	>	else:
967	>	return self.executable
968
969		def executableArgs(self):
970	<	return " -p pset.cfg"
970	>	# FUTURE: This function tests the CMSSW version. Can be simplified as we drop support for old versions
971	>	if self.scriptExe:#CarlosDaniele
972	>	return   self.scriptExe + " $NJob"
973	>	else:
974	>	version_array = self.scram.getSWVersion().split('_')
975	>	major = 0
976	>	minor = 0
977	>	try:
978	>	major = int(version_array[1])
979	>	minor = int(version_array[2])
980	>	except:
981	>	msg = "Cannot parse CMSSW version string: " + "_".join(version_array) + " for major and minor release number!"
982	>	raise CrabException(msg)
983	>
984	>	ex_args = ""
985	>	# FUTURE: This tests the CMSSW version. Can remove code as versions deprecated
986	>	# Framework job report
987	>	if major >= 1 and minor >= 5 :
988	>	ex_args += " -j $RUNTIME_AREA/crab_fjr_$NJob.xml"
989	>	# Type of cfg file
990	>	if major >= 2 :
991	>	ex_args += " -p pset.py"
992	>	else:
993	>	ex_args += " -p pset.cfg"
994	>	return ex_args
995
996		def inputSandbox(self, nj):
997		"""
998		Returns a list of filenames to be put in JDL input sandbox.
999		"""
1000		inp_box = []
1001	<	# dict added to delete duplicate from input sandbox file list
1002	<	seen = {}
1001	>	# # dict added to delete duplicate from input sandbox file list
1002	>	# seen = {}
1003		## code
1004		if os.path.isfile(self.tgzNameWithPath):
1005		inp_box.append(self.tgzNameWithPath)
1006	+	if os.path.isfile(self.MLtgzfile):
1007	+	inp_box.append(self.MLtgzfile)
1008		## config
1009	<	inp_box.append(common.job_list[nj].configFilename())
1009	>	if not self.pset is None:
1010	>	inp_box.append(common.work_space.pathForTgz() + 'job/' + self.configFilename())
1011		## additional input files
1012	<	#for file in self.additional_inbox_files:
1013	<	#    inp_box.append(common.work_space.cwdDir()+file)
1012	>	tgz = self.additionalInputFileTgz()
1013	>	inp_box.append(tgz)
1014	>	## executable
1015	>	wrapper = os.path.basename(str(common._db.queryTask('scriptName')))
1016	>	inp_box.append(common.work_space.pathForTgz() +'job/'+ wrapper)
1017		return inp_box
1018
1019		def outputSandbox(self, nj):
#	Line 682 | Line 1022 | class Cmssw(JobType):
1022		"""
1023		out_box = []
1024
685	–	stdout=common.job_list[nj].stdout()
686	–	stderr=common.job_list[nj].stderr()
687	–
1025		## User Declared output files
1026	<	for out in self.output_file:
1027	<	n_out = nj + 1
1026	>	for out in (self.output_file+self.output_file_sandbox):
1027	>	n_out = nj + 1
1028		out_box.append(self.numberFile_(out,str(n_out)))
1029		return out_box
693	–	return []
1030
1031		def prepareSteeringCards(self):
1032		"""
#	Line 703 | Line 1039 | class Cmssw(JobType):
1039		Returns part of a job script which renames the produced files.
1040		"""
1041
1042	<	txt = '\n'
1043	<	txt += '# directory content\n'
1042	>	txt = '\n#Written by cms_cmssw::wsRenameOutput\n'
1043	>	txt += 'echo ">>> current directory (SOFTWARE_DIR): $SOFTWARE_DIR" \n'
1044	>	txt += 'echo ">>> current directory content:"\n'
1045		txt += 'ls \n'
1046	<	file_list = ''
1047	<	for fileWithSuffix in self.output_file:
1046	>	txt += '\n'
1047	>
1048	>	for fileWithSuffix in (self.output_file):
1049		output_file_num = self.numberFile_(fileWithSuffix, '$NJob')
712	–	file_list=file_list+output_file_num+' '
1050		txt += '\n'
1051		txt += '# check output file\n'
1052	<	txt += 'ls '+fileWithSuffix+'\n'
1053	<	txt += 'exe_result=$?\n'
1054	<	txt += 'if [ $exe_result -ne 0 ] ; then\n'
1055	<	txt += '   echo "ERROR: No output file to manage"\n'
1056	<	txt += '   echo "JOB_EXIT_STATUS = $exe_result"\n'
1057	<	txt += '   echo "JobExitCode=60302" | tee -a $RUNTIME_AREA/$repo\n'
1058	<	txt += '   dumpStatus $RUNTIME_AREA/$repo\n'
1059	<	### OLI_DANIELE
1060	<	if common.scheduler.boss_scheduler_name == 'condor_g':
1052	>	txt += 'if [ -e ./'+fileWithSuffix+' ] ; then\n'
1053	>	if (self.copy_data == 1):  # For OSG nodes, file is in $WORKING_DIR, should not be moved to $RUNTIME_AREA
1054	>	txt += '    mv '+fileWithSuffix+' '+output_file_num+'\n'
1055	>	txt += '    ln -s `pwd`/'+output_file_num+' $RUNTIME_AREA/'+fileWithSuffix+'\n'
1056	>	else:
1057	>	txt += '    mv '+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n'
1058	>	txt += '    ln -s $RUNTIME_AREA/'+output_file_num+' $RUNTIME_AREA/'+fileWithSuffix+'\n'
1059	>	txt += 'else\n'
1060	>	txt += '    job_exit_code=60302\n'
1061	>	txt += '    echo "WARNING: Output file '+fileWithSuffix+' not found"\n'
1062	>	if common.scheduler.name().upper() == 'CONDOR_G':
1063		txt += '    if [ $middleware == OSG ]; then \n'
1064		txt += '        echo "prepare dummy output file"\n'
1065		txt += '        echo "Processing of job output failed" > $RUNTIME_AREA/'+output_file_num+'\n'
1066		txt += '    fi \n'
728	–	txt += 'else\n'
729	–	txt += '   cp '+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n'
1067		txt += 'fi\n'
1068	<
1069	<	txt += 'cd $RUNTIME_AREA\n'
1070	<	file_list=file_list[:-1]
1071	<	txt += 'file_list="'+file_list+'"\n'
1072	<	### OLI_DANIELE
736	<	txt += 'if [ $middleware == OSG ]; then\n'
737	<	txt += '    cd $RUNTIME_AREA\n'
738	<	txt += '    echo "Remove working directory: $WORKING_DIR"\n'
739	<	txt += '    /bin/rm -rf $WORKING_DIR\n'
740	<	txt += '    if [ -d $WORKING_DIR ] ;then\n'
741	<	txt += '        echo "SET_EXE 60999 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after cleanup of WN"\n'
742	<	txt += '        echo "JOB_EXIT_STATUS = 60999"\n'
743	<	txt += '        echo "JobExitCode=60999" | tee -a $RUNTIME_AREA/$repo\n'
744	<	txt += '        dumpStatus $RUNTIME_AREA/$repo\n'
745	<	txt += '    fi\n'
746	<	txt += 'fi\n'
1068	>	file_list = []
1069	>	for fileWithSuffix in (self.output_file):
1070	>	file_list.append(self.numberFile_(fileWithSuffix, '$NJob'))
1071	>
1072	>	txt += 'file_list="'+string.join(file_list,' ')+'"\n'
1073		txt += '\n'
1074	+	txt += 'echo ">>> current directory (SOFTWARE_DIR): $SOFTWARE_DIR" \n'
1075	+	txt += 'echo ">>> current directory content:"\n'
1076	+	txt += 'ls \n'
1077	+	txt += '\n'
1078	+	txt += 'cd $RUNTIME_AREA\n'
1079	+	txt += 'echo ">>> current directory (RUNTIME_AREA):  $RUNTIME_AREA"\n'
1080		return txt
1081
1082		def numberFile_(self, file, txt):
#	Line 755 | Line 1087 | class Cmssw(JobType):
1087		# take away last extension
1088		name = p[0]
1089		for x in p[1:-1]:
1090	<	name=name+"."+x
1090	>	name=name+"."+x
1091		# add "_txt"
1092		if len(p)>1:
1093	<	ext = p[len(p)-1]
1094	<	#result = name + '_' + str(txt) + "." + ext
1095	<	result = name + '_' + txt + "." + ext
1096	<	else:
1097	<	#result = name + '_' + str(txt)
766	<	result = name + '_' + txt
767	<
1093	>	ext = p[len(p)-1]
1094	>	result = name + '_' + txt + "." + ext
1095	>	else:
1096	>	result = name + '_' + txt
1097	>
1098		return result
1099
1100	<	def getRequirements(self):
1100	>	def getRequirements(self, nj=[]):
1101		"""
1102	<	return job requirements to add to jdl files
1102	>	return job requirements to add to jdl files
1103		"""
1104		req = ''
1105	<	if common.analisys_common_info['sites']:
1106	<	if common.analisys_common_info['sw_version']:
1107	<	req='Member("VO-cms-' + \
1108	<	common.analisys_common_info['sw_version'] + \
1109	<	'", other.GlueHostApplicationSoftwareRunTimeEnvironment)'
1110	<	if len(common.analisys_common_info['sites'])>0:
1111	<	req = req + ' && ('
1112	<	for i in range(len(common.analisys_common_info['sites'])):
1113	<	req = req + 'other.GlueCEInfoHostName == "' \
1114	<	+ common.analisys_common_info['sites'][i] + '"'
1115	<	if ( i < (int(len(common.analisys_common_info['sites']) - 1)) ):
1116	<	req = req + ' || '
1117	<	req = req + ')'
1118	<	#print "req = ", req
1105	>	if self.version:
1106	>	req='Member("VO-cms-' + \
1107	>	self.version + \
1108	>	'", other.GlueHostApplicationSoftwareRunTimeEnvironment)'
1109	>	## SL add requirement for OS version only if SL4
1110	>	#reSL4 = re.compile( r'slc4' )
1111	>	if self.executable_arch: # and reSL4.search(self.executable_arch):
1112	>	req+=' && Member("VO-cms-' + \
1113	>	self.executable_arch + \
1114	>	'", other.GlueHostApplicationSoftwareRunTimeEnvironment)'
1115	>
1116	>	req = req + ' && (other.GlueHostNetworkAdapterOutboundIP)'
1117	>	if common.scheduler.name() == "glitecoll":
1118	>	req += ' && other.GlueCEStateStatus == "Production" '
1119	>
1120		return req
1121
1122		def configFilename(self):
1123		""" return the config filename """
1124		return self.name()+'.cfg'
1125
795	–	### OLI_DANIELE
1126		def wsSetupCMSOSGEnvironment_(self):
1127		"""
1128		Returns part of a job script which is prepares
1129		the execution environment and which is common for all CMS jobs.
1130		"""
1131	<	txt = '\n'
1132	<	txt += '   echo "### SETUP CMS OSG  ENVIRONMENT ###"\n'
1133	<	txt += '   if [ -f $GRID3_APP_DIR/cmssoft/cmsset_default.sh ] ;then\n'
1134	<	txt += '      # Use $GRID3_APP_DIR/cmssoft/cmsset_default.sh to setup cms software\n'
1135	<	txt += '       source $GRID3_APP_DIR/cmssoft/cmsset_default.sh '+self.version+'\n'
1136	<	txt += '   elif [ -f $OSG_APP/cmssoft/cmsset_default.sh ] ;then\n'
1137	<	txt += '      # Use $OSG_APP/cmssoft/cmsset_default.sh to setup cms software\n'
1138	<	txt += '       source $OSG_APP/cmssoft/cmsset_default.sh '+self.version+'\n'
1139	<	txt += '   else\n'
1140	<	txt += '       echo "SET_CMS_ENV 10020 ==> ERROR $GRID3_APP_DIR/cmssoft/cmsset_default.sh and $OSG_APP/cmssoft/cmsset_default.sh file not found"\n'
1141	<	txt += '       echo "JOB_EXIT_STATUS = 10020"\n'
1142	<	txt += '       echo "JobExitCode=10020" | tee -a $RUNTIME_AREA/$repo\n'
1143	<	txt += '       dumpStatus $RUNTIME_AREA/$repo\n'
814	<	txt += '       exit 1\n'
815	<	txt += '\n'
816	<	txt += '       echo "Remove working directory: $WORKING_DIR"\n'
817	<	txt += '       cd $RUNTIME_AREA\n'
818	<	txt += '       /bin/rm -rf $WORKING_DIR\n'
819	<	txt += '       if [ -d $WORKING_DIR ] ;then\n'
820	<	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/cmsset_default.sh file not found"\n'
821	<	txt += '            echo "JOB_EXIT_STATUS = 10017"\n'
822	<	txt += '            echo "JobExitCode=10017" | tee -a $RUNTIME_AREA/$repo\n'
823	<	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
824	<	txt += '       fi\n'
825	<	txt += '\n'
826	<	txt += '       exit 1\n'
827	<	txt += '   fi\n'
1131	>	txt = '\n#Written by cms_cmssw::wsSetupCMSOSGEnvironment_\n'
1132	>	txt += '    echo ">>> setup CMS OSG environment:"\n'
1133	>	txt += '    echo "set SCRAM ARCH to ' + self.executable_arch + '"\n'
1134	>	txt += '    export SCRAM_ARCH='+self.executable_arch+'\n'
1135	>	txt += '    echo "SCRAM_ARCH = $SCRAM_ARCH"\n'
1136	>	txt += '    if [ -f $OSG_APP/cmssoft/cms/cmsset_default.sh ] ;then\n'
1137	>	txt += '      # Use $OSG_APP/cmssoft/cms/cmsset_default.sh to setup cms software\n'
1138	>	txt += '        source $OSG_APP/cmssoft/cms/cmsset_default.sh '+self.version+'\n'
1139	>	txt += '    else\n'
1140	>	txt += '        echo "ERROR ==> $OSG_APP/cmssoft/cms/cmsset_default.sh file not found"\n'
1141	>	txt += '        job_exit_code=10020\n'
1142	>	txt += '        func_exit\n'
1143	>	txt += '    fi\n'
1144		txt += '\n'
1145	<	txt += '   echo "SET_CMS_ENV 0 ==> setup cms environment ok"\n'
1146	<	txt += '   echo " END SETUP CMS OSG  ENVIRONMENT "\n'
1145	>	txt += '    echo "==> setup cms environment ok"\n'
1146	>	txt += '    echo "SCRAM_ARCH = $SCRAM_ARCH"\n'
1147
1148		return txt
1149	<
1149	>
1150		### OLI_DANIELE
1151		def wsSetupCMSLCGEnvironment_(self):
1152		"""
1153		Returns part of a job script which is prepares
1154		the execution environment and which is common for all CMS jobs.
1155		"""
1156	<	txt  = '   \n'
1157	<	txt += '   echo " ### SETUP CMS LCG  ENVIRONMENT ### "\n'
1158	<	txt += '   if [ ! $VO_CMS_SW_DIR ] ;then\n'
1159	<	txt += '       echo "SET_CMS_ENV 10031 ==> ERROR CMS software dir not found on WN `hostname`"\n'
1160	<	txt += '       echo "JOB_EXIT_STATUS = 10031" \n'
1161	<	txt += '       echo "JobExitCode=10031" | tee -a $RUNTIME_AREA/$repo\n'
1162	<	txt += '       dumpStatus $RUNTIME_AREA/$repo\n'
1163	<	txt += '       exit 1\n'
1164	<	txt += '   else\n'
1165	<	txt += '       echo "Sourcing environment... "\n'
1166	<	txt += '       if [ ! -s $VO_CMS_SW_DIR/cmsset_default.sh ] ;then\n'
1167	<	txt += '           echo "SET_CMS_ENV 10020 ==> ERROR cmsset_default.sh file not found into dir $VO_CMS_SW_DIR"\n'
1168	<	txt += '           echo "JOB_EXIT_STATUS = 10020"\n'
1169	<	txt += '           echo "JobExitCode=10020" | tee -a $RUNTIME_AREA/$repo\n'
1170	<	txt += '           dumpStatus $RUNTIME_AREA/$repo\n'
1171	<	txt += '           exit 1\n'
1172	<	txt += '       fi\n'
1173	<	txt += '       echo "sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
1174	<	txt += '       source $VO_CMS_SW_DIR/cmsset_default.sh\n'
1175	<	txt += '       result=$?\n'
1176	<	txt += '       if [ $result -ne 0 ]; then\n'
1177	<	txt += '           echo "SET_CMS_ENV 10032 ==> ERROR problem sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
1178	<	txt += '           echo "JOB_EXIT_STATUS = 10032"\n'
1179	<	txt += '           echo "JobExitCode=10032" | tee -a $RUNTIME_AREA/$repo\n'
1180	<	txt += '           dumpStatus $RUNTIME_AREA/$repo\n'
1181	<	txt += '           exit 1\n'
1182	<	txt += '       fi\n'
1183	<	txt += '   fi\n'
1184	<	txt += '   \n'
1185	<	txt += '   string=`cat /etc/redhat-release`\n'
1186	<	txt += '   echo $string\n'
1187	<	txt += '   if [[ $string = *alhalla* ]]; then\n'
1188	<	txt += '       echo "SCRAM_ARCH= $SCRAM_ARCH"\n'
1189	<	txt += '   elif [[ $string = *Enterprise* ]] || [[ $string = *cientific* ]]; then\n'
1190	<	txt += '       export SCRAM_ARCH=slc3_ia32_gcc323\n'
1191	<	txt += '       echo "SCRAM_ARCH= $SCRAM_ARCH"\n'
1192	<	txt += '   else\n'
1193	<	txt += '       echo "SET_CMS_ENV 10033 ==> ERROR OS unknown, LCG environment not initialized"\n'
1194	<	txt += '       echo "JOB_EXIT_STATUS = 10033"\n'
1195	<	txt += '       echo "JobExitCode=10033" | tee -a $RUNTIME_AREA/$repo\n'
1196	<	txt += '       dumpStatus $RUNTIME_AREA/$repo\n'
1197	<	txt += '       exit 1\n'
1198	<	txt += '   fi\n'
1199	<	txt += '   echo "SET_CMS_ENV 0 ==> setup cms environment ok"\n'
1200	<	txt += '   echo "### END SETUP CMS LCG ENVIRONMENT ###"\n'
1156	>	txt = '\n#Written by cms_cmssw::wsSetupCMSLCGEnvironment_\n'
1157	>	txt += '    echo ">>> setup CMS LCG environment:"\n'
1158	>	txt += '    echo "set SCRAM ARCH and BUILD_ARCH to ' + self.executable_arch + ' ###"\n'
1159	>	txt += '    export SCRAM_ARCH='+self.executable_arch+'\n'
1160	>	txt += '    export BUILD_ARCH='+self.executable_arch+'\n'
1161	>	txt += '    if [ ! $VO_CMS_SW_DIR ] ;then\n'
1162	>	txt += '        echo "ERROR ==> CMS software dir not found on WN `hostname`"\n'
1163	>	txt += '        job_exit_code=10031\n'
1164	>	txt += '        func_exit\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 "ERROR ==> cmsset_default.sh file not found into dir $VO_CMS_SW_DIR"\n'
1169	>	txt += '            job_exit_code=10020\n'
1170	>	txt += '            func_exit\n'
1171	>	txt += '        fi\n'
1172	>	txt += '        echo "sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
1173	>	txt += '        source $VO_CMS_SW_DIR/cmsset_default.sh\n'
1174	>	txt += '        result=$?\n'
1175	>	txt += '        if [ $result -ne 0 ]; then\n'
1176	>	txt += '            echo "ERROR ==> problem sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
1177	>	txt += '            job_exit_code=10032\n'
1178	>	txt += '            func_exit\n'
1179	>	txt += '        fi\n'
1180	>	txt += '    fi\n'
1181	>	txt += '    \n'
1182	>	txt += '    echo "==> setup cms environment ok"\n'
1183	>	return txt
1184	>
1185	>	### FEDE FOR DBS OUTPUT PUBLICATION
1186	>	def modifyReport(self, nj):
1187	>	"""
1188	>	insert the part of the script that modifies the FrameworkJob Report
1189	>	"""
1190	>
1191	>	txt = '\n#Written by cms_cmssw::modifyReport\n'
1192	>	publish_data = int(self.cfg_params.get('USER.publish_data',0))
1193	>	if (publish_data == 1):
1194	>	processedDataset = self.cfg_params['USER.publish_data_name']
1195	>	LFNBaseName = LFNBase(processedDataset)
1196	>
1197	>	txt += 'if [ $copy_exit_status -eq 0 ]; then\n'
1198	>	txt += '    FOR_LFN=%s_${PSETHASH}/\n'%(LFNBaseName)
1199	>	txt += 'else\n'
1200	>	txt += '    FOR_LFN=/copy_problems/ \n'
1201	>	txt += '    SE=""\n'
1202	>	txt += '    SE_PATH=""\n'
1203	>	txt += 'fi\n'
1204	>
1205	>	txt += 'echo ">>> Modify Job Report:" \n'
1206	>	txt += 'chmod a+x $SOFTWARE_DIR/ProdCommon/ProdCommon/FwkJobRep/ModifyJobReport.py\n'
1207	>	txt += 'ProcessedDataset='+processedDataset+'\n'
1208	>	txt += 'echo "ProcessedDataset = $ProcessedDataset"\n'
1209	>	txt += 'echo "SE = $SE"\n'
1210	>	txt += 'echo "SE_PATH = $SE_PATH"\n'
1211	>	txt += 'echo "FOR_LFN = $FOR_LFN" \n'
1212	>	txt += 'echo "CMSSW_VERSION = $CMSSW_VERSION"\n\n'
1213	>	txt += 'echo "$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 $SE_PATH"\n'
1214	>	txt += '$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 $SE_PATH\n'
1215	>	txt += 'modifyReport_result=$?\n'
1216	>	txt += 'if [ $modifyReport_result -ne 0 ]; then\n'
1217	>	txt += '    modifyReport_result=70500\n'
1218	>	txt += '    job_exit_code=$modifyReport_result\n'
1219	>	txt += '    echo "ModifyReportResult=$modifyReport_result" | tee -a $RUNTIME_AREA/$repo\n'
1220	>	txt += '    echo "WARNING: Problem with ModifyJobReport"\n'
1221	>	txt += 'else\n'
1222	>	txt += '    mv NewFrameworkJobReport.xml $RUNTIME_AREA/crab_fjr_$NJob.xml\n'
1223	>	txt += 'fi\n'
1224		return txt
1225
1226		def setParam_(self, param, value):
#	Line 889 | Line 1228 | class Cmssw(JobType):
1228
1229		def getParams(self):
1230		return self._params
1231	+
1232	+	def uniquelist(self, old):
1233	+	"""
1234	+	remove duplicates from a list
1235	+	"""
1236	+	nd={}
1237	+	for e in old:
1238	+	nd[e]=0
1239	+	return nd.keys()
1240	+
1241	+	def outList(self):
1242	+	"""
1243	+	check the dimension of the output files
1244	+	"""
1245	+	txt = ''
1246	+	txt += 'echo ">>> list of expected files on output sandbox"\n'
1247	+	listOutFiles = []
1248	+	stdout = 'CMSSW_$NJob.stdout'
1249	+	stderr = 'CMSSW_$NJob.stderr'
1250	+	if (self.return_data == 1):
1251	+	for file in (self.output_file+self.output_file_sandbox):
1252	+	listOutFiles.append(self.numberFile_(file, '$NJob'))
1253	+	listOutFiles.append(stdout)
1254	+	listOutFiles.append(stderr)
1255	+	else:
1256	+	for file in (self.output_file_sandbox):
1257	+	listOutFiles.append(self.numberFile_(file, '$NJob'))
1258	+	listOutFiles.append(stdout)
1259	+	listOutFiles.append(stderr)
1260	+	txt += 'echo "output files: '+string.join(listOutFiles,' ')+'"\n'
1261	+	txt += 'filesToCheck="'+string.join(listOutFiles,' ')+'"\n'
1262	+	txt += 'export filesToCheck\n'
1263	+	return txt

Diff Legend

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