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 1.33 by mkirn, Fri Jul 28 18:19:34 2006 UTC vs.
Revision 1.191 by spiga, Tue May 27 22:14:26 2008 UTC

#	Line 2 | Line 2 | from JobType import JobType
2		from crab_logger import Logger
3		from crab_exceptions import *
4		from crab_util import *
5	<	import math
5	>	from BlackWhiteListParser import BlackWhiteListParser
6		import common
7	–	import PsetManipulator
8	–
9	–	import DBSInfo_EDM
10	–	import DataDiscovery_EDM
11	–	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.scriptName = 'CMSSW.sh'
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	<	self.setParam_('application', self.version)
46	<	common.analisys_common_info['sw_version'] = self.version
47	<	### FEDE
48	<	common.analisys_common_info['copy_input_data'] = 0
40	<	common.analisys_common_info['events_management'] = 1
41	<
42	<	### collect Data cards
45	>	version_array = self.version.split('_')
46	>	self.CMSSW_major = 0
47	>	self.CMSSW_minor = 0
48	>	self.CMSSW_patch = 0
49		try:
50	<	tmp =  cfg_params['CMSSW.datasetpath']
51	<	log.debug(6, "CMSSW::CMSSW(): datasetPath = "+tmp)
52	<	if string.lower(tmp)=='none':
53	<	self.datasetPath = None
54	<	self.selectNoInput = 1
49	<	else:
50	<	self.datasetPath = tmp
51	<	self.selectNoInput = 0
52	<	except KeyError:
53	<	msg = "Error: datasetpath not defined "
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	<	# ML monitoring
57	<	# split dataset path style: /PreProdR3Minbias/SIM/GEN-SIM
58	<	if not self.datasetPath:
59	<	self.setParam_('dataset', 'None')
60	<	self.setParam_('owner', 'None')
61	<	else:
62	<	datasetpath_split = self.datasetPath.split("/")
63	<	self.setParam_('dataset', datasetpath_split[1])
64	<	self.setParam_('owner', datasetpath_split[-1])
57	>	### collect Data cards
58
59	<	self.setTaskid_()
60	<	self.setParam_('taskId', self.cfg_params['taskId'])
59	>	if not cfg_params.has_key('CMSSW.datasetpath'):
60	>	msg = "Error: datasetpath not defined "
61	>	raise CrabException(msg)
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	>	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']
74	<	self.setParam_('exe', self.executable)
75	<	log.debug(6, "CMSSW::CMSSW(): executable = "+self.executable)
76	<	msg = "Default executable cmsRun overridden. Switch to " + self.executable
77	<	log.debug(3,msg)
78	<	except KeyError:
79	<	self.executable = 'cmsRun'
80	<	self.setParam_('exe', self.executable)
81	<	msg = "User executable not defined. Use cmsRun"
82	<	log.debug(3,msg)
83	<	pass
76	>	self.executable = cfg_params.get('CMSSW.executable','cmsRun')
77	>	log.debug(6, "CMSSW::CMSSW(): executable = "+self.executable)
78
79	<	try:
80	<	self.pset = cfg_params['CMSSW.pset']
81	<	log.debug(6, "Cmssw::Cmssw(): PSet file = "+self.pset)
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	<	except KeyError:
87	<	raise CrabException("PSet file missing. Cannot run cmsRun ")
86	>	else:
87	>	self.pset = None
88
89		# output files
90	<	try:
91	<	self.output_file = []
90	>	## stuff which must be returned always via sandbox
91	>	self.output_file_sandbox = []
92
93	<	tmp = cfg_params['CMSSW.output_file']
94	<	if tmp != '':
95	<	tmpOutFiles = string.split(cfg_params['CMSSW.output_file'],',')
96	<	log.debug(7, 'cmssw::cmssw(): output files '+str(tmpOutFiles))
97	<	for tmp in tmpOutFiles:
98	<	tmp=string.strip(tmp)
99	<	self.output_file.append(tmp)
100	<	pass
101	<	else:
102	<	log.message("No output file defined: only stdout/err will be available")
93	>	# add fjr report by default via sandbox
94	>	self.output_file_sandbox.append(self.fjrFileName)
95	>
96	>	# other output files to be returned via sandbox or copied to SE
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 will be available")
111	<	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	<	self.additional_inbox_files.append(self.scriptExe)
114	<	if self.scriptExe != '':
115	<	if not os.path.isfile(self.scriptExe):
116	<	msg ="WARNING. file "+self.scriptExe+" not found"
117	<	raise CrabException(msg)
118	<	except KeyError:
119	<	pass
120	<
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	>
118	>	if self.datasetPath == None and self.pset == None and self.scriptExe == '' :
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	<	if not os.path.exists(tmp):
127	<	raise CrabException("Additional input file not found: "+tmp)
128	<	self.additional_inbox_files.append(string.strip(tmp))
126	>	tmp = string.strip(tmp)
127	>	dirname = ''
128	>	if not tmp[0]=="/": dirname = "."
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	>	self.additional_inbox_files.append(string.strip(file))
141		pass
142		pass
143	<	except KeyError:
144	<	pass
135	<
136	<	# files per job
137	<	try:
138	<	self.filesPerJob = int(cfg_params['CMSSW.files_per_jobs']) #Daniele
139	<	self.selectFilesPerJob = 1
140	<	except KeyError:
141	<	self.filesPerJob = 0
142	<	self.selectFilesPerJob = 0
143	>	common.logger.debug(5,"Additional input files: "+str(self.additional_inbox_files))
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	<	## source seed for pythia
161	<	try:
162	<	self.sourceSeed = int(cfg_params['CMSSW.pythia_seed'])
163	<	except KeyError:
164	<	self.sourceSeed = None
165	<	common.logger.debug(5,"No seed given")
166	<
167	<	try:
168	<	self.sourceSeedVtx = int(cfg_params['CMSSW.vtx_seed'])
169	<	except KeyError:
170	<	self.sourceSeedVtx = None
171	<	common.logger.debug(5,"No vertex seed given")
172	<
173	<	if not (self.selectFilesPerJob + self.selectEventsPerJob + self.selectNumberOfJobs == 1 ):
174	<	msg = 'Must define either files_per_jobs or events_per_job or number_of_jobs'
175	<	raise CrabException(msg)
176	<
177	<	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	<	except KeyError:
165	<	msg = 'Must define total_number_of_events'
166	<	raise CrabException(msg)
167	<
183	<	CEBlackList = []
184	<	try:
185	<	tmpBad = string.split(cfg_params['EDG.ce_black_list'],',')
186	<	for tmp in tmpBad:
187	<	tmp=string.strip(tmp)
188	<	CEBlackList.append(tmp)
189	<	except KeyError:
190	<	pass
191	<
192	<	self.reCEBlackList=[]
193	<	for bad in CEBlackList:
194	<	self.reCEBlackList.append(re.compile( bad ))
195	<
196	<	common.logger.debug(5,'CEBlackList: '+str(CEBlackList))
197	<
198	<	CEWhiteList = []
199	<	try:
200	<	tmpGood = string.split(cfg_params['EDG.ce_white_list'],',')
201	<	for tmp in tmpGood:
202	<	tmp=string.strip(tmp)
203	<	CEWhiteList.append(tmp)
204	<	except KeyError:
205	<	pass
164	>	self.selectTotalNumberEvents = 1
165	>	else:
166	>	self.total_number_of_events = 0
167	>	self.selectTotalNumberEvents = 0
168
169	<	#print 'CEWhiteList: ',CEWhiteList
170	<	self.reCEWhiteList=[]
171	<	for Good in CEWhiteList:
172	<	self.reCEWhiteList.append(re.compile( Good ))
169	>	if self.pset != None:
170	>	if ( (self.selectTotalNumberEvents + self.selectEventsPerJob + self.selectNumberOfJobs) != 2 ):
171	>	msg = 'Must define exactly two of total_number_of_events, events_per_job, or number_of_jobs.'
172	>	raise CrabException(msg)
173	>	else:
174	>	if (self.selectNumberOfJobs == 0):
175	>	msg = 'Must specify  number_of_jobs.'
176	>	raise CrabException(msg)
177	>
178	>	## New method of dealing with seeds
179	>	self.incrementSeeds = []
180	>	self.preserveSeeds = []
181	>	if cfg_params.has_key('CMSSW.preserve_seeds'):
182	>	tmpList = cfg_params['CMSSW.preserve_seeds'].split(',')
183	>	for tmp in tmpList:
184	>	tmp.strip()
185	>	self.preserveSeeds.append(tmp)
186	>	if cfg_params.has_key('CMSSW.increment_seeds'):
187	>	tmpList = cfg_params['CMSSW.increment_seeds'].split(',')
188	>	for tmp in tmpList:
189	>	tmp.strip()
190	>	self.incrementSeeds.append(tmp)
191	>
192	>	## Old method of dealing with seeds
193	>	## FUTURE: This is for old CMSSW and old CRAB. Can throw exceptions after a couple of CRAB releases and then
194	>	## remove
195	>	self.sourceSeed = cfg_params.get('CMSSW.pythia_seed',None)
196	>	if self.sourceSeed:
197	>	print "pythia_seed is a deprecated parameter. Use preserve_seeds or increment_seeds in the future.\n","Added to increment_seeds."
198	>	self.incrementSeeds.append('sourceSeed')
199	>	self.incrementSeeds.append('theSource')
200	>
201	>	self.sourceSeedVtx = cfg_params.get('CMSSW.vtx_seed',None)
202	>	if self.sourceSeedVtx:
203	>	print "vtx_seed is a deprecated parameter. Use preserve_seeds or increment_seeds in the future.\n","Added to increment_seeds."
204	>	self.incrementSeeds.append('VtxSmeared')
205	>
206	>	self.sourceSeedG4 = cfg_params.get('CMSSW.g4_seed',None)
207	>	if self.sourceSeedG4:
208	>	print "g4_seed is a deprecated parameter. Use preserve_seeds or increment_seeds in the future.\n","Added to increment_seeds."
209	>	self.incrementSeeds.append('g4SimHits')
210	>
211	>	self.sourceSeedMix = cfg_params.get('CMSSW.mix_seed',None)
212	>	if self.sourceSeedMix:
213	>	print "mix_seed is a deprecated parameter. Use preserve_seeds or increment_seeds in the future.\n","Added to increment_seeds."
214	>	self.incrementSeeds.append('mix')
215	>
216	>	self.firstRun = cfg_params.get('CMSSW.first_run',None)
217	>
218	>	if self.pset != None: #CarlosDaniele
219	>	import PsetManipulator as pp
220	>	PsetEdit = pp.PsetManipulator(self.pset) #Daniele Pset
221
222	<	common.logger.debug(5,'CEWhiteList: '+str(CEWhiteList))
222	>	# Copy/return
223
224	<	self.PsetEdit = PsetManipulator.PsetManipulator(self.pset) #Daniele Pset
224	>	self.copy_data = int(cfg_params.get('USER.copy_data',0))
225	>	self.return_data = int(cfg_params.get('USER.return_data',0))
226
227		#DBSDLS-start
228	<	## Initialize the variables that are extracted from DBS/DLS and needed in other places of the code
228	>	## Initialize the variables that are extracted from DBS/DLS and needed in other places of the code
229		self.maxEvents=0  # max events available   ( --> check the requested nb. of evts in Creator.py)
230		self.DBSPaths={}  # all dbs paths requested ( --> input to the site local discovery script)
231	+	self.jobDestination=[]  # Site destination(s) for each job (list of lists)
232		## Perform the data location and discovery (based on DBS/DLS)
233		## SL: Don't if NONE is specified as input (pythia use case)
234	<	common.analisys_common_info['sites']=None
234	>	blockSites = {}
235		if self.datasetPath:
236	<	self.DataDiscoveryAndLocation(cfg_params)
237	<	#DBSDLS-end
236	>	blockSites = self.DataDiscoveryAndLocation(cfg_params)
237	>	#DBSDLS-end
238	>
239
227	–	self.tgzNameWithPath = self.getTarBall(self.executable)
228	–
240		## Select Splitting
241	<	if self.selectNoInput: self.jobSplittingNoInput()
242	<	elif self.selectFilesPerJob or self.selectEventsPerJob or self.selectNumberOfJobs: self.jobSplittingPerFiles()
241	>	if self.selectNoInput:
242	>	if self.pset == None:
243	>	self.jobSplittingForScript()
244	>	else:
245	>	self.jobSplittingNoInput()
246		else:
247	<	msg = 'Don\'t know how to split...'
234	<	raise CrabException(msg)
247	>	self.jobSplittingByBlocks(blockSites)
248
249		# modify Pset
250	<	try:
251	<	if (self.datasetPath): # standard job
252	<	#self.PsetEdit.maxEvent(self.eventsPerJob)
253	<	# always process all events in a file
254	<	self.PsetEdit.maxEvent("-1")
255	<	self.PsetEdit.inputModule("INPUT")
256	<
257	<	else:  # pythia like job
258	<	self.PsetEdit.maxEvent(self.eventsPerJob)
259	<	if (self.sourceSeed) :
260	<	self.PsetEdit.pythiaSeed("INPUT")
248	<	if (self.sourceSeedVtx) :
249	<	self.PsetEdit.pythiaSeedVtx("INPUTVTX")
250	<	self.PsetEdit.psetWriter(self.configFilename())
251	<	except:
252	<	msg='Error while manipuliating ParameterSet: exiting...'
253	<	raise CrabException(msg)
250	>	if self.pset != None:
251	>	try:
252	>	# Add FrameworkJobReport to parameter-set, set max events.
253	>	# Reset later for data jobs by writeCFG which does all modifications
254	>	PsetEdit.addCrabFJR(self.fjrFileName) # FUTURE: Job report addition not needed by CMSSW>1.5
255	>	PsetEdit.maxEvent(self.eventsPerJob)
256	>	PsetEdit.psetWriter(self.configFilename())
257	>	except:
258	>	msg='Error while manipulating ParameterSet: exiting...'
259	>	raise CrabException(msg)
260	>	self.tgzNameWithPath = self.getTarBall(self.executable)
261
262		def DataDiscoveryAndLocation(self, cfg_params):
263
264	+	import DataDiscovery
265	+	import DataLocation
266		common.logger.debug(10,"CMSSW::DataDiscoveryAndLocation()")
267
268		datasetPath=self.datasetPath
269
261	–	## TODO
262	–	dataTiersList = ""
263	–	dataTiers = dataTiersList.split(',')
264	–
270		## Contact the DBS
271	+	common.logger.message("Contacting Data Discovery Services ...")
272		try:
273	<	self.pubdata=DataDiscovery_EDM.DataDiscovery_EDM(datasetPath, dataTiers, cfg_params)
273	>	self.pubdata=DataDiscovery.DataDiscovery(datasetPath, cfg_params)
274		self.pubdata.fetchDBSInfo()
275
276	<	except DataDiscovery_EDM.NotExistingDatasetError, ex :
276	>	except DataDiscovery.NotExistingDatasetError, ex :
277		msg = 'ERROR ***: failed Data Discovery in DBS : %s'%ex.getErrorMessage()
278		raise CrabException(msg)
279	<
274	<	except DataDiscovery_EDM.NoDataTierinProvenanceError, ex :
279	>	except DataDiscovery.NoDataTierinProvenanceError, ex :
280		msg = 'ERROR ***: failed Data Discovery in DBS : %s'%ex.getErrorMessage()
281		raise CrabException(msg)
282	<	except DataDiscovery_EDM.DataDiscoveryError, ex:
283	<	msg = 'ERROR ***: failed Data Discovery in DBS  %s'%ex.getErrorMessage()
282	>	except DataDiscovery.DataDiscoveryError, ex:
283	>	msg = 'ERROR ***: failed Data Discovery in DBS :  %s'%ex.getErrorMessage()
284		raise CrabException(msg)
285
286	<	## get list of all required data in the form of dbs paths  (dbs path = /dataset/datatier/owner)
287	<	## self.DBSPaths=self.pubdata.getDBSPaths()
288	<	common.logger.message("Required data are :"+self.datasetPath)
284	<
285	<	filesbyblock=self.pubdata.getFiles()
286	<	#        print filesbyblock
287	<	self.AllInputFiles=filesbyblock.values()
288	<	self.files = self.AllInputFiles
286	>	self.filesbyblock=self.pubdata.getFiles()
287	>	self.eventsbyblock=self.pubdata.getEventsPerBlock()
288	>	self.eventsbyfile=self.pubdata.getEventsPerFile()
289
290		## get max number of events
291	<	#common.logger.debug(10,"number of events for primary fileblocks %i"%self.pubdata.getMaxEvents())
292	<	self.maxEvents=self.pubdata.getMaxEvents() ##  self.maxEvents used in Creator.py
293	<	common.logger.message("\nThe number of available events is %s"%self.maxEvents)
291	>	self.maxEvents=self.pubdata.getMaxEvents()
292
293		## Contact the DLS and build a list of sites hosting the fileblocks
294		try:
295	<	dataloc=DataLocation_EDM.DataLocation_EDM(filesbyblock.keys(),cfg_params)
295	>	dataloc=DataLocation.DataLocation(self.filesbyblock.keys(),cfg_params)
296		dataloc.fetchDLSInfo()
297	<	except DataLocation_EDM.DataLocationError , ex:
297	>	except DataLocation.DataLocationError , ex:
298		msg = 'ERROR ***: failed Data Location in DLS \n %s '%ex.getErrorMessage()
299		raise CrabException(msg)
302	–
303	–	allsites=dataloc.getSites()
304	–	common.logger.debug(5,"sites are %s"%allsites)
305	–	sites=self.checkBlackList(allsites)
306	–	common.logger.debug(5,"sites are (after black list) %s"%sites)
307	–	sites=self.checkWhiteList(sites)
308	–	common.logger.debug(5,"sites are (after white list) %s"%sites)
300
310	–	if len(sites)==0:
311	–	msg = 'No sites hosting all the needed data! Exiting... '
312	–	raise CrabException(msg)
301
302	<	common.logger.message("List of Sites ("+str(len(sites))+") hosting the data : "+str(sites))
303	<	common.logger.debug(6, "List of Sites: "+str(sites))
304	<	common.analisys_common_info['sites']=sites    ## used in SchedulerEdg.py in createSchScript
305	<	self.setParam_('TargetCE', ','.join(sites))
306	<	return
307	<
308	<	def jobSplittingPerFiles(self):
321	<	"""
322	<	Perform job splitting based on number of files to be accessed per job
323	<	"""
324	<	common.logger.debug(5,'Splitting per input files')
325	<	common.logger.message('Required '+str(self.total_number_of_events)+' events in total ')
326	<	common.logger.message('Available '+str(self.maxEvents)+' events in total ')
327	<	common.logger.message('Required '+str(self.filesPerJob)+' files per job ')
328	<	common.logger.message('Required '+str(self.theNumberOfJobs)+' jobs in total ')
329	<	common.logger.message('Required '+str(self.eventsPerJob)+' events per job')
302	>	sites = dataloc.getSites()
303	>	allSites = []
304	>	listSites = sites.values()
305	>	for listSite in listSites:
306	>	for oneSite in listSite:
307	>	allSites.append(oneSite)
308	>	allSites = self.uniquelist(allSites)
309
310	<	## if asked to process all events, do it
311	<	if self.total_number_of_events == -1:
312	<	self.total_number_of_events=self.maxEvents
310	>	# screen output
311	>	common.logger.message("Requested dataset: " + datasetPath + " has " + str(self.maxEvents) + " events in " + str(len(self.filesbyblock.keys())) + " blocks.\n")
312	>
313	>	return sites
314	>
315	>	def jobSplittingByBlocks(self, blockSites):
316	>	"""
317	>	Perform job splitting. Jobs run over an integer number of files
318	>	and no more than one block.
319	>	ARGUMENT: blockSites: dictionary with blocks as keys and list of host sites as values
320	>	REQUIRES: self.selectTotalNumberEvents, self.selectEventsPerJob, self.selectNumberofJobs,
321	>	self.total_number_of_events, self.eventsPerJob, self.theNumberOfJobs,
322	>	self.maxEvents, self.filesbyblock
323	>	SETS: self.jobDestination - Site destination(s) for each job (a list of lists)
324	>	self.total_number_of_jobs - Total # of jobs
325	>	self.list_of_args - File(s) job will run on (a list of lists)
326	>	"""
327	>
328	>	# ---- Handle the possible job splitting configurations ---- #
329	>	if (self.selectTotalNumberEvents):
330	>	totalEventsRequested = self.total_number_of_events
331	>	if (self.selectEventsPerJob):
332	>	eventsPerJobRequested = self.eventsPerJob
333	>	if (self.selectNumberOfJobs):
334	>	totalEventsRequested = self.theNumberOfJobs * self.eventsPerJob
335	>
336	>	# If user requested all the events in the dataset
337	>	if (totalEventsRequested == -1):
338	>	eventsRemaining=self.maxEvents
339	>	# If user requested more events than are in the dataset
340	>	elif (totalEventsRequested > self.maxEvents):
341	>	eventsRemaining = self.maxEvents
342	>	common.logger.message("Requested "+str(self.total_number_of_events)+ " events, but only "+str(self.maxEvents)+" events are available.")
343	>	# If user requested less events than are in the dataset
344		else:
345	<	if self.total_number_of_events>self.maxEvents:
336	<	common.logger.message("Asked "+str(self.total_number_of_events)+" but only "+str(self.maxEvents)+" available.")
337	<	self.total_number_of_events=self.maxEvents
338	<	pass
345	>	eventsRemaining = totalEventsRequested
346
347	<	## TODO: SL need to have (from DBS) a detailed list of how many events per each file
348	<	n_tot_files = (len(self.files[0]))
349	<	## SL: this is wrong if the files have different number of events
343	<	evPerFile = int(self.maxEvents)/n_tot_files
344	<
345	<	common.logger.debug(5,'Events per File '+str(evPerFile))
346	<
347	<	## compute job splitting parameters: filesPerJob, eventsPerJob and theNumberOfJobs
348	<	if self.selectFilesPerJob:
349	<	## user define files per event.
350	<	filesPerJob = self.filesPerJob
351	<	eventsPerJob = filesPerJob*evPerFile
352	<	theNumberOfJobs = int(self.total_number_of_events*1./eventsPerJob)
353	<	check = int(self.total_number_of_events) - (theNumberOfJobs*eventsPerJob)
354	<	if check > 0:
355	<	theNumberOfJobs +=1
356	<	filesLastJob = int(check*1./evPerFile+0.5)
357	<	common.logger.message('Warning: last job will be created with '+str(check)+' files')
358	<	else:
359	<	filesLastJob = filesPerJob
347	>	# If user requested more events per job than are in the dataset
348	>	if (self.selectEventsPerJob and eventsPerJobRequested > self.maxEvents):
349	>	eventsPerJobRequested = self.maxEvents
350
351	<	elif self.selectNumberOfJobs:
352	<	## User select the number of jobs: last might be bigger to match request of events
363	<	theNumberOfJobs =  self.theNumberOfJobs
364	<
365	<	eventsPerJob = self.total_number_of_events/theNumberOfJobs
366	<	filesPerJob = int(eventsPerJob/evPerFile)
367	<	if (filesPerJob==0) : filesPerJob=1
368	<	check = int(self.total_number_of_events) - (int(theNumberOfJobs)*filesPerJob*evPerFile)
369	<	if not check == 0:
370	<	if check<0:
371	<	missingFiles = int(check/evPerFile)
372	<	additionalJobs = int(missingFiles/filesPerJob)
373	<	#print missingFiles, additionalJobs
374	<	theNumberOfJobs+=additionalJobs
375	<	common.logger.message('Warning: will create only '+str(theNumberOfJobs)+' jobs')
376	<	check = int(self.total_number_of_events) - (int(theNumberOfJobs)*filesPerJob*evPerFile)
377	<
378	<	if check >0 :
379	<	filesLastJob = filesPerJob+int(check*1./evPerFile+0.5)
380	<	common.logger.message('Warning: last job will be created with '+str(filesLastJob*evPerFile)+' events')
381	<	else:
382	<	filesLastJob = filesPerJob
383	<	else:
384	<	filesLastJob = filesPerJob
385	<	elif self.selectEventsPerJob:
386	<	# SL case if asked events per job
387	<	## estimate the number of files per job to match the user requirement
388	<	filesPerJob = int(float(self.eventsPerJob)/float(evPerFile))
389	<	if filesPerJob==0: filesPerJob=1
390	<	common.logger.debug(5,"filesPerJob "+str(filesPerJob))
391	<	if (filesPerJob==0): filesPerJob=1
392	<	eventsPerJob=filesPerJob*evPerFile
393	<	theNumberOfJobs = int(self.total_number_of_events)/int(eventsPerJob)
394	<	check = int(self.total_number_of_events) - (int(theNumberOfJobs)*eventsPerJob)
395	<	if not check == 0:
396	<	missingFiles = int(check/evPerFile)
397	<	additionalJobs = int(missingFiles/filesPerJob)
398	<	if ( additionalJobs>0) : theNumberOfJobs+=additionalJobs
399	<	check = int(self.total_number_of_events) - (int(theNumberOfJobs)*eventsPerJob)
400	<	if not check == 0:
401	<	if (check <0 ):
402	<	filesLastJob = filesPerJob+int(check*1./evPerFile-0.5)
403	<	else:
404	<	theNumberOfJobs+=1
405	<	filesLastJob = int(check*1./evPerFile+0.5)
351	>	# For user info at end
352	>	totalEventCount = 0
353
354	<	common.logger.message('Warning: last job will be created with '+str(filesLastJob*evPerFile)+' events')
355	<	else:
409	<	filesLastJob = filesPerJob
410	<	else:
411	<	filesLastJob = filesPerJob
412	<
413	<	self.total_number_of_jobs = theNumberOfJobs
354	>	if (self.selectTotalNumberEvents and self.selectNumberOfJobs):
355	>	eventsPerJobRequested = int(eventsRemaining/self.theNumberOfJobs)
356
357	<	totalEventsToBeUsed=theNumberOfJobs*filesPerJob*evPerFile
358	<	if not check == 0:
417	<	#    print (theNumberOfJobs-1)*filesPerJob*evPerFile,filesLastJob*evPerFile
418	<	totalEventsToBeUsed=(theNumberOfJobs-1)*filesPerJob*evPerFile+filesLastJob*evPerFile
357	>	if (self.selectNumberOfJobs):
358	>	common.logger.message("May not create the exact number_of_jobs requested.")
359
360	<	common.logger.message(str(self.total_number_of_jobs)+' jobs will be created, each for '+str(filesPerJob*evPerFile)+' events, for a total of '+str(totalEventsToBeUsed)+' events')
360	>	if ( self.ncjobs == 'all' ) :
361	>	totalNumberOfJobs = 999999999
362	>	else :
363	>	totalNumberOfJobs = self.ncjobs
364
365	<	totalFilesToBeUsed=filesPerJob*(theNumberOfJobs-1)+filesLastJob
365	>	blocks = blockSites.keys()
366	>	blockCount = 0
367	>	# Backup variable in case self.maxEvents counted events in a non-included block
368	>	numBlocksInDataset = len(blocks)
369
370	<	## set job arguments (files)
370	>	jobCount = 0
371		list_of_lists = []
426	–	lastFile=0
427	–	for i in range(0, int(totalFilesToBeUsed), filesPerJob)[:-1]:
428	–	parString = "\\{"
429	–
430	–	lastFile=i+filesPerJob
431	–	params = self.files[0][i: lastFile]
432	–	for i in range(len(params) - 1):
433	–	parString += '\\\"' + params[i] + '\\\"\,'
434	–
435	–	parString += '\\\"' + params[len(params) - 1] + '\\\"\\}'
436	–	list_of_lists.append([parString])
437	–	pass
372
373	<	## last job
374	<	parString = "\\{"
375	<
376	<	params = self.files[0][lastFile: lastFile+filesLastJob]
377	<	for i in range(len(params) - 1):
378	<	parString += '\\\"' + params[i] + '\\\"\,'
379	<
380	<	parString += '\\\"' + params[len(params) - 1] + '\\\"\\}'
381	<	list_of_lists.append([parString])
382	<	pass
373	>	# list tracking which jobs are in which jobs belong to which block
374	>	jobsOfBlock = {}
375	>
376	>	# ---- Iterate over the blocks in the dataset until ---- #
377	>	# ---- we've met the requested total # of events    ---- #
378	>	while ( (eventsRemaining > 0) and (blockCount < numBlocksInDataset) and (jobCount < totalNumberOfJobs)):
379	>	block = blocks[blockCount]
380	>	blockCount += 1
381	>	if block not in jobsOfBlock.keys() :
382	>	jobsOfBlock[block] = []
383	>
384	>	if self.eventsbyblock.has_key(block) :
385	>	numEventsInBlock = self.eventsbyblock[block]
386	>	common.logger.debug(5,'Events in Block File '+str(numEventsInBlock))
387	>
388	>	files = self.filesbyblock[block]
389	>	numFilesInBlock = len(files)
390	>	if (numFilesInBlock <= 0):
391	>	continue
392	>	fileCount = 0
393	>
394	>	# ---- New block => New job ---- #
395	>	parString = ""
396	>	# counter for number of events in files currently worked on
397	>	filesEventCount = 0
398	>	# flag if next while loop should touch new file
399	>	newFile = 1
400	>	# job event counter
401	>	jobSkipEventCount = 0
402	>
403	>	# ---- Iterate over the files in the block until we've met the requested ---- #
404	>	# ---- total # of events or we've gone over all the files in this block  ---- #
405	>	while ( (eventsRemaining > 0) and (fileCount < numFilesInBlock) and (jobCount < totalNumberOfJobs) ):
406	>	file = files[fileCount]
407	>	if newFile :
408	>	try:
409	>	numEventsInFile = self.eventsbyfile[file]
410	>	common.logger.debug(6, "File "+str(file)+" has "+str(numEventsInFile)+" events")
411	>	# increase filesEventCount
412	>	filesEventCount += numEventsInFile
413	>	# Add file to current job
414	>	parString += '\\\"' + file + '\\\"\,'
415	>	newFile = 0
416	>	except KeyError:
417	>	common.logger.message("File "+str(file)+" has unknown number of events: skipping")
418	>
419	>	eventsPerJobRequested = min(eventsPerJobRequested, eventsRemaining)
420	>	# if less events in file remain than eventsPerJobRequested
421	>	if ( filesEventCount - jobSkipEventCount < eventsPerJobRequested):
422	>	# if last file in block
423	>	if ( fileCount == numFilesInBlock-1 ) :
424	>	# end job using last file, use remaining events in block
425	>	# close job and touch new file
426	>	fullString = parString[:-2]
427	>	list_of_lists.append([fullString,str(-1),str(jobSkipEventCount)])
428	>	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(filesEventCount - jobSkipEventCount)+" events (last file in block).")
429	>	self.jobDestination.append(blockSites[block])
430	>	common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
431	>	# fill jobs of block dictionary
432	>	jobsOfBlock[block].append(jobCount+1)
433	>	# reset counter
434	>	jobCount = jobCount + 1
435	>	totalEventCount = totalEventCount + filesEventCount - jobSkipEventCount
436	>	eventsRemaining = eventsRemaining - filesEventCount + jobSkipEventCount
437	>	jobSkipEventCount = 0
438	>	# reset file
439	>	parString = ""
440	>	filesEventCount = 0
441	>	newFile = 1
442	>	fileCount += 1
443	>	else :
444	>	# go to next file
445	>	newFile = 1
446	>	fileCount += 1
447	>	# if events in file equal to eventsPerJobRequested
448	>	elif ( filesEventCount - jobSkipEventCount == eventsPerJobRequested ) :
449	>	# close job and touch new file
450	>	fullString = parString[:-2]
451	>	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
452	>	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.")
453	>	self.jobDestination.append(blockSites[block])
454	>	common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
455	>	jobsOfBlock[block].append(jobCount+1)
456	>	# reset counter
457	>	jobCount = jobCount + 1
458	>	totalEventCount = totalEventCount + eventsPerJobRequested
459	>	eventsRemaining = eventsRemaining - eventsPerJobRequested
460	>	jobSkipEventCount = 0
461	>	# reset file
462	>	parString = ""
463	>	filesEventCount = 0
464	>	newFile = 1
465	>	fileCount += 1
466	>
467	>	# if more events in file remain than eventsPerJobRequested
468	>	else :
469	>	# close job but don't touch new file
470	>	fullString = parString[:-2]
471	>	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
472	>	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.")
473	>	self.jobDestination.append(blockSites[block])
474	>	common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
475	>	jobsOfBlock[block].append(jobCount+1)
476	>	# increase counter
477	>	jobCount = jobCount + 1
478	>	totalEventCount = totalEventCount + eventsPerJobRequested
479	>	eventsRemaining = eventsRemaining - eventsPerJobRequested
480	>	# calculate skip events for last file
481	>	# use filesEventCount (contains several files), jobSkipEventCount and eventsPerJobRequest
482	>	jobSkipEventCount = eventsPerJobRequested - (filesEventCount - jobSkipEventCount - self.eventsbyfile[file])
483	>	# remove all but the last file
484	>	filesEventCount = self.eventsbyfile[file]
485	>	parString = '\\\"' + file + '\\\"\,'
486	>	pass # END if
487	>	pass # END while (iterate over files in the block)
488	>	pass # END while (iterate over blocks in the dataset)
489	>	self.ncjobs = self.total_number_of_jobs = jobCount
490	>	if (eventsRemaining > 0 and jobCount < totalNumberOfJobs ):
491	>	common.logger.message("Could not run on all requested events because some blocks not hosted at allowed sites.")
492	>	common.logger.message(str(jobCount)+" job(s) can run on "+str(totalEventCount)+" events.\n")
493	>
494	>	# screen output
495	>	screenOutput = "List of jobs and available destination sites:\n\n"
496	>
497	>	# keep trace of block with no sites to print a warning at the end
498	>	noSiteBlock = []
499	>	bloskNoSite = []
500	>
501	>	blockCounter = 0
502	>	for block in blocks:
503	>	if block in jobsOfBlock.keys() :
504	>	blockCounter += 1
505	>	screenOutput += "Block %5i: jobs %20s: sites: %s\n" % (blockCounter,spanRanges(jobsOfBlock[block]),
506	>	','.join(self.blackWhiteListParser.checkWhiteList(self.blackWhiteListParser.checkBlackList(blockSites[block],block),block)))
507	>	if len(self.blackWhiteListParser.checkWhiteList(self.blackWhiteListParser.checkBlackList(blockSites[block],block),block)) == 0:
508	>	noSiteBlock.append( spanRanges(jobsOfBlock[block]) )
509	>	bloskNoSite.append( blockCounter )
510	>
511	>	common.logger.message(screenOutput)
512	>	if len(noSiteBlock) > 0 and len(bloskNoSite) > 0:
513	>	msg = 'WARNING: No sites are hosting any part of data for block:\n                '
514	>	virgola = ""
515	>	if len(bloskNoSite) > 1:
516	>	virgola = ","
517	>	for block in bloskNoSite:
518	>	msg += ' ' + str(block) + virgola
519	>	msg += '\n               Related jobs:\n                 '
520	>	virgola = ""
521	>	if len(noSiteBlock) > 1:
522	>	virgola = ","
523	>	for range_jobs in noSiteBlock:
524	>	msg += str(range_jobs) + virgola
525	>	msg += '\n               will not be submitted and this block of data can not be analyzed!\n'
526	>	if self.cfg_params.has_key('EDG.se_white_list'):
527	>	msg += 'WARNING: SE White List: '+self.cfg_params['EDG.se_white_list']+'\n'
528	>	msg += '(Hint: By whitelisting you force the job to run at this particular site(s).\n'
529	>	msg += 'Please check if the dataset is available at this site!)\n'
530	>	if self.cfg_params.has_key('EDG.ce_white_list'):
531	>	msg += 'WARNING: CE White List: '+self.cfg_params['EDG.ce_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	>
535	>	common.logger.message(msg)
536
537		self.list_of_args = list_of_lists
451	–	# print self.list_of_args[0]
538		return
539
540		def jobSplittingNoInput(self):
#	Line 456 | Line 542 | class Cmssw(JobType):
542		Perform job splitting based on number of event per job
543		"""
544		common.logger.debug(5,'Splitting per events')
545	<	common.logger.message('Required '+str(self.eventsPerJob)+' events per job ')
546	<	common.logger.message('Required '+str(self.theNumberOfJobs)+' jobs in total ')
547	<	common.logger.message('Required '+str(self.total_number_of_events)+' events in total ')
545	>
546	>	if (self.selectEventsPerJob):
547	>	common.logger.message('Required '+str(self.eventsPerJob)+' events per job ')
548	>	if (self.selectNumberOfJobs):
549	>	common.logger.message('Required '+str(self.theNumberOfJobs)+' jobs in total ')
550	>	if (self.selectTotalNumberEvents):
551	>	common.logger.message('Required '+str(self.total_number_of_events)+' events in total ')
552
553		if (self.total_number_of_events < 0):
554		msg='Cannot split jobs per Events with "-1" as total number of events'
555		raise CrabException(msg)
556
557		if (self.selectEventsPerJob):
558	<	self.total_number_of_jobs = int(self.total_number_of_events/self.eventsPerJob)
558	>	if (self.selectTotalNumberEvents):
559	>	self.total_number_of_jobs = int(self.total_number_of_events/self.eventsPerJob)
560	>	elif(self.selectNumberOfJobs) :
561	>	self.total_number_of_jobs =self.theNumberOfJobs
562	>	self.total_number_of_events =int(self.theNumberOfJobs*self.eventsPerJob)
563	>
564		elif (self.selectNumberOfJobs) :
565		self.total_number_of_jobs = self.theNumberOfJobs
566		self.eventsPerJob = int(self.total_number_of_events/self.total_number_of_jobs)
#	Line 477 | Line 572 | class Cmssw(JobType):
572
573		common.logger.debug(5,'Check  '+str(check))
574
575	<	common.logger.message(str(self.total_number_of_jobs)+' jobs will be created, each for '+str(self.eventsPerJob)+' for a total of '+str(self.total_number_of_jobs*self.eventsPerJob)+' events')
575	>	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')
576		if check > 0:
577	<	common.logger.message('Warning: asked '+str(self.total_number_of_events)+' but will do only '+str(int(self.total_number_of_jobs)*self.eventsPerJob))
483	<
577	>	common.logger.message('Warning: asked '+str(self.total_number_of_events)+' but can do only '+str(int(self.total_number_of_jobs)*self.eventsPerJob))
578
579		# argument is seed number.$i
580		self.list_of_args = []
581		for i in range(self.total_number_of_jobs):
582	<	if (self.sourceSeed):
583	<	if (self.sourceSeedVtx):
584	<	## pythia + vtx random seed
585	<	self.list_of_args.append([
586	<	str(self.sourceSeed)+str(i),
587	<	str(self.sourceSeedVtx)+str(i)
588	<	])
495	<	else:
496	<	## only pythia random seed
497	<	self.list_of_args.append([(str(self.sourceSeed)+str(i))])
498	<	else:
499	<	## no random seed
500	<	self.list_of_args.append([str(i)])
501	<	#print self.list_of_args
582	>	## Since there is no input, any site is good
583	>	self.jobDestination.append([""]) #must be empty to write correctly the xml
584	>	args=[]
585	>	if (self.firstRun):
586	>	## pythia first run
587	>	args.append(str(self.firstRun)+str(i))
588	>	self.list_of_args.append(args)
589
590		return
591
592	+
593	+	def jobSplittingForScript(self):
594	+	"""
595	+	Perform job splitting based on number of job
596	+	"""
597	+	common.logger.debug(5,'Splitting per job')
598	+	common.logger.message('Required '+str(self.theNumberOfJobs)+' jobs in total ')
599	+
600	+	self.total_number_of_jobs = self.theNumberOfJobs
601	+
602	+	common.logger.debug(5,'N jobs  '+str(self.total_number_of_jobs))
603	+
604	+	common.logger.message(str(self.total_number_of_jobs)+' jobs can be created')
605	+
606	+	# argument is seed number.$i
607	+	self.list_of_args = []
608	+	for i in range(self.total_number_of_jobs):
609	+	self.jobDestination.append([""])
610	+	self.list_of_args.append([str(i)])
611	+	return
612	+
613		def split(self, jobParams):
614	<
507	<	common.jobDB.load()
508	<	#### Fabio
614	>
615		njobs = self.total_number_of_jobs
616		arglist = self.list_of_args
617		# create the empty structure
618		for i in range(njobs):
619		jobParams.append("")
620	<
620	>
621	>	listID=[]
622	>	listField=[]
623		for job in range(njobs):
624		jobParams[job] = arglist[job]
625	<	# print str(arglist[job])
626	<	# print jobParams[job]
627	<	common.jobDB.setArguments(job, jobParams[job])
625	>	listID.append(job+1)
626	>	job_ToSave ={}
627	>	concString = ' '
628	>	argu=''
629	>	if len(jobParams[job]):
630	>	argu +=   concString.join(jobParams[job] )
631	>	job_ToSave['arguments']= str(job+1)+' '+argu
632	>	job_ToSave['dlsDestination']= self.jobDestination[job]
633	>	listField.append(job_ToSave)
634	>	msg="Job "+str(job)+" Arguments:   "+str(job+1)+" "+argu+"\n"  \
635	>	+"                     Destination: "+str(self.jobDestination[job])
636	>	common.logger.debug(5,msg)
637	>	common._db.updateJob_(listID,listField)
638	>	self.argsList = (len(jobParams[0])+1)
639
521	–	common.jobDB.save()
640		return
641	<
524	<	def getJobTypeArguments(self, nj, sched):
525	<	result = ''
526	<	for i in common.jobDB.arguments(nj):
527	<	result=result+str(i)+" "
528	<	return result
529	<
641	>
642		def numberOfJobs(self):
531	–	# Fabio
643		return self.total_number_of_jobs
644
534	–	def checkBlackList(self, allSites):
535	–	if len(self.reCEBlackList)==0: return allSites
536	–	sites = []
537	–	for site in allSites:
538	–	common.logger.debug(10,'Site '+site)
539	–	good=1
540	–	for re in self.reCEBlackList:
541	–	if re.search(site):
542	–	common.logger.message('CE in black list, skipping site '+site)
543	–	good=0
544	–	pass
545	–	if good: sites.append(site)
546	–	if len(sites) == 0:
547	–	common.logger.debug(3,"No sites found after BlackList")
548	–	return sites
549	–
550	–	def checkWhiteList(self, allSites):
551	–
552	–	if len(self.reCEWhiteList)==0: return allSites
553	–	sites = []
554	–	for site in allSites:
555	–	good=0
556	–	for re in self.reCEWhiteList:
557	–	if re.search(site):
558	–	common.logger.debug(5,'CE in white list, adding site '+site)
559	–	good=1
560	–	if not good: continue
561	–	sites.append(site)
562	–	if len(sites) == 0:
563	–	common.logger.message("No sites found after WhiteList\n")
564	–	else:
565	–	common.logger.debug(5,"Selected sites via WhiteList are "+str(sites)+"\n")
566	–	return sites
567	–
645		def getTarBall(self, exe):
646		"""
647		Return the TarBall with lib and exe
648		"""
649	<
573	<	# if it exist, just return it
574	<	self.tgzNameWithPath = common.work_space.shareDir()+self.tgz_name
649	>	self.tgzNameWithPath = common.work_space.pathForTgz()+'share/'+self.tgz_name
650		if os.path.exists(self.tgzNameWithPath):
651		return self.tgzNameWithPath
652
#	Line 584 | Line 659 | class Cmssw(JobType):
659
660		# First of all declare the user Scram area
661		swArea = self.scram.getSWArea_()
587	–	#print "swArea = ", swArea
588	–	swVersion = self.scram.getSWVersion()
589	–	#print "swVersion = ", swVersion
662		swReleaseTop = self.scram.getReleaseTop_()
663	<	#print "swReleaseTop = ", swReleaseTop
592	<
663	>
664		## check if working area is release top
665		if swReleaseTop == '' or swArea == swReleaseTop:
666	+	common.logger.debug(3,"swArea = "+swArea+" swReleaseTop ="+swReleaseTop)
667		return
668
669	<	filesToBeTarred = []
670	<	## First find the executable
671	<	if (self.executable != ''):
672	<	exeWithPath = self.scram.findFile_(executable)
673	<	#           print exeWithPath
674	<	if ( not exeWithPath ):
675	<	raise CrabException('User executable '+executable+' not found')
676	<
677	<	## then check if it's private or not
678	<	if exeWithPath.find(swReleaseTop) == -1:
679	<	# the exe is private, so we must ship
680	<	common.logger.debug(5,"Exe "+exeWithPath+" to be tarred")
681	<	path = swArea+'/'
682	<	exe = string.replace(exeWithPath, path,'')
683	<	filesToBeTarred.append(exe)
684	<	pass
685	<	else:
686	<	# the exe is from release, we'll find it on WN
687	<	pass
688	<
689	<	## Now get the libraries: only those in local working area
690	<	libDir = 'lib'
691	<	lib = swArea+'/' +libDir
692	<	common.logger.debug(5,"lib "+lib+" to be tarred")
693	<	if os.path.exists(lib):
694	<	filesToBeTarred.append(libDir)
695	<
696	<	## Now check if module dir is present
697	<	moduleDir = 'module'
698	<	if os.path.isdir(swArea+'/'+moduleDir):
699	<	filesToBeTarred.append(moduleDir)
700	<
701	<	## Now check if the Data dir is present
702	<	dataDir = 'src/Data/'
703	<	if os.path.isdir(swArea+'/'+dataDir):
704	<	filesToBeTarred.append(dataDir)
705	<
706	<	## Create the tar-ball
707	<	if len(filesToBeTarred)>0:
708	<	cwd = os.getcwd()
709	<	os.chdir(swArea)
710	<	tarcmd = 'tar zcvf ' + self.tgzNameWithPath + ' '
711	<	for line in filesToBeTarred:
712	<	tarcmd = tarcmd + line + ' '
713	<	cout = runCommand(tarcmd)
714	<	if not cout:
715	<	raise CrabException('Could not create tar-ball')
716	<	os.chdir(cwd)
717	<	else:
718	<	common.logger.debug(5,"No files to be to be tarred")
719	<
720	<	return
721	<
722	<	def wsSetupEnvironment(self, nj):
669	>	import tarfile
670	>	try: # create tar ball
671	>	tar = tarfile.open(self.tgzNameWithPath, "w:gz")
672	>	## First find the executable
673	>	if (self.executable != ''):
674	>	exeWithPath = self.scram.findFile_(executable)
675	>	if ( not exeWithPath ):
676	>	raise CrabException('User executable '+executable+' not found')
677	>
678	>	## then check if it's private or not
679	>	if exeWithPath.find(swReleaseTop) == -1:
680	>	# the exe is private, so we must ship
681	>	common.logger.debug(5,"Exe "+exeWithPath+" to be tarred")
682	>	path = swArea+'/'
683	>	# distinguish case when script is in user project area or given by full path somewhere else
684	>	if exeWithPath.find(path) >= 0 :
685	>	exe = string.replace(exeWithPath, path,'')
686	>	tar.add(path+exe,exe)
687	>	else :
688	>	tar.add(exeWithPath,os.path.basename(executable))
689	>	pass
690	>	else:
691	>	# the exe is from release, we'll find it on WN
692	>	pass
693	>
694	>	## Now get the libraries: only those in local working area
695	>	libDir = 'lib'
696	>	lib = swArea+'/' +libDir
697	>	common.logger.debug(5,"lib "+lib+" to be tarred")
698	>	if os.path.exists(lib):
699	>	tar.add(lib,libDir)
700	>
701	>	## Now check if module dir is present
702	>	moduleDir = 'module'
703	>	module = swArea + '/' + moduleDir
704	>	if os.path.isdir(module):
705	>	tar.add(module,moduleDir)
706	>
707	>	## Now check if any data dir(s) is present
708	>	swAreaLen=len(swArea)
709	>	self.dataExist = False
710	>	for root, dirs, files in os.walk(swArea):
711	>	if "data" in dirs:
712	>	self.dataExist=True
713	>	common.logger.debug(5,"data "+root+"/data"+" to be tarred")
714	>	tar.add(root+"/data",root[swAreaLen:]+"/data")
715	>
716	>	### CMSSW ParameterSet
717	>	if not self.pset is None:
718	>	cfg_file = common.work_space.jobDir()+self.configFilename()
719	>	tar.add(cfg_file,self.configFilename())
720	>	common.logger.debug(5,"File added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
721	>
722	>
723	>	## Add ProdCommon dir to tar
724	>	prodcommonDir = 'ProdCommon'
725	>	prodcommonPath = os.environ['CRABDIR'] + '/' + 'ProdCommon'
726	>	if os.path.isdir(prodcommonPath):
727	>	tar.add(prodcommonPath,prodcommonDir)
728	>	common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
729	>
730	>	##### ML stuff
731	>	ML_file_list=['report.py', 'DashboardAPI.py', 'Logger.py', 'ProcInfo.py', 'apmon.py']
732	>	path=os.environ['CRABDIR'] + '/python/'
733	>	for file in ML_file_list:
734	>	tar.add(path+file,file)
735	>	common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
736	>
737	>	##### Utils
738	>	Utils_file_list=['parseCrabFjr.py','writeCfg.py', 'JobReportErrorCode.py']
739	>	for file in Utils_file_list:
740	>	tar.add(path+file,file)
741	>	common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
742	>
743	>	##### AdditionalFiles
744	>	for file in self.additional_inbox_files:
745	>	tar.add(file,string.split(file,'/')[-1])
746	>	common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
747	>
748	>	tar.close()
749	>	except :
750	>	raise CrabException('Could not create tar-ball')
751	>
752	>	## check for tarball size
753	>	tarballinfo = os.stat(self.tgzNameWithPath)
754	>	if ( tarballinfo.st_size > self.MaxTarBallSize*1024*1024 ) :
755	>	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.')
756	>
757	>	## create tar-ball with ML stuff
758	>
759	>	def wsSetupEnvironment(self, nj=0):
760		"""
761		Returns part of a job script which prepares
762		the execution environment for the job 'nj'.
763		"""
764	+	if (self.CMSSW_major >= 2 and self.CMSSW_minor >= 1) or (self.CMSSW_major >= 3):
765	+	psetName = 'pset.py'
766	+	else:
767	+	psetName = 'pset.cfg'
768		# Prepare JobType-independent part
769	<	txt = ''
770	<
771	<	## OLI_Daniele at this level  middleware already known
659	<
660	<	txt += 'if [ $middleware == LCG ]; then \n'
769	>	txt = '\n#Written by cms_cmssw::wsSetupEnvironment\n'
770	>	txt += 'echo ">>> setup environment"\n'
771	>	txt += 'if [ $middleware == LCG ]; then \n'
772		txt += self.wsSetupCMSLCGEnvironment_()
773		txt += 'elif [ $middleware == OSG ]; then\n'
774	<	txt += '    time=`date -u +"%s"`\n'
775	<	txt += '    WORKING_DIR=$OSG_WN_TMP/cms_$time\n'
776	<	txt += '    echo "Creating working directory: $WORKING_DIR"\n'
777	<	txt += '    /bin/mkdir -p $WORKING_DIR\n'
778	<	txt += '    if [ ! -d $WORKING_DIR ] ;then\n'
668	<	txt += '        echo "SET_CMS_ENV 10016 ==> OSG $WORKING_DIR could not be created on WN `hostname`"\n'
669	<	txt += '        echo "JOB_EXIT_STATUS = 10016"\n'
670	<	txt += '        echo "JobExitCode=10016" | tee -a $RUNTIME_AREA/$repo\n'
671	<	txt += '        dumpStatus $RUNTIME_AREA/$repo\n'
672	<	txt += '        rm -f $RUNTIME_AREA/$repo \n'
673	<	txt += '        echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
674	<	txt += '        echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
675	<	txt += '        exit 1\n'
774	>	txt += '    WORKING_DIR=`/bin/mktemp  -d $OSG_WN_TMP/cms_XXXXXXXXXXXX`\n'
775	>	txt += '    if [ ! $? == 0 ] ;then\n'
776	>	txt += '        echo "ERROR ==> OSG $WORKING_DIR could not be created on WN `hostname`"\n'
777	>	txt += '        job_exit_code=10016\n'
778	>	txt += '        func_exit\n'
779		txt += '    fi\n'
780	+	txt += '    echo ">>> Created working directory: $WORKING_DIR"\n'
781		txt += '\n'
782		txt += '    echo "Change to working directory: $WORKING_DIR"\n'
783		txt += '    cd $WORKING_DIR\n'
784	<	txt += self.wsSetupCMSOSGEnvironment_()
784	>	txt += '    echo ">>> current directory (WORKING_DIR): $WORKING_DIR"\n'
785	>	txt += self.wsSetupCMSOSGEnvironment_()
786		txt += 'fi\n'
787
788		# Prepare JobType-specific part
789		scram = self.scram.commandName()
790		txt += '\n\n'
791	<	txt += 'echo "### SPECIFIC JOB SETUP ENVIRONMENT ###"\n'
791	>	txt += 'echo ">>> specific cmssw setup environment:"\n'
792	>	txt += 'echo "CMSSW_VERSION =  '+self.version+'"\n'
793		txt += scram+' project CMSSW '+self.version+'\n'
794		txt += 'status=$?\n'
795		txt += 'if [ $status != 0 ] ; then\n'
796	<	txt += '   echo "SET_EXE_ENV 10034 ==>ERROR CMSSW '+self.version+' not found on `hostname`" \n'
797	<	txt += '   echo "JOB_EXIT_STATUS = 10034"\n'
798	<	txt += '   echo "JobExitCode=10034" | tee -a $RUNTIME_AREA/$repo\n'
693	<	txt += '   dumpStatus $RUNTIME_AREA/$repo\n'
694	<	txt += '   rm -f $RUNTIME_AREA/$repo \n'
695	<	txt += '   echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
696	<	txt += '   echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
697	<	## OLI_Daniele
698	<	txt += '    if [ $middleware == OSG ]; then \n'
699	<	txt += '        echo "Remove working directory: $WORKING_DIR"\n'
700	<	txt += '        cd $RUNTIME_AREA\n'
701	<	txt += '        /bin/rm -rf $WORKING_DIR\n'
702	<	txt += '        if [ -d $WORKING_DIR ] ;then\n'
703	<	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'
704	<	txt += '            echo "JOB_EXIT_STATUS = 10018"\n'
705	<	txt += '            echo "JobExitCode=10018" | tee -a $RUNTIME_AREA/$repo\n'
706	<	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
707	<	txt += '            rm -f $RUNTIME_AREA/$repo \n'
708	<	txt += '            echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
709	<	txt += '            echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
710	<	txt += '        fi\n'
711	<	txt += '    fi \n'
712	<	txt += '   exit 1 \n'
796	>	txt += '    echo "ERROR ==> CMSSW '+self.version+' not found on `hostname`" \n'
797	>	txt += '    job_exit_code=10034\n'
798	>	txt += '    func_exit\n'
799		txt += 'fi \n'
714	–	txt += 'echo "CMSSW_VERSION =  '+self.version+'"\n'
800		txt += 'cd '+self.version+'\n'
801	<	### needed grep for bug in scramv1 ###
801	>	txt += 'SOFTWARE_DIR=`pwd`\n'
802	>	txt += 'echo ">>> current directory (SOFTWARE_DIR): $SOFTWARE_DIR" \n'
803		txt += 'eval `'+scram+' runtime -sh | grep -v SCRAMRT_LSB_JOBNAME`\n'
804	<
804	>	txt += 'if [ $? != 0 ] ; then\n'
805	>	txt += '    echo "ERROR ==> Problem with the command: "\n'
806	>	txt += '    echo "eval \`'+scram+' runtime -sh | grep -v SCRAMRT_LSB_JOBNAME \` at `hostname`"\n'
807	>	txt += '    job_exit_code=10034\n'
808	>	txt += '    func_exit\n'
809	>	txt += 'fi \n'
810		# Handle the arguments:
811		txt += "\n"
812		txt += "## number of arguments (first argument always jobnumber)\n"
813		txt += "\n"
814	<	#        txt += "narg=$#\n"
724	<	txt += "if [ $nargs -lt 2 ]\n"
814	>	txt += "if [ $nargs -lt "+str(self.argsList)+" ]\n"
815		txt += "then\n"
816	<	txt += "    echo 'SET_EXE_ENV 1 ==> ERROR Too few arguments' +$nargs+ \n"
817	<	txt += '    echo "JOB_EXIT_STATUS = 50113"\n'
818	<	txt += '    echo "JobExitCode=50113" | tee -a $RUNTIME_AREA/$repo\n'
729	<	txt += '    dumpStatus $RUNTIME_AREA/$repo\n'
730	<	txt += '    rm -f $RUNTIME_AREA/$repo \n'
731	<	txt += '    echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
732	<	txt += '    echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
733	<	## OLI_Daniele
734	<	txt += '    if [ $middleware == OSG ]; then \n'
735	<	txt += '        echo "Remove working directory: $WORKING_DIR"\n'
736	<	txt += '        cd $RUNTIME_AREA\n'
737	<	txt += '        /bin/rm -rf $WORKING_DIR\n'
738	<	txt += '        if [ -d $WORKING_DIR ] ;then\n'
739	<	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'
740	<	txt += '            echo "JOB_EXIT_STATUS = 50114"\n'
741	<	txt += '            echo "JobExitCode=50114" | tee -a $RUNTIME_AREA/$repo\n'
742	<	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
743	<	txt += '            rm -f $RUNTIME_AREA/$repo \n'
744	<	txt += '            echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
745	<	txt += '            echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
746	<	txt += '        fi\n'
747	<	txt += '    fi \n'
748	<	txt += "    exit 1\n"
816	>	txt += "    echo 'ERROR ==> Too few arguments' +$nargs+ \n"
817	>	txt += '    job_exit_code=50113\n'
818	>	txt += "    func_exit\n"
819		txt += "fi\n"
820		txt += "\n"
821
822		# Prepare job-specific part
823		job = common.job_list[nj]
824	<	pset = os.path.basename(job.configFilename())
825	<	txt += '\n'
826	<	if (self.datasetPath): # standard job
827	<	#txt += 'InputFiles=$2\n'
828	<	txt += 'InputFiles=${args[1]}\n'
759	<	txt += 'echo "Inputfiles:<$InputFiles>"\n'
760	<	txt += 'sed "s#{\'INPUT\'}#$InputFiles#" $RUNTIME_AREA/'+pset+' > pset.cfg\n'
761	<	else:  # pythia like job
762	<	if (self.sourceSeed):
763	<	txt += 'Seed=$2\n'
764	<	txt += 'echo "Seed: <$Seed>"\n'
765	<	txt += 'sed "s#\<INPUT\>#$Seed#" $RUNTIME_AREA/'+pset+' > tmp.cfg\n'
766	<	if (self.sourceSeedVtx):
767	<	txt += 'VtxSeed=$3\n'
768	<	txt += 'echo "VtxSeed: <$VtxSeed>"\n'
769	<	txt += 'sed "s#INPUTVTX#$VtxSeed#" tmp.cfg > pset.cfg\n'
770	<	else:
771	<	txt += 'mv tmp.cfg pset.cfg\n'
772	<	else:
773	<	txt += '# Copy untouched pset\n'
774	<	txt += 'cp $RUNTIME_AREA/'+pset+' pset.cfg\n'
824	>	if (self.datasetPath):
825	>	txt += '\n'
826	>	txt += 'DatasetPath='+self.datasetPath+'\n'
827	>
828	>	datasetpath_split = self.datasetPath.split("/")
829
830	+	txt += 'PrimaryDataset='+datasetpath_split[1]+'\n'
831	+	txt += 'DataTier='+datasetpath_split[2]+'\n'
832	+	txt += 'ApplicationFamily=cmsRun\n'
833
834	<	if len(self.additional_inbox_files) > 0:
835	<	for file in self.additional_inbox_files:
836	<	relFile = file.split("/")[-1]
837	<	txt += 'if [ -e $RUNTIME_AREA/'+relFile+' ] ; then\n'
838	<	txt += '   cp $RUNTIME_AREA/'+relFile+' .\n'
839	<	txt += '   chmod +x '+relFile+'\n'
840	<	txt += 'fi\n'
841	<	pass
834	>	else:
835	>	txt += 'DatasetPath=MCDataTier\n'
836	>	txt += 'PrimaryDataset=null\n'
837	>	txt += 'DataTier=null\n'
838	>	txt += 'ApplicationFamily=MCDataTier\n'
839	>	if self.pset != None:
840	>	pset = os.path.basename(job.configFilename())
841	>	txt += '\n'
842	>	txt += 'cp  $RUNTIME_AREA/'+pset+' .\n'
843	>	if (self.datasetPath): # standard job
844	>	txt += 'InputFiles=${args[1]}; export InputFiles\n'
845	>	txt += 'MaxEvents=${args[2]}; export MaxEvents\n'
846	>	txt += 'SkipEvents=${args[3]}; export SkipEvents\n'
847	>	txt += 'echo "Inputfiles:<$InputFiles>"\n'
848	>	txt += 'echo "MaxEvents:<$MaxEvents>"\n'
849	>	txt += 'echo "SkipEvents:<$SkipEvents>"\n'
850	>	else:  # pythia like job
851	>	txt += 'PreserveSeeds='  + ','.join(self.preserveSeeds)  + '; export PreserveSeeds\n'
852	>	txt += 'IncrementSeeds=' + ','.join(self.incrementSeeds) + '; export IncrementSeeds\n'
853	>	txt += 'echo "PreserveSeeds: <$PreserveSeeds>"\n'
854	>	txt += 'echo "IncrementSeeds:<$IncrementSeeds>"\n'
855	>	if (self.firstRun):
856	>	txt += 'FirstRun=${args[1]}; export FirstRun\n'
857	>	txt += 'echo "FirstRun: <$FirstRun>"\n'
858
859	<	txt += 'echo "### END JOB SETUP ENVIRONMENT ###"\n\n'
859	>	txt += 'mv -f ' + pset + ' ' + psetName + '\n'
860
861	<	txt += '\n'
862	<	txt += 'echo "***** cat pset.cfg *********"\n'
863	<	txt += 'cat pset.cfg\n'
864	<	txt += 'echo "****** end pset.cfg ********"\n'
865	<	txt += '\n'
866	<	# txt += 'echo "***** cat pset1.cfg *********"\n'
867	<	# txt += 'cat pset1.cfg\n'
868	<	# txt += 'echo "****** end pset1.cfg ********"\n'
861	>
862	>	if self.pset != None:
863	>	# FUTURE: Can simply for 2_1_x and higher
864	>	txt += '\n'
865	>	if self.debug_pset==True:
866	>	txt += 'echo "***** cat ' + psetName + ' *********"\n'
867	>	txt += 'cat ' + psetName + '\n'
868	>	txt += 'echo "****** end ' + psetName + ' ********"\n'
869	>	txt += '\n'
870	>	txt += 'PSETHASH=`edmConfigHash < ' + psetName + '` \n'
871	>	txt += 'echo "PSETHASH = $PSETHASH" \n'
872	>	txt += '\n'
873		return txt
874
875	<	def wsBuildExe(self, nj):
875	>	def wsUntarSoftware(self, nj=0):
876		"""
877		Put in the script the commands to build an executable
878		or a library.
879		"""
880
881	<	txt = ""
881	>	txt = '\n#Written by cms_cmssw::wsUntarSoftware\n'
882
883		if os.path.isfile(self.tgzNameWithPath):
884	<	txt += 'echo "tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+'"\n'
884	>	txt += 'echo ">>> tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+' :" \n'
885		txt += 'tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+'\n'
886	+	txt += 'ls -Al \n'
887		txt += 'untar_status=$? \n'
888		txt += 'if [ $untar_status -ne 0 ]; then \n'
889	<	txt += '   echo "SET_EXE 1 ==> ERROR Untarring .tgz file failed"\n'
890	<	txt += '   echo "JOB_EXIT_STATUS = $untar_status" \n'
891	<	txt += '   echo "JobExitCode=$untar_status" | tee -a $RUNTIME_AREA/$repo\n'
814	<	txt += '   if [ $middleware == OSG ]; then \n'
815	<	txt += '       echo "Remove working directory: $WORKING_DIR"\n'
816	<	txt += '       cd $RUNTIME_AREA\n'
817	<	txt += '       /bin/rm -rf $WORKING_DIR\n'
818	<	txt += '       if [ -d $WORKING_DIR ] ;then\n'
819	<	txt += '           echo "SET_EXE 50999 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after Untarring .tgz file failed"\n'
820	<	txt += '           echo "JOB_EXIT_STATUS = 50999"\n'
821	<	txt += '           echo "JobExitCode=50999" | tee -a $RUNTIME_AREA/$repo\n'
822	<	txt += '           dumpStatus $RUNTIME_AREA/$repo\n'
823	<	txt += '           rm -f $RUNTIME_AREA/$repo \n'
824	<	txt += '           echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
825	<	txt += '           echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
826	<	txt += '       fi\n'
827	<	txt += '   fi \n'
828	<	txt += '   \n'
829	<	txt += '   exit 1 \n'
889	>	txt += '   echo "ERROR ==> Untarring .tgz file failed"\n'
890	>	txt += '   job_exit_code=$untar_status\n'
891	>	txt += '   func_exit\n'
892		txt += 'else \n'
893		txt += '   echo "Successful untar" \n'
894		txt += 'fi \n'
895	+	txt += '\n'
896	+	txt += 'echo ">>> Include ProdCommon in PYTHONPATH:"\n'
897	+	txt += 'if [ -z "$PYTHONPATH" ]; then\n'
898	+	txt += '   export PYTHONPATH=$RUNTIME_AREA/ProdCommon\n'
899	+	txt += 'else\n'
900	+	txt += '   export PYTHONPATH=$RUNTIME_AREA/ProdCommon:${PYTHONPATH}\n'
901	+	txt += 'echo "PYTHONPATH=$PYTHONPATH"\n'
902	+	txt += 'fi\n'
903	+	txt += '\n'
904	+
905		pass
906	<
906	>
907	>	return txt
908	>
909	>	def wsBuildExe(self, nj=0):
910	>	"""
911	>	Put in the script the commands to build an executable
912	>	or a library.
913	>	"""
914	>
915	>	txt = '\n#Written by cms_cmssw::wsBuildExe\n'
916	>	txt += 'echo ">>> moving CMSSW software directories in `pwd`" \n'
917	>
918	>	txt += 'rm -r lib/ module/ \n'
919	>	txt += 'mv $RUNTIME_AREA/lib/ . \n'
920	>	txt += 'mv $RUNTIME_AREA/module/ . \n'
921	>	if self.dataExist == True:
922	>	txt += 'rm -r src/ \n'
923	>	txt += 'mv $RUNTIME_AREA/src/ . \n'
924	>	if len(self.additional_inbox_files)>0:
925	>	for file in self.additional_inbox_files:
926	>	txt += 'mv $RUNTIME_AREA/'+os.path.basename(file)+' . \n'
927	>	txt += 'mv $RUNTIME_AREA/ProdCommon/ . \n'
928	>
929	>	txt += 'if [ -z "$PYTHONPATH" ]; then\n'
930	>	txt += '   export PYTHONPATH=$SOFTWARE_DIR/ProdCommon\n'
931	>	txt += 'else\n'
932	>	txt += '   export PYTHONPATH=$SOFTWARE_DIR/ProdCommon:${PYTHONPATH}\n'
933	>	txt += 'echo "PYTHONPATH=$PYTHONPATH"\n'
934	>	txt += 'fi\n'
935	>	txt += '\n'
936	>
937		return txt
938
939		def modifySteeringCards(self, nj):
940		"""
941	<	modify the card provided by the user,
941	>	modify the card provided by the user,
942		writing a new card into share dir
943		"""
944	<
944	>
945		def executableName(self):
946	<	return self.executable
946	>	if self.scriptExe:
947	>	return "sh "
948	>	else:
949	>	return self.executable
950
951		def executableArgs(self):
952	<	return " -p pset.cfg"
952	>	# FUTURE: This function tests the CMSSW version. Can be simplified as we drop support for old versions
953	>	if self.scriptExe:#CarlosDaniele
954	>	return   self.scriptExe + " $NJob"
955	>	else:
956	>	ex_args = ""
957	>	# FUTURE: This tests the CMSSW version. Can remove code as versions deprecated
958	>	# Framework job report
959	>	if (self.CMSSW_major >= 1 and self.CMSSW_minor >= 5) or (self.CMSSW_major >= 2):
960	>	ex_args += " -j $RUNTIME_AREA/crab_fjr_$NJob.xml"
961	>	# Type of config file
962	>	if self.CMSSW_major >= 2 :
963	>	ex_args += " -p pset.py"
964	>	else:
965	>	ex_args += " -p pset.cfg"
966	>	return ex_args
967
968		def inputSandbox(self, nj):
969		"""
970		Returns a list of filenames to be put in JDL input sandbox.
971		"""
972		inp_box = []
854	–	# dict added to delete duplicate from input sandbox file list
855	–	seen = {}
856	–	## code
973		if os.path.isfile(self.tgzNameWithPath):
974		inp_box.append(self.tgzNameWithPath)
975	<	## config
976	<	inp_box.append(common.job_list[nj].configFilename())
861	<	## additional input files
862	<	#for file in self.additional_inbox_files:
863	<	#    inp_box.append(common.work_space.cwdDir()+file)
975	>	wrapper = os.path.basename(str(common._db.queryTask('scriptName')))
976	>	inp_box.append(common.work_space.pathForTgz() +'job/'+ wrapper)
977		return inp_box
978
979		def outputSandbox(self, nj):
#	Line 869 | Line 982 | class Cmssw(JobType):
982		"""
983		out_box = []
984
872	–	stdout=common.job_list[nj].stdout()
873	–	stderr=common.job_list[nj].stderr()
874	–
985		## User Declared output files
986	<	for out in self.output_file:
987	<	n_out = nj + 1
986	>	for out in (self.output_file+self.output_file_sandbox):
987	>	n_out = nj + 1
988		out_box.append(self.numberFile_(out,str(n_out)))
989		return out_box
880	–	return []
990
991		def prepareSteeringCards(self):
992		"""
#	Line 890 | Line 999 | class Cmssw(JobType):
999		Returns part of a job script which renames the produced files.
1000		"""
1001
1002	<	txt = '\n'
1003	<	txt += '# directory content\n'
1002	>	txt = '\n#Written by cms_cmssw::wsRenameOutput\n'
1003	>	txt += 'echo ">>> current directory (SOFTWARE_DIR): $SOFTWARE_DIR" \n'
1004	>	txt += 'echo ">>> current directory content:"\n'
1005		txt += 'ls \n'
1006	<	file_list = ''
1007	<	for fileWithSuffix in self.output_file:
1006	>	txt += '\n'
1007	>
1008	>	for fileWithSuffix in (self.output_file):
1009		output_file_num = self.numberFile_(fileWithSuffix, '$NJob')
899	–	file_list=file_list+output_file_num+' '
1010		txt += '\n'
1011		txt += '# check output file\n'
1012	<	txt += 'ls '+fileWithSuffix+'\n'
1013	<	txt += 'ls_result=$?\n'
1014	<	#txt += 'exe_result=$?\n'
1015	<	txt += 'if [ $ls_result -ne 0 ] ; then\n'
1016	<	txt += '   echo "ERROR: Problem with output file"\n'
1017	<	#txt += '   echo "JOB_EXIT_STATUS = $exe_result"\n'
1018	<	#txt += '   echo "JobExitCode=60302" | tee -a $RUNTIME_AREA/$repo\n'
1019	<	#txt += '   dumpStatus $RUNTIME_AREA/$repo\n'
1020	<	### OLI_DANIELE
1021	<	if common.scheduler.boss_scheduler_name == 'condor_g':
1012	>	txt += 'if [ -e ./'+fileWithSuffix+' ] ; then\n'
1013	>	if (self.copy_data == 1):  # For OSG nodes, file is in $WORKING_DIR, should not be moved to $RUNTIME_AREA
1014	>	txt += '    mv '+fileWithSuffix+' '+output_file_num+'\n'
1015	>	txt += '    ln -s `pwd`/'+output_file_num+' $RUNTIME_AREA/'+fileWithSuffix+'\n'
1016	>	else:
1017	>	txt += '    mv '+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n'
1018	>	txt += '    ln -s $RUNTIME_AREA/'+output_file_num+' $RUNTIME_AREA/'+fileWithSuffix+'\n'
1019	>	txt += 'else\n'
1020	>	txt += '    job_exit_code=60302\n'
1021	>	txt += '    echo "WARNING: Output file '+fileWithSuffix+' not found"\n'
1022	>	if common.scheduler.name().upper() == 'CONDOR_G':
1023		txt += '    if [ $middleware == OSG ]; then \n'
1024		txt += '        echo "prepare dummy output file"\n'
1025		txt += '        echo "Processing of job output failed" > $RUNTIME_AREA/'+output_file_num+'\n'
1026		txt += '    fi \n'
916	–	txt += 'else\n'
917	–	txt += '   cp '+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n'
1027		txt += 'fi\n'
1028	<
1029	<	txt += 'cd $RUNTIME_AREA\n'
1030	<	file_list=file_list[:-1]
1031	<	txt += 'file_list="'+file_list+'"\n'
1032	<	txt += 'cd $RUNTIME_AREA\n'
1033	<	### OLI_DANIELE
1034	<	txt += 'if [ $middleware == OSG ]; then\n'
1035	<	txt += '    cd $RUNTIME_AREA\n'
1036	<	txt += '    echo "Remove working directory: $WORKING_DIR"\n'
928	<	txt += '    /bin/rm -rf $WORKING_DIR\n'
929	<	txt += '    if [ -d $WORKING_DIR ] ;then\n'
930	<	txt += '        echo "SET_EXE 60999 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after cleanup of WN"\n'
931	<	txt += '        echo "JOB_EXIT_STATUS = 60999"\n'
932	<	txt += '        echo "JobExitCode=60999" | tee -a $RUNTIME_AREA/$repo\n'
933	<	txt += '        dumpStatus $RUNTIME_AREA/$repo\n'
934	<	txt += '        rm -f $RUNTIME_AREA/$repo \n'
935	<	txt += '        echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
936	<	txt += '        echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
937	<	txt += '    fi\n'
938	<	txt += 'fi\n'
1028	>	file_list = []
1029	>	for fileWithSuffix in (self.output_file):
1030	>	file_list.append(self.numberFile_(fileWithSuffix, '$NJob'))
1031	>
1032	>	txt += 'file_list="'+string.join(file_list,' ')+'"\n'
1033	>	txt += '\n'
1034	>	txt += 'echo ">>> current directory (SOFTWARE_DIR): $SOFTWARE_DIR" \n'
1035	>	txt += 'echo ">>> current directory content:"\n'
1036	>	txt += 'ls \n'
1037		txt += '\n'
1038	+	txt += 'cd $RUNTIME_AREA\n'
1039	+	txt += 'echo ">>> current directory (RUNTIME_AREA):  $RUNTIME_AREA"\n'
1040		return txt
1041
1042		def numberFile_(self, file, txt):
#	Line 947 | Line 1047 | class Cmssw(JobType):
1047		# take away last extension
1048		name = p[0]
1049		for x in p[1:-1]:
1050	<	name=name+"."+x
1050	>	name=name+"."+x
1051		# add "_txt"
1052		if len(p)>1:
1053	<	ext = p[len(p)-1]
1054	<	#result = name + '_' + str(txt) + "." + ext
955	<	result = name + '_' + txt + "." + ext
1053	>	ext = p[len(p)-1]
1054	>	result = name + '_' + txt + "." + ext
1055		else:
1056	<	#result = name + '_' + str(txt)
1057	<	result = name + '_' + txt
959	<
1056	>	result = name + '_' + txt
1057	>
1058		return result
1059
1060	<	def getRequirements(self):
1060	>	def getRequirements(self, nj=[]):
1061		"""
1062	<	return job requirements to add to jdl files
1062	>	return job requirements to add to jdl files
1063		"""
1064		req = ''
1065	<	if common.analisys_common_info['sw_version']:
1065	>	if self.version:
1066		req='Member("VO-cms-' + \
1067	<	common.analisys_common_info['sw_version'] + \
1067	>	self.version + \
1068	>	'", other.GlueHostApplicationSoftwareRunTimeEnvironment)'
1069	>	if self.executable_arch:
1070	>	req+=' && Member("VO-cms-' + \
1071	>	self.executable_arch + \
1072		'", other.GlueHostApplicationSoftwareRunTimeEnvironment)'
1073	<	if common.analisys_common_info['sites']:
1074	<	if len(common.analisys_common_info['sites'])>0:
1075	<	req = req + ' && ('
1076	<	for i in range(len(common.analisys_common_info['sites'])):
1077	<	req = req + 'other.GlueCEInfoHostName == "' \
976	<	+ common.analisys_common_info['sites'][i] + '"'
977	<	if ( i < (int(len(common.analisys_common_info['sites']) - 1)) ):
978	<	req = req + ' || '
979	<	req = req + ')'
980	<	#print "req = ", req
1073	>
1074	>	req = req + ' && (other.GlueHostNetworkAdapterOutboundIP)'
1075	>	if common.scheduler.name() == "glitecoll":
1076	>	req += ' && other.GlueCEStateStatus == "Production" '
1077	>
1078		return req
1079
1080		def configFilename(self):
1081		""" return the config filename """
1082	<	return self.name()+'.cfg'
1082	>	# FUTURE: Can remove cfg mode for CMSSW >= 2_1_x
1083	>	if (self.CMSSW_major >= 2 and self.CMSSW_minor >= 1) or (self.CMSSW_major >= 3):
1084	>	return self.name()+'.py'
1085	>	else:
1086	>	return self.name()+'.cfg'
1087
987	–	### OLI_DANIELE
1088		def wsSetupCMSOSGEnvironment_(self):
1089		"""
1090		Returns part of a job script which is prepares
1091		the execution environment and which is common for all CMS jobs.
1092		"""
1093	<	txt = '\n'
1094	<	txt += '   echo "### SETUP CMS OSG  ENVIRONMENT ###"\n'
1095	<	txt += '   if [ -f $GRID3_APP_DIR/cmssoft/cmsset_default.sh ] ;then\n'
1096	<	txt += '      # Use $GRID3_APP_DIR/cmssoft/cmsset_default.sh to setup cms software\n'
1097	<	txt += '       source $GRID3_APP_DIR/cmssoft/cmsset_default.sh '+self.version+'\n'
1098	<	txt += '   elif [ -f $OSG_APP/cmssoft/cmsset_default.sh ] ;then\n'
1099	<	txt += '      # Use $OSG_APP/cmssoft/cmsset_default.sh to setup cms software\n'
1100	<	txt += '       source $OSG_APP/cmssoft/cmsset_default.sh '+self.version+'\n'
1101	<	txt += '   else\n'
1102	<	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'
1103	<	txt += '       echo "JOB_EXIT_STATUS = 10020"\n'
1104	<	txt += '       echo "JobExitCode=10020" | tee -a $RUNTIME_AREA/$repo\n'
1105	<	txt += '       dumpStatus $RUNTIME_AREA/$repo\n'
1006	<	txt += '       rm -f $RUNTIME_AREA/$repo \n'
1007	<	txt += '       echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1008	<	txt += '       echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1009	<	txt += '       exit 1\n'
1010	<	txt += '\n'
1011	<	txt += '       echo "Remove working directory: $WORKING_DIR"\n'
1012	<	txt += '       cd $RUNTIME_AREA\n'
1013	<	txt += '       /bin/rm -rf $WORKING_DIR\n'
1014	<	txt += '       if [ -d $WORKING_DIR ] ;then\n'
1015	<	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'
1016	<	txt += '            echo "JOB_EXIT_STATUS = 10017"\n'
1017	<	txt += '            echo "JobExitCode=10017" | tee -a $RUNTIME_AREA/$repo\n'
1018	<	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
1019	<	txt += '            rm -f $RUNTIME_AREA/$repo \n'
1020	<	txt += '            echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1021	<	txt += '            echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1022	<	txt += '       fi\n'
1023	<	txt += '\n'
1024	<	txt += '       exit 1\n'
1025	<	txt += '   fi\n'
1093	>	txt = '\n#Written by cms_cmssw::wsSetupCMSOSGEnvironment_\n'
1094	>	txt += '    echo ">>> setup CMS OSG environment:"\n'
1095	>	txt += '    echo "set SCRAM ARCH to ' + self.executable_arch + '"\n'
1096	>	txt += '    export SCRAM_ARCH='+self.executable_arch+'\n'
1097	>	txt += '    echo "SCRAM_ARCH = $SCRAM_ARCH"\n'
1098	>	txt += '    if [ -f $OSG_APP/cmssoft/cms/cmsset_default.sh ] ;then\n'
1099	>	txt += '      # Use $OSG_APP/cmssoft/cms/cmsset_default.sh to setup cms software\n'
1100	>	txt += '        source $OSG_APP/cmssoft/cms/cmsset_default.sh '+self.version+'\n'
1101	>	txt += '    else\n'
1102	>	txt += '        echo "ERROR ==> $OSG_APP/cmssoft/cms/cmsset_default.sh file not found"\n'
1103	>	txt += '        job_exit_code=10020\n'
1104	>	txt += '        func_exit\n'
1105	>	txt += '    fi\n'
1106		txt += '\n'
1107	<	txt += '   echo "SET_CMS_ENV 0 ==> setup cms environment ok"\n'
1108	<	txt += '   echo " END SETUP CMS OSG  ENVIRONMENT "\n'
1107	>	txt += '    echo "==> setup cms environment ok"\n'
1108	>	txt += '    echo "SCRAM_ARCH = $SCRAM_ARCH"\n'
1109
1110		return txt
1111	<
1032	<	### OLI_DANIELE
1111	>
1112		def wsSetupCMSLCGEnvironment_(self):
1113		"""
1114		Returns part of a job script which is prepares
1115		the execution environment and which is common for all CMS jobs.
1116		"""
1117	<	txt  = '   \n'
1118	<	txt += '   echo " ### SETUP CMS LCG  ENVIRONMENT ### "\n'
1119	<	txt += '   if [ ! $VO_CMS_SW_DIR ] ;then\n'
1120	<	txt += '       echo "SET_CMS_ENV 10031 ==> ERROR CMS software dir not found on WN `hostname`"\n'
1121	<	txt += '       echo "JOB_EXIT_STATUS = 10031" \n'
1122	<	txt += '       echo "JobExitCode=10031" | tee -a $RUNTIME_AREA/$repo\n'
1123	<	txt += '       dumpStatus $RUNTIME_AREA/$repo\n'
1124	<	txt += '       rm -f $RUNTIME_AREA/$repo \n'
1125	<	txt += '       echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1126	<	txt += '       echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1127	<	txt += '       exit 1\n'
1128	<	txt += '   else\n'
1129	<	txt += '       echo "Sourcing environment... "\n'
1130	<	txt += '       if [ ! -s $VO_CMS_SW_DIR/cmsset_default.sh ] ;then\n'
1131	<	txt += '           echo "SET_CMS_ENV 10020 ==> ERROR cmsset_default.sh file not found into dir $VO_CMS_SW_DIR"\n'
1132	<	txt += '           echo "JOB_EXIT_STATUS = 10020"\n'
1133	<	txt += '           echo "JobExitCode=10020" | tee -a $RUNTIME_AREA/$repo\n'
1134	<	txt += '           dumpStatus $RUNTIME_AREA/$repo\n'
1135	<	txt += '           rm -f $RUNTIME_AREA/$repo \n'
1136	<	txt += '           echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1137	<	txt += '           echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1138	<	txt += '           exit 1\n'
1139	<	txt += '       fi\n'
1140	<	txt += '       echo "sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
1141	<	txt += '       source $VO_CMS_SW_DIR/cmsset_default.sh\n'
1142	<	txt += '       result=$?\n'
1143	<	txt += '       if [ $result -ne 0 ]; then\n'
1144	<	txt += '           echo "SET_CMS_ENV 10032 ==> ERROR problem sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
1145	<	txt += '           echo "JOB_EXIT_STATUS = 10032"\n'
1146	<	txt += '           echo "JobExitCode=10032" | tee -a $RUNTIME_AREA/$repo\n'
1147	<	txt += '           dumpStatus $RUNTIME_AREA/$repo\n'
1148	<	txt += '           rm -f $RUNTIME_AREA/$repo \n'
1149	<	txt += '           echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1150	<	txt += '           echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1151	<	txt += '           exit 1\n'
1152	<	txt += '       fi\n'
1153	<	txt += '   fi\n'
1154	<	txt += '   \n'
1155	<	txt += '   string=`cat /etc/redhat-release`\n'
1156	<	txt += '   echo $string\n'
1157	<	txt += '   if [[ $string = *alhalla* ]]; then\n'
1158	<	txt += '       echo "SCRAM_ARCH= $SCRAM_ARCH"\n'
1159	<	txt += '   elif [[ $string = *Enterprise* ]] || [[ $string = *cientific* ]]; then\n'
1160	<	txt += '       export SCRAM_ARCH=slc3_ia32_gcc323\n'
1161	<	txt += '       echo "SCRAM_ARCH= $SCRAM_ARCH"\n'
1162	<	txt += '   else\n'
1163	<	txt += '       echo "SET_CMS_ENV 10033 ==> ERROR OS unknown, LCG environment not initialized"\n'
1164	<	txt += '       echo "JOB_EXIT_STATUS = 10033"\n'
1165	<	txt += '       echo "JobExitCode=10033" | tee -a $RUNTIME_AREA/$repo\n'
1166	<	txt += '       dumpStatus $RUNTIME_AREA/$repo\n'
1167	<	txt += '       rm -f $RUNTIME_AREA/$repo \n'
1168	<	txt += '       echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1169	<	txt += '       echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1170	<	txt += '       exit 1\n'
1171	<	txt += '   fi\n'
1172	<	txt += '   echo "SET_CMS_ENV 0 ==> setup cms environment ok"\n'
1173	<	txt += '   echo "### END SETUP CMS LCG ENVIRONMENT ###"\n'
1117	>	txt = '\n#Written by cms_cmssw::wsSetupCMSLCGEnvironment_\n'
1118	>	txt += '    echo ">>> setup CMS LCG environment:"\n'
1119	>	txt += '    echo "set SCRAM ARCH and BUILD_ARCH to ' + self.executable_arch + ' ###"\n'
1120	>	txt += '    export SCRAM_ARCH='+self.executable_arch+'\n'
1121	>	txt += '    export BUILD_ARCH='+self.executable_arch+'\n'
1122	>	txt += '    if [ ! $VO_CMS_SW_DIR ] ;then\n'
1123	>	txt += '        echo "ERROR ==> CMS software dir not found on WN `hostname`"\n'
1124	>	txt += '        job_exit_code=10031\n'
1125	>	txt += '        func_exit\n'
1126	>	txt += '    else\n'
1127	>	txt += '        echo "Sourcing environment... "\n'
1128	>	txt += '        if [ ! -s $VO_CMS_SW_DIR/cmsset_default.sh ] ;then\n'
1129	>	txt += '            echo "ERROR ==> cmsset_default.sh file not found into dir $VO_CMS_SW_DIR"\n'
1130	>	txt += '            job_exit_code=10020\n'
1131	>	txt += '            func_exit\n'
1132	>	txt += '        fi\n'
1133	>	txt += '        echo "sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
1134	>	txt += '        source $VO_CMS_SW_DIR/cmsset_default.sh\n'
1135	>	txt += '        result=$?\n'
1136	>	txt += '        if [ $result -ne 0 ]; then\n'
1137	>	txt += '            echo "ERROR ==> problem sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
1138	>	txt += '            job_exit_code=10032\n'
1139	>	txt += '            func_exit\n'
1140	>	txt += '        fi\n'
1141	>	txt += '    fi\n'
1142	>	txt += '    \n'
1143	>	txt += '    echo "==> setup cms environment ok"\n'
1144	>	return txt
1145	>
1146	>	def modifyReport(self, nj):
1147	>	"""
1148	>	insert the part of the script that modifies the FrameworkJob Report
1149	>	"""
1150	>	txt = '\n#Written by cms_cmssw::modifyReport\n'
1151	>	publish_data = int(self.cfg_params.get('USER.publish_data',0))
1152	>	if (publish_data == 1):
1153	>	processedDataset = self.cfg_params['USER.publish_data_name']
1154	>	LFNBaseName = LFNBase(processedDataset)
1155	>
1156	>	txt += 'if [ $copy_exit_status -eq 0 ]; then\n'
1157	>	txt += '    FOR_LFN=%s_${PSETHASH}/\n'%(LFNBaseName)
1158	>	txt += 'else\n'
1159	>	txt += '    FOR_LFN=/copy_problems/ \n'
1160	>	txt += '    SE=""\n'
1161	>	txt += '    SE_PATH=""\n'
1162	>	txt += 'fi\n'
1163	>
1164	>	txt += 'echo ">>> Modify Job Report:" \n'
1165	>	txt += 'chmod a+x $SOFTWARE_DIR/ProdCommon/ProdCommon/FwkJobRep/ModifyJobReport.py\n'
1166	>	txt += 'ProcessedDataset='+processedDataset+'\n'
1167	>	txt += 'echo "ProcessedDataset = $ProcessedDataset"\n'
1168	>	txt += 'echo "SE = $SE"\n'
1169	>	txt += 'echo "SE_PATH = $SE_PATH"\n'
1170	>	txt += 'echo "FOR_LFN = $FOR_LFN" \n'
1171	>	txt += 'echo "CMSSW_VERSION = $CMSSW_VERSION"\n\n'
1172	>	txt += 'echo "$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 $SE_PATH"\n'
1173	>	txt += '$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 $SE_PATH\n'
1174	>	txt += 'modifyReport_result=$?\n'
1175	>	txt += 'if [ $modifyReport_result -ne 0 ]; then\n'
1176	>	txt += '    modifyReport_result=70500\n'
1177	>	txt += '    job_exit_code=$modifyReport_result\n'
1178	>	txt += '    echo "ModifyReportResult=$modifyReport_result" | tee -a $RUNTIME_AREA/$repo\n'
1179	>	txt += '    echo "WARNING: Problem with ModifyJobReport"\n'
1180	>	txt += 'else\n'
1181	>	txt += '    mv NewFrameworkJobReport.xml $RUNTIME_AREA/crab_fjr_$NJob.xml\n'
1182	>	txt += 'fi\n'
1183	>	return txt
1184	>
1185	>	def wsParseFJR(self):
1186	>	"""
1187	>	Parse the FrameworkJobReport to obtain useful infos
1188	>	"""
1189	>	txt = '\n#Written by cms_cmssw::wsParseFJR\n'
1190	>	txt += 'echo ">>> Parse FrameworkJobReport crab_fjr.xml"\n'
1191	>	txt += 'if [ -s $RUNTIME_AREA/crab_fjr_$NJob.xml ]; then\n'
1192	>	txt += '    if [ -s $RUNTIME_AREA/parseCrabFjr.py ]; then\n'
1193	>	txt += '        cmd_out=`python $RUNTIME_AREA/parseCrabFjr.py --input $RUNTIME_AREA/crab_fjr_$NJob.xml --MonitorID $MonitorID --MonitorJobID $MonitorJobID`\n'
1194	>	txt += '        echo "Result of parsing the FrameworkJobReport crab_fjr.xml: $cmd_out"\n'
1195	>	txt += '        executable_exit_status=`echo $cmd_out | awk -F\; "{print $1}" | awk -F ' ' "{print $NF}"`\n'
1196	>	txt += '        if [ $executable_exit_status -eq 50115 ];then\n'
1197	>	txt += '            echo ">>> crab_fjr.xml contents: "\n'
1198	>	txt += '            cat $RUNTIME_AREA/crab_fjr_NJob.xml\n'
1199	>	txt += '            echo "Wrong FrameworkJobReport --> does not contain useful info. ExitStatus: $executable_exit_status"\n'
1200	>	txt += '        else\n'
1201	>	txt += '            echo "Extracted ExitStatus from FrameworkJobReport parsing output: $executable_exit_status"\n'
1202	>	txt += '        fi\n'
1203	>	txt += '    else\n'
1204	>	txt += '        echo "CRAB python script to parse CRAB FrameworkJobReport crab_fjr.xml is not available, using exit code of executable from command line."\n'
1205	>	txt += '    fi\n'
1206	>	#### Patch to check input data reading for CMSSW16x Hopefully we-ll remove it asap
1207	>
1208	>	if self.datasetPath:
1209	>	# VERIFY PROCESSED DATA
1210	>	txt += '    if [ $executable_exit_status -eq 0 ];then\n'
1211	>	txt += '      echo ">>> Verify list of processed files:"\n'
1212	>	txt += '      echo $InputFiles |tr -d "\\" |tr "," \n"|tr -d "\"" > input-files.txt\n'
1213	>	txt += '      grep LFN $RUNTIME_AREA/crab_fjr_$NJob.xml |cut -d">" -f2|cut -d"<" -f1|grep "/" > processed-files.txt\n'
1214	>	txt += '      cat input-files.txt  | sort | uniq > tmp.txt\n'
1215	>	txt += '      mv tmp.txt input-files.txt\n'
1216	>	txt += '      echo "cat input-files.txt"\n'
1217	>	txt += '      echo "----------------------"\n'
1218	>	txt += '      cat input-files.txt\n'
1219	>	txt += '      cat processed-files.txt | sort | uniq > tmp.txt\n'
1220	>	txt += '      mv tmp.txt processed-files.txt\n'
1221	>	txt += '      echo "----------------------"\n'
1222	>	txt += '      echo "cat processed-files.txt"\n'
1223	>	txt += '      echo "----------------------"\n'
1224	>	txt += '      cat processed-files.txt\n'
1225	>	txt += '      echo "----------------------"\n'
1226	>	txt += '      diff -q input-files.txt processed-files.txt\n'
1227	>	txt += '      fileverify_status=$?\n'
1228	>	txt += '      if [ $fileverify_status -ne 0 ]; then\n'
1229	>	txt += '         executable_exit_status=30001\n'
1230	>	txt += '         echo "ERROR ==> not all input files processed"\n'
1231	>	txt += '         echo "      ==> list of processed files from crab_fjr.xml differs from list in pset.cfg"\n'
1232	>	txt += '         echo "      ==> diff input-files.txt processed-files.txt"\n'
1233	>	txt += '      fi\n'
1234	>	txt += '    fi\n'
1235	>	txt += '\n'
1236	>	txt += 'else\n'
1237	>	txt += '    echo "CRAB FrameworkJobReport crab_fjr.xml is not available, using exit code of executable from command line."\n'
1238	>	txt += 'fi\n'
1239	>	txt += '\n'
1240	>	txt += 'echo "ExeExitCode=$executable_exit_status" | tee -a $RUNTIME_AREA/$repo\n'
1241	>	txt += 'echo "EXECUTABLE_EXIT_STATUS = $executable_exit_status"\n'
1242	>	txt += 'job_exit_code=$executable_exit_status\n'
1243	>
1244		return txt
1245
1246		def setParam_(self, param, value):
#	Line 1100 | Line 1249 | class Cmssw(JobType):
1249		def getParams(self):
1250		return self._params
1251
1252	<	def setTaskid_(self):
1253	<	self._taskId = self.cfg_params['taskId']
1254	<
1255	<	def getTaskid(self):
1256	<	return self._taskId
1252	>	def uniquelist(self, old):
1253	>	"""
1254	>	remove duplicates from a list
1255	>	"""
1256	>	nd={}
1257	>	for e in old:
1258	>	nd[e]=0
1259	>	return nd.keys()
1260	>
1261	>	def outList(self):
1262	>	"""
1263	>	check the dimension of the output files
1264	>	"""
1265	>	txt = ''
1266	>	txt += 'echo ">>> list of expected files on output sandbox"\n'
1267	>	listOutFiles = []
1268	>	stdout = 'CMSSW_$NJob.stdout'
1269	>	stderr = 'CMSSW_$NJob.stderr'
1270	>	if (self.return_data == 1):
1271	>	for file in (self.output_file+self.output_file_sandbox):
1272	>	listOutFiles.append(self.numberFile_(file, '$NJob'))
1273	>	listOutFiles.append(stdout)
1274	>	listOutFiles.append(stderr)
1275	>	else:
1276	>	for file in (self.output_file_sandbox):
1277	>	listOutFiles.append(self.numberFile_(file, '$NJob'))
1278	>	listOutFiles.append(stdout)
1279	>	listOutFiles.append(stderr)
1280	>	txt += 'echo "output files: '+string.join(listOutFiles,' ')+'"\n'
1281	>	txt += 'filesToCheck="'+string.join(listOutFiles,' ')+'"\n'
1282	>	txt += 'export filesToCheck\n'
1283	>	return txt

Diff Legend

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