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 1.4 by afanfani, Tue May 30 00:48:37 2006 UTC vs.
Revision 1.183 by spiga, Wed Apr 30 18:21:07 2008 UTC

#	Line 2 | Line 2 | from JobType import JobType
2		from crab_logger import Logger
3		from crab_exceptions import *
4		from crab_util import *
5	+	from BlackWhiteListParser import BlackWhiteListParser
6		import common
6	–	import PsetManipulator
7	–
8	–	import DBSInfo_EDM
9	–	#from DataDiscovery_EDM import DataDiscovery_EDM
10	–	import DataDiscovery_EDM
11	–	#from DataLocation_EDM import DataLocation_EDM
12	–	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)
30	–	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 = ''      #scrip use case Da
39	+	self.datasetPath = '' #scrip use case Da
40
41	+	# set FJR file name
42	+	self.fjrFileName = 'crab_fjr.xml'
43
44		self.version = self.scram.getSWVersion()
45	<	common.analisys_common_info['sw_version'] = self.version
46	<	### FEDE
47	<	common.analisys_common_info['copy_input_data'] = 0
48	<	common.analisys_common_info['events_management'] = 1
45	>	version_array = self.version.split('_')
46	>	self.major_version = 0
47	>	self.minor_version = 0
48	>	try:
49	>	self.major_version = int(version_array[1])
50	>	self.minor_version = int(version_array[2])
51	>	except:
52	>	msg = "Cannot parse CMSSW version string: " + "_".join(version_array) + " for major and minor release number!"
53	>	raise CrabException(msg)
54	>
55	>
56	>	#
57	>	# Try to block creation in case of arch/version mismatch
58	>	#
59	>
60	>	#        a = string.split(self.version, "_")
61	>	#
62	>	#        if int(a[1]) == 1 and (int(a[2]) < 5 and self.executable_arch.find('slc4') == 0):
63	>	#            msg = "Warning: You are using %s version of CMSSW  with %s architecture. \n--> Did you compile your libraries with SLC3? Otherwise you can find some problems running on SLC4 Grid nodes.\n"%(self.version, self.executable_arch)
64	>	#            common.logger.message(msg)
65	>	#        if int(a[1]) == 1 and (int(a[2]) >= 5 and self.executable_arch.find('slc3') == 0):
66	>	#            msg = "Error: CMS does not support %s with %s architecture"%(self.version, self.executable_arch)
67	>	#            raise CrabException(msg)
68	>	#
69	>
70
71		### collect Data cards
72	<	try:
73	<	#   self.owner = cfg_params['CMSSW.owner']
74	<	#   log.debug(6, "CMSSW::CMSSW(): owner = "+self.owner)
47	<	#   self.dataset = cfg_params['CMSSW.dataset']
48	<	self.datasetPath = cfg_params['CMSSW.datasetpath']
49	<	log.debug(6, "CMSSW::CMSSW(): datasetPath = "+self.datasetPath)
50	<	except KeyError:
51	<	#    msg = "Error: owner and/or dataset not defined "
52	<	msg = "Error: datasetpath not defined "
72	>
73	>	if not cfg_params.has_key('CMSSW.datasetpath'):
74	>	msg = "Error: datasetpath not defined "
75		raise CrabException(msg)
76	+	tmp =  cfg_params['CMSSW.datasetpath']
77	+	log.debug(6, "CMSSW::CMSSW(): datasetPath = "+tmp)
78	+	if string.lower(tmp)=='none':
79	+	self.datasetPath = None
80	+	self.selectNoInput = 1
81	+	else:
82	+	self.datasetPath = tmp
83	+	self.selectNoInput = 0
84	+
85		self.dataTiers = []
55	–	#       try:
56	–	#           tmpDataTiers = string.split(cfg_params['CMSSW.data_tier'],',')
57	–	#           for tmp in tmpDataTiers:
58	–	#               tmp=string.strip(tmp)
59	–	#               self.dataTiers.append(tmp)
60	–	#               pass
61	–	#           pass
62	–	#       except KeyError:
63	–	#           pass
64	–	#       log.debug(6, "Cmssw::Cmssw(): dataTiers = "+str(self.dataTiers))
86
87		## now the application
88	<	try:
89	<	self.executable = cfg_params['CMSSW.executable']
69	<	log.debug(6, "CMSSW::CMSSW(): executable = "+self.executable)
70	<	msg = "Default executable cmsRun overridden. Switch to " + self.executable
71	<	log.debug(3,msg)
72	<	except KeyError:
73	<	self.executable = 'cmsRun'
74	<	msg = "User executable not defined. Use cmsRun"
75	<	log.debug(3,msg)
76	<	pass
88	>	self.executable = cfg_params.get('CMSSW.executable','cmsRun')
89	>	log.debug(6, "CMSSW::CMSSW(): executable = "+self.executable)
90
91	<	try:
92	<	self.pset = cfg_params['CMSSW.pset']
93	<	log.debug(6, "Cmssw::Cmssw(): PSet file = "+self.pset)
91	>	if not cfg_params.has_key('CMSSW.pset'):
92	>	raise CrabException("PSet file missing. Cannot run cmsRun ")
93	>	self.pset = cfg_params['CMSSW.pset']
94	>	log.debug(6, "Cmssw::Cmssw(): PSet file = "+self.pset)
95	>	if self.pset.lower() != 'none' :
96		if (not os.path.exists(self.pset)):
97		raise CrabException("User defined PSet file "+self.pset+" does not exist")
98	<	except KeyError:
99	<	raise CrabException("PSet file missing. Cannot run cmsRun ")
98	>	else:
99	>	self.pset = None
100
101		# output files
102	<	try:
103	<	self.output_file = []
102	>	## stuff which must be returned always via sandbox
103	>	self.output_file_sandbox = []
104
105	<	tmp = cfg_params['CMSSW.output_file']
106	<	if tmp != '':
107	<	tmpOutFiles = string.split(cfg_params['CMSSW.output_file'],',')
108	<	log.debug(7, 'cmssw::cmssw(): output files '+str(tmpOutFiles))
109	<	for tmp in tmpOutFiles:
110	<	tmp=string.strip(tmp)
111	<	self.output_file.append(tmp)
112	<	pass
113	<	else:
114	<	log.message("No output file defined: only stdout/err will be available")
105	>	# add fjr report by default via sandbox
106	>	self.output_file_sandbox.append(self.fjrFileName)
107	>
108	>	# other output files to be returned via sandbox or copied to SE
109	>	self.output_file = []
110	>	tmp = cfg_params.get('CMSSW.output_file',None)
111	>	if tmp :
112	>	tmpOutFiles = string.split(tmp,',')
113	>	log.debug(7, 'cmssw::cmssw(): output files '+str(tmpOutFiles))
114	>	for tmp in tmpOutFiles:
115	>	tmp=string.strip(tmp)
116	>	self.output_file.append(tmp)
117		pass
118	<	pass
119	<	except KeyError:
120	<	log.message("No output file defined: only stdout/err will be available")
104	<	pass
118	>	else:
119	>	log.message("No output file defined: only stdout/err and the CRAB Framework Job Report will be available\n")
120	>	pass
121
122		# script_exe file as additional file in inputSandbox
123	<	try:
124	<	self.scriptExe = cfg_params['USER.script_exe']
125	<	self.additional_inbox_files.append(self.scriptExe)
126	<	except KeyError:
127	<	pass
128	<	if self.scriptExe != '':
129	<	if os.path.isfile(self.scriptExe):
130	<	pass
131	<	else:
132	<	log.message("WARNING. file "+self.scriptExe+" not found")
133	<	sys.exit()
134	<
123	>	self.scriptExe = cfg_params.get('USER.script_exe',None)
124	>	if self.scriptExe :
125	>	if not os.path.isfile(self.scriptExe):
126	>	msg ="ERROR. file "+self.scriptExe+" not found"
127	>	raise CrabException(msg)
128	>	self.additional_inbox_files.append(string.strip(self.scriptExe))
129	>
130	>	#CarlosDaniele
131	>	if self.datasetPath == None and self.pset == None and self.scriptExe == '' :
132	>	msg ="Error. script_exe  not defined"
133	>	raise CrabException(msg)
134	>
135		## additional input files
136	<	try:
137	<	tmpAddFiles = string.split(cfg_params['CMSSW.additional_input_files'],',')
136	>	if cfg_params.has_key('USER.additional_input_files'):
137	>	tmpAddFiles = string.split(cfg_params['USER.additional_input_files'],',')
138		for tmp in tmpAddFiles:
139	<	if not os.path.exists(tmp):
140	<	raise CrabException("Additional input file not found: "+tmp)
141	<	tmp=string.strip(tmp)
142	<	self.additional_inbox_files.append(tmp)
139	>	tmp = string.strip(tmp)
140	>	dirname = ''
141	>	if not tmp[0]=="/": dirname = "."
142	>	files = []
143	>	if string.find(tmp,"*")>-1:
144	>	files = glob.glob(os.path.join(dirname, tmp))
145	>	if len(files)==0:
146	>	raise CrabException("No additional input file found with this pattern: "+tmp)
147	>	else:
148	>	files.append(tmp)
149	>	for file in files:
150	>	if not os.path.exists(file):
151	>	raise CrabException("Additional input file not found: "+file)
152	>	pass
153	>	# fname = string.split(file, '/')[-1]
154	>	# storedFile = common.work_space.pathForTgz()+'share/'+fname
155	>	# shutil.copyfile(file, storedFile)
156	>	self.additional_inbox_files.append(string.strip(file))
157		pass
158		pass
159	<	except KeyError:
160	<	pass
131	<
132	<	try:
133	<	self.filesPerJob = int(cfg_params['CMSSW.files_per_jobs']) #Daniele
134	<	except KeyError:
135	<	self.filesPerJob = 1
136	<
137	<	## Max event   will be total_number_of_events ???  Daniele
138	<	try:
139	<	self.maxEv = cfg_params['CMSSW.event_per_job']
140	<	except KeyError:
141	<	self.maxEv = "-1"
142	<	##
143	<	try:
144	<	self.total_number_of_events = int(cfg_params['CMSSW.total_number_of_events'])
145	<	except KeyError:
146	<	msg = 'Must define total_number_of_events'
147	<	raise CrabException(msg)
148	<
149	<	CEBlackList = []
150	<	try:
151	<	tmpBad = string.split(cfg_params['EDG.ce_black_list'],',')
152	<	for tmp in tmpBad:
153	<	tmp=string.strip(tmp)
154	<	CEBlackList.append(tmp)
155	<	except KeyError:
156	<	pass
159	>	common.logger.debug(5,"Additional input files: "+str(self.additional_inbox_files))
160	>	pass
161
162	<	self.reCEBlackList=[]
163	<	for bad in CEBlackList:
164	<	self.reCEBlackList.append(re.compile( bad ))
162	>	## Events per job
163	>	if cfg_params.has_key('CMSSW.events_per_job'):
164	>	self.eventsPerJob =int( cfg_params['CMSSW.events_per_job'])
165	>	self.selectEventsPerJob = 1
166	>	else:
167	>	self.eventsPerJob = -1
168	>	self.selectEventsPerJob = 0
169
170	<	common.logger.debug(5,'CEBlackList: '+str(CEBlackList))
170	>	## number of jobs
171	>	if cfg_params.has_key('CMSSW.number_of_jobs'):
172	>	self.theNumberOfJobs =int( cfg_params['CMSSW.number_of_jobs'])
173	>	self.selectNumberOfJobs = 1
174	>	else:
175	>	self.theNumberOfJobs = 0
176	>	self.selectNumberOfJobs = 0
177
178	<	CEWhiteList = []
179	<	try:
180	<	tmpGood = string.split(cfg_params['EDG.ce_white_list'],',')
181	<	for tmp in tmpGood:
182	<	tmp=string.strip(tmp)
183	<	CEWhiteList.append(tmp)
170	<	except KeyError:
171	<	pass
178	>	if cfg_params.has_key('CMSSW.total_number_of_events'):
179	>	self.total_number_of_events = int(cfg_params['CMSSW.total_number_of_events'])
180	>	self.selectTotalNumberEvents = 1
181	>	else:
182	>	self.total_number_of_events = 0
183	>	self.selectTotalNumberEvents = 0
184
185	<	#print 'CEWhiteList: ',CEWhiteList
186	<	self.reCEWhiteList=[]
187	<	for Good in CEWhiteList:
188	<	self.reCEWhiteList.append(re.compile( Good ))
185	>	if self.pset != None: #CarlosDaniele
186	>	if ( (self.selectTotalNumberEvents + self.selectEventsPerJob + self.selectNumberOfJobs) != 2 ):
187	>	msg = 'Must define exactly two of total_number_of_events, events_per_job, or number_of_jobs.'
188	>	raise CrabException(msg)
189	>	else:
190	>	if (self.selectNumberOfJobs == 0):
191	>	msg = 'Must specify  number_of_jobs.'
192	>	raise CrabException(msg)
193	>
194	>	## New method of dealing with seeds
195	>	self.incrementSeeds = []
196	>	self.preserveSeeds = []
197	>	if cfg_params.has_key('CMSSW.preserve_seeds'):
198	>	tmpList = cfg_params['CMSSW.preserve_seeds'].split(',')
199	>	for tmp in tmpList:
200	>	tmp.strip()
201	>	self.preserveSeeds.append(tmp)
202	>	if cfg_params.has_key('CMSSW.increment_seeds'):
203	>	tmpList = cfg_params['CMSSW.increment_seeds'].split(',')
204	>	for tmp in tmpList:
205	>	tmp.strip()
206	>	self.incrementSeeds.append(tmp)
207	>
208	>	## Old method of dealing with seeds
209	>	## FUTURE: This is for old CMSSW and old CRAB. Can throw exceptions after a couple of CRAB releases and then
210	>	## remove
211	>	self.sourceSeed = cfg_params.get('CMSSW.pythia_seed',None)
212	>	if self.sourceSeed:
213	>	print "pythia_seed is a deprecated parameter. Use preserve_seeds or increment_seeds in the future.\n","Added to increment_seeds."
214	>	self.incrementSeeds.append('sourceSeed')
215	>
216	>	self.sourceSeedVtx = cfg_params.get('CMSSW.vtx_seed',None)
217	>	if self.sourceSeedVtx:
218	>	print "vtx_seed is a deprecated parameter. Use preserve_seeds or increment_seeds in the future.\n","Added to increment_seeds."
219	>	self.incrementSeeds.append('VtxSmeared')
220	>
221	>	self.sourceSeedG4 = cfg_params.get('CMSSW.g4_seed',None)
222	>	if self.sourceSeedG4:
223	>	print "g4_seed is a deprecated parameter. Use preserve_seeds or increment_seeds in the future.\n","Added to increment_seeds."
224	>	self.incrementSeeds.append('g4SimHits')
225	>
226	>	self.sourceSeedMix = cfg_params.get('CMSSW.mix_seed',None)
227	>	if self.sourceSeedMix:
228	>	print "mix_seed is a deprecated parameter. Use preserve_seeds or increment_seeds in the future.\n","Added to increment_seeds."
229	>	self.incrementSeeds.append('mix')
230	>
231	>	self.firstRun = cfg_params.get('CMSSW.first_run',None)
232	>
233	>	if self.pset != None: #CarlosDaniele
234	>	import PsetManipulator as pp
235	>	PsetEdit = pp.PsetManipulator(self.pset) #Daniele Pset
236
237	<	common.logger.debug(5,'CEWhiteList: '+str(CEWhiteList))
237	>	# Copy/return
238
239	<	self.PsetEdit = PsetManipulator.PsetManipulator(self.pset) #Daniele Pset
239	>	self.copy_data = int(cfg_params.get('USER.copy_data',0))
240	>	self.return_data = int(cfg_params.get('USER.return_data',0))
241
242		#DBSDLS-start
243	<	## Initialize the variables that are extracted from DBS/DLS and needed in other places of the code
243	>	## Initialize the variables that are extracted from DBS/DLS and needed in other places of the code
244		self.maxEvents=0  # max events available   ( --> check the requested nb. of evts in Creator.py)
245		self.DBSPaths={}  # all dbs paths requested ( --> input to the site local discovery script)
246	+	self.jobDestination=[]  # Site destination(s) for each job (list of lists)
247		## Perform the data location and discovery (based on DBS/DLS)
248	<	self.DataDiscoveryAndLocation(cfg_params)
249	<	#DBSDLS-end
248	>	## SL: Don't if NONE is specified as input (pythia use case)
249	>	blockSites = {}
250	>	if self.datasetPath:
251	>	blockSites = self.DataDiscoveryAndLocation(cfg_params)
252	>	#DBSDLS-end
253
190	–	self.tgzNameWithPath = self.getTarBall(self.executable)
254
255	<	self.jobSplitting()  #Daniele job Splitting
256	<	self.PsetEdit.maxEvent(self.maxEv) #Daniele
257	<	self.PsetEdit.inputModule("INPUT") #Daniele
258	<	self.PsetEdit.psetWriter(self.configFilename())
255	>	## Select Splitting
256	>	if self.selectNoInput:
257	>	if self.pset == None: #CarlosDaniele
258	>	self.jobSplittingForScript()
259	>	else:
260	>	self.jobSplittingNoInput()
261	>	else:
262	>	self.jobSplittingByBlocks(blockSites)
263	>
264	>	# modify Pset
265	>	if self.pset != None: #CarlosDaniele
266	>	try:
267	>	# Add FrameworkJobReport to parameter-set, set max events.
268	>	# Reset later for data jobs by writeCFG which does all modifications
269	>	PsetEdit.addCrabFJR(self.fjrFileName) # FUTURE: Job report addition not needed by CMSSW>1.5
270	>	PsetEdit.maxEvent(self.eventsPerJob)
271	>	PsetEdit.psetWriter(self.configFilename())
272	>	except:
273	>	msg='Error while manipuliating ParameterSet: exiting...'
274	>	raise CrabException(msg)
275	>	self.tgzNameWithPath = self.getTarBall(self.executable)
276
277		def DataDiscoveryAndLocation(self, cfg_params):
278
279	+	import DataDiscovery
280	+	import DataLocation
281		common.logger.debug(10,"CMSSW::DataDiscoveryAndLocation()")
282
201	–	#datasetPath = "/"+self.owner+"/"+self.dataTiers[0]+"/"+self.dataset
202	–
283		datasetPath=self.datasetPath
284
205	–	## TODO
206	–	dataTiersList = ""
207	–	dataTiers = dataTiersList.split(',')
208	–
285		## Contact the DBS
286	+	common.logger.message("Contacting Data Discovery Services ...")
287		try:
288	<	self.pubdata=DataDiscovery_EDM.DataDiscovery_EDM(datasetPath, dataTiers, cfg_params)
288	>	self.pubdata=DataDiscovery.DataDiscovery(datasetPath, cfg_params)
289		self.pubdata.fetchDBSInfo()
290
291	<	except DataDiscovery_EDM.NotExistingDatasetError, ex :
291	>	except DataDiscovery.NotExistingDatasetError, ex :
292		msg = 'ERROR ***: failed Data Discovery in DBS : %s'%ex.getErrorMessage()
293		raise CrabException(msg)
294	<
218	<	except DataDiscovery_EDM.NoDataTierinProvenanceError, ex :
294	>	except DataDiscovery.NoDataTierinProvenanceError, ex :
295		msg = 'ERROR ***: failed Data Discovery in DBS : %s'%ex.getErrorMessage()
296		raise CrabException(msg)
297	<	except DataDiscovery_EDM.DataDiscoveryError, ex:
298	<	msg = 'ERROR ***: failed Data Discovery in DBS  %s'%ex.getErrorMessage()
297	>	except DataDiscovery.DataDiscoveryError, ex:
298	>	msg = 'ERROR ***: failed Data Discovery in DBS :  %s'%ex.getErrorMessage()
299		raise CrabException(msg)
300
301	<	## get list of all required data in the form of dbs paths  (dbs path = /dataset/datatier/owner)
302	<	## self.DBSPaths=self.pubdata.getDBSPaths()
303	<	common.logger.message("Required data are :"+self.datasetPath)
228	<
229	<	filesbyblock=self.pubdata.getFiles()
230	<	self.AllInputFiles=filesbyblock.values()
231	<	self.files = self.AllInputFiles
232	<
233	<	## TEMP
234	<	#    self.filesTmp = filesbyblock.values()
235	<	#    self.files = []
236	<	#    locPath='rfio:cmsbose2.bo.infn.it:/flatfiles/SE00/cms/fanfani/ProdTest/'
237	<	#    locPath=''
238	<	#    tmp = []
239	<	#    for file in self.filesTmp[0]:
240	<	#        tmp.append(locPath+file)
241	<	#    self.files.append(tmp)
242	<	## END TEMP
301	>	self.filesbyblock=self.pubdata.getFiles()
302	>	self.eventsbyblock=self.pubdata.getEventsPerBlock()
303	>	self.eventsbyfile=self.pubdata.getEventsPerFile()
304
305		## get max number of events
306	<	#common.logger.debug(10,"number of events for primary fileblocks %i"%self.pubdata.getMaxEvents())
246	<	self.maxEvents=self.pubdata.getMaxEvents() ##  self.maxEvents used in Creator.py
247	<	common.logger.message("\nThe number of available events is %s"%self.maxEvents)
306	>	self.maxEvents=self.pubdata.getMaxEvents() ##  self.maxEvents used in Creator.py
307
308		## Contact the DLS and build a list of sites hosting the fileblocks
309		try:
310	<	dataloc=DataLocation_EDM.DataLocation_EDM(filesbyblock.keys(),cfg_params)
311	<	dataloc.fetchDLSInfo()
312	<
254	<	except DataLocation_EDM.DataLocationError , ex:
310	>	dataloc=DataLocation.DataLocation(self.filesbyblock.keys(),cfg_params)
311	>	dataloc.fetchDLSInfo()
312	>	except DataLocation.DataLocationError , ex:
313		msg = 'ERROR ***: failed Data Location in DLS \n %s '%ex.getErrorMessage()
314		raise CrabException(msg)
257	–
258	–	allsites=dataloc.getSites()
259	–	common.logger.debug(5,"sites are %s"%allsites)
260	–	sites=self.checkBlackList(allsites)
261	–	common.logger.debug(5,"sites are (after black list) %s"%sites)
262	–	sites=self.checkWhiteList(sites)
263	–	common.logger.debug(5,"sites are (after white list) %s"%sites)
315
316	<	if len(sites)==0:
317	<	msg = 'No sites hosting all the needed data! Exiting... '
316	>
317	>	sites = dataloc.getSites()
318	>	allSites = []
319	>	listSites = sites.values()
320	>	for listSite in listSites:
321	>	for oneSite in listSite:
322	>	allSites.append(oneSite)
323	>	allSites = self.uniquelist(allSites)
324	>
325	>	# screen output
326	>	common.logger.message("Requested dataset: " + datasetPath + " has " + str(self.maxEvents) + " events in " + str(len(self.filesbyblock.keys())) + " blocks.\n")
327	>
328	>	return sites
329	>
330	>	# to Be Removed  DS -- BL
331	>	#  def setArgsList(self, argsList):
332	>	#      self.argsList = argsList
333	>
334	>	def jobSplittingByBlocks(self, blockSites):
335	>	"""
336	>	Perform job splitting. Jobs run over an integer number of files
337	>	and no more than one block.
338	>	ARGUMENT: blockSites: dictionary with blocks as keys and list of host sites as values
339	>	REQUIRES: self.selectTotalNumberEvents, self.selectEventsPerJob, self.selectNumberofJobs,
340	>	self.total_number_of_events, self.eventsPerJob, self.theNumberOfJobs,
341	>	self.maxEvents, self.filesbyblock
342	>	SETS: self.jobDestination - Site destination(s) for each job (a list of lists)
343	>	self.total_number_of_jobs - Total # of jobs
344	>	self.list_of_args - File(s) job will run on (a list of lists)
345	>	"""
346	>
347	>	# ---- Handle the possible job splitting configurations ---- #
348	>	if (self.selectTotalNumberEvents):
349	>	totalEventsRequested = self.total_number_of_events
350	>	if (self.selectEventsPerJob):
351	>	eventsPerJobRequested = self.eventsPerJob
352	>	if (self.selectNumberOfJobs):
353	>	totalEventsRequested = self.theNumberOfJobs * self.eventsPerJob
354	>
355	>	# If user requested all the events in the dataset
356	>	if (totalEventsRequested == -1):
357	>	eventsRemaining=self.maxEvents
358	>	# If user requested more events than are in the dataset
359	>	elif (totalEventsRequested > self.maxEvents):
360	>	eventsRemaining = self.maxEvents
361	>	common.logger.message("Requested "+str(self.total_number_of_events)+ " events, but only "+str(self.maxEvents)+" events are available.")
362	>	# If user requested less events than are in the dataset
363	>	else:
364	>	eventsRemaining = totalEventsRequested
365	>
366	>	# If user requested more events per job than are in the dataset
367	>	if (self.selectEventsPerJob and eventsPerJobRequested > self.maxEvents):
368	>	eventsPerJobRequested = self.maxEvents
369	>
370	>	# For user info at end
371	>	totalEventCount = 0
372	>
373	>	if (self.selectTotalNumberEvents and self.selectNumberOfJobs):
374	>	eventsPerJobRequested = int(eventsRemaining/self.theNumberOfJobs)
375	>
376	>	if (self.selectNumberOfJobs):
377	>	common.logger.message("May not create the exact number_of_jobs requested.")
378	>
379	>	if ( self.ncjobs == 'all' ) :
380	>	totalNumberOfJobs = 999999999
381	>	else :
382	>	totalNumberOfJobs = self.ncjobs
383	>
384	>	blocks = blockSites.keys()
385	>	blockCount = 0
386	>	# Backup variable in case self.maxEvents counted events in a non-included block
387	>	numBlocksInDataset = len(blocks)
388	>
389	>	jobCount = 0
390	>	list_of_lists = []
391	>
392	>	# list tracking which jobs are in which jobs belong to which block
393	>	jobsOfBlock = {}
394	>
395	>	# ---- Iterate over the blocks in the dataset until ---- #
396	>	# ---- we've met the requested total # of events    ---- #
397	>	while ( (eventsRemaining > 0) and (blockCount < numBlocksInDataset) and (jobCount < totalNumberOfJobs)):
398	>	block = blocks[blockCount]
399	>	blockCount += 1
400	>	if block not in jobsOfBlock.keys() :
401	>	jobsOfBlock[block] = []
402	>
403	>	if self.eventsbyblock.has_key(block) :
404	>	numEventsInBlock = self.eventsbyblock[block]
405	>	common.logger.debug(5,'Events in Block File '+str(numEventsInBlock))
406	>
407	>	files = self.filesbyblock[block]
408	>	numFilesInBlock = len(files)
409	>	if (numFilesInBlock <= 0):
410	>	continue
411	>	fileCount = 0
412	>
413	>	# ---- New block => New job ---- #
414	>	parString = ""
415	>	# counter for number of events in files currently worked on
416	>	filesEventCount = 0
417	>	# flag if next while loop should touch new file
418	>	newFile = 1
419	>	# job event counter
420	>	jobSkipEventCount = 0
421	>
422	>	# ---- Iterate over the files in the block until we've met the requested ---- #
423	>	# ---- total # of events or we've gone over all the files in this block  ---- #
424	>	while ( (eventsRemaining > 0) and (fileCount < numFilesInBlock) and (jobCount < totalNumberOfJobs) ):
425	>	file = files[fileCount]
426	>	if newFile :
427	>	try:
428	>	numEventsInFile = self.eventsbyfile[file]
429	>	common.logger.debug(6, "File "+str(file)+" has "+str(numEventsInFile)+" events")
430	>	# increase filesEventCount
431	>	filesEventCount += numEventsInFile
432	>	# Add file to current job
433	>	parString += '\\\"' + file + '\\\"\,'
434	>	newFile = 0
435	>	except KeyError:
436	>	common.logger.message("File "+str(file)+" has unknown number of events: skipping")
437	>
438	>	eventsPerJobRequested = min(eventsPerJobRequested, eventsRemaining)
439	>	# if less events in file remain than eventsPerJobRequested
440	>	if ( filesEventCount - jobSkipEventCount < eventsPerJobRequested):
441	>	# if last file in block
442	>	if ( fileCount == numFilesInBlock-1 ) :
443	>	# end job using last file, use remaining events in block
444	>	# close job and touch new file
445	>	fullString = parString[:-2]
446	>	list_of_lists.append([fullString,str(-1),str(jobSkipEventCount)])
447	>	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(filesEventCount - jobSkipEventCount)+" events (last file in block).")
448	>	self.jobDestination.append(blockSites[block])
449	>	common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
450	>	# fill jobs of block dictionary
451	>	jobsOfBlock[block].append(jobCount+1)
452	>	# reset counter
453	>	jobCount = jobCount + 1
454	>	totalEventCount = totalEventCount + filesEventCount - jobSkipEventCount
455	>	eventsRemaining = eventsRemaining - filesEventCount + jobSkipEventCount
456	>	jobSkipEventCount = 0
457	>	# reset file
458	>	parString = ""
459	>	filesEventCount = 0
460	>	newFile = 1
461	>	fileCount += 1
462	>	else :
463	>	# go to next file
464	>	newFile = 1
465	>	fileCount += 1
466	>	# if events in file equal to eventsPerJobRequested
467	>	elif ( filesEventCount - jobSkipEventCount == eventsPerJobRequested ) :
468	>	# close job and touch new file
469	>	fullString = parString[:-2]
470	>	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
471	>	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.")
472	>	self.jobDestination.append(blockSites[block])
473	>	common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
474	>	jobsOfBlock[block].append(jobCount+1)
475	>	# reset counter
476	>	jobCount = jobCount + 1
477	>	totalEventCount = totalEventCount + eventsPerJobRequested
478	>	eventsRemaining = eventsRemaining - eventsPerJobRequested
479	>	jobSkipEventCount = 0
480	>	# reset file
481	>	parString = ""
482	>	filesEventCount = 0
483	>	newFile = 1
484	>	fileCount += 1
485	>
486	>	# if more events in file remain than eventsPerJobRequested
487	>	else :
488	>	# close job but don't touch new file
489	>	fullString = parString[:-2]
490	>	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
491	>	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.")
492	>	self.jobDestination.append(blockSites[block])
493	>	common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
494	>	jobsOfBlock[block].append(jobCount+1)
495	>	# increase counter
496	>	jobCount = jobCount + 1
497	>	totalEventCount = totalEventCount + eventsPerJobRequested
498	>	eventsRemaining = eventsRemaining - eventsPerJobRequested
499	>	# calculate skip events for last file
500	>	# use filesEventCount (contains several files), jobSkipEventCount and eventsPerJobRequest
501	>	jobSkipEventCount = eventsPerJobRequested - (filesEventCount - jobSkipEventCount - self.eventsbyfile[file])
502	>	# remove all but the last file
503	>	filesEventCount = self.eventsbyfile[file]
504	>	parString = '\\\"' + file + '\\\"\,'
505	>	pass # END if
506	>	pass # END while (iterate over files in the block)
507	>	pass # END while (iterate over blocks in the dataset)
508	>	self.ncjobs = self.total_number_of_jobs = jobCount
509	>	if (eventsRemaining > 0 and jobCount < totalNumberOfJobs ):
510	>	common.logger.message("Could not run on all requested events because some blocks not hosted at allowed sites.")
511	>	common.logger.message(str(jobCount)+" job(s) can run on "+str(totalEventCount)+" events.\n")
512	>
513	>	# screen output
514	>	screenOutput = "List of jobs and available destination sites:\n\n"
515	>
516	>	# keep trace of block with no sites to print a warning at the end
517	>	noSiteBlock = []
518	>	bloskNoSite = []
519	>
520	>	blockCounter = 0
521	>	for block in blocks:
522	>	if block in jobsOfBlock.keys() :
523	>	blockCounter += 1
524	>	screenOutput += "Block %5i: jobs %20s: sites: %s\n" % (blockCounter,spanRanges(jobsOfBlock[block]),
525	>	','.join(self.blackWhiteListParser.checkWhiteList(self.blackWhiteListParser.checkBlackList(blockSites[block],block),block)))
526	>	if len(self.blackWhiteListParser.checkWhiteList(self.blackWhiteListParser.checkBlackList(blockSites[block],block),block)) == 0:
527	>	noSiteBlock.append( spanRanges(jobsOfBlock[block]) )
528	>	bloskNoSite.append( blockCounter )
529	>
530	>	common.logger.message(screenOutput)
531	>	if len(noSiteBlock) > 0 and len(bloskNoSite) > 0:
532	>	msg = 'WARNING: No sites are hosting any part of data for block:\n                '
533	>	virgola = ""
534	>	if len(bloskNoSite) > 1:
535	>	virgola = ","
536	>	for block in bloskNoSite:
537	>	msg += ' ' + str(block) + virgola
538	>	msg += '\n               Related jobs:\n                 '
539	>	virgola = ""
540	>	if len(noSiteBlock) > 1:
541	>	virgola = ","
542	>	for range_jobs in noSiteBlock:
543	>	msg += str(range_jobs) + virgola
544	>	msg += '\n               will not be submitted and this block of data can not be analyzed!\n'
545	>	if self.cfg_params.has_key('EDG.se_white_list'):
546	>	msg += 'WARNING: SE White List: '+self.cfg_params['EDG.se_white_list']+'\n'
547	>	msg += '(Hint: By whitelisting you force the job to run at this particular site(s).\n'
548	>	msg += 'Please check if the dataset is available at this site!)\n'
549	>	if self.cfg_params.has_key('EDG.ce_white_list'):
550	>	msg += 'WARNING: CE White List: '+self.cfg_params['EDG.ce_white_list']+'\n'
551	>	msg += '(Hint: By whitelisting you force the job to run at this particular site(s).\n'
552	>	msg += 'Please check if the dataset is available at this site!)\n'
553	>
554	>	common.logger.message(msg)
555	>
556	>	self.list_of_args = list_of_lists
557	>	return
558	>
559	>	def jobSplittingNoInput(self):
560	>	"""
561	>	Perform job splitting based on number of event per job
562	>	"""
563	>	common.logger.debug(5,'Splitting per events')
564	>
565	>	if (self.selectEventsPerJob):
566	>	common.logger.message('Required '+str(self.eventsPerJob)+' events per job ')
567	>	if (self.selectNumberOfJobs):
568	>	common.logger.message('Required '+str(self.theNumberOfJobs)+' jobs in total ')
569	>	if (self.selectTotalNumberEvents):
570	>	common.logger.message('Required '+str(self.total_number_of_events)+' events in total ')
571	>
572	>	if (self.total_number_of_events < 0):
573	>	msg='Cannot split jobs per Events with "-1" as total number of events'
574		raise CrabException(msg)
575
576	<	common.logger.message("List of Sites hosting the data : "+str(sites))
577	<	common.logger.debug(6, "List of Sites: "+str(sites))
578	<	common.analisys_common_info['sites']=sites    ## used in SchedulerEdg.py in createSchScript
576	>	if (self.selectEventsPerJob):
577	>	if (self.selectTotalNumberEvents):
578	>	self.total_number_of_jobs = int(self.total_number_of_events/self.eventsPerJob)
579	>	elif(self.selectNumberOfJobs) :
580	>	self.total_number_of_jobs =self.theNumberOfJobs
581	>	self.total_number_of_events =int(self.theNumberOfJobs*self.eventsPerJob)
582	>
583	>	elif (self.selectNumberOfJobs) :
584	>	self.total_number_of_jobs = self.theNumberOfJobs
585	>	self.eventsPerJob = int(self.total_number_of_events/self.total_number_of_jobs)
586	>
587	>	common.logger.debug(5,'N jobs  '+str(self.total_number_of_jobs))
588	>
589	>	# is there any remainder?
590	>	check = int(self.total_number_of_events) - (int(self.total_number_of_jobs)*self.eventsPerJob)
591	>
592	>	common.logger.debug(5,'Check  '+str(check))
593	>
594	>	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')
595	>	if check > 0:
596	>	common.logger.message('Warning: asked '+str(self.total_number_of_events)+' but can do only '+str(int(self.total_number_of_jobs)*self.eventsPerJob))
597	>
598	>	# argument is seed number.$i
599	>	self.list_of_args = []
600	>	for i in range(self.total_number_of_jobs):
601	>	## Since there is no input, any site is good
602	>	self.jobDestination.append([""]) #must be empty to write correctly the xml
603	>	args=[]
604	>	if (self.firstRun):
605	>	## pythia first run
606	>	args.append(str(self.firstRun)+str(i))
607	>	self.list_of_args.append(args)
608	>
609		return
610	<
611	<	def jobSplitting(self):
610	>
611	>
612	>	def jobSplittingForScript(self):#CarlosDaniele
613		"""
614	<	first implemntation for job splitting
615	<	"""
616	<	#  print 'eventi totali '+str(self.maxEvents)
617	<	#  print 'eventi totali richiesti dallo user '+str(self.total_number_of_events)
280	<	#print 'files per job '+str(self.filesPerJob)
281	<	common.logger.message('Required '+str(self.filesPerJob)+' files per job ')
282	<	common.logger.message('Required '+str(self.total_number_of_events)+' events in total ')
283	<
284	<	## TODO: SL need to have (from DBS) a detailed list of how many events per each file
285	<	n_tot_files = (len(self.files[0]))
286	<	## SL: this is wrong if the files have different number of events
287	<	evPerFile = int(self.maxEvents)/n_tot_files
288	<
289	<	common.logger.debug(5,'Events per File '+str(evPerFile))
290	<
291	<	## if asked to process all events, do it
292	<	if self.total_number_of_events == -1:
293	<	self.total_number_of_events=self.maxEvents
294	<	self.total_number_of_jobs = int(n_tot_files)*1/int(self.filesPerJob)
295	<	common.logger.message(str(self.total_number_of_jobs)+' jobs will be created for all available events '+str(self.total_number_of_events)+' events')
296	<
297	<	else:
298	<	self.total_number_of_files = int(self.total_number_of_events/evPerFile)
299	<	## SL: if ask for less event than what is computed to be available on a
300	<	##     file, process the first file anyhow.
301	<	if self.total_number_of_files == 0:
302	<	self.total_number_of_files = self.total_number_of_files + 1
303	<
304	<	common.logger.debug(5,'N files  '+str(self.total_number_of_files))
305	<
306	<	check = 0
307	<
308	<	## Compute the number of jobs
309	<	#self.total_number_of_jobs = int(n_tot_files)*1/int(self.filesPerJob)
310	<	self.total_number_of_jobs = int(self.total_number_of_files/self.filesPerJob)
311	<	common.logger.debug(5,'N jobs  '+str(self.total_number_of_jobs))
312	<
313	<	## is there any remainder?
314	<	check = int(self.total_number_of_files) - (int(self.total_number_of_jobs)*self.filesPerJob)
315	<
316	<	common.logger.debug(5,'Check  '+str(check))
317	<
318	<	if check > 0:
319	<	self.total_number_of_jobs =  self.total_number_of_jobs + 1
320	<	common.logger.message('Warning: last job will be created with '+str(check)+' files')
614	>	Perform job splitting based on number of job
615	>	"""
616	>	common.logger.debug(5,'Splitting per job')
617	>	common.logger.message('Required '+str(self.theNumberOfJobs)+' jobs in total ')
618
619	<	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')
323	<	pass
619	>	self.total_number_of_jobs = self.theNumberOfJobs
620
621	<	list_of_lists = []
326	<	for i in xrange(0, int(n_tot_files), self.filesPerJob):
327	<	list_of_lists.append(self.files[0][i: i+self.filesPerJob])
621	>	common.logger.debug(5,'N jobs  '+str(self.total_number_of_jobs))
622
623	<	self.list_of_files = list_of_lists
624	<
623	>	common.logger.message(str(self.total_number_of_jobs)+' jobs can be created')
624	>
625	>	# argument is seed number.$i
626	>	self.list_of_args = []
627	>	for i in range(self.total_number_of_jobs):
628	>	## Since there is no input, any site is good
629	>	# self.jobDestination.append(["Any"])
630	>	self.jobDestination.append([""])
631	>	## no random seed
632	>	self.list_of_args.append([str(i)])
633		return
634
635		def split(self, jobParams):
636	<
335	<	common.jobDB.load()
636	>
637		#### Fabio
638		njobs = self.total_number_of_jobs
639	<	filelist = self.list_of_files
639	>	arglist = self.list_of_args
640		# create the empty structure
641		for i in range(njobs):
642		jobParams.append("")
643	<
643	>
644	>	listID=[]
645	>	listField=[]
646		for job in range(njobs):
647	<	jobParams[job] = filelist[job]
648	<	common.jobDB.setArguments(job, jobParams[job])
647	>	jobParams[job] = arglist[job]
648	>	listID.append(job+1)
649	>	job_ToSave ={}
650	>	concString = ' '
651	>	argu=''
652	>	if len(jobParams[job]):
653	>	argu +=   concString.join(jobParams[job] )
654	>	job_ToSave['arguments']= str(job+1)+' '+argu## new BL--DS
655	>	job_ToSave['dlsDestination']= self.jobDestination[job]## new BL--DS
656	>	#common._db.updateJob_(job,job_ToSave)## new BL--DS
657	>	listField.append(job_ToSave)
658	>	msg="Job "+str(job)+" Arguments:   "+str(job+1)+" "+argu+"\n"  \
659	>	+"                     Destination: "+str(self.jobDestination[job])
660	>	common.logger.debug(5,msg)
661	>	#common.logger.debug(5,"Job "+str(job)+" Destination: "+str(self.jobDestination[job]))
662	>	common._db.updateJob_(listID,listField)## new BL--DS
663	>	## Pay Attention Here....DS--BL
664	>	self.argsList = (len(jobParams[0])+1)
665
347	–	common.jobDB.save()
666		return
667	<
350	<	def getJobTypeArguments(self, nj, sched):
351	<	params = common.jobDB.arguments(nj)
352	<	#print params
353	<	parString = "\\{"
354	<
355	<	for i in range(len(params) - 1):
356	<	parString += '\\\"' + params[i] + '\\\"\,'
357	<
358	<	parString += '\\\"' + params[len(params) - 1] + '\\\"\\}'
359	<	return parString
360	<
667	>
668		def numberOfJobs(self):
669		# Fabio
363	–
670		return self.total_number_of_jobs
365	–
366	–
367	–
368	–	def checkBlackList(self, allSites):
369	–	if len(self.reCEBlackList)==0: return allSites
370	–	sites = []
371	–	for site in allSites:
372	–	common.logger.debug(10,'Site '+site)
373	–	good=1
374	–	for re in self.reCEBlackList:
375	–	if re.search(site):
376	–	common.logger.message('CE in black list, skipping site '+site)
377	–	good=0
378	–	pass
379	–	if good: sites.append(site)
380	–	if len(sites) == 0:
381	–	common.logger.debug(3,"No sites found after BlackList")
382	–	return sites
383	–
384	–	def checkWhiteList(self, allSites):
385	–
386	–	if len(self.reCEWhiteList)==0: return allSites
387	–	sites = []
388	–	for site in allSites:
389	–	good=0
390	–	for re in self.reCEWhiteList:
391	–	if re.search(site):
392	–	common.logger.debug(5,'CE in white list, adding site '+site)
393	–	good=1
394	–	if not good: continue
395	–	sites.append(site)
396	–	if len(sites) == 0:
397	–	common.logger.message("No sites found after WhiteList\n")
398	–	else:
399	–	common.logger.debug(5,"Selected sites via WhiteList are "+str(sites)+"\n")
400	–	return sites
671
672		def getTarBall(self, exe):
673		"""
674		Return the TarBall with lib and exe
675		"""
676	<
676	>
677		# if it exist, just return it
678	<	self.tgzNameWithPath = common.work_space.shareDir()+self.tgz_name
678	>	#
679	>	# Marco. Let's start to use relative path for Boss XML files
680	>	#
681	>	self.tgzNameWithPath = common.work_space.pathForTgz()+'share/'+self.tgz_name
682		if os.path.exists(self.tgzNameWithPath):
683		return self.tgzNameWithPath
684
#	Line 419 | Line 692 | class Cmssw(JobType):
692		# First of all declare the user Scram area
693		swArea = self.scram.getSWArea_()
694		#print "swArea = ", swArea
695	<	swVersion = self.scram.getSWVersion()
696	<	#print "swVersion = ", swVersion
695	>	# swVersion = self.scram.getSWVersion()
696	>	# print "swVersion = ", swVersion
697		swReleaseTop = self.scram.getReleaseTop_()
698		#print "swReleaseTop = ", swReleaseTop
699	<
699	>
700		## check if working area is release top
701		if swReleaseTop == '' or swArea == swReleaseTop:
702	+	common.logger.debug(3,"swArea = "+swArea+" swReleaseTop ="+swReleaseTop)
703		return
704
705	<	filesToBeTarred = []
706	<	## First find the executable
707	<	if (self.executable != ''):
708	<	exeWithPath = self.scram.findFile_(executable)
709	<	#           print exeWithPath
710	<	if ( not exeWithPath ):
711	<	raise CrabException('User executable '+executable+' not found')
712	<
713	<	## then check if it's private or not
714	<	if exeWithPath.find(swReleaseTop) == -1:
715	<	# the exe is private, so we must ship
716	<	common.logger.debug(5,"Exe "+exeWithPath+" to be tarred")
717	<	path = swArea+'/'
718	<	exe = string.replace(exeWithPath, path,'')
719	<	filesToBeTarred.append(exe)
720	<	pass
721	<	else:
722	<	# the exe is from release, we'll find it on WN
723	<	pass
724	<
725	<	## Now get the libraries: only those in local working area
726	<	libDir = 'lib'
727	<	lib = swArea+'/' +libDir
728	<	common.logger.debug(5,"lib "+lib+" to be tarred")
729	<	if os.path.exists(lib):
730	<	filesToBeTarred.append(libDir)
731	<
732	<	## Now check if module dir is present
733	<	moduleDir = 'module'
734	<	if os.path.isdir(swArea+'/'+moduleDir):
735	<	filesToBeTarred.append(moduleDir)
736	<
737	<	## Now check if the Data dir is present
738	<	dataDir = 'src/Data/'
739	<	if os.path.isdir(swArea+'/'+dataDir):
740	<	filesToBeTarred.append(dataDir)
741	<
742	<	## Create the tar-ball
743	<	if len(filesToBeTarred)>0:
744	<	cwd = os.getcwd()
745	<	os.chdir(swArea)
746	<	tarcmd = 'tar zcvf ' + self.tgzNameWithPath + ' '
747	<	for line in filesToBeTarred:
748	<	tarcmd = tarcmd + line + ' '
749	<	cout = runCommand(tarcmd)
750	<	if not cout:
751	<	raise CrabException('Could not create tar-ball')
752	<	os.chdir(cwd)
753	<	else:
754	<	common.logger.debug(5,"No files to be to be tarred")
755	<
756	<	return
757	<
758	<	def wsSetupEnvironment(self, nj):
705	>	import tarfile
706	>	try: # create tar ball
707	>	tar = tarfile.open(self.tgzNameWithPath, "w:gz")
708	>	## First find the executable
709	>	if (self.executable != ''):
710	>	exeWithPath = self.scram.findFile_(executable)
711	>	if ( not exeWithPath ):
712	>	raise CrabException('User executable '+executable+' not found')
713	>
714	>	## then check if it's private or not
715	>	if exeWithPath.find(swReleaseTop) == -1:
716	>	# the exe is private, so we must ship
717	>	common.logger.debug(5,"Exe "+exeWithPath+" to be tarred")
718	>	path = swArea+'/'
719	>	# distinguish case when script is in user project area or given by full path somewhere else
720	>	if exeWithPath.find(path) >= 0 :
721	>	exe = string.replace(exeWithPath, path,'')
722	>	tar.add(path+exe,exe)
723	>	else :
724	>	tar.add(exeWithPath,os.path.basename(executable))
725	>	pass
726	>	else:
727	>	# the exe is from release, we'll find it on WN
728	>	pass
729	>
730	>	## Now get the libraries: only those in local working area
731	>	libDir = 'lib'
732	>	lib = swArea+'/' +libDir
733	>	common.logger.debug(5,"lib "+lib+" to be tarred")
734	>	if os.path.exists(lib):
735	>	tar.add(lib,libDir)
736	>
737	>	## Now check if module dir is present
738	>	moduleDir = 'module'
739	>	module = swArea + '/' + moduleDir
740	>	if os.path.isdir(module):
741	>	tar.add(module,moduleDir)
742	>
743	>	## Now check if any data dir(s) is present
744	>	swAreaLen=len(swArea)
745	>	self.dataExist = False
746	>	for root, dirs, files in os.walk(swArea):
747	>	if "data" in dirs:
748	>	self.dataExist=True
749	>	common.logger.debug(5,"data "+root+"/data"+" to be tarred")
750	>	tar.add(root+"/data",root[swAreaLen:]+"/data")
751	>
752	>	### CMSSW ParameterSet
753	>	if not self.pset is None:
754	>	cfg_file = common.work_space.jobDir()+self.configFilename()
755	>	tar.add(cfg_file,self.configFilename())
756	>	common.logger.debug(5,"File added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
757	>
758	>
759	>	## Add ProdCommon dir to tar
760	>	prodcommonDir = 'ProdCommon'
761	>	prodcommonPath = os.environ['CRABDIR'] + '/' + 'ProdCommon'
762	>	if os.path.isdir(prodcommonPath):
763	>	tar.add(prodcommonPath,prodcommonDir)
764	>	common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
765	>
766	>	##### ML stuff
767	>	ML_file_list=['report.py', 'DashboardAPI.py', 'Logger.py', 'ProcInfo.py', 'apmon.py']
768	>	path=os.environ['CRABDIR'] + '/python/'
769	>	for file in ML_file_list:
770	>	tar.add(path+file,file)
771	>	common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
772	>
773	>	##### Utils
774	>	Utils_file_list=['parseCrabFjr.py','writeCfg.py', 'JobReportErrorCode.py']
775	>	for file in Utils_file_list:
776	>	tar.add(path+file,file)
777	>	common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
778	>
779	>	##### AdditionalFiles
780	>	for file in self.additional_inbox_files:
781	>	tar.add(file,string.split(file,'/')[-1])
782	>	common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
783	>
784	>	tar.close()
785	>	except :
786	>	raise CrabException('Could not create tar-ball')
787	>
788	>	## check for tarball size
789	>	tarballinfo = os.stat(self.tgzNameWithPath)
790	>	if ( tarballinfo.st_size > self.MaxTarBallSize*1024*1024 ) :
791	>	raise CrabException('Input sandbox size of ' + str(float(tarballinfo.st_size)/1024.0/1024.0) + ' MB is larger than the allowed ' + str(self.MaxTarBallSize) + ' MB input sandbox limit and not supported by the used GRID submission system. Please make sure that no unnecessary files are in all data directories in your local CMSSW project area as they are automatically packed into the input sandbox.')
792	>
793	>	## create tar-ball with ML stuff
794	>
795	>	def wsSetupEnvironment(self, nj=0):
796		"""
797		Returns part of a job script which prepares
798		the execution environment for the job 'nj'.
799		"""
800		# Prepare JobType-independent part
801	<	txt = ''
802	<
803	<	## OLI_Daniele at this level  middleware already known
493	<
494	<	txt += 'if [ $middleware == LCG ]; then \n'
801	>	txt = '\n#Written by cms_cmssw::wsSetupEnvironment\n'
802	>	txt += 'echo ">>> setup environment"\n'
803	>	txt += 'if [ $middleware == LCG ]; then \n'
804		txt += self.wsSetupCMSLCGEnvironment_()
805		txt += 'elif [ $middleware == OSG ]; then\n'
806	<	txt += '    time=`date -u +"%s"`\n'
807	<	txt += '    WORKING_DIR=$OSG_WN_TMP/cms_$time\n'
808	<	txt += '    echo "Creating working directory: $WORKING_DIR"\n'
809	<	txt += '    /bin/mkdir -p $WORKING_DIR\n'
810	<	txt += '    if [ ! -d $WORKING_DIR ] ;then\n'
502	<	txt += '        echo "OSG WORKING DIR ==> $WORKING_DIR could not be created on on WN `hostname`"\n'
503	<
504	<	txt += '        echo "JOB_EXIT_STATUS = 1"\n'
505	<	txt += '        exit 1\n'
806	>	txt += '    WORKING_DIR=`/bin/mktemp  -d $OSG_WN_TMP/cms_XXXXXXXXXXXX`\n'
807	>	txt += '    if [ ! $? == 0 ] ;then\n'
808	>	txt += '        echo "ERROR ==> OSG $WORKING_DIR could not be created on WN `hostname`"\n'
809	>	txt += '        job_exit_code=10016\n'
810	>	txt += '        func_exit\n'
811		txt += '    fi\n'
812	+	txt += '    echo ">>> Created working directory: $WORKING_DIR"\n'
813		txt += '\n'
814		txt += '    echo "Change to working directory: $WORKING_DIR"\n'
815		txt += '    cd $WORKING_DIR\n'
816	<	txt += self.wsSetupCMSOSGEnvironment_()
816	>	txt += '    echo ">>> current directory (WORKING_DIR): $WORKING_DIR"\n'
817	>	txt += self.wsSetupCMSOSGEnvironment_()
818	>	#txt += '    echo "### Set SCRAM ARCH to ' + self.executable_arch + ' ###"\n'
819	>	#txt += '    export SCRAM_ARCH='+self.executable_arch+'\n'
820		txt += 'fi\n'
821
822		# Prepare JobType-specific part
823		scram = self.scram.commandName()
824		txt += '\n\n'
825	<	txt += 'echo "### SPECIFIC JOB SETUP ENVIRONMENT ###"\n'
825	>	txt += 'echo ">>> specific cmssw setup environment:"\n'
826	>	txt += 'echo "CMSSW_VERSION =  '+self.version+'"\n'
827		txt += scram+' project CMSSW '+self.version+'\n'
828		txt += 'status=$?\n'
829		txt += 'if [ $status != 0 ] ; then\n'
830	<	txt += '   echo "SET_EXE_ENV 1 ==>ERROR CMSSW '+self.version+' not found on `hostname`" \n'
831	<	txt += '   echo "JOB_EXIT_STATUS = 10034"\n'
832	<	txt += '   echo "SanityCheckCode = 10034" | tee -a $RUNTIME_AREA/$repo\n'
523	<	txt += '   dumpStatus $RUNTIME_AREA/$repo\n'
524	<	## OLI_Daniele
525	<	txt += '    if [ $middleware == OSG ]; then \n'
526	<	txt += '        echo "Remove working directory: $WORKING_DIR"\n'
527	<	txt += '        cd $RUNTIME_AREA\n'
528	<	txt += '        /bin/rm -rf $WORKING_DIR\n'
529	<	txt += '        if [ -d $WORKING_DIR ] ;then\n'
530	<	txt += '            echo "OSG WORKING DIR ==> $WORKING_DIR could not be deleted on on WN `hostname`"\n'
531	<	txt += '        fi\n'
532	<	txt += '    fi \n'
533	<	txt += '   exit 1 \n'
830	>	txt += '    echo "ERROR ==> CMSSW '+self.version+' not found on `hostname`" \n'
831	>	txt += '    job_exit_code=10034\n'
832	>	txt += '    func_exit\n'
833		txt += 'fi \n'
535	–	txt += 'echo "CMSSW_VERSION =  '+self.version+'"\n'
834		txt += 'cd '+self.version+'\n'
835	<	### needed grep for bug in scramv1 ###
835	>	txt += 'SOFTWARE_DIR=`pwd`\n'
836	>	txt += 'echo ">>> current directory (SOFTWARE_DIR): $SOFTWARE_DIR" \n'
837		txt += 'eval `'+scram+' runtime -sh | grep -v SCRAMRT_LSB_JOBNAME`\n'
838	<
838	>	txt += 'if [ $? != 0 ] ; then\n'
839	>	txt += '    echo "ERROR ==> Problem with the command: "\n'
840	>	txt += '    echo "eval \`'+scram+' runtime -sh | grep -v SCRAMRT_LSB_JOBNAME \` at `hostname`"\n'
841	>	txt += '    job_exit_code=10034\n'
842	>	txt += '    func_exit\n'
843	>	txt += 'fi \n'
844		# Handle the arguments:
845		txt += "\n"
846	<	txt += "## ARGUMNETS: $1 Job Number\n"
543	<	# txt += "## ARGUMNETS: $2 First Event for this job\n"
544	<	# txt += "## ARGUMNETS: $3 Max Event for this job\n"
846	>	txt += "## number of arguments (first argument always jobnumber)\n"
847		txt += "\n"
848	<	txt += "narg=$#\n"
547	<	txt += "if [ $narg -lt 2 ]\n"
848	>	txt += "if [ $nargs -lt "+str(self.argsList)+" ]\n"
849		txt += "then\n"
850	<	txt += "    echo 'SET_EXE_ENV 1 ==> ERROR Too few arguments' +$narg+ \n"
851	<	txt += '    echo "JOB_EXIT_STATUS = 50113"\n'
852	<	txt += '    echo "SanityCheckCode = 50113" | tee -a $RUNTIME_AREA/$repo\n'
552	<	txt += '    dumpStatus $RUNTIME_AREA/$repo\n'
553	<	## OLI_Daniele
554	<	txt += '    if [ $middleware == OSG ]; then \n'
555	<	txt += '        echo "Remove working directory: $WORKING_DIR"\n'
556	<	txt += '        cd $RUNTIME_AREA\n'
557	<	txt += '        /bin/rm -rf $WORKING_DIR\n'
558	<	txt += '        if [ -d $WORKING_DIR ] ;then\n'
559	<	txt += '            echo "OSG WORKING DIR ==> $WORKING_DIR could not be deleted on on WN `hostname`"\n'
560	<	txt += '        fi\n'
561	<	txt += '    fi \n'
562	<	txt += "    exit 1\n"
850	>	txt += "    echo 'ERROR ==> Too few arguments' +$nargs+ \n"
851	>	txt += '    job_exit_code=50113\n'
852	>	txt += "    func_exit\n"
853		txt += "fi\n"
854		txt += "\n"
565	–	txt += "NJob=$1\n"
566	–	txt += "InputFiles=$2\n"
567	–	txt += "echo \"<$InputFiles>\"\n"
568	–	# txt += "Args = ` cat $2 |  sed -e \'s/\\\\//g\' -e \'s/\"/\\x27/g\' `"
569	–
570	–	### OLI_DANIELE
571	–	txt += 'if [ $middleware == LCG ]; then \n'
572	–	txt += '    echo "MonitorJobID=`echo ${NJob}_$EDG_WL_JOBID`" | tee -a $RUNTIME_AREA/$repo\n'
573	–	txt += '    echo "SyncGridJobId=`echo $EDG_WL_JOBID`" | tee -a $RUNTIME_AREA/$repo\n'
574	–	txt += '    echo "SyncCE=`edg-brokerinfo getCE`" | tee -a $RUNTIME_AREA/$repo\n'
575	–	txt += 'elif [ $middleware == OSG ]; then\n'
576	–
577	–	# OLI: added monitoring for dashbord, use hash of crab.cfg
578	–	if common.scheduler.boss_scheduler_name == 'condor_g':
579	–	# create hash of cfg file
580	–	hash = makeCksum(common.work_space.cfgFileName())
581	–	txt += '    echo "MonitorJobID=`echo ${NJob}_'+hash+'_$GLOBUS_GRAM_JOB_CONTACT`" | tee -a $RUNTIME_AREA/$repo\n'
582	–	txt += '    echo "SyncGridJobId=`echo $GLOBUS_GRAM_JOB_CONTACT`" | tee -a $RUNTIME_AREA/$repo\n'
583	–	txt += '    echo "SyncCE=`echo $hostname`" | tee -a $RUNTIME_AREA/$repo\n'
584	–	else :
585	–	txt += '    echo "MonitorJobID=`echo ${NJob}_$EDG_WL_JOBID`" | tee -a $RUNTIME_AREA/$repo\n'
586	–	txt += '    echo "SyncGridJobId=`echo $EDG_WL_JOBID`" | tee -a $RUNTIME_AREA/$repo\n'
587	–	txt += '    echo "SyncCE=`$EDG_WL_LOG_DESTINATION`" | tee -a $RUNTIME_AREA/$repo\n'
588	–
589	–	txt += 'fi\n'
590	–	txt += 'dumpStatus $RUNTIME_AREA/$repo\n'
855
856		# Prepare job-specific part
857		job = common.job_list[nj]
858	<	pset = os.path.basename(job.configFilename())
859	<	txt += '\n'
860	<	#txt += 'echo sed "s#{\'INPUT\'}#$InputFiles#" $RUNTIME_AREA/'+pset+' \n'
861	<	txt += 'sed "s#{\'INPUT\'}#$InputFiles#" $RUNTIME_AREA/'+pset+' > pset.cfg\n'
598	<	#txt += 'sed "s#{\'INPUT\'}#${InputFiles}#" $RUNTIME_AREA/'+pset+' > pset1.cfg\n'
858	>	### FEDE FOR DBS OUTPUT PUBLICATION
859	>	if (self.datasetPath):
860	>	txt += '\n'
861	>	txt += 'DatasetPath='+self.datasetPath+'\n'
862
863	<	if len(self.additional_inbox_files) > 0:
601	<	for file in self.additional_inbox_files:
602	<	txt += 'if [ -e $RUNTIME_AREA/'+file+' ] ; then\n'
603	<	txt += '   cp $RUNTIME_AREA/'+file+' .\n'
604	<	txt += '   chmod +x '+file+'\n'
605	<	txt += 'fi\n'
606	<	pass
863	>	datasetpath_split = self.datasetPath.split("/")
864
865	<	txt += 'echo "### END JOB SETUP ENVIRONMENT ###"\n\n'
865	>	txt += 'PrimaryDataset='+datasetpath_split[1]+'\n'
866	>	txt += 'DataTier='+datasetpath_split[2]+'\n'
867	>	txt += 'ApplicationFamily=cmsRun\n'
868
869	<	txt += '\n'
870	<	txt += 'echo "***** cat pset.cfg *********"\n'
871	<	txt += 'cat pset.cfg\n'
872	<	txt += 'echo "****** end pset.cfg ********"\n'
873	<	txt += '\n'
874	<	# txt += 'echo "***** cat pset1.cfg *********"\n'
875	<	# txt += 'cat pset1.cfg\n'
876	<	# txt += 'echo "****** end pset1.cfg ********"\n'
869	>	else:
870	>	txt += 'DatasetPath=MCDataTier\n'
871	>	txt += 'PrimaryDataset=null\n'
872	>	txt += 'DataTier=null\n'
873	>	txt += 'ApplicationFamily=MCDataTier\n'
874	>	if self.pset != None:
875	>	pset = os.path.basename(job.configFilename())
876	>	txt += '\n'
877	>	txt += 'cp  $RUNTIME_AREA/'+pset+' .\n'
878	>	if (self.datasetPath): # standard job
879	>	txt += 'InputFiles=${args[1]}; export InputFiles\n'
880	>	txt += 'MaxEvents=${args[2]}; export MaxEvents\n'
881	>	txt += 'SkipEvents=${args[3]}; export SkipEvents\n'
882	>	txt += 'echo "Inputfiles:<$InputFiles>"\n'
883	>	txt += 'echo "MaxEvents:<$MaxEvents>"\n'
884	>	txt += 'echo "SkipEvents:<$SkipEvents>"\n'
885	>	else:  # pythia like job
886	>	txt += 'PreserveSeeds='  + ','.join(self.preserveSeeds)  + '; export PreserveSeeds\n'
887	>	txt += 'IncrementSeeds=' + ','.join(self.incrementSeeds) + '; export IncrementSeeds\n'
888	>	txt += 'echo "PreserveSeeds: <$PreserveSeeds>"\n'
889	>	txt += 'echo "IncrementSeeds:<$IncrementSeeds>"\n'
890	>	if (self.firstRun):
891	>	txt += 'FirstRun=${args[1]}; export FirstRun\n'
892	>	txt += 'echo "FirstRun: <$FirstRun>"\n'
893	>
894	>	txt += 'mv -f '+pset+' pset.cfg\n'
895	>
896	>
897	>	if self.pset != None:
898	>	txt += '\n'
899	>	txt += 'echo "***** cat pset.cfg *********"\n'
900	>	txt += 'cat pset.cfg\n'
901	>	txt += 'echo "****** end pset.cfg ********"\n'
902	>	txt += '\n'
903	>	txt += 'PSETHASH=`EdmConfigHash < pset.cfg` \n'
904	>	txt += 'echo "PSETHASH = $PSETHASH" \n'
905	>	txt += '\n'
906		return txt
907
908	<	def wsBuildExe(self, nj):
908	>	def wsUntarSoftware(self, nj=0):
909		"""
910		Put in the script the commands to build an executable
911		or a library.
912		"""
913
914	<	txt = ""
914	>	txt = '\n#Written by cms_cmssw::wsUntarSoftware\n'
915
916		if os.path.isfile(self.tgzNameWithPath):
917	<	txt += 'echo "tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+'"\n'
917	>	txt += 'echo ">>> tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+' :" \n'
918		txt += 'tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+'\n'
919	+	txt += 'ls -Al \n'
920		txt += 'untar_status=$? \n'
921		txt += 'if [ $untar_status -ne 0 ]; then \n'
922	<	txt += '   echo "SET_EXE 1 ==> ERROR Untarring .tgz file failed"\n'
923	<	txt += '   echo "JOB_EXIT_STATUS = $untar_status" \n'
924	<	txt += '   echo "SanityCheckCode = $untar_status" | tee -a $repo\n'
636	<	txt += '   if [ $middleware == OSG ]; then \n'
637	<	txt += '       echo "Remove working directory: $WORKING_DIR"\n'
638	<	txt += '       cd $RUNTIME_AREA\n'
639	<	txt += '       /bin/rm -rf $WORKING_DIR\n'
640	<	txt += '       if [ -d $WORKING_DIR ] ;then\n'
641	<	txt += '           echo "OSG WORKING DIR ==> $WORKING_DIR could not be deleted on on WN `hostname`"\n'
642	<	txt += '       fi\n'
643	<	txt += '   fi \n'
644	<	txt += '   \n'
645	<	txt += '   exit $untar_status \n'
922	>	txt += '   echo "ERROR ==> Untarring .tgz file failed"\n'
923	>	txt += '   job_exit_code=$untar_status\n'
924	>	txt += '   func_exit\n'
925		txt += 'else \n'
926		txt += '   echo "Successful untar" \n'
927		txt += 'fi \n'
928	+	txt += '\n'
929	+	txt += 'echo ">>> Include ProdCommon in PYTHONPATH:"\n'
930	+	txt += 'if [ -z "$PYTHONPATH" ]; then\n'
931	+	txt += '   export PYTHONPATH=$RUNTIME_AREA/ProdCommon\n'
932	+	txt += 'else\n'
933	+	txt += '   export PYTHONPATH=$RUNTIME_AREA/ProdCommon:${PYTHONPATH}\n'
934	+	txt += 'echo "PYTHONPATH=$PYTHONPATH"\n'
935	+	txt += 'fi\n'
936	+	txt += '\n'
937	+
938		pass
939	<
939	>
940	>	return txt
941	>
942	>	def wsBuildExe(self, nj=0):
943	>	"""
944	>	Put in the script the commands to build an executable
945	>	or a library.
946	>	"""
947	>
948	>	txt = '\n#Written by cms_cmssw::wsBuildExe\n'
949	>	txt += 'echo ">>> moving CMSSW software directories in `pwd`" \n'
950	>
951	>	txt += 'rm -r lib/ module/ \n'
952	>	txt += 'mv $RUNTIME_AREA/lib/ . \n'
953	>	txt += 'mv $RUNTIME_AREA/module/ . \n'
954	>	if self.dataExist == True: txt += 'mv $RUNTIME_AREA/src/ . \n'
955	>	if len(self.additional_inbox_files)>0:
956	>	for file in self.additional_inbox_files:
957	>	txt += 'mv $RUNTIME_AREA/'+file+' . \n'
958	>	txt += 'mv $RUNTIME_AREA/ProdCommon/ . \n'
959	>
960	>	txt += 'if [ -z "$PYTHONPATH" ]; then\n'
961	>	txt += '   export PYTHONPATH=$SOFTWARE_DIR/ProdCommon\n'
962	>	txt += 'else\n'
963	>	txt += '   export PYTHONPATH=$SOFTWARE_DIR/ProdCommon:${PYTHONPATH}\n'
964	>	txt += 'echo "PYTHONPATH=$PYTHONPATH"\n'
965	>	txt += 'fi\n'
966	>	txt += '\n'
967	>
968		return txt
969
970		def modifySteeringCards(self, nj):
971		"""
972	<	modify the card provided by the user,
972	>	modify the card provided by the user,
973		writing a new card into share dir
974		"""
975	<
975	>
976		def executableName(self):
977	<	return self.executable
977	>	if self.scriptExe: #CarlosDaniele
978	>	return "sh "
979	>	else:
980	>	return self.executable
981
982		def executableArgs(self):
983	<	return " -p pset.cfg"
983	>	# FUTURE: This function tests the CMSSW version. Can be simplified as we drop support for old versions
984	>	if self.scriptExe:#CarlosDaniele
985	>	return   self.scriptExe + " $NJob"
986	>	else:
987	>	ex_args = ""
988	>	# FUTURE: This tests the CMSSW version. Can remove code as versions deprecated
989	>	# Framework job report
990	>	if (self.major_version >= 1 and self.minor_version >= 5) or (self.major_version >= 2):
991	>	ex_args += " -j $RUNTIME_AREA/crab_fjr_$NJob.xml"
992	>	# Type of cfg file
993	>	if self.major_version >= 2 :
994	>	ex_args += " -p pset.py"
995	>	else:
996	>	ex_args += " -p pset.cfg"
997	>	return ex_args
998
999		def inputSandbox(self, nj):
1000		"""
1001		Returns a list of filenames to be put in JDL input sandbox.
1002		"""
1003		inp_box = []
1004	<	# dict added to delete duplicate from input sandbox file list
1005	<	seen = {}
1004	>	# # dict added to delete duplicate from input sandbox file list
1005	>	# seen = {}
1006		## code
1007		if os.path.isfile(self.tgzNameWithPath):
1008		inp_box.append(self.tgzNameWithPath)
1009	<	## config
1010	<	inp_box.append(common.job_list[nj].configFilename())
677	<	## additional input files
678	<	#for file in self.additional_inbox_files:
679	<	#    inp_box.append(common.work_space.cwdDir()+file)
1009	>	wrapper = os.path.basename(str(common._db.queryTask('scriptName')))
1010	>	inp_box.append(common.work_space.pathForTgz() +'job/'+ wrapper)
1011		return inp_box
1012
1013		def outputSandbox(self, nj):
#	Line 685 | Line 1016 | class Cmssw(JobType):
1016		"""
1017		out_box = []
1018
688	–	stdout=common.job_list[nj].stdout()
689	–	stderr=common.job_list[nj].stderr()
690	–
1019		## User Declared output files
1020	<	for out in self.output_file:
1021	<	n_out = nj + 1
1020	>	for out in (self.output_file+self.output_file_sandbox):
1021	>	n_out = nj + 1
1022		out_box.append(self.numberFile_(out,str(n_out)))
1023		return out_box
696	–	return []
1024
1025		def prepareSteeringCards(self):
1026		"""
#	Line 706 | Line 1033 | class Cmssw(JobType):
1033		Returns part of a job script which renames the produced files.
1034		"""
1035
1036	<	txt = '\n'
1037	<	file_list = ''
1038	<	check = len(self.output_file)
1039	<	i = 0
1040	<	for fileWithSuffix in self.output_file:
1041	<	i= i + 1
1036	>	txt = '\n#Written by cms_cmssw::wsRenameOutput\n'
1037	>	txt += 'echo ">>> current directory (SOFTWARE_DIR): $SOFTWARE_DIR" \n'
1038	>	txt += 'echo ">>> current directory content:"\n'
1039	>	txt += 'ls \n'
1040	>	txt += '\n'
1041	>
1042	>	for fileWithSuffix in (self.output_file):
1043		output_file_num = self.numberFile_(fileWithSuffix, '$NJob')
716	–	file_list=file_list+output_file_num+''
717	–	txt += '\n'
718	–	txt += 'ls \n'
1044		txt += '\n'
1045	<	txt += 'ls '+fileWithSuffix+'\n'
1046	<	txt += 'exe_result=$?\n'
1047	<	txt += 'if [ $exe_result -ne 0 ] ; then\n'
1048	<	txt += '   echo "ERROR: No output file to manage"\n'
1049	<	### OLI_DANIELE
1050	<	txt += '    if [ $middleware == OSG ]; then \n'
1051	<	txt += '        echo "prepare dummy output file"\n'
1052	<	txt += '        cp '+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n'
728	<	txt += '    fi \n'
1045	>	txt += '# check output file\n'
1046	>	txt += 'if [ -e ./'+fileWithSuffix+' ] ; then\n'
1047	>	if (self.copy_data == 1):  # For OSG nodes, file is in $WORKING_DIR, should not be moved to $RUNTIME_AREA
1048	>	txt += '    mv '+fileWithSuffix+' '+output_file_num+'\n'
1049	>	txt += '    ln -s `pwd`/'+output_file_num+' $RUNTIME_AREA/'+fileWithSuffix+'\n'
1050	>	else:
1051	>	txt += '    mv '+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n'
1052	>	txt += '    ln -s $RUNTIME_AREA/'+output_file_num+' $RUNTIME_AREA/'+fileWithSuffix+'\n'
1053		txt += 'else\n'
1054	<	txt += '   cp '+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n'
1054	>	txt += '    job_exit_code=60302\n'
1055	>	txt += '    echo "WARNING: Output file '+fileWithSuffix+' not found"\n'
1056	>	if common.scheduler.name().upper() == 'CONDOR_G':
1057	>	txt += '    if [ $middleware == OSG ]; then \n'
1058	>	txt += '        echo "prepare dummy output file"\n'
1059	>	txt += '        echo "Processing of job output failed" > $RUNTIME_AREA/'+output_file_num+'\n'
1060	>	txt += '    fi \n'
1061		txt += 'fi\n'
1062	<	if i == check:
1063	<	txt += 'cd $RUNTIME_AREA\n'
1064	<	pass
1065	<	pass
1066	<
1067	<	file_list=file_list[:-1]
1068	<	txt += 'file_list="'+file_list+'"\n'
1069	<	### OLI_DANIELE
1070	<	txt += 'if [ $middleware == OSG ]; then\n'
741	<	txt += '    cd $RUNTIME_AREA\n'
742	<	txt += '    echo "Remove working directory: $WORKING_DIR"\n'
743	<	txt += '    /bin/rm -rf $WORKING_DIR\n'
744	<	txt += '    if [ -d $WORKING_DIR ] ;then\n'
745	<	txt += '        echo "OSG WORKING DIR ==> $WORKING_DIR could not be deleted on on WN `hostname`"\n'
746	<	txt += '    fi\n'
747	<	txt += 'fi\n'
1062	>	file_list = []
1063	>	for fileWithSuffix in (self.output_file):
1064	>	file_list.append(self.numberFile_(fileWithSuffix, '$NJob'))
1065	>
1066	>	txt += 'file_list="'+string.join(file_list,' ')+'"\n'
1067	>	txt += '\n'
1068	>	txt += 'echo ">>> current directory (SOFTWARE_DIR): $SOFTWARE_DIR" \n'
1069	>	txt += 'echo ">>> current directory content:"\n'
1070	>	txt += 'ls \n'
1071		txt += '\n'
1072	+	txt += 'cd $RUNTIME_AREA\n'
1073	+	txt += 'echo ">>> current directory (RUNTIME_AREA):  $RUNTIME_AREA"\n'
1074		return txt
1075
1076		def numberFile_(self, file, txt):
#	Line 756 | Line 1081 | class Cmssw(JobType):
1081		# take away last extension
1082		name = p[0]
1083		for x in p[1:-1]:
1084	<	name=name+"."+x
1084	>	name=name+"."+x
1085		# add "_txt"
1086		if len(p)>1:
1087	<	ext = p[len(p)-1]
1088	<	#result = name + '_' + str(txt) + "." + ext
1089	<	result = name + '_' + txt + "." + ext
1090	<	else:
1091	<	#result = name + '_' + str(txt)
767	<	result = name + '_' + txt
768	<
1087	>	ext = p[len(p)-1]
1088	>	result = name + '_' + txt + "." + ext
1089	>	else:
1090	>	result = name + '_' + txt
1091	>
1092		return result
1093
1094	<	def getRequirements(self):
1094	>	def getRequirements(self, nj=[]):
1095		"""
1096	<	return job requirements to add to jdl files
1096	>	return job requirements to add to jdl files
1097		"""
1098		req = ''
1099	<	if common.analisys_common_info['sites']:
1100	<	if common.analisys_common_info['sw_version']:
1101	<	req='Member("VO-cms-' + \
1102	<	common.analisys_common_info['sw_version'] + \
1103	<	'", other.GlueHostApplicationSoftwareRunTimeEnvironment)'
1104	<	if len(common.analisys_common_info['sites'])>0:
1105	<	req = req + ' && ('
1106	<	for i in range(len(common.analisys_common_info['sites'])):
1107	<	req = req + 'other.GlueCEInfoHostName == "' \
1108	<	+ common.analisys_common_info['sites'][i] + '"'
1109	<	if ( i < (int(len(common.analisys_common_info['sites']) - 1)) ):
1110	<	req = req + ' || '
1111	<	req = req + ')'
1112	<	#print "req = ", req
1099	>	if self.version:
1100	>	req='Member("VO-cms-' + \
1101	>	self.version + \
1102	>	'", other.GlueHostApplicationSoftwareRunTimeEnvironment)'
1103	>	## SL add requirement for OS version only if SL4
1104	>	#reSL4 = re.compile( r'slc4' )
1105	>	if self.executable_arch: # and reSL4.search(self.executable_arch):
1106	>	req+=' && Member("VO-cms-' + \
1107	>	self.executable_arch + \
1108	>	'", other.GlueHostApplicationSoftwareRunTimeEnvironment)'
1109	>
1110	>	req = req + ' && (other.GlueHostNetworkAdapterOutboundIP)'
1111	>	if common.scheduler.name() == "glitecoll":
1112	>	req += ' && other.GlueCEStateStatus == "Production" '
1113	>
1114		return req
1115
1116		def configFilename(self):
1117		""" return the config filename """
1118	<	return self.name()+'.cfg'
1118	>	# FUTURE: Can remove cfg mode for CMSSW >= 2_1_x
1119	>	if (self.major_version >= 2 and self.minor_version >= 1) or (self.major_version >= 3):
1120	>	return self.name()+'.py'
1121	>	else:
1122	>	return self.name()+'.cfg'
1123
796	–	### OLI_DANIELE
1124		def wsSetupCMSOSGEnvironment_(self):
1125		"""
1126		Returns part of a job script which is prepares
1127		the execution environment and which is common for all CMS jobs.
1128		"""
1129	<	txt = '\n'
1130	<	txt += '   echo "### SETUP CMS OSG  ENVIRONMENT ###"\n'
1131	<	txt += '   if [ -f $GRID3_APP_DIR/cmssoft/cmsset_default.sh ] ;then\n'
1132	<	txt += '      # Use $GRID3_APP_DIR/cmssoft/cmsset_default.sh to setup cms software\n'
1133	<	txt += '       source $GRID3_APP_DIR/cmssoft/cmsset_default.sh '+self.version+'\n'
1134	<	txt += '   elif [ -f $OSG_APP/cmssoft/cmsset_default.sh ] ;then\n'
1135	<	txt += '      # Use $OSG_APP/cmssoft/cmsset_default.sh to setup cms software\n'
1136	<	txt += '       source $OSG_APP/cmssoft/cmsset_default.sh '+self.version+'\n'
1137	<	txt += '   else\n'
1138	<	txt += '       echo "SET_CMS_ENV 10020 ==> ERROR $GRID3_APP_DIR/cmssoft/cmsset_default.sh and $OSG_APP/cmssoft/cmsset_default.sh file not found"\n'
1139	<	txt += '       echo "JOB_EXIT_STATUS = 10020"\n'
1140	<	txt += '       echo "JobExitCode=10020" | tee -a $RUNTIME_AREA/$repo\n'
1141	<	txt += '       dumpStatus $RUNTIME_AREA/$repo\n'
815	<	txt += '       exit\n'
816	<	txt += '\n'
817	<	txt += '       echo "Remove working directory: $WORKING_DIR"\n'
818	<	txt += '       cd $RUNTIME_AREA\n'
819	<	txt += '       /bin/rm -rf $WORKING_DIR\n'
820	<	txt += '       if [ -d $WORKING_DIR ] ;then\n'
821	<	txt += '           echo "OSG WORKING DIR ==> $WORKING_DIR could not be deleted on on WN `hostname`"\n'
822	<	txt += '       fi\n'
823	<	txt += '\n'
824	<	txt += '       exit\n'
825	<	txt += '   fi\n'
1129	>	txt = '\n#Written by cms_cmssw::wsSetupCMSOSGEnvironment_\n'
1130	>	txt += '    echo ">>> setup CMS OSG environment:"\n'
1131	>	txt += '    echo "set SCRAM ARCH to ' + self.executable_arch + '"\n'
1132	>	txt += '    export SCRAM_ARCH='+self.executable_arch+'\n'
1133	>	txt += '    echo "SCRAM_ARCH = $SCRAM_ARCH"\n'
1134	>	txt += '    if [ -f $OSG_APP/cmssoft/cms/cmsset_default.sh ] ;then\n'
1135	>	txt += '      # Use $OSG_APP/cmssoft/cms/cmsset_default.sh to setup cms software\n'
1136	>	txt += '        source $OSG_APP/cmssoft/cms/cmsset_default.sh '+self.version+'\n'
1137	>	txt += '    else\n'
1138	>	txt += '        echo "ERROR ==> $OSG_APP/cmssoft/cms/cmsset_default.sh file not found"\n'
1139	>	txt += '        job_exit_code=10020\n'
1140	>	txt += '        func_exit\n'
1141	>	txt += '    fi\n'
1142		txt += '\n'
1143	<	txt += '   echo "SET_CMS_ENV 0 ==> setup cms environment ok"\n'
1144	<	txt += '   echo " END SETUP CMS OSG  ENVIRONMENT "\n'
1143	>	txt += '    echo "==> setup cms environment ok"\n'
1144	>	txt += '    echo "SCRAM_ARCH = $SCRAM_ARCH"\n'
1145
1146		return txt
1147	<
1147	>
1148		### OLI_DANIELE
1149		def wsSetupCMSLCGEnvironment_(self):
1150		"""
1151		Returns part of a job script which is prepares
1152		the execution environment and which is common for all CMS jobs.
1153		"""
1154	<	txt  = '   \n'
1155	<	txt += '   echo " ### SETUP CMS LCG  ENVIRONMENT ### "\n'
1156	<	txt += '      echo "JOB_EXIT_STATUS = 0"\n'
1157	<	txt += '   if [ ! $VO_CMS_SW_DIR ] ;then\n'
1158	<	txt += '       echo "SET_CMS_ENV 10031 ==> ERROR CMS software dir not found on WN `hostname`"\n'
1159	<	txt += '       echo "JOB_EXIT_STATUS = 10031" \n'
1160	<	txt += '       echo "JobExitCode=10031" | tee -a $RUNTIME_AREA/$repo\n'
1161	<	txt += '       dumpStatus $RUNTIME_AREA/$repo\n'
1162	<	txt += '       exit\n'
1163	<	txt += '   else\n'
1164	<	txt += '       echo "Sourcing environment... "\n'
1165	<	txt += '       if [ ! -s $VO_CMS_SW_DIR/cmsset_default.sh ] ;then\n'
1166	<	txt += '           echo "SET_CMS_ENV 10020 ==> ERROR cmsset_default.sh file not found into dir $VO_CMS_SW_DIR"\n'
1167	<	txt += '           echo "JOB_EXIT_STATUS = 10020"\n'
1168	<	txt += '           echo "JobExitCode=10020" | tee -a $RUNTIME_AREA/$repo\n'
1169	<	txt += '           dumpStatus $RUNTIME_AREA/$repo\n'
1170	<	txt += '           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 "SET_CMS_ENV 10032 ==> ERROR problem sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
1177	<	txt += '           echo "JOB_EXIT_STATUS = 10032"\n'
1178	<	txt += '           echo "JobExitCode=10032" | tee -a $RUNTIME_AREA/$repo\n'
1179	<	txt += '           dumpStatus $RUNTIME_AREA/$repo\n'
1180	<	txt += '           exit\n'
1181	<	txt += '       fi\n'
1182	<	txt += '   fi\n'
1183	<	txt += '   \n'
1184	<	txt += '   string=`cat /etc/redhat-release`\n'
1185	<	txt += '   echo $string\n'
1186	<	txt += '   if [[ $string = *alhalla* ]]; then\n'
1187	<	txt += '       echo "SCRAM_ARCH= $SCRAM_ARCH"\n'
1188	<	txt += '   elif [[ $string = *Enterprise* ]] || [[ $string = *cientific* ]]; then\n'
1189	<	txt += '       export SCRAM_ARCH=slc3_ia32_gcc323\n'
1190	<	txt += '       echo "SCRAM_ARCH= $SCRAM_ARCH"\n'
1191	<	txt += '   else\n'
1192	<	txt += '       echo "SET_CMS_ENV 1 ==> ERROR OS unknown, LCG environment not initialized"\n'
1193	<	txt += '       echo "JOB_EXIT_STATUS = 10033"\n'
1194	<	txt += '       echo "JobExitCode=10033" | tee -a $RUNTIME_AREA/$repo\n'
1195	<	txt += '       dumpStatus $RUNTIME_AREA/$repo\n'
1196	<	txt += '       exit\n'
1197	<	txt += '   fi\n'
1198	<	txt += '   echo "SET_CMS_ENV 0 ==> setup cms environment ok"\n'
1199	<	txt += '   echo "### END SETUP CMS LCG ENVIRONMENT ###"\n'
1154	>	txt = '\n#Written by cms_cmssw::wsSetupCMSLCGEnvironment_\n'
1155	>	txt += '    echo ">>> setup CMS LCG environment:"\n'
1156	>	txt += '    echo "set SCRAM ARCH and BUILD_ARCH to ' + self.executable_arch + ' ###"\n'
1157	>	txt += '    export SCRAM_ARCH='+self.executable_arch+'\n'
1158	>	txt += '    export BUILD_ARCH='+self.executable_arch+'\n'
1159	>	txt += '    if [ ! $VO_CMS_SW_DIR ] ;then\n'
1160	>	txt += '        echo "ERROR ==> CMS software dir not found on WN `hostname`"\n'
1161	>	txt += '        job_exit_code=10031\n'
1162	>	txt += '        func_exit\n'
1163	>	txt += '    else\n'
1164	>	txt += '        echo "Sourcing environment... "\n'
1165	>	txt += '        if [ ! -s $VO_CMS_SW_DIR/cmsset_default.sh ] ;then\n'
1166	>	txt += '            echo "ERROR ==> cmsset_default.sh file not found into dir $VO_CMS_SW_DIR"\n'
1167	>	txt += '            job_exit_code=10020\n'
1168	>	txt += '            func_exit\n'
1169	>	txt += '        fi\n'
1170	>	txt += '        echo "sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
1171	>	txt += '        source $VO_CMS_SW_DIR/cmsset_default.sh\n'
1172	>	txt += '        result=$?\n'
1173	>	txt += '        if [ $result -ne 0 ]; then\n'
1174	>	txt += '            echo "ERROR ==> problem sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
1175	>	txt += '            job_exit_code=10032\n'
1176	>	txt += '            func_exit\n'
1177	>	txt += '        fi\n'
1178	>	txt += '    fi\n'
1179	>	txt += '    \n'
1180	>	txt += '    echo "==> setup cms environment ok"\n'
1181	>	return txt
1182	>
1183	>	### FEDE FOR DBS OUTPUT PUBLICATION
1184	>	def modifyReport(self, nj):
1185	>	"""
1186	>	insert the part of the script that modifies the FrameworkJob Report
1187	>	"""
1188	>
1189	>	txt = '\n#Written by cms_cmssw::modifyReport\n'
1190	>	publish_data = int(self.cfg_params.get('USER.publish_data',0))
1191	>	if (publish_data == 1):
1192	>	processedDataset = self.cfg_params['USER.publish_data_name']
1193	>	LFNBaseName = LFNBase(processedDataset)
1194	>
1195	>	txt += 'if [ $copy_exit_status -eq 0 ]; then\n'
1196	>	txt += '    FOR_LFN=%s_${PSETHASH}/\n'%(LFNBaseName)
1197	>	txt += 'else\n'
1198	>	txt += '    FOR_LFN=/copy_problems/ \n'
1199	>	txt += '    SE=""\n'
1200	>	txt += '    SE_PATH=""\n'
1201	>	txt += 'fi\n'
1202	>
1203	>	txt += 'echo ">>> Modify Job Report:" \n'
1204	>	txt += 'chmod a+x $SOFTWARE_DIR/ProdCommon/ProdCommon/FwkJobRep/ModifyJobReport.py\n'
1205	>	txt += 'ProcessedDataset='+processedDataset+'\n'
1206	>	txt += 'echo "ProcessedDataset = $ProcessedDataset"\n'
1207	>	txt += 'echo "SE = $SE"\n'
1208	>	txt += 'echo "SE_PATH = $SE_PATH"\n'
1209	>	txt += 'echo "FOR_LFN = $FOR_LFN" \n'
1210	>	txt += 'echo "CMSSW_VERSION = $CMSSW_VERSION"\n\n'
1211	>	txt += 'echo "$SOFTWARE_DIR/ProdCommon/ProdCommon/FwkJobRep/ModifyJobReport.py $RUNTIME_AREA/crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH"\n'
1212	>	txt += '$SOFTWARE_DIR/ProdCommon/ProdCommon/FwkJobRep/ModifyJobReport.py $RUNTIME_AREA/crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH\n'
1213	>	txt += 'modifyReport_result=$?\n'
1214	>	txt += 'if [ $modifyReport_result -ne 0 ]; then\n'
1215	>	txt += '    modifyReport_result=70500\n'
1216	>	txt += '    job_exit_code=$modifyReport_result\n'
1217	>	txt += '    echo "ModifyReportResult=$modifyReport_result" | tee -a $RUNTIME_AREA/$repo\n'
1218	>	txt += '    echo "WARNING: Problem with ModifyJobReport"\n'
1219	>	txt += 'else\n'
1220	>	txt += '    mv NewFrameworkJobReport.xml $RUNTIME_AREA/crab_fjr_$NJob.xml\n'
1221	>	txt += 'fi\n'
1222	>	return txt
1223	>
1224	>	def setParam_(self, param, value):
1225	>	self._params[param] = value
1226	>
1227	>	def getParams(self):
1228	>	return self._params
1229	>
1230	>	def uniquelist(self, old):
1231	>	"""
1232	>	remove duplicates from a list
1233	>	"""
1234	>	nd={}
1235	>	for e in old:
1236	>	nd[e]=0
1237	>	return nd.keys()
1238	>
1239	>	def outList(self):
1240	>	"""
1241	>	check the dimension of the output files
1242	>	"""
1243	>	txt = ''
1244	>	txt += 'echo ">>> list of expected files on output sandbox"\n'
1245	>	listOutFiles = []
1246	>	stdout = 'CMSSW_$NJob.stdout'
1247	>	stderr = 'CMSSW_$NJob.stderr'
1248	>	if (self.return_data == 1):
1249	>	for file in (self.output_file+self.output_file_sandbox):
1250	>	listOutFiles.append(self.numberFile_(file, '$NJob'))
1251	>	listOutFiles.append(stdout)
1252	>	listOutFiles.append(stderr)
1253	>	else:
1254	>	for file in (self.output_file_sandbox):
1255	>	listOutFiles.append(self.numberFile_(file, '$NJob'))
1256	>	listOutFiles.append(stdout)
1257	>	listOutFiles.append(stderr)
1258	>	txt += 'echo "output files: '+string.join(listOutFiles,' ')+'"\n'
1259	>	txt += 'filesToCheck="'+string.join(listOutFiles,' ')+'"\n'
1260	>	txt += 'export filesToCheck\n'
1261		return txt

Diff Legend

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