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 1.145 by fanzago, Wed Dec 5 14:40:06 2007 UTC vs.
Revision 1.314 by spiga, Thu Jun 18 17:23:02 2009 UTC

#	Line 1 | Line 1
1		from JobType import JobType
2	–	from crab_logger import Logger
2		from crab_exceptions import *
3		from crab_util import *
5	–	from BlackWhiteListParser import BlackWhiteListParser
4		import common
5		import Scram
6	+	from Splitter import JobSplitter
7
8	+	from IMProv.IMProvNode import IMProvNode
9		import os, string, glob
10
11		class Cmssw(JobType):
12	<	def __init__(self, cfg_params, ncjobs):
12	>	def __init__(self, cfg_params, ncjobs,skip_blocks, isNew):
13		JobType.__init__(self, 'CMSSW')
14	<	common.logger.debug(3,'CMSSW::__init__')
15	<
16	<	self.argsList = []
14	>	common.logger.debug('CMSSW::__init__')
15	>	self.skip_blocks = skip_blocks
16	>	self.argsList = 1
17
18		self._params = {}
19		self.cfg_params = cfg_params
20	–	# init BlackWhiteListParser
21	–	self.blackWhiteListParser = BlackWhiteListParser(cfg_params)
20
21	<	try:
22	<	self.MaxTarBallSize = float(self.cfg_params['EDG.maxtarballsize'])
23	<	except KeyError:
24	<	self.MaxTarBallSize = 9.5
21	>	### Temporary patch to automatically skip the ISB size check:
22	>	server=self.cfg_params.get('CRAB.server_name',None)
23	>	size = 9.5
24	>	if server or common.scheduler.name().upper() in ['LSF','CAF']: size = 99999
25	>	### D.S.
26	>	self.MaxTarBallSize = float(self.cfg_params.get('GRID.maxtarballsize',size))
27
28		# number of jobs requested to be created, limit obj splitting
29		self.ncjobs = ncjobs
30
31	–	log = common.logger
32	–
31		self.scram = Scram.Scram(cfg_params)
32		self.additional_inbox_files = []
33		self.scriptExe = ''
34		self.executable = ''
35		self.executable_arch = self.scram.getArch()
36	<	self.tgz_name = 'default.tgz'
37	<	self.additional_tgz_name = 'additional.tgz'
36	>	self.tgz_name = 'default.tar.gz'
37	>	self.tar_name = 'default.tar'
38		self.scriptName = 'CMSSW.sh'
39	<	self.pset = ''      #scrip use case Da
40	<	self.datasetPath = '' #scrip use case Da
39	>	self.pset = ''
40	>	self.datasetPath = ''
41
42	+	self.tgzNameWithPath = common.work_space.pathForTgz()+self.tgz_name
43		# set FJR file name
44		self.fjrFileName = 'crab_fjr.xml'
45
46		self.version = self.scram.getSWVersion()
47	+	common.logger.log(10-1,"CMSSW version is: "+str(self.version))
48
49	<	#
50	<	# Try to block creation in case of arch/version mismatch
51	<	#
52	<
53	<	a = string.split(self.version, "_")
54	<
55	<	if int(a[1]) == 1 and (int(a[2]) < 5 and self.executable_arch.find('slc4') == 0):
56	<	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)
57	<	common.logger.message(msg)
58	<	if int(a[1]) == 1 and (int(a[2]) >= 5 and self.executable_arch.find('slc3') == 0):
59	<	msg = "Error: CMS does not support %s with %s architecture"%(self.version, self.executable_arch)
49	>	version_array = self.version.split('_')
50	>	self.CMSSW_major = 0
51	>	self.CMSSW_minor = 0
52	>	self.CMSSW_patch = 0
53	>	try:
54	>	self.CMSSW_major = int(version_array[1])
55	>	self.CMSSW_minor = int(version_array[2])
56	>	self.CMSSW_patch = int(version_array[3])
57	>	except:
58	>	msg = "Cannot parse CMSSW version string: " + self.version + " for major and minor release number!"
59		raise CrabException(msg)
60
61	<	common.taskDB.setDict('codeVersion',self.version)
62	<	self.setParam_('application', self.version)
61	>	if self.CMSSW_major < 1 or (self.CMSSW_major == 1 and self.CMSSW_minor < 5):
62	>	msg = "CRAB supports CMSSW >= 1_5_x only. Use an older CRAB version."
63	>	raise CrabException(msg)
64	>	"""
65	>	As CMSSW versions are dropped we can drop more code:
66	>	1.X dropped: drop support for running .cfg on WN
67	>	2.0 dropped: drop all support for cfg here and in writeCfg
68	>	2.0 dropped: Recheck the random number seed support
69	>	"""
70
71		### collect Data cards
72
73	<	try:
74	<	tmp =  cfg_params['CMSSW.datasetpath']
75	<	log.debug(6, "CMSSW::CMSSW(): datasetPath = "+tmp)
76	<	if string.lower(tmp)=='none':
77	<	self.datasetPath = None
78	<	self.selectNoInput = 1
79	<	else:
80	<	self.datasetPath = tmp
75	<	self.selectNoInput = 0
76	<	except KeyError:
73	>
74	>	### Temporary: added to remove input file control in the case of PU
75	>	self.dataset_pu = cfg_params.get('CMSSW.dataset_pu', None)
76	>
77	>	tmp =  cfg_params['CMSSW.datasetpath']
78	>	common.logger.log(10-1, "CMSSW::CMSSW(): datasetPath = "+tmp)
79	>
80	>	if tmp =='':
81		msg = "Error: datasetpath not defined "
82		raise CrabException(msg)
83	<
84	<	# ML monitoring
85	<	# split dataset path style: /PreProdR3Minbias/SIM/GEN-SIM
82	<	if not self.datasetPath:
83	<	self.setParam_('dataset', 'None')
84	<	self.setParam_('owner', 'None')
83	>	elif string.lower(tmp)=='none':
84	>	self.datasetPath = None
85	>	self.selectNoInput = 1
86		else:
87	<	try:
88	<	datasetpath_split = self.datasetPath.split("/")
88	<	# standard style
89	<	self.setParam_('datasetFull', self.datasetPath)
90	<	self.setParam_('dataset', datasetpath_split[1])
91	<	self.setParam_('owner', datasetpath_split[2])
92	<	except:
93	<	self.setParam_('dataset', self.datasetPath)
94	<	self.setParam_('owner', self.datasetPath)
95	<
96	<	self.setTaskid_()
97	<	self.setParam_('taskId', self.cfg_params['taskId'])
87	>	self.datasetPath = tmp
88	>	self.selectNoInput = 0
89
90		self.dataTiers = []
91
92	+	self.debugWrap=''
93	+	self.debug_wrapper = int(cfg_params.get('USER.debug_wrapper',0))
94	+	if self.debug_wrapper == 1: self.debugWrap='--debug'
95	+
96		## now the application
97	<	try:
98	<	self.executable = cfg_params['CMSSW.executable']
99	<	self.setParam_('exe', self.executable)
100	<	log.debug(6, "CMSSW::CMSSW(): executable = "+self.executable)
106	<	msg = "Default executable cmsRun overridden. Switch to " + self.executable
107	<	log.debug(3,msg)
108	<	except KeyError:
109	<	self.executable = 'cmsRun'
110	<	self.setParam_('exe', self.executable)
111	<	msg = "User executable not defined. Use cmsRun"
112	<	log.debug(3,msg)
113	<	pass
97	>	self.managedGenerators = ['madgraph', 'comphep', 'lhe']
98	>	self.generator = cfg_params.get('CMSSW.generator','pythia').lower()
99	>	self.executable = cfg_params.get('CMSSW.executable','cmsRun')
100	>	common.logger.log(10-1, "CMSSW::CMSSW(): executable = "+self.executable)
101
102	<	try:
116	<	self.pset = cfg_params['CMSSW.pset']
117	<	log.debug(6, "Cmssw::Cmssw(): PSet file = "+self.pset)
118	<	if self.pset.lower() != 'none' :
119	<	if (not os.path.exists(self.pset)):
120	<	raise CrabException("User defined PSet file "+self.pset+" does not exist")
121	<	else:
122	<	self.pset = None
123	<	except KeyError:
102	>	if not cfg_params.has_key('CMSSW.pset'):
103		raise CrabException("PSet file missing. Cannot run cmsRun ")
104	+	self.pset = cfg_params['CMSSW.pset']
105	+	common.logger.log(10-1, "Cmssw::Cmssw(): PSet file = "+self.pset)
106	+	if self.pset.lower() != 'none' :
107	+	if (not os.path.exists(self.pset)):
108	+	raise CrabException("User defined PSet file "+self.pset+" does not exist")
109	+	else:
110	+	self.pset = None
111
112		# output files
113		## stuff which must be returned always via sandbox
#	Line 131 | Line 117 | class Cmssw(JobType):
117		self.output_file_sandbox.append(self.fjrFileName)
118
119		# other output files to be returned via sandbox or copied to SE
120	<	try:
121	<	self.output_file = []
122	<	tmp = cfg_params['CMSSW.output_file']
123	<	if tmp != '':
124	<	tmpOutFiles = string.split(cfg_params['CMSSW.output_file'],',')
125	<	log.debug(7, 'cmssw::cmssw(): output files '+str(tmpOutFiles))
126	<	for tmp in tmpOutFiles:
127	<	tmp=string.strip(tmp)
142	<	self.output_file.append(tmp)
143	<	pass
144	<	else:
145	<	log.message("No output file defined: only stdout/err and the CRAB Framework Job Report will be available\n")
146	<	pass
147	<	pass
148	<	except KeyError:
149	<	log.message("No output file defined: only stdout/err and the CRAB Framework Job Report will be available\n")
150	<	pass
120	>	outfileflag = False
121	>	self.output_file = []
122	>	tmp = cfg_params.get('CMSSW.output_file',None)
123	>	if tmp :
124	>	self.output_file = [x.strip() for x in tmp.split(',')]
125	>	outfileflag = True #output found
126	>	#else:
127	>	#    log.message("No output file defined: only stdout/err and the CRAB Framework Job Report will be available\n")
128
129		# script_exe file as additional file in inputSandbox
130	<	try:
131	<	self.scriptExe = cfg_params['USER.script_exe']
132	<	if self.scriptExe != '':
133	<	if not os.path.isfile(self.scriptExe):
134	<	msg ="ERROR. file "+self.scriptExe+" not found"
135	<	raise CrabException(msg)
136	<	self.additional_inbox_files.append(string.strip(self.scriptExe))
137	<	except KeyError:
138	<	self.scriptExe = ''
130	>	self.scriptExe = cfg_params.get('USER.script_exe',None)
131	>	if self.scriptExe :
132	>	if not os.path.isfile(self.scriptExe):
133	>	msg ="ERROR. file "+self.scriptExe+" not found"
134	>	raise CrabException(msg)
135	>	self.additional_inbox_files.append(string.strip(self.scriptExe))
136	>
137	>	self.AdditionalArgs = cfg_params.get('USER.script_arguments',None)
138	>	if self.AdditionalArgs : self.AdditionalArgs = string.replace(self.AdditionalArgs,',',' ')
139
163	–	#CarlosDaniele
140		if self.datasetPath == None and self.pset == None and self.scriptExe == '' :
141	<	msg ="Error. script_exe  not defined"
142	<	raise CrabException(msg)
141	>	msg ="Error. script_exe  not defined"
142	>	raise CrabException(msg)
143	>
144	>	# use parent files...
145	>	self.useParent = int(self.cfg_params.get('CMSSW.use_parent',0))
146
147		## additional input files
148	<	try:
148	>	if cfg_params.has_key('USER.additional_input_files'):
149		tmpAddFiles = string.split(cfg_params['USER.additional_input_files'],',')
150		for tmp in tmpAddFiles:
151		tmp = string.strip(tmp)
#	Line 183 | Line 162 | class Cmssw(JobType):
162		if not os.path.exists(file):
163		raise CrabException("Additional input file not found: "+file)
164		pass
186	–	# fname = string.split(file, '/')[-1]
187	–	# storedFile = common.work_space.pathForTgz()+'share/'+fname
188	–	# shutil.copyfile(file, storedFile)
165		self.additional_inbox_files.append(string.strip(file))
166		pass
167		pass
168	<	common.logger.debug(5,"Additional input files: "+str(self.additional_inbox_files))
169	<	except KeyError:
194	<	pass
195	<
196	<	# files per job
197	<	try:
198	<	if (cfg_params['CMSSW.files_per_jobs']):
199	<	raise CrabException("files_per_jobs no longer supported.  Quitting.")
200	<	except KeyError:
201	<	pass
202	<
203	<	## Events per job
204	<	try:
205	<	self.eventsPerJob =int( cfg_params['CMSSW.events_per_job'])
206	<	self.selectEventsPerJob = 1
207	<	except KeyError:
208	<	self.eventsPerJob = -1
209	<	self.selectEventsPerJob = 0
210	<
211	<	## number of jobs
212	<	try:
213	<	self.theNumberOfJobs =int( cfg_params['CMSSW.number_of_jobs'])
214	<	self.selectNumberOfJobs = 1
215	<	except KeyError:
216	<	self.theNumberOfJobs = 0
217	<	self.selectNumberOfJobs = 0
218	<
219	<	try:
220	<	self.total_number_of_events = int(cfg_params['CMSSW.total_number_of_events'])
221	<	self.selectTotalNumberEvents = 1
222	<	except KeyError:
223	<	self.total_number_of_events = 0
224	<	self.selectTotalNumberEvents = 0
225	<
226	<	if self.pset != None: #CarlosDaniele
227	<	if ( (self.selectTotalNumberEvents + self.selectEventsPerJob + self.selectNumberOfJobs) != 2 ):
228	<	msg = 'Must define exactly two of total_number_of_events, events_per_job, or number_of_jobs.'
229	<	raise CrabException(msg)
230	<	else:
231	<	if (self.selectNumberOfJobs == 0):
232	<	msg = 'Must specify  number_of_jobs.'
233	<	raise CrabException(msg)
234	<
235	<	## source seed for pythia
236	<	try:
237	<	self.sourceSeed = int(cfg_params['CMSSW.pythia_seed'])
238	<	except KeyError:
239	<	self.sourceSeed = None
240	<	common.logger.debug(5,"No seed given")
241	<
242	<	try:
243	<	self.sourceSeedVtx = int(cfg_params['CMSSW.vtx_seed'])
244	<	except KeyError:
245	<	self.sourceSeedVtx = None
246	<	common.logger.debug(5,"No vertex seed given")
247	<
248	<	try:
249	<	self.sourceSeedG4 = int(cfg_params['CMSSW.g4_seed'])
250	<	except KeyError:
251	<	self.sourceSeedG4 = None
252	<	common.logger.debug(5,"No g4 sim hits seed given")
253	<
254	<	try:
255	<	self.sourceSeedMix = int(cfg_params['CMSSW.mix_seed'])
256	<	except KeyError:
257	<	self.sourceSeedMix = None
258	<	common.logger.debug(5,"No mix seed given")
168	>	common.logger.debug("Additional input files: "+str(self.additional_inbox_files))
169	>	pass
170
260	–	try:
261	–	self.firstRun = int(cfg_params['CMSSW.first_run'])
262	–	except KeyError:
263	–	self.firstRun = None
264	–	common.logger.debug(5,"No first run given")
265	–	if self.pset != None: #CarlosDaniele
266	–	import PsetManipulator as pp
267	–	PsetEdit = pp.PsetManipulator(self.pset) #Daniele Pset
171
172	+	## New method of dealing with seeds
173	+	self.incrementSeeds = []
174	+	self.preserveSeeds = []
175	+	if cfg_params.has_key('CMSSW.preserve_seeds'):
176	+	tmpList = cfg_params['CMSSW.preserve_seeds'].split(',')
177	+	for tmp in tmpList:
178	+	tmp.strip()
179	+	self.preserveSeeds.append(tmp)
180	+	if cfg_params.has_key('CMSSW.increment_seeds'):
181	+	tmpList = cfg_params['CMSSW.increment_seeds'].split(',')
182	+	for tmp in tmpList:
183	+	tmp.strip()
184	+	self.incrementSeeds.append(tmp)
185	+
186	+	self.firstRun = cfg_params.get('CMSSW.first_run',None)
187	+
188	+	# Copy/return
189	+	self.copy_data = int(cfg_params.get('USER.copy_data',0))
190	+	self.return_data = int(cfg_params.get('USER.return_data',0))
191	+
192	+	self.conf = {}
193	+	self.conf['pubdata'] = None
194	+	# number of jobs requested to be created, limit obj splitting DD
195		#DBSDLS-start
196		## Initialize the variables that are extracted from DBS/DLS and needed in other places of the code
197		self.maxEvents=0  # max events available   ( --> check the requested nb. of evts in Creator.py)
#	Line 277 | Line 203 | class Cmssw(JobType):
203		if self.datasetPath:
204		blockSites = self.DataDiscoveryAndLocation(cfg_params)
205		#DBSDLS-end
206	<
281	<	self.tgzNameWithPath = self.getTarBall(self.executable)
206	>	self.conf['blockSites']=blockSites
207
208		## Select Splitting
209	+	splitByRun = int(cfg_params.get('CMSSW.split_by_run',0))
210	+
211		if self.selectNoInput:
212	<	if self.pset == None: #CarlosDaniele
213	<	self.jobSplittingForScript()
212	>	if self.pset == None:
213	>	self.algo = 'ForScript'
214		else:
215	<	self.jobSplittingNoInput()
215	>	self.algo = 'NoInput'
216	>	self.conf['managedGenerators']=self.managedGenerators
217	>	self.conf['generator']=self.generator
218	>	elif splitByRun ==1:
219	>	self.algo = 'RunBased'
220		else:
221	<	self.jobSplittingByBlocks(blockSites)
221	>	self.algo = 'EventBased'
222	>
223	>	#        self.algo = 'LumiBased'
224	>	splitter = JobSplitter(self.cfg_params,self.conf)
225	>	self.dict = splitter.Algos()[self.algo]()
226	>
227	>	self.argsFile= '%s/arguments.xml'%common.work_space.shareDir()
228	>	self.rootArgsFilename= 'arguments'
229	>	# modify Pset only the first time
230	>	if (isNew and self.pset != None): self.ModifyPset()
231	>
232	>	## Prepare inputSandbox TarBall (only the first time)
233	>	self.tarNameWithPath = self.getTarBall(self.executable)
234	>
235	>
236	>	def ModifyPset(self):
237	>	import PsetManipulator as pp
238	>	PsetEdit = pp.PsetManipulator(self.pset)
239	>	try:
240	>	# Add FrameworkJobReport to parameter-set, set max events.
241	>	# Reset later for data jobs by writeCFG which does all modifications
242	>	PsetEdit.maxEvent(1)
243	>	PsetEdit.skipEvent(0)
244	>	PsetEdit.psetWriter(self.configFilename())
245	>	## If present, add TFileService to output files
246	>	if not int(self.cfg_params.get('CMSSW.skip_TFileService_output',0)):
247	>	tfsOutput = PsetEdit.getTFileService()
248	>	if tfsOutput:
249	>	if tfsOutput in self.output_file:
250	>	common.logger.debug("Output from TFileService "+tfsOutput+" already in output files")
251	>	else:
252	>	outfileflag = True #output found
253	>	self.output_file.append(tfsOutput)
254	>	common.logger.info("Adding "+tfsOutput+" (from TFileService) to list of output files")
255	>	pass
256	>	pass
257	>	## If present and requested, add PoolOutputModule to output files
258	>	edmOutput = PsetEdit.getPoolOutputModule()
259	>	if int(self.cfg_params.get('CMSSW.get_edm_output',0)):
260	>	if edmOutput:
261	>	if edmOutput in self.output_file:
262	>	common.logger.debug("Output from PoolOutputModule "+edmOutput+" already in output files")
263	>	else:
264	>	self.output_file.append(edmOutput)
265	>	common.logger.info("Adding "+edmOutput+" (from PoolOutputModule) to list of output files")
266	>	pass
267	>	pass
268	>	# not required: check anyhow if present, to avoid accidental T2 overload
269	>	else:
270	>	if edmOutput and (edmOutput not in self.output_file):
271	>	msg = "ERROR: a PoolOutputModule is present in your ParameteSet %s \n"%self.pset
272	>	msg +="         but the file produced ( %s ) is not in the list of output files\n"%edmOutput
273	>	msg += "WARNING: please remove it. If you want to keep it, add the file to output_files or use CMSSW.get_edm_output\n"
274	>	if int(self.cfg_params.get('CMSSW.ignore_edm_output',0)):
275	>	msg +=" CMSSW.ignore_edm_output==True : Hope you know what you are doing...\n"
276	>	common.logger.info(msg)
277	>	else:
278	>	raise CrabException(msg)
279	>	pass
280	>	pass
281	>
282	>	if (PsetEdit.getBadFilesSetting()):
283	>	msg = "WARNING: You have set skipBadFiles to True. This will continue processing on some errors and you may not be notified."
284	>	common.logger.info(msg)
285	>
286	>	except CrabException, msg:
287	>	common.logger.info(str(msg))
288	>	msg='Error while manipulating ParameterSet (see previous message, if any): exiting...'
289	>	raise CrabException(msg)
290
292	–	# modify Pset
293	–	if self.pset != None: #CarlosDaniele
294	–	try:
295	–	if (self.datasetPath): # standard job
296	–	# allow to processa a fraction of events in a file
297	–	PsetEdit.inputModule("INPUTFILE")
298	–	PsetEdit.maxEvent(0)
299	–	PsetEdit.skipEvent(0)
300	–	else:  # pythia like job
301	–	PsetEdit.maxEvent(self.eventsPerJob)
302	–	if (self.firstRun):
303	–	PsetEdit.pythiaFirstRun(0)  #First Run
304	–	if (self.sourceSeed) :
305	–	PsetEdit.pythiaSeed(0)
306	–	if (self.sourceSeedVtx) :
307	–	PsetEdit.vtxSeed(0)
308	–	if (self.sourceSeedG4) :
309	–	PsetEdit.g4Seed(0)
310	–	if (self.sourceSeedMix) :
311	–	PsetEdit.mixSeed(0)
312	–	# add FrameworkJobReport to parameter-set
313	–	PsetEdit.addCrabFJR(self.fjrFileName)
314	–	PsetEdit.psetWriter(self.configFilename())
315	–	except:
316	–	msg='Error while manipuliating ParameterSet: exiting...'
317	–	raise CrabException(msg)
291
292		def DataDiscoveryAndLocation(self, cfg_params):
293
294		import DataDiscovery
295		import DataLocation
296	<	common.logger.debug(10,"CMSSW::DataDiscoveryAndLocation()")
296	>	common.logger.log(10-1,"CMSSW::DataDiscoveryAndLocation()")
297
298		datasetPath=self.datasetPath
299
300		## Contact the DBS
301	<	common.logger.message("Contacting Data Discovery Services ...")
301	>	common.logger.info("Contacting Data Discovery Services ...")
302		try:
303	<
331	<	self.pubdata=DataDiscovery.DataDiscovery(datasetPath, cfg_params)
303	>	self.pubdata=DataDiscovery.DataDiscovery(datasetPath, cfg_params,self.skip_blocks)
304		self.pubdata.fetchDBSInfo()
305
306		except DataDiscovery.NotExistingDatasetError, ex :
#	Line 342 | Line 314 | class Cmssw(JobType):
314		raise CrabException(msg)
315
316		self.filesbyblock=self.pubdata.getFiles()
317	<	self.eventsbyblock=self.pubdata.getEventsPerBlock()
318	<	self.eventsbyfile=self.pubdata.getEventsPerFile()
317	>	#print self.filesbyblock
318	>	self.conf['pubdata']=self.pubdata
319
320		## get max number of events
321	<	self.maxEvents=self.pubdata.getMaxEvents() ##  self.maxEvents used in Creator.py
321	>	self.maxEvents=self.pubdata.getMaxEvents()
322
323		## Contact the DLS and build a list of sites hosting the fileblocks
324		try:
325		dataloc=DataLocation.DataLocation(self.filesbyblock.keys(),cfg_params)
326		dataloc.fetchDLSInfo()
327	+
328		except DataLocation.DataLocationError , ex:
329		msg = 'ERROR ***: failed Data Location in DLS \n %s '%ex.getErrorMessage()
330		raise CrabException(msg)
331
332
333	<	sites = dataloc.getSites()
333	>	unsorted_sites = dataloc.getSites()
334	>	sites = self.filesbyblock.fromkeys(self.filesbyblock,'')
335	>	for lfn in self.filesbyblock.keys():
336	>	if unsorted_sites.has_key(lfn):
337	>	sites[lfn]=unsorted_sites[lfn]
338	>	else:
339	>	sites[lfn]=[]
340	>
341	>	if len(sites)==0:
342	>	msg = 'ERROR ***: no location for any of the blocks of this dataset: \n\t %s \n'%datasetPath
343	>	msg += "\tMaybe the dataset is located only at T1's (or at T0), where analysis jobs are not allowed\n"
344	>	msg += "\tPlease check DataDiscovery page https://cmsweb.cern.ch/dbs_discovery/\n"
345	>	raise CrabException(msg)
346	>
347		allSites = []
348		listSites = sites.values()
349		for listSite in listSites:
350		for oneSite in listSite:
351		allSites.append(oneSite)
352	<	allSites = self.uniquelist(allSites)
352	>	[allSites.append(it) for it in allSites if not allSites.count(it)]
353	>
354
355		# screen output
356	<	common.logger.message("Requested dataset: " + datasetPath + " has " + str(self.maxEvents) + " events in " + str(len(self.filesbyblock.keys())) + " blocks.\n")
356	>	common.logger.info("Requested dataset: " + datasetPath + " has " + str(self.maxEvents) + " events in " + str(len(self.filesbyblock.keys())) + " blocks.\n")
357
358		return sites
359
373	–	def setArgsList(self, argsList):
374	–	self.argsList = argsList
375	–
376	–	def jobSplittingByBlocks(self, blockSites):
377	–	"""
378	–	Perform job splitting. Jobs run over an integer number of files
379	–	and no more than one block.
380	–	ARGUMENT: blockSites: dictionary with blocks as keys and list of host sites as values
381	–	REQUIRES: self.selectTotalNumberEvents, self.selectEventsPerJob, self.selectNumberofJobs,
382	–	self.total_number_of_events, self.eventsPerJob, self.theNumberOfJobs,
383	–	self.maxEvents, self.filesbyblock
384	–	SETS: self.jobDestination - Site destination(s) for each job (a list of lists)
385	–	self.total_number_of_jobs - Total # of jobs
386	–	self.list_of_args - File(s) job will run on (a list of lists)
387	–	"""
388	–
389	–	# ---- Handle the possible job splitting configurations ---- #
390	–	if (self.selectTotalNumberEvents):
391	–	totalEventsRequested = self.total_number_of_events
392	–	if (self.selectEventsPerJob):
393	–	eventsPerJobRequested = self.eventsPerJob
394	–	if (self.selectNumberOfJobs):
395	–	totalEventsRequested = self.theNumberOfJobs * self.eventsPerJob
396	–
397	–	# If user requested all the events in the dataset
398	–	if (totalEventsRequested == -1):
399	–	eventsRemaining=self.maxEvents
400	–	# If user requested more events than are in the dataset
401	–	elif (totalEventsRequested > self.maxEvents):
402	–	eventsRemaining = self.maxEvents
403	–	common.logger.message("Requested "+str(self.total_number_of_events)+ " events, but only "+str(self.maxEvents)+" events are available.")
404	–	# If user requested less events than are in the dataset
405	–	else:
406	–	eventsRemaining = totalEventsRequested
360
361	<	# If user requested more events per job than are in the dataset
409	<	if (self.selectEventsPerJob and eventsPerJobRequested > self.maxEvents):
410	<	eventsPerJobRequested = self.maxEvents
361	>	def split(self, jobParams,firstJobID):
362
363	<	# For user info at end
364	<	totalEventCount = 0
363	>	jobParams = self.dict['args']
364	>	njobs = self.dict['njobs']
365	>	self.jobDestination = self.dict['jobDestination']
366
367	<	if (self.selectTotalNumberEvents and self.selectNumberOfJobs):
368	<	eventsPerJobRequested = int(eventsRemaining/self.theNumberOfJobs)
367	>	if njobs==0:
368	>	raise CrabException("Ask to split "+str(njobs)+" jobs: aborting")
369
370	<	if (self.selectNumberOfJobs):
371	<	common.logger.message("May not create the exact number_of_jobs requested.")
372	<
421	<	if ( self.ncjobs == 'all' ) :
422	<	totalNumberOfJobs = 999999999
423	<	else :
424	<	totalNumberOfJobs = self.ncjobs
425	<
426	<
427	<	blocks = blockSites.keys()
428	<	blockCount = 0
429	<	# Backup variable in case self.maxEvents counted events in a non-included block
430	<	numBlocksInDataset = len(blocks)
431	<
432	<	jobCount = 0
433	<	list_of_lists = []
434	<
435	<	# list tracking which jobs are in which jobs belong to which block
436	<	jobsOfBlock = {}
437	<
438	<	# ---- Iterate over the blocks in the dataset until ---- #
439	<	# ---- we've met the requested total # of events    ---- #
440	<	while ( (eventsRemaining > 0) and (blockCount < numBlocksInDataset) and (jobCount < totalNumberOfJobs)):
441	<	block = blocks[blockCount]
442	<	blockCount += 1
443	<	if block not in jobsOfBlock.keys() :
444	<	jobsOfBlock[block] = []
445	<
446	<	if self.eventsbyblock.has_key(block) :
447	<	numEventsInBlock = self.eventsbyblock[block]
448	<	common.logger.debug(5,'Events in Block File '+str(numEventsInBlock))
449	<
450	<	files = self.filesbyblock[block]
451	<	numFilesInBlock = len(files)
452	<	if (numFilesInBlock <= 0):
453	<	continue
454	<	fileCount = 0
455	<
456	<	# ---- New block => New job ---- #
457	<	parString = ""
458	<	# counter for number of events in files currently worked on
459	<	filesEventCount = 0
460	<	# flag if next while loop should touch new file
461	<	newFile = 1
462	<	# job event counter
463	<	jobSkipEventCount = 0
464	<
465	<	# ---- Iterate over the files in the block until we've met the requested ---- #
466	<	# ---- total # of events or we've gone over all the files in this block  ---- #
467	<	while ( (eventsRemaining > 0) and (fileCount < numFilesInBlock) and (jobCount < totalNumberOfJobs) ):
468	<	file = files[fileCount]
469	<	if newFile :
470	<	try:
471	<	numEventsInFile = self.eventsbyfile[file]
472	<	common.logger.debug(6, "File "+str(file)+" has "+str(numEventsInFile)+" events")
473	<	# increase filesEventCount
474	<	filesEventCount += numEventsInFile
475	<	# Add file to current job
476	<	parString += '\\\"' + file + '\\\"\,'
477	<	newFile = 0
478	<	except KeyError:
479	<	common.logger.message("File "+str(file)+" has unknown number of events: skipping")
480	<
481	<
482	<	# if less events in file remain than eventsPerJobRequested
483	<	if ( filesEventCount - jobSkipEventCount < eventsPerJobRequested ) :
484	<	# if last file in block
485	<	if ( fileCount == numFilesInBlock-1 ) :
486	<	# end job using last file, use remaining events in block
487	<	# close job and touch new file
488	<	fullString = parString[:-2]
489	<	list_of_lists.append([fullString,str(-1),str(jobSkipEventCount)])
490	<	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(filesEventCount - jobSkipEventCount)+" events (last file in block).")
491	<	self.jobDestination.append(blockSites[block])
492	<	common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
493	<	# fill jobs of block dictionary
494	<	jobsOfBlock[block].append(jobCount+1)
495	<	# reset counter
496	<	jobCount = jobCount + 1
497	<	totalEventCount = totalEventCount + filesEventCount - jobSkipEventCount
498	<	eventsRemaining = eventsRemaining - filesEventCount + jobSkipEventCount
499	<	jobSkipEventCount = 0
500	<	# reset file
501	<	parString = ""
502	<	filesEventCount = 0
503	<	newFile = 1
504	<	fileCount += 1
505	<	else :
506	<	# go to next file
507	<	newFile = 1
508	<	fileCount += 1
509	<	# if events in file equal to eventsPerJobRequested
510	<	elif ( filesEventCount - jobSkipEventCount == eventsPerJobRequested ) :
511	<	# close job and touch new file
512	<	fullString = parString[:-2]
513	<	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
514	<	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.")
515	<	self.jobDestination.append(blockSites[block])
516	<	common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
517	<	jobsOfBlock[block].append(jobCount+1)
518	<	# reset counter
519	<	jobCount = jobCount + 1
520	<	totalEventCount = totalEventCount + eventsPerJobRequested
521	<	eventsRemaining = eventsRemaining - eventsPerJobRequested
522	<	jobSkipEventCount = 0
523	<	# reset file
524	<	parString = ""
525	<	filesEventCount = 0
526	<	newFile = 1
527	<	fileCount += 1
528	<
529	<	# if more events in file remain than eventsPerJobRequested
530	<	else :
531	<	# close job but don't touch new file
532	<	fullString = parString[:-2]
533	<	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
534	<	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.")
535	<	self.jobDestination.append(blockSites[block])
536	<	common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
537	<	jobsOfBlock[block].append(jobCount+1)
538	<	# increase counter
539	<	jobCount = jobCount + 1
540	<	totalEventCount = totalEventCount + eventsPerJobRequested
541	<	eventsRemaining = eventsRemaining - eventsPerJobRequested
542	<	# calculate skip events for last file
543	<	# use filesEventCount (contains several files), jobSkipEventCount and eventsPerJobRequest
544	<	jobSkipEventCount = eventsPerJobRequested - (filesEventCount - jobSkipEventCount - self.eventsbyfile[file])
545	<	# remove all but the last file
546	<	filesEventCount = self.eventsbyfile[file]
547	<	parString = ""
548	<	parString += '\\\"' + file + '\\\"\,'
549	<	pass # END if
550	<	pass # END while (iterate over files in the block)
551	<	pass # END while (iterate over blocks in the dataset)
552	<	self.ncjobs = self.total_number_of_jobs = jobCount
553	<	if (eventsRemaining > 0 and jobCount < totalNumberOfJobs ):
554	<	common.logger.message("Could not run on all requested events because some blocks not hosted at allowed sites.")
555	<	common.logger.message(str(jobCount)+" job(s) can run on "+str(totalEventCount)+" events.\n")
556	<
557	<	# screen output
558	<	screenOutput = "List of jobs and available destination sites:\n\n"
559	<
560	<	# keep trace of block with no sites to print a warning at the end
561	<	noSiteBlock = []
562	<	bloskNoSite = []
563	<
564	<	blockCounter = 0
565	<	for block in blocks:
566	<	if block in jobsOfBlock.keys() :
567	<	blockCounter += 1
568	<	screenOutput += "Block %5i: jobs %20s: sites: %s\n" % (blockCounter,spanRanges(jobsOfBlock[block]),','.join(self.blackWhiteListParser.checkWhiteList(self.blackWhiteListParser.checkBlackList(blockSites[block],block),block)))
569	<	if len(self.blackWhiteListParser.checkWhiteList(self.blackWhiteListParser.checkBlackList(blockSites[block],block),block)) == 0:
570	<	noSiteBlock.append( spanRanges(jobsOfBlock[block]) )
571	<	bloskNoSite.append( blockCounter )
572	<
573	<	common.logger.message(screenOutput)
574	<	if len(noSiteBlock) > 0 and len(bloskNoSite) > 0:
575	<	msg = 'WARNING: No sites are hosting any part of data for block:\n                '
576	<	virgola = ""
577	<	if len(bloskNoSite) > 1:
578	<	virgola = ","
579	<	for block in bloskNoSite:
580	<	msg += ' ' + str(block) + virgola
581	<	msg += '\n               Related jobs:\n                 '
582	<	virgola = ""
583	<	if len(noSiteBlock) > 1:
584	<	virgola = ","
585	<	for range_jobs in noSiteBlock:
586	<	msg += str(range_jobs) + virgola
587	<	msg += '\n               will not be submitted and this block of data can not be analyzed!\n'
588	<	common.logger.message(msg)
370	>	# create the empty structure
371	>	for i in range(njobs):
372	>	jobParams.append("")
373
374	<	self.list_of_args = list_of_lists
374	>	listID=[]
375	>	listField=[]
376	>	listDictions=[]
377	>	exist= os.path.exists(self.argsFile)
378	>	for id in range(njobs):
379	>	job = id + int(firstJobID)
380	>	listID.append(job+1)
381	>	job_ToSave ={}
382	>	concString = ' '
383	>	argu=''
384	>	str_argu = str(job+1)
385	>	if len(jobParams[id]):
386	>	argu = {'JobID': job+1}
387	>	for i in range(len(jobParams[id])):
388	>	argu[self.dict['params'][i]]=jobParams[id][i]
389	>	# just for debug
390	>	str_argu += concString.join(jobParams[id])
391	>	if argu != '': listDictions.append(argu)
392	>	job_ToSave['arguments']= str(job+1)
393	>	job_ToSave['dlsDestination']= self.jobDestination[id]
394	>	listField.append(job_ToSave)
395	>	from ProdCommon.SiteDB.CmsSiteMapper import CmsSEMap
396	>	cms_se = CmsSEMap()
397	>	msg="Job  %s  Arguments:  %s\n"%(str(job+1),str_argu)
398	>	msg+="\t  Destination: %s "%(str(self.jobDestination[id]))
399	>	SEDestination = [cms_se[dest] for dest in self.jobDestination[id]]
400	>	msg+="\t  CMSDestination: %s "%(str(SEDestination))
401	>	common.logger.log(10-1,msg)
402	>	# write xml
403	>	if len(listDictions):
404	>	if exist==False: self.CreateXML()
405	>	self.addEntry(listDictions)
406	>	self.addXMLfile()
407	>	common._db.updateJob_(listID,listField)
408	>	self.zipTarFile()
409		return
410
411	<	def jobSplittingNoInput(self):
594	<	"""
595	<	Perform job splitting based on number of event per job
596	<	"""
597	<	common.logger.debug(5,'Splitting per events')
598	<
599	<	if (self.selectEventsPerJob):
600	<	common.logger.message('Required '+str(self.eventsPerJob)+' events per job ')
601	<	if (self.selectNumberOfJobs):
602	<	common.logger.message('Required '+str(self.theNumberOfJobs)+' jobs in total ')
603	<	if (self.selectTotalNumberEvents):
604	<	common.logger.message('Required '+str(self.total_number_of_events)+' events in total ')
411	>	def addXMLfile(self):
412
413	<	if (self.total_number_of_events < 0):
414	<	msg='Cannot split jobs per Events with "-1" as total number of events'
413	>	import tarfile
414	>	try:
415	>	tar = tarfile.open(self.tarNameWithPath, "a")
416	>	tar.add(self.argsFile, os.path.basename(self.argsFile))
417	>	tar.close()
418	>	except IOError, exc:
419	>	msg = 'Could not add %s to %s \n'%(self.argsFile,self.tarNameWithPath)
420	>	msg += str(exc)
421	>	raise CrabException(msg)
422	>	except tarfile.TarError, exc:
423	>	msg = 'Could not add %s to %s \n'%(self.argsFile,self.tarNameWithPath)
424	>	msg += str(exc)
425		raise CrabException(msg)
426
427	<	if (self.selectEventsPerJob):
611	<	if (self.selectTotalNumberEvents):
612	<	self.total_number_of_jobs = int(self.total_number_of_events/self.eventsPerJob)
613	<	elif(self.selectNumberOfJobs) :
614	<	self.total_number_of_jobs =self.theNumberOfJobs
615	<	self.total_number_of_events =int(self.theNumberOfJobs*self.eventsPerJob)
616	<
617	<	elif (self.selectNumberOfJobs) :
618	<	self.total_number_of_jobs = self.theNumberOfJobs
619	<	self.eventsPerJob = int(self.total_number_of_events/self.total_number_of_jobs)
620	<
621	<	common.logger.debug(5,'N jobs  '+str(self.total_number_of_jobs))
622	<
623	<	# is there any remainder?
624	<	check = int(self.total_number_of_events) - (int(self.total_number_of_jobs)*self.eventsPerJob)
625	<
626	<	common.logger.debug(5,'Check  '+str(check))
627	<
628	<	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')
629	<	if check > 0:
630	<	common.logger.message('Warning: asked '+str(self.total_number_of_events)+' but can do only '+str(int(self.total_number_of_jobs)*self.eventsPerJob))
631	<
632	<	# argument is seed number.$i
633	<	self.list_of_args = []
634	<	for i in range(self.total_number_of_jobs):
635	<	## Since there is no input, any site is good
636	<	self.jobDestination.append([""]) #must be empty to write correctly the xml
637	<	args=[]
638	<	if (self.firstRun):
639	<	## pythia first run
640	<	args.append(str(self.firstRun)+str(i))
641	<	if (self.sourceSeed):
642	<	args.append(str(self.sourceSeed)+str(i))
643	<	if (self.sourceSeedVtx):
644	<	## + vtx random seed
645	<	args.append(str(self.sourceSeedVtx)+str(i))
646	<	if (self.sourceSeedG4):
647	<	## + G4 random seed
648	<	args.append(str(self.sourceSeedG4)+str(i))
649	<	if (self.sourceSeedMix):
650	<	## + Mix random seed
651	<	args.append(str(self.sourceSeedMix)+str(i))
652	<	pass
653	<	pass
654	<	self.list_of_args.append(args)
655	<	pass
656	<
657	<	return
658	<
659	<
660	<	def jobSplittingForScript(self):#CarlosDaniele
427	>	def CreateXML(self):
428		"""
662	–	Perform job splitting based on number of job
429		"""
430	<	common.logger.debug(5,'Splitting per job')
431	<	common.logger.message('Required '+str(self.theNumberOfJobs)+' jobs in total ')
666	<
667	<	self.total_number_of_jobs = self.theNumberOfJobs
668	<
669	<	common.logger.debug(5,'N jobs  '+str(self.total_number_of_jobs))
670	<
671	<	common.logger.message(str(self.total_number_of_jobs)+' jobs can be created')
672	<
673	<	# argument is seed number.$i
674	<	self.list_of_args = []
675	<	for i in range(self.total_number_of_jobs):
676	<	## Since there is no input, any site is good
677	<	# self.jobDestination.append(["Any"])
678	<	self.jobDestination.append([""])
679	<	## no random seed
680	<	self.list_of_args.append([str(i)])
430	>	result = IMProvNode( self.rootArgsFilename )
431	>	outfile = file( self.argsFile, 'w').write(str(result))
432		return
433
434	<	def split(self, jobParams):
435	<
436	<	common.jobDB.load()
686	<	#### Fabio
687	<	njobs = self.total_number_of_jobs
688	<	arglist = self.list_of_args
689	<	# create the empty structure
690	<	for i in range(njobs):
691	<	jobParams.append("")
692	<
693	<	for job in range(njobs):
694	<	jobParams[job] = arglist[job]
695	<	# print str(arglist[job])
696	<	# print jobParams[job]
697	<	common.jobDB.setArguments(job, jobParams[job])
698	<	common.logger.debug(5,"Job "+str(job)+" Destination: "+str(self.jobDestination[job]))
699	<	common.jobDB.setDestination(job, self.jobDestination[job])
434	>	def addEntry(self, listDictions):
435	>	"""
436	>	_addEntry_
437
438	<	common.jobDB.save()
438	>	add an entry to the xml file
439	>	"""
440	>	from IMProv.IMProvLoader import loadIMProvFile
441	>	## load xml
442	>	improvDoc = loadIMProvFile(self.argsFile)
443	>	entrname= 'Job'
444	>	for dictions in listDictions:
445	>	report = IMProvNode(entrname , None, **dictions)
446	>	improvDoc.addNode(report)
447	>	outfile = file( self.argsFile, 'w').write(str(improvDoc))
448		return
449
704	–	def getJobTypeArguments(self, nj, sched):
705	–	result = ''
706	–	for i in common.jobDB.arguments(nj):
707	–	result=result+str(i)+" "
708	–	return result
709	–
450		def numberOfJobs(self):
451	<	# Fabio
712	<	return self.total_number_of_jobs
451	>	return self.dict['njobs']
452
453		def getTarBall(self, exe):
454		"""
455		Return the TarBall with lib and exe
456		"""
457	<
458	<	# if it exist, just return it
459	<	#
721	<	# Marco. Let's start to use relative path for Boss XML files
722	<	#
723	<	self.tgzNameWithPath = common.work_space.pathForTgz()+'share/'+self.tgz_name
724	<	if os.path.exists(self.tgzNameWithPath):
725	<	return self.tgzNameWithPath
457	>	self.tarNameWithPath = common.work_space.pathForTgz()+self.tar_name
458	>	if os.path.exists(self.tarNameWithPath):
459	>	return self.tarNameWithPath
460
461		# Prepare a tar gzipped file with user binaries.
462		self.buildTar_(exe)
463
464	<	return string.strip(self.tgzNameWithPath)
464	>	return string.strip(self.tarNameWithPath)
465
466		def buildTar_(self, executable):
467
468		# First of all declare the user Scram area
469		swArea = self.scram.getSWArea_()
736	–	#print "swArea = ", swArea
737	–	# swVersion = self.scram.getSWVersion()
738	–	# print "swVersion = ", swVersion
470		swReleaseTop = self.scram.getReleaseTop_()
740	–	#print "swReleaseTop = ", swReleaseTop
471
472		## check if working area is release top
473		if swReleaseTop == '' or swArea == swReleaseTop:
474	+	common.logger.debug("swArea = "+swArea+" swReleaseTop ="+swReleaseTop)
475		return
476
477		import tarfile
478		try: # create tar ball
479	<	tar = tarfile.open(self.tgzNameWithPath, "w:gz")
479	>	#tar = tarfile.open(self.tgzNameWithPath, "w:gz")
480	>	tar = tarfile.open(self.tarNameWithPath, "w")
481		## First find the executable
482		if (self.executable != ''):
483		exeWithPath = self.scram.findFile_(executable)
#	Line 755 | Line 487 | class Cmssw(JobType):
487		## then check if it's private or not
488		if exeWithPath.find(swReleaseTop) == -1:
489		# the exe is private, so we must ship
490	<	common.logger.debug(5,"Exe "+exeWithPath+" to be tarred")
490	>	common.logger.debug("Exe "+exeWithPath+" to be tarred")
491		path = swArea+'/'
492		# distinguish case when script is in user project area or given by full path somewhere else
493		if exeWithPath.find(path) >= 0 :
#	Line 769 | Line 501 | class Cmssw(JobType):
501		pass
502
503		## Now get the libraries: only those in local working area
504	+	tar.dereference=True
505		libDir = 'lib'
506		lib = swArea+'/' +libDir
507	<	common.logger.debug(5,"lib "+lib+" to be tarred")
507	>	common.logger.debug("lib "+lib+" to be tarred")
508		if os.path.exists(lib):
509		tar.add(lib,libDir)
510
#	Line 780 | Line 513 | class Cmssw(JobType):
513		module = swArea + '/' + moduleDir
514		if os.path.isdir(module):
515		tar.add(module,moduleDir)
516	+	tar.dereference=False
517
518		## Now check if any data dir(s) is present
519	<	swAreaLen=len(swArea)
520	<	for root, dirs, files in os.walk(swArea):
521	<	if "data" in dirs:
522	<	common.logger.debug(5,"data "+root+"/data"+" to be tarred")
523	<	tar.add(root+"/data",root[swAreaLen:]+"/data")
524	<
525	<	## Add ProdAgent dir to tar
526	<	paDir = 'ProdAgentApi'
527	<	pa = os.environ['CRABDIR'] + '/' + 'ProdAgentApi'
528	<	if os.path.isdir(pa):
529	<	tar.add(pa,paDir)
530	<
531	<	### FEDE FOR DBS PUBLICATION
532	<	## Add PRODCOMMON dir to tar
533	<	prodcommonDir = 'ProdCommon'
800	<	prodcommonPath = os.environ['CRABDIR'] + '/' + 'ProdCommon'
801	<	if os.path.isdir(prodcommonPath):
802	<	tar.add(prodcommonPath,prodcommonDir)
803	<	#############################
519	>	self.dataExist = False
520	>	todo_list = [(i, i) for i in  os.listdir(swArea+"/src")]
521	>	while len(todo_list):
522	>	entry, name = todo_list.pop()
523	>	if name.startswith('crab_0_') or  name.startswith('.') or name == 'CVS':
524	>	continue
525	>	if os.path.isdir(swArea+"/src/"+entry):
526	>	entryPath = entry + '/'
527	>	todo_list += [(entryPath + i, i) for i in  os.listdir(swArea+"/src/"+entry)]
528	>	if name == 'data':
529	>	self.dataExist=True
530	>	common.logger.debug("data "+entry+" to be tarred")
531	>	tar.add(swArea+"/src/"+entry,"src/"+entry)
532	>	pass
533	>	pass
534
535	<	common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
536	<	tar.close()
537	<	except :
538	<	raise CrabException('Could not create tar-ball')
535	>	### CMSSW ParameterSet
536	>	if not self.pset is None:
537	>	cfg_file = common.work_space.jobDir()+self.configFilename()
538	>	tar.add(cfg_file,self.configFilename())
539
540	<	## check for tarball size
541	<	tarballinfo = os.stat(self.tgzNameWithPath)
542	<	if ( tarballinfo.st_size > self.MaxTarBallSize*1024*1024 ) :
543	<	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.')
540	>	try:
541	>	crab_cfg_file = common.work_space.shareDir()+'/crab.cfg'
542	>	tar.add(crab_cfg_file,'crab.cfg')
543	>	except:
544	>	pass
545
546	<	## create tar-ball with ML stuff
547	<	self.MLtgzfile =  common.work_space.pathForTgz()+'share/MLfiles.tgz'
548	<	try:
549	<	tar = tarfile.open(self.MLtgzfile, "w:gz")
546	>	## Add ProdCommon dir to tar
547	>	prodcommonDir = './'
548	>	prodcommonPath = os.environ['CRABDIR'] + '/' + 'external/'
549	>	neededStuff = ['ProdCommon/__init__.py','ProdCommon/FwkJobRep', 'ProdCommon/CMSConfigTools', \
550	>	'ProdCommon/Core', 'ProdCommon/MCPayloads', 'IMProv', 'ProdCommon/Storage', \
551	>	'WMCore/__init__.py','WMCore/Algorithms']
552	>	for file in neededStuff:
553	>	tar.add(prodcommonPath+file,prodcommonDir+file)
554	>
555	>	##### ML stuff
556	>	ML_file_list=['report.py', 'DashboardAPI.py', 'Logger.py', 'ProcInfo.py', 'apmon.py']
557		path=os.environ['CRABDIR'] + '/python/'
558	<	for file in ['report.py', 'DashboardAPI.py', 'Logger.py', 'ProcInfo.py', 'apmon.py', 'parseCrabFjr.py']:
558	>	for file in ML_file_list:
559		tar.add(path+file,file)
560	<	common.logger.debug(5,"Files added to "+self.MLtgzfile+" : "+str(tar.getnames()))
560	>
561	>	##### Utils
562	>	Utils_file_list=['parseCrabFjr.py','writeCfg.py', 'fillCrabFjr.py','cmscp.py']
563	>	for file in Utils_file_list:
564	>	tar.add(path+file,file)
565	>
566	>	##### AdditionalFiles
567	>	tar.dereference=True
568	>	for file in self.additional_inbox_files:
569	>	tar.add(file,string.split(file,'/')[-1])
570	>	tar.dereference=False
571	>	common.logger.log(10-1,"Files in "+self.tarNameWithPath+" : "+str(tar.getnames()))
572	>
573		tar.close()
574	<	except :
575	<	raise CrabException('Could not create ML files tar-ball')
574	>	except IOError, exc:
575	>	msg = 'Could not create tar-ball %s \n'%self.tarNameWithPath
576	>	msg += str(exc)
577	>	raise CrabException(msg)
578	>	except tarfile.TarError, exc:
579	>	msg = 'Could not create tar-ball %s \n'%self.tarNameWithPath
580	>	msg += str(exc)
581	>	raise CrabException(msg)
582
583	<	return
583	>	def zipTarFile(self):
584
585	<	def additionalInputFileTgz(self):
586	<	"""
587	<	Put all additional files into a tar ball and return its name
588	<	"""
589	<	import tarfile
590	<	tarName=  common.work_space.pathForTgz()+'share/'+self.additional_tgz_name
591	<	tar = tarfile.open(tarName, "w:gz")
592	<	for file in self.additional_inbox_files:
593	<	tar.add(file,string.split(file,'/')[-1])
594	<	common.logger.debug(5,"Files added to "+self.additional_tgz_name+" : "+str(tar.getnames()))
595	<	tar.close()
596	<	return tarName
585	>	cmd = "gzip -c %s > %s "%(self.tarNameWithPath,self.tgzNameWithPath)
586	>	res=runCommand(cmd)
587	>
588	>	tarballinfo = os.stat(self.tgzNameWithPath)
589	>	if ( tarballinfo.st_size > self.MaxTarBallSize*1024*1024 ) :
590	>	msg  = 'Input sandbox size of ' + str(float(tarballinfo.st_size)/1024.0/1024.0) + ' MB is larger than the allowed ' + str(self.MaxTarBallSize) \
591	>	+'MB input sandbox limit \n'
592	>	msg += '      and not supported by the direct GRID submission system.\n'
593	>	msg += '      Please use the CRAB server mode by setting server_name=<NAME> in section [CRAB] of your crab.cfg.\n'
594	>	msg += '      For further infos please see https://twiki.cern.ch/twiki/bin/view/CMS/CrabServer#CRABSERVER_for_Users'
595	>	raise CrabException(msg)
596	>
597	>	## create tar-ball with ML stuff
598
599	<	def wsSetupEnvironment(self, nj):
599	>	def wsSetupEnvironment(self, nj=0):
600		"""
601		Returns part of a job script which prepares
602		the execution environment for the job 'nj'.
603		"""
604	+	# FUTURE: Drop support for .cfg when possible
605	+	if (self.CMSSW_major >= 2 and self.CMSSW_minor >= 1) or (self.CMSSW_major >= 3):
606	+	psetName = 'pset.py'
607	+	else:
608	+	psetName = 'pset.cfg'
609		# Prepare JobType-independent part
610	<	txt = ''
610	>	txt = '\n#Written by cms_cmssw::wsSetupEnvironment\n'
611		txt += 'echo ">>> setup environment"\n'
612	<	txt += 'if [ $middleware == LCG ]; then \n'
612	>	txt += 'if [ $middleware == LCG ] || [ $middleware == CAF ] || [ $middleware == LSF ]; then \n'
613		txt += self.wsSetupCMSLCGEnvironment_()
614		txt += 'elif [ $middleware == OSG ]; then\n'
615		txt += '    WORKING_DIR=`/bin/mktemp  -d $OSG_WN_TMP/cms_XXXXXXXXXXXX`\n'
616		txt += '    if [ ! $? == 0 ] ;then\n'
617	<	txt += '        echo "SET_CMS_ENV 10016 ==> OSG $WORKING_DIR could not be created on WN `hostname`"\n'
618	<	txt += '        echo "JOB_EXIT_STATUS = 10016"\n'
619	<	txt += '        echo "JobExitCode=10016" | tee -a $RUNTIME_AREA/$repo\n'
858	<	txt += '        dumpStatus $RUNTIME_AREA/$repo\n'
859	<	txt += '        exit 1\n'
617	>	txt += '        echo "ERROR ==> OSG $WORKING_DIR could not be created on WN `hostname`"\n'
618	>	txt += '        job_exit_code=10016\n'
619	>	txt += '        func_exit\n'
620		txt += '    fi\n'
621		txt += '    echo ">>> Created working directory: $WORKING_DIR"\n'
622		txt += '\n'
#	Line 864 | Line 624 | class Cmssw(JobType):
624		txt += '    cd $WORKING_DIR\n'
625		txt += '    echo ">>> current directory (WORKING_DIR): $WORKING_DIR"\n'
626		txt += self.wsSetupCMSOSGEnvironment_()
627	<	#txt += '    echo "### Set SCRAM ARCH to ' + self.executable_arch + ' ###"\n'
628	<	#txt += '    export SCRAM_ARCH='+self.executable_arch+'\n'
627	>	#Setup SGE Environment
628	>	txt += 'elif [ $middleware == SGE ]; then\n'
629	>	txt += self.wsSetupCMSLCGEnvironment_()
630	>
631	>	txt += 'elif [ $middleware == ARC ]; then\n'
632	>	txt += self.wsSetupCMSLCGEnvironment_()
633	>
634		txt += 'fi\n'
635
636		# Prepare JobType-specific part
#	Line 876 | Line 641 | class Cmssw(JobType):
641		txt += scram+' project CMSSW '+self.version+'\n'
642		txt += 'status=$?\n'
643		txt += 'if [ $status != 0 ] ; then\n'
644	<	txt += '    echo "SET_EXE_ENV 10034 ==>ERROR CMSSW '+self.version+' not found on `hostname`" \n'
645	<	txt += '    echo "JOB_EXIT_STATUS = 10034"\n'
646	<	txt += '    echo "JobExitCode=10034" | tee -a $RUNTIME_AREA/$repo\n'
882	<	txt += '    dumpStatus $RUNTIME_AREA/$repo\n'
883	<	txt += '    if [ $middleware == OSG ]; then \n'
884	<	txt += '        cd $RUNTIME_AREA\n'
885	<	txt += '        echo ">>> current directory (RUNTIME_AREA): $RUNTIME_AREA"\n'
886	<	txt += '        echo ">>> Remove working directory: $WORKING_DIR"\n'
887	<	txt += '        /bin/rm -rf $WORKING_DIR\n'
888	<	txt += '        if [ -d $WORKING_DIR ] ;then\n'
889	<	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'
890	<	txt += '            echo "JOB_EXIT_STATUS = 10018"\n'
891	<	txt += '            echo "JobExitCode=10018" | tee -a $RUNTIME_AREA/$repo\n'
892	<	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
893	<	txt += '        fi\n'
894	<	txt += '    fi \n'
895	<	txt += '    exit 1 \n'
644	>	txt += '    echo "ERROR ==> CMSSW '+self.version+' not found on `hostname`" \n'
645	>	txt += '    job_exit_code=10034\n'
646	>	txt += '    func_exit\n'
647		txt += 'fi \n'
648		txt += 'cd '+self.version+'\n'
649	<	########## FEDE FOR DBS2 ######################
899	<	txt += 'SOFTWARE_DIR=`pwd`\n'
649	>	txt += 'SOFTWARE_DIR=`pwd`; export SOFTWARE_DIR\n'
650		txt += 'echo ">>> current directory (SOFTWARE_DIR): $SOFTWARE_DIR" \n'
901	–	###############################################
902	–	### needed grep for bug in scramv1 ###
903	–	txt += scram+' runtime -sh\n'
651		txt += 'eval `'+scram+' runtime -sh | grep -v SCRAMRT_LSB_JOBNAME`\n'
652	<	txt += 'echo $PATH\n'
652	>	txt += 'if [ $? != 0 ] ; then\n'
653	>	txt += '    echo "ERROR ==> Problem with the command: "\n'
654	>	txt += '    echo "eval \`'+scram+' runtime -sh | grep -v SCRAMRT_LSB_JOBNAME \` at `hostname`"\n'
655	>	txt += '    job_exit_code=10034\n'
656	>	txt += '    func_exit\n'
657	>	txt += 'fi \n'
658		# Handle the arguments:
659		txt += "\n"
660		txt += "## number of arguments (first argument always jobnumber)\n"
661		txt += "\n"
662	<	txt += "if [ $nargs -lt "+str(len(self.argsList[nj].split()))+" ]\n"
662	>	txt += "if [ $nargs -lt "+str(self.argsList)+" ]\n"
663		txt += "then\n"
664	<	txt += "    echo 'SET_EXE_ENV 1 ==> ERROR Too few arguments' +$nargs+ \n"
665	<	txt += '    echo "JOB_EXIT_STATUS = 50113"\n'
666	<	txt += '    echo "JobExitCode=50113" | tee -a $RUNTIME_AREA/$repo\n'
915	<	txt += '    dumpStatus $RUNTIME_AREA/$repo\n'
916	<	txt += '    if [ $middleware == OSG ]; then \n'
917	<	txt += '        cd $RUNTIME_AREA\n'
918	<	txt += '        echo ">>> current directory (RUNTIME_AREA): $RUNTIME_AREA"\n'
919	<	txt += '        echo ">>> Remove working directory: $WORKING_DIR"\n'
920	<	txt += '        /bin/rm -rf $WORKING_DIR\n'
921	<	txt += '        if [ -d $WORKING_DIR ] ;then\n'
922	<	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'
923	<	txt += '            echo "JOB_EXIT_STATUS = 50114"\n'
924	<	txt += '            echo "JobExitCode=50114" | tee -a $RUNTIME_AREA/$repo\n'
925	<	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
926	<	txt += '        fi\n'
927	<	txt += '    fi \n'
928	<	txt += "    exit 1\n"
664	>	txt += "    echo 'ERROR ==> Too few arguments' +$nargs+ \n"
665	>	txt += '    job_exit_code=50113\n'
666	>	txt += "    func_exit\n"
667		txt += "fi\n"
668		txt += "\n"
669
670		# Prepare job-specific part
671		job = common.job_list[nj]
934	–	### FEDE FOR DBS OUTPUT PUBLICATION
672		if (self.datasetPath):
673	+	self.primaryDataset = self.datasetPath.split("/")[1]
674	+	DataTier = self.datasetPath.split("/")[2]
675		txt += '\n'
676		txt += 'DatasetPath='+self.datasetPath+'\n'
677
678	<	datasetpath_split = self.datasetPath.split("/")
679	<
941	<	txt += 'PrimaryDataset='+datasetpath_split[1]+'\n'
942	<	txt += 'DataTier='+datasetpath_split[2]+'\n'
678	>	txt += 'PrimaryDataset='+self.primaryDataset +'\n'
679	>	txt += 'DataTier='+DataTier+'\n'
680		txt += 'ApplicationFamily=cmsRun\n'
681
682		else:
683	+	self.primaryDataset = 'null'
684		txt += 'DatasetPath=MCDataTier\n'
685		txt += 'PrimaryDataset=null\n'
686		txt += 'DataTier=null\n'
687		txt += 'ApplicationFamily=MCDataTier\n'
688	<	if self.pset != None: #CarlosDaniele
688	>	if self.pset != None:
689		pset = os.path.basename(job.configFilename())
690		txt += '\n'
691		txt += 'cp  $RUNTIME_AREA/'+pset+' .\n'
954	–	if (self.datasetPath): # standard job
955	–	txt += 'InputFiles=${args[1]}\n'
956	–	txt += 'MaxEvents=${args[2]}\n'
957	–	txt += 'SkipEvents=${args[3]}\n'
958	–	txt += 'echo "Inputfiles:<$InputFiles>"\n'
959	–	txt += 'sed "s#\'INPUTFILE\'#$InputFiles#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
960	–	txt += 'echo "MaxEvents:<$MaxEvents>"\n'
961	–	txt += 'sed "s#int32 input = 0#int32 input = $MaxEvents#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
962	–	txt += 'echo "SkipEvents:<$SkipEvents>"\n'
963	–	txt += 'sed "s#uint32 skipEvents = 0#uint32 skipEvents = $SkipEvents#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
964	–	else:  # pythia like job
965	–	seedIndex=1
966	–	if (self.firstRun):
967	–	txt += 'FirstRun=${args['+str(seedIndex)+']}\n'
968	–	txt += 'echo "FirstRun: <$FirstRun>"\n'
969	–	txt += 'sed "s#uint32 firstRun = 0#uint32 firstRun = $FirstRun#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
970	–	seedIndex=seedIndex+1
971	–
972	–	if (self.sourceSeed):
973	–	txt += 'Seed=${args['+str(seedIndex)+']}\n'
974	–	txt += 'sed "s#uint32 sourceSeed = 0#uint32 sourceSeed = $Seed#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
975	–	seedIndex=seedIndex+1
976	–	## the following seeds are not always present
977	–	if (self.sourceSeedVtx):
978	–	txt += 'VtxSeed=${args['+str(seedIndex)+']}\n'
979	–	txt += 'echo "VtxSeed: <$VtxSeed>"\n'
980	–	txt += 'sed "s#uint32 VtxSmeared = 0#uint32 VtxSmeared = $VtxSeed#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
981	–	seedIndex += 1
982	–	if (self.sourceSeedG4):
983	–	txt += 'G4Seed=${args['+str(seedIndex)+']}\n'
984	–	txt += 'echo "G4Seed: <$G4Seed>"\n'
985	–	txt += 'sed "s#uint32 g4SimHits = 0#uint32 g4SimHits = $G4Seed#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
986	–	seedIndex += 1
987	–	if (self.sourceSeedMix):
988	–	txt += 'mixSeed=${args['+str(seedIndex)+']}\n'
989	–	txt += 'echo "MixSeed: <$mixSeed>"\n'
990	–	txt += 'sed "s#uint32 mix = 0#uint32 mix = $mixSeed#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
991	–	seedIndex += 1
992	–	pass
993	–	pass
994	–	txt += 'mv -f '+pset+' pset.cfg\n'
692
693	<	if len(self.additional_inbox_files) > 0:
694	<	txt += 'if [ -e $RUNTIME_AREA/'+self.additional_tgz_name+' ] ; then\n'
695	<	txt += '  tar xzvf $RUNTIME_AREA/'+self.additional_tgz_name+'\n'
696	<	txt += 'fi\n'
1000	<	pass
693	>	txt += 'PreserveSeeds='  + ','.join(self.preserveSeeds)  + '; export PreserveSeeds\n'
694	>	txt += 'IncrementSeeds=' + ','.join(self.incrementSeeds) + '; export IncrementSeeds\n'
695	>	txt += 'echo "PreserveSeeds: <$PreserveSeeds>"\n'
696	>	txt += 'echo "IncrementSeeds:<$IncrementSeeds>"\n'
697
698	<	if self.pset != None: #CarlosDaniele
699	<	txt += '\n'
1004	<	txt += 'echo "***** cat pset.cfg *********"\n'
1005	<	txt += 'cat pset.cfg\n'
1006	<	txt += 'echo "****** end pset.cfg ********"\n'
1007	<	txt += '\n'
1008	<	### FEDE FOR DBS OUTPUT PUBLICATION
1009	<	txt += 'PSETHASH=`EdmConfigHash < pset.cfg` \n'
1010	<	txt += 'echo "PSETHASH = $PSETHASH" \n'
1011	<	##############
698	>	txt += 'mv -f ' + pset + ' ' + psetName + '\n'
699	>	else:
700		txt += '\n'
701	+	txt += 'export AdditionalArgs=%s\n'%(self.AdditionalArgs)
702	+
703		return txt
704
705	<	def wsBuildExe(self, nj=0):
705	>	def wsUntarSoftware(self, nj=0):
706		"""
707		Put in the script the commands to build an executable
708		or a library.
709		"""
710
711	<	txt = ""
711	>	txt = '\n#Written by cms_cmssw::wsUntarSoftware\n'
712
713		if os.path.isfile(self.tgzNameWithPath):
714		txt += 'echo ">>> tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+' :" \n'
715	<	txt += 'tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+'\n'
715	>	txt += 'tar zxvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+'\n'
716	>	if  self.debug_wrapper==1 :
717	>	txt += 'ls -Al \n'
718		txt += 'untar_status=$? \n'
719		txt += 'if [ $untar_status -ne 0 ]; then \n'
720	<	txt += '   echo "SET_EXE 1 ==> ERROR Untarring .tgz file failed"\n'
721	<	txt += '   echo "JOB_EXIT_STATUS = $untar_status" \n'
722	<	txt += '   echo "JobExitCode=$untar_status" | tee -a $RUNTIME_AREA/$repo\n'
1031	<	txt += '   if [ $middleware == OSG ]; then \n'
1032	<	txt += '       cd $RUNTIME_AREA\n'
1033	<	txt += '        echo ">>> current directory (RUNTIME_AREA): $RUNTIME_AREA"\n'
1034	<	txt += '        echo ">>> Remove working directory: $WORKING_DIR"\n'
1035	<	txt += '       /bin/rm -rf $WORKING_DIR\n'
1036	<	txt += '       if [ -d $WORKING_DIR ] ;then\n'
1037	<	txt += '           echo "SET_EXE 50999 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after Untarring .tgz file failed"\n'
1038	<	txt += '           echo "JOB_EXIT_STATUS = 50999"\n'
1039	<	txt += '           echo "JobExitCode=50999" | tee -a $RUNTIME_AREA/$repo\n'
1040	<	txt += '           dumpStatus $RUNTIME_AREA/$repo\n'
1041	<	txt += '       fi\n'
1042	<	txt += '   fi \n'
1043	<	txt += '   \n'
1044	<	txt += '   exit 1 \n'
720	>	txt += '   echo "ERROR ==> Untarring .tgz file failed"\n'
721	>	txt += '   job_exit_code=$untar_status\n'
722	>	txt += '   func_exit\n'
723		txt += 'else \n'
724		txt += '   echo "Successful untar" \n'
725		txt += 'fi \n'
726		txt += '\n'
727	<	txt += 'echo ">>> Include ProdAgentApi and PRODCOMMON in PYTHONPATH:"\n'
727	>	txt += 'echo ">>> Include $RUNTIME_AREA in PYTHONPATH:"\n'
728		txt += 'if [ -z "$PYTHONPATH" ]; then\n'
729	<	#### FEDE FOR DBS OUTPUT PUBLICATION
1052	<	txt += '   export PYTHONPATH=$SOFTWARE_DIR/ProdAgentApi:$SOFTWARE_DIR/ProdCommon\n'
729	>	txt += '   export PYTHONPATH=$RUNTIME_AREA/\n'
730		txt += 'else\n'
731	<	txt += '   export PYTHONPATH=$SOFTWARE_DIR/ProdAgentApi:$SOFTWARE_DIR/ProdCommon:${PYTHONPATH}\n'
731	>	txt += '   export PYTHONPATH=$RUNTIME_AREA/:${PYTHONPATH}\n'
732		txt += 'echo "PYTHONPATH=$PYTHONPATH"\n'
1056	–	###################
733		txt += 'fi\n'
734		txt += '\n'
735
#	Line 1061 | Line 737 | class Cmssw(JobType):
737
738		return txt
739
740	<	def modifySteeringCards(self, nj):
740	>	def wsBuildExe(self, nj=0):
741		"""
742	<	modify the card provided by the user,
743	<	writing a new card into share dir
742	>	Put in the script the commands to build an executable
743	>	or a library.
744		"""
745
746	+	txt = '\n#Written by cms_cmssw::wsBuildExe\n'
747	+	txt += 'echo ">>> moving CMSSW software directories in `pwd`" \n'
748	+
749	+	txt += 'rm -r lib/ module/ \n'
750	+	txt += 'mv $RUNTIME_AREA/lib/ . \n'
751	+	txt += 'mv $RUNTIME_AREA/module/ . \n'
752	+	if self.dataExist == True:
753	+	txt += 'rm -r src/ \n'
754	+	txt += 'mv $RUNTIME_AREA/src/ . \n'
755	+	if len(self.additional_inbox_files)>0:
756	+	for file in self.additional_inbox_files:
757	+	txt += 'mv $RUNTIME_AREA/'+os.path.basename(file)+' . \n'
758	+	# txt += 'mv $RUNTIME_AREA/ProdCommon/ . \n'
759	+	# txt += 'mv $RUNTIME_AREA/IMProv/ . \n'
760	+
761	+	txt += 'echo ">>> Include $RUNTIME_AREA in PYTHONPATH:"\n'
762	+	txt += 'if [ -z "$PYTHONPATH" ]; then\n'
763	+	txt += '   export PYTHONPATH=$RUNTIME_AREA/\n'
764	+	txt += 'else\n'
765	+	txt += '   export PYTHONPATH=$RUNTIME_AREA/:${PYTHONPATH}\n'
766	+	txt += 'echo "PYTHONPATH=$PYTHONPATH"\n'
767	+	txt += 'fi\n'
768	+	txt += '\n'
769	+
770	+	if self.pset != None:
771	+	# FUTURE: Drop support for .cfg when possible
772	+	if (self.CMSSW_major >= 2 and self.CMSSW_minor >= 1) or (self.CMSSW_major >= 3):
773	+	psetName = 'pset.py'
774	+	else:
775	+	psetName = 'pset.cfg'
776	+	# FUTURE: Can simply for 2_1_x and higher
777	+	txt += '\n'
778	+	if self.debug_wrapper == 1:
779	+	txt += 'echo "***** cat ' + psetName + ' *********"\n'
780	+	txt += 'cat ' + psetName + '\n'
781	+	txt += 'echo "****** end ' + psetName + ' ********"\n'
782	+	txt += '\n'
783	+	txt += 'echo "***********************" \n'
784	+	txt += 'which edmConfigHash \n'
785	+	txt += 'echo "***********************" \n'
786	+	if (self.CMSSW_major >= 2 and self.CMSSW_minor >= 1) or (self.CMSSW_major >= 3):
787	+	txt += 'edmConfigHash ' + psetName + ' \n'
788	+	txt += 'PSETHASH=`edmConfigHash ' + psetName + '` \n'
789	+	else:
790	+	txt += 'PSETHASH=`edmConfigHash < ' + psetName + '` \n'
791	+	txt += 'echo "PSETHASH = $PSETHASH" \n'
792	+	#### FEDE temporary fix for noEdm files #####
793	+	txt += 'if [ -z "$PSETHASH" ]; then \n'
794	+	txt += '   export PSETHASH=null\n'
795	+	txt += 'fi \n'
796	+	#############################################
797	+	txt += '\n'
798	+	return txt
799	+
800	+
801		def executableName(self):
802	<	if self.scriptExe: #CarlosDaniele
802	>	if self.scriptExe:
803		return "sh "
804		else:
805		return self.executable
806
807		def executableArgs(self):
808	<	if self.scriptExe:#CarlosDaniele
809	<	return   self.scriptExe + " $NJob"
808	>	# FUTURE: This function tests the CMSSW version. Can be simplified as we drop support for old versions
809	>	if self.scriptExe:
810	>	return self.scriptExe + " $NJob $AdditionalArgs"
811		else:
812	<	# if >= CMSSW_1_5_X, add -j crab_fjr.xml
813	<	version_array = self.scram.getSWVersion().split('_')
814	<	major = 0
815	<	minor = 0
816	<	try:
1085	<	major = int(version_array[1])
1086	<	minor = int(version_array[2])
1087	<	except:
1088	<	msg = "Cannot parse CMSSW version string: " + "_".join(version_array) + " for major and minor release number!"
1089	<	raise CrabException(msg)
1090	<	if major >= 1 and minor >= 5 :
1091	<	return " -j " + self.fjrFileName + " -p pset.cfg"
812	>	ex_args = ""
813	>	ex_args += " -j $RUNTIME_AREA/crab_fjr_$NJob.xml"
814	>	# Type of config file depends on CMSSW version
815	>	if self.CMSSW_major >= 2 :
816	>	ex_args += " -p pset.py"
817		else:
818	<	return " -p pset.cfg"
818	>	ex_args += " -p pset.cfg"
819	>	return ex_args
820
821		def inputSandbox(self, nj):
822		"""
823		Returns a list of filenames to be put in JDL input sandbox.
824		"""
825		inp_box = []
1100	–	# # dict added to delete duplicate from input sandbox file list
1101	–	# seen = {}
1102	–	## code
826		if os.path.isfile(self.tgzNameWithPath):
827		inp_box.append(self.tgzNameWithPath)
828	<	if os.path.isfile(self.MLtgzfile):
1106	<	inp_box.append(self.MLtgzfile)
1107	<	## config
1108	<	if not self.pset is None:
1109	<	inp_box.append(common.work_space.pathForTgz() + 'job/' + self.configFilename())
1110	<	## additional input files
1111	<	tgz = self.additionalInputFileTgz()
1112	<	inp_box.append(tgz)
828	>	inp_box.append(common.work_space.jobDir() + self.scriptName)
829		return inp_box
830
831		def outputSandbox(self, nj):
#	Line 1121 | Line 837 | class Cmssw(JobType):
837		## User Declared output files
838		for out in (self.output_file+self.output_file_sandbox):
839		n_out = nj + 1
840	<	out_box.append(self.numberFile_(out,str(n_out)))
840	>	out_box.append(numberFile(out,str(n_out)))
841		return out_box
842
1127	–	def prepareSteeringCards(self):
1128	–	"""
1129	–	Make initial modifications of the user's steering card file.
1130	–	"""
1131	–	return
843
844		def wsRenameOutput(self, nj):
845		"""
846		Returns part of a job script which renames the produced files.
847		"""
848
849	<	txt = '\n'
850	<	txt += 'echo" >>> directory content:"\n'
851	<	txt += 'ls \n'
849	>	txt = '\n#Written by cms_cmssw::wsRenameOutput\n'
850	>	txt += 'echo ">>> current directory (SOFTWARE_DIR): $SOFTWARE_DIR" \n'
851	>	txt += 'echo ">>> current directory content:"\n'
852	>	if self.debug_wrapper==1:
853	>	txt += 'ls -Al\n'
854		txt += '\n'
855
1143	–	txt += 'output_exit_status=0\n'
1144	–
1145	–	for fileWithSuffix in (self.output_file_sandbox):
1146	–	output_file_num = self.numberFile_(fileWithSuffix, '$NJob')
1147	–	txt += '\n'
1148	–	txt += '# check output file\n'
1149	–	txt += 'if [ -e ./'+fileWithSuffix+' ] ; then\n'
1150	–	#txt += '    mv '+fileWithSuffix+' $RUNTIME_AREA\n'
1151	–	txt += '    mv '+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n'
1152	–	#txt += '    cp $RUNTIME_AREA/'+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n'
1153	–	txt += '    ln -s $RUNTIME_AREA/'+output_file_num+' $RUNTIME_AREA/'+fileWithSuffix+'\n'
1154	–	txt += 'else\n'
1155	–	txt += '    exit_status=60302\n'
1156	–	txt += '    echo "ERROR: Problem with output file '+fileWithSuffix+'"\n'
1157	–	if common.scheduler.boss_scheduler_name == 'condor_g':
1158	–	txt += '    if [ $middleware == OSG ]; then \n'
1159	–	txt += '        echo "prepare dummy output file"\n'
1160	–	txt += '        echo "Processing of job output failed" > $RUNTIME_AREA/'+output_file_num+'\n'
1161	–	txt += '    fi \n'
1162	–	txt += 'fi\n'
1163	–
856		for fileWithSuffix in (self.output_file):
857	<	output_file_num = self.numberFile_(fileWithSuffix, '$NJob')
857	>	output_file_num = numberFile(fileWithSuffix, '$NJob')
858		txt += '\n'
859		txt += '# check output file\n'
860		txt += 'if [ -e ./'+fileWithSuffix+' ] ; then\n'
861	<	#txt += '    mv '+fileWithSuffix+' $RUNTIME_AREA\n'
862	<	txt += '    mv '+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n'
863	<	#txt += '    cp $RUNTIME_AREA/'+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n'
864	<	txt += '    ln -s $RUNTIME_AREA/'+output_file_num+' $RUNTIME_AREA/'+fileWithSuffix+'\n'
861	>	if (self.copy_data == 1):  # For OSG nodes, file is in $WORKING_DIR, should not be moved to $RUNTIME_AREA
862	>	txt += '    mv '+fileWithSuffix+' '+output_file_num+'\n'
863	>	txt += '    ln -s `pwd`/'+output_file_num+' $RUNTIME_AREA/'+fileWithSuffix+'\n'
864	>	else:
865	>	txt += '    mv '+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n'
866	>	txt += '    ln -s $RUNTIME_AREA/'+output_file_num+' $RUNTIME_AREA/'+fileWithSuffix+'\n'
867		txt += 'else\n'
868	<	txt += '    exit_status=60302\n'
869	<	txt += '    echo "ERROR: Problem with output file '+fileWithSuffix+'"\n'
870	<	txt += '    echo "JOB_EXIT_STATUS = $exit_status"\n'
1177	<	txt += '    output_exit_status=$exit_status\n'
1178	<	if common.scheduler.boss_scheduler_name == 'condor_g':
868	>	txt += '    job_exit_code=60302\n'
869	>	txt += '    echo "WARNING: Output file '+fileWithSuffix+' not found"\n'
870	>	if common.scheduler.name().upper() == 'CONDOR_G':
871		txt += '    if [ $middleware == OSG ]; then \n'
872		txt += '        echo "prepare dummy output file"\n'
873		txt += '        echo "Processing of job output failed" > $RUNTIME_AREA/'+output_file_num+'\n'
#	Line 1183 | Line 875 | class Cmssw(JobType):
875		txt += 'fi\n'
876		file_list = []
877		for fileWithSuffix in (self.output_file):
878	<	file_list.append(self.numberFile_(fileWithSuffix, '$NJob'))
878	>	file_list.append(numberFile('$SOFTWARE_DIR/'+fileWithSuffix, '$NJob'))
879
880	<	txt += 'file_list="'+string.join(file_list,' ')+'"\n'
880	>	txt += 'file_list="'+string.join(file_list,',')+'"\n'
881	>	txt += '\n'
882	>	txt += 'echo ">>> current directory (SOFTWARE_DIR): $SOFTWARE_DIR" \n'
883	>	txt += 'echo ">>> current directory content:"\n'
884	>	if self.debug_wrapper==1:
885	>	txt += 'ls -Al\n'
886	>	txt += '\n'
887		txt += 'cd $RUNTIME_AREA\n'
888		txt += 'echo ">>> current directory (RUNTIME_AREA):  $RUNTIME_AREA"\n'
889		return txt
890
1193	–	def numberFile_(self, file, txt):
1194	–	"""
1195	–	append _'txt' before last extension of a file
1196	–	"""
1197	–	p = string.split(file,".")
1198	–	# take away last extension
1199	–	name = p[0]
1200	–	for x in p[1:-1]:
1201	–	name=name+"."+x
1202	–	# add "_txt"
1203	–	if len(p)>1:
1204	–	ext = p[len(p)-1]
1205	–	result = name + '_' + txt + "." + ext
1206	–	else:
1207	–	result = name + '_' + txt
1208	–
1209	–	return result
1210	–
891		def getRequirements(self, nj=[]):
892		"""
893		return job requirements to add to jdl files
#	Line 1217 | Line 897 | class Cmssw(JobType):
897		req='Member("VO-cms-' + \
898		self.version + \
899		'", other.GlueHostApplicationSoftwareRunTimeEnvironment)'
900	<	## SL add requirement for OS version only if SL4
1221	<	#reSL4 = re.compile( r'slc4' )
1222	<	if self.executable_arch: # and reSL4.search(self.executable_arch):
900	>	if self.executable_arch:
901		req+=' && Member("VO-cms-' + \
902		self.executable_arch + \
903		'", other.GlueHostApplicationSoftwareRunTimeEnvironment)'
904
905		req = req + ' && (other.GlueHostNetworkAdapterOutboundIP)'
906	+	if ( common.scheduler.name() == "glitecoll" ) or ( common.scheduler.name() == "glite"):
907	+	req += ' && other.GlueCEStateStatus == "Production" '
908
909		return req
910
911		def configFilename(self):
912		""" return the config filename """
913	<	return self.name()+'.cfg'
913	>	# FUTURE: Can remove cfg mode for CMSSW >= 2_1_x
914	>	if (self.CMSSW_major >= 2 and self.CMSSW_minor >= 1) or (self.CMSSW_major >= 3):
915	>	return self.name()+'.py'
916	>	else:
917	>	return self.name()+'.cfg'
918
919		def wsSetupCMSOSGEnvironment_(self):
920		"""
921		Returns part of a job script which is prepares
922		the execution environment and which is common for all CMS jobs.
923		"""
924	<	txt = '    echo ">>> setup CMS OSG environment:"\n'
924	>	txt = '\n#Written by cms_cmssw::wsSetupCMSOSGEnvironment_\n'
925	>	txt += '    echo ">>> setup CMS OSG environment:"\n'
926		txt += '    echo "set SCRAM ARCH to ' + self.executable_arch + '"\n'
927		txt += '    export SCRAM_ARCH='+self.executable_arch+'\n'
928		txt += '    echo "SCRAM_ARCH = $SCRAM_ARCH"\n'
#	Line 1245 | Line 930 | class Cmssw(JobType):
930		txt += '      # Use $OSG_APP/cmssoft/cms/cmsset_default.sh to setup cms software\n'
931		txt += '        source $OSG_APP/cmssoft/cms/cmsset_default.sh '+self.version+'\n'
932		txt += '    else\n'
933	<	txt += '        echo "SET_CMS_ENV 10020 ==> ERROR $OSG_APP/cmssoft/cms/cmsset_default.sh file not found"\n'
934	<	txt += '        echo "JOB_EXIT_STATUS = 10020"\n'
935	<	txt += '        echo "JobExitCode=10020" | tee -a $RUNTIME_AREA/$repo\n'
1251	<	txt += '        dumpStatus $RUNTIME_AREA/$repo\n'
1252	<	txt += '\n'
1253	<	txt += '        cd $RUNTIME_AREA\n'
1254	<	txt += '        echo ">>> current directory (RUNTIME_AREA): $RUNTIME_AREA"\n'
1255	<	txt += '        echo ">>> Remove working directory: $WORKING_DIR"\n'
1256	<	txt += '        /bin/rm -rf $WORKING_DIR\n'
1257	<	txt += '        if [ -d $WORKING_DIR ] ;then\n'
1258	<	txt += '            echo "SET_CMS_ENV 10017 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after $OSG_APP/cmssoft/cms/cmsset_default.sh file not found"\n'
1259	<	txt += '            echo "JOB_EXIT_STATUS = 10017"\n'
1260	<	txt += '            echo "JobExitCode=10017" | tee -a $RUNTIME_AREA/$repo\n'
1261	<	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
1262	<	txt += '        fi\n'
1263	<	txt += '\n'
1264	<	txt += '        exit 1\n'
933	>	txt += '        echo "ERROR ==> $OSG_APP/cmssoft/cms/cmsset_default.sh file not found"\n'
934	>	txt += '        job_exit_code=10020\n'
935	>	txt += '        func_exit\n'
936		txt += '    fi\n'
937		txt += '\n'
938	<	txt += '    echo "SET_CMS_ENV 0 ==> setup cms environment ok"\n'
938	>	txt += '    echo "==> setup cms environment ok"\n'
939		txt += '    echo "SCRAM_ARCH = $SCRAM_ARCH"\n'
940
941		return txt
942
1272	–	### OLI_DANIELE
943		def wsSetupCMSLCGEnvironment_(self):
944		"""
945		Returns part of a job script which is prepares
946		the execution environment and which is common for all CMS jobs.
947		"""
948	<	txt = '    echo ">>> setup CMS LCG environment:"\n'
948	>	txt = '\n#Written by cms_cmssw::wsSetupCMSLCGEnvironment_\n'
949	>	txt += '    echo ">>> setup CMS LCG environment:"\n'
950		txt += '    echo "set SCRAM ARCH and BUILD_ARCH to ' + self.executable_arch + ' ###"\n'
951		txt += '    export SCRAM_ARCH='+self.executable_arch+'\n'
952		txt += '    export BUILD_ARCH='+self.executable_arch+'\n'
953		txt += '    if [ ! $VO_CMS_SW_DIR ] ;then\n'
954	<	txt += '        echo "SET_CMS_ENV 10031 ==> ERROR CMS software dir not found on WN `hostname`"\n'
955	<	txt += '        echo "JOB_EXIT_STATUS = 10031" \n'
956	<	txt += '        echo "JobExitCode=10031" | tee -a $RUNTIME_AREA/$repo\n'
1286	<	txt += '        dumpStatus $RUNTIME_AREA/$repo\n'
1287	<	txt += '        exit 1\n'
954	>	txt += '        echo "ERROR ==> CMS software dir not found on WN `hostname`"\n'
955	>	txt += '        job_exit_code=10031\n'
956	>	txt += '        func_exit\n'
957		txt += '    else\n'
958		txt += '        echo "Sourcing environment... "\n'
959		txt += '        if [ ! -s $VO_CMS_SW_DIR/cmsset_default.sh ] ;then\n'
960	<	txt += '            echo "SET_CMS_ENV 10020 ==> ERROR cmsset_default.sh file not found into dir $VO_CMS_SW_DIR"\n'
961	<	txt += '            echo "JOB_EXIT_STATUS = 10020"\n'
962	<	txt += '            echo "JobExitCode=10020" | tee -a $RUNTIME_AREA/$repo\n'
1294	<	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
1295	<	txt += '            exit 1\n'
960	>	txt += '            echo "ERROR ==> cmsset_default.sh file not found into dir $VO_CMS_SW_DIR"\n'
961	>	txt += '            job_exit_code=10020\n'
962	>	txt += '            func_exit\n'
963		txt += '        fi\n'
964		txt += '        echo "sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
965		txt += '        source $VO_CMS_SW_DIR/cmsset_default.sh\n'
966		txt += '        result=$?\n'
967		txt += '        if [ $result -ne 0 ]; then\n'
968	<	txt += '            echo "SET_CMS_ENV 10032 ==> ERROR problem sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
969	<	txt += '            echo "JOB_EXIT_STATUS = 10032"\n'
970	<	txt += '            echo "JobExitCode=10032" | tee -a $RUNTIME_AREA/$repo\n'
1304	<	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
1305	<	txt += '            exit 1\n'
968	>	txt += '            echo "ERROR ==> problem sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
969	>	txt += '            job_exit_code=10032\n'
970	>	txt += '            func_exit\n'
971		txt += '        fi\n'
972		txt += '    fi\n'
973		txt += '    \n'
974	<	txt += '    echo "SET_CMS_ENV 0 ==> setup cms environment ok"\n'
974	>	txt += '    echo "==> setup cms environment ok"\n'
975		return txt
976
977	<	### FEDE FOR DBS OUTPUT PUBLICATION
1313	<	def modifyReport(self, nj):
977	>	def wsModifyReport(self, nj):
978		"""
979		insert the part of the script that modifies the FrameworkJob Report
980		"""
981
982		txt = ''
983	<	try:
984	<	publish_data = int(self.cfg_params['USER.publish_data'])
985	<	except KeyError:
986	<	publish_data = 0
987	<	if (publish_data == 1):
1324	<	txt += 'echo ">>> Modify Job Report:" \n'
1325	<	################ FEDE FOR DBS2 #############################################
1326	<	txt += 'chmod a+x $SOFTWARE_DIR/ProdAgentApi/FwkJobRep/ModifyJobReport.py\n'
1327	<	#############################################################################
983	>	publish_data = int(self.cfg_params.get('USER.publish_data',0))
984	>	#if (publish_data == 1):
985	>	if (self.copy_data == 1):
986	>	txt = '\n#Written by cms_cmssw::wsModifyReport\n'
987	>	publish_data = int(self.cfg_params.get('USER.publish_data',0))
988
1329	–	txt += 'if [ -z "$SE" ]; then\n'
1330	–	txt += '    SE="" \n'
1331	–	txt += 'fi \n'
1332	–	txt += 'if [ -z "$SE_PATH" ]; then\n'
1333	–	txt += '    SE_PATH="" \n'
1334	–	txt += 'fi \n'
1335	–	txt += 'echo "SE = $SE"\n'
1336	–	txt += 'echo "SE_PATH = $SE_PATH"\n'
989
990	<	processedDataset = self.cfg_params['USER.publish_data_name']
991	<	txt += 'ProcessedDataset='+processedDataset+'\n'
992	<	#### LFN=/store/user/<user>/processedDataset_PSETHASH
1341	<	txt += 'if [ "$SE_PATH" == "" ]; then\n'
1342	<	#### FEDE: added slash in LFN ##############
990	>	txt += 'if [ $StageOutExitStatus -eq 0 ]; then\n'
991	>	txt += '    FOR_LFN=$LFNBaseName\n'
992	>	txt += 'else\n'
993		txt += '    FOR_LFN=/copy_problems/ \n'
994	<	txt += 'else \n'
995	<	txt += '    tmp=`echo $SE_PATH | awk -F \'store\' \'{print$2}\'` \n'
996	<	#####  FEDE TO BE CHANGED, BECAUSE STORE IS HARDCODED!!!! ########
997	<	txt += '    FOR_LFN=/store$tmp \n'
998	<	txt += 'fi \n'
999	<	txt += 'echo "ProcessedDataset = $ProcessedDataset"\n'
994	>	txt += 'fi\n'
995	>
996	>	txt += 'echo ">>> Modify Job Report:" \n'
997	>	txt += 'chmod a+x $RUNTIME_AREA/ProdCommon/FwkJobRep/ModifyJobReport.py\n'
998	>	txt += 'echo "SE = $SE"\n'
999	>	txt += 'echo "SE_PATH = $SE_PATH"\n'
1000		txt += 'echo "FOR_LFN = $FOR_LFN" \n'
1001		txt += 'echo "CMSSW_VERSION = $CMSSW_VERSION"\n\n'
1352	–	#txt += 'echo "$RUNTIME_AREA/'+self.version+'/ProdAgentApi/FwkJobRep/ModifyJobReport.py crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH"\n'
1353	–	txt += 'echo "$SOFTWARE_DIR/ProdAgentApi/FwkJobRep/ModifyJobReport.py crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH"\n'
1354	–	txt += '$SOFTWARE_DIR/ProdAgentApi/FwkJobRep/ModifyJobReport.py crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH\n'
1355	–	#txt += '$RUNTIME_AREA/'+self.version+'/ProdAgentApi/FwkJobRep/ModifyJobReport.py crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH\n'
1002
1003	+
1004	+	args = 'fjr $RUNTIME_AREA/crab_fjr_$NJob.xml n_job $NJob for_lfn $FOR_LFN PrimaryDataset $PrimaryDataset  ApplicationFamily $ApplicationFamily ApplicationName $executable cmssw_version $CMSSW_VERSION psethash $PSETHASH se_name $SE se_path $SE_PATH'
1005	+	if (publish_data == 1):
1006	+	processedDataset = self.cfg_params['USER.publish_data_name']
1007	+	txt += 'ProcessedDataset='+processedDataset+'\n'
1008	+	txt += 'echo "ProcessedDataset = $ProcessedDataset"\n'
1009	+	args += ' UserProcessedDataset $USER-$ProcessedDataset-$PSETHASH'
1010	+
1011	+	txt += 'echo "$RUNTIME_AREA/ProdCommon/FwkJobRep/ModifyJobReport.py '+str(args)+'"\n'
1012	+	txt += '$RUNTIME_AREA/ProdCommon/FwkJobRep/ModifyJobReport.py '+str(args)+'\n'
1013		txt += 'modifyReport_result=$?\n'
1358	–	txt += 'echo modifyReport_result = $modifyReport_result\n'
1014		txt += 'if [ $modifyReport_result -ne 0 ]; then\n'
1015	<	txt += '    exit_status=1\n'
1016	<	txt += '    echo "ERROR: Problem with ModifyJobReport"\n'
1015	>	txt += '    modifyReport_result=70500\n'
1016	>	txt += '    job_exit_code=$modifyReport_result\n'
1017	>	txt += '    echo "ModifyReportResult=$modifyReport_result" | tee -a $RUNTIME_AREA/$repo\n'
1018	>	txt += '    echo "WARNING: Problem with ModifyJobReport"\n'
1019		txt += 'else\n'
1020	<	txt += '    mv NewFrameworkJobReport.xml crab_fjr_$NJob.xml\n'
1020	>	txt += '    mv NewFrameworkJobReport.xml $RUNTIME_AREA/crab_fjr_$NJob.xml\n'
1021		txt += 'fi\n'
1365	–	else:
1366	–	txt += 'echo "no data publication required"\n'
1022		return txt
1023
1024	<	def cleanEnv(self):
1025	<	txt = ''
1026	<	txt += 'if [ $middleware == OSG ]; then\n'
1027	<	txt += '    cd $RUNTIME_AREA\n'
1028	<	txt += '    echo ">>> current directory (RUNTIME_AREA): $RUNTIME_AREA"\n'
1029	<	txt += '    echo ">>> Remove working directory: $WORKING_DIR"\n'
1030	<	txt += '    /bin/rm -rf $WORKING_DIR\n'
1031	<	txt += '    if [ -d $WORKING_DIR ] ;then\n'
1032	<	txt += '        echo "SET_EXE 60999 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after cleanup of WN"\n'
1033	<	txt += '        echo "JOB_EXIT_STATUS = 60999"\n'
1034	<	txt += '        echo "JobExitCode=60999" | tee -a $RUNTIME_AREA/$repo\n'
1035	<	txt += '        dumpStatus $RUNTIME_AREA/$repo\n'
1024	>	def wsParseFJR(self):
1025	>	"""
1026	>	Parse the FrameworkJobReport to obtain useful infos
1027	>	"""
1028	>	txt = '\n#Written by cms_cmssw::wsParseFJR\n'
1029	>	txt += 'echo ">>> Parse FrameworkJobReport crab_fjr.xml"\n'
1030	>	txt += 'if [ -s $RUNTIME_AREA/crab_fjr_$NJob.xml ]; then\n'
1031	>	txt += '    if [ -s $RUNTIME_AREA/parseCrabFjr.py ]; then\n'
1032	>	txt += '        cmd_out=`python $RUNTIME_AREA/parseCrabFjr.py --input $RUNTIME_AREA/crab_fjr_$NJob.xml --dashboard $MonitorID,$MonitorJobID '+self.debugWrap+'`\n'
1033	>	if self.debug_wrapper==1 :
1034	>	txt += '        echo "Result of parsing the FrameworkJobReport crab_fjr.xml: $cmd_out"\n'
1035	>	txt += '        executable_exit_status=`python $RUNTIME_AREA/parseCrabFjr.py --input $RUNTIME_AREA/crab_fjr_$NJob.xml --exitcode`\n'
1036	>	txt += '        if [ $executable_exit_status -eq 50115 ];then\n'
1037	>	txt += '            echo ">>> crab_fjr.xml contents: "\n'
1038	>	txt += '            cat $RUNTIME_AREA/crab_fjr_$NJob.xml\n'
1039	>	txt += '            echo "Wrong FrameworkJobReport --> does not contain useful info. ExitStatus: $executable_exit_status"\n'
1040	>	txt += '        elif [ $executable_exit_status -eq -999 ];then\n'
1041	>	txt += '            echo "ExitStatus from FrameworkJobReport not available. not available. Using exit code of executable from command line."\n'
1042	>	txt += '        else\n'
1043	>	txt += '            echo "Extracted ExitStatus from FrameworkJobReport parsing output: $executable_exit_status"\n'
1044	>	txt += '        fi\n'
1045	>	txt += '    else\n'
1046	>	txt += '        echo "CRAB python script to parse CRAB FrameworkJobReport crab_fjr.xml is not available, using exit code of executable from command line."\n'
1047		txt += '    fi\n'
1048	+	#### Patch to check input data reading for CMSSW16x Hopefully we-ll remove it asap
1049	+	txt += '    if [ $executable_exit_status -eq 0 ];then\n'
1050	+	txt += '        echo ">>> Executable succeded  $executable_exit_status"\n'
1051	+	## This cannot more work given the changes on the Job argumentsJob
1052	+	"""
1053	+	if (self.datasetPath and not (self.dataset_pu or self.useParent==1)) :
1054	+	# VERIFY PROCESSED DATA
1055	+	txt += '        echo ">>> Verify list of processed files:"\n'
1056	+	txt += '        echo $InputFiles |tr -d \'\\\\\' |tr \',\' \'\\n\'|tr -d \'"\' > input-files.txt\n'
1057	+	txt += '        python $RUNTIME_AREA/parseCrabFjr.py --input $RUNTIME_AREA/crab_fjr_$NJob.xml --lfn > processed-files.txt\n'
1058	+	txt += '        cat input-files.txt  | sort | uniq > tmp.txt\n'
1059	+	txt += '        mv tmp.txt input-files.txt\n'
1060	+	txt += '        echo "cat input-files.txt"\n'
1061	+	txt += '        echo "----------------------"\n'
1062	+	txt += '        cat input-files.txt\n'
1063	+	txt += '        cat processed-files.txt | sort | uniq > tmp.txt\n'
1064	+	txt += '        mv tmp.txt processed-files.txt\n'
1065	+	txt += '        echo "----------------------"\n'
1066	+	txt += '        echo "cat processed-files.txt"\n'
1067	+	txt += '        echo "----------------------"\n'
1068	+	txt += '        cat processed-files.txt\n'
1069	+	txt += '        echo "----------------------"\n'
1070	+	txt += '        diff -qbB input-files.txt processed-files.txt\n'
1071	+	txt += '        fileverify_status=$?\n'
1072	+	txt += '        if [ $fileverify_status -ne 0 ]; then\n'
1073	+	txt += '            executable_exit_status=30001\n'
1074	+	txt += '            echo "ERROR ==> not all input files processed"\n'
1075	+	txt += '            echo "      ==> list of processed files from crab_fjr.xml differs from list in pset.cfg"\n'
1076	+	txt += '            echo "      ==> diff input-files.txt processed-files.txt"\n'
1077	+	txt += '        fi\n'
1078	+	"""
1079	+	txt += '    fi\n'
1080	+	txt += 'else\n'
1081	+	txt += '    echo "CRAB FrameworkJobReport crab_fjr.xml is not available, using exit code of executable from command line."\n'
1082		txt += 'fi\n'
1083		txt += '\n'
1084	+	txt += 'if [ $executable_exit_status -ne 0 ] && [ $executable_exit_status -ne 50115 ] && [ $executable_exit_status -ne 50117 ] && [ $executable_exit_status -ne 30001 ];then\n'
1085	+	txt += '    echo ">>> Executable failed  $executable_exit_status"\n'
1086	+	txt += '    echo "ExeExitCode=$executable_exit_status" | tee -a $RUNTIME_AREA/$repo\n'
1087	+	txt += '    echo "EXECUTABLE_EXIT_STATUS = $executable_exit_status"\n'
1088	+	txt += '    job_exit_code=$executable_exit_status\n'
1089	+	txt += '    func_exit\n'
1090	+	txt += 'fi\n\n'
1091	+	txt += 'echo "ExeExitCode=$executable_exit_status" | tee -a $RUNTIME_AREA/$repo\n'
1092	+	txt += 'echo "EXECUTABLE_EXIT_STATUS = $executable_exit_status"\n'
1093	+	txt += 'job_exit_code=$executable_exit_status\n'
1094	+
1095		return txt
1096
1097		def setParam_(self, param, value):
#	Line 1389 | Line 1100 | class Cmssw(JobType):
1100		def getParams(self):
1101		return self._params
1102
1103	<	def setTaskid_(self):
1393	<	self._taskId = self.cfg_params['taskId']
1394	<
1395	<	def getTaskid(self):
1396	<	return self._taskId
1397	<
1398	<	def uniquelist(self, old):
1399	<	"""
1400	<	remove duplicates from a list
1401	<	"""
1402	<	nd={}
1403	<	for e in old:
1404	<	nd[e]=0
1405	<	return nd.keys()
1406	<
1407	<
1408	<	def checkOut(self, limit):
1103	>	def outList(self,list=False):
1104		"""
1105		check the dimension of the output files
1106		"""
1107	<	txt = 'echo ">>> Starting output sandbox limit check :"\n'
1108	<	allOutFiles = ""
1107	>	txt = ''
1108	>	txt += 'echo ">>> list of expected files on output sandbox"\n'
1109		listOutFiles = []
1110	<	txt += 'stdoutFile=`ls | grep *stdout` \n'
1111	<	txt += 'stderrFile=`ls | grep *stderr` \n'
1112	<	for fileOut in (self.output_file+self.output_file_sandbox):
1113	<	if fileOut.find('crab_fjr') == -1:
1114	<	allOutFiles = allOutFiles + " " + self.numberFile_(fileOut, '$NJob') + " $stdoutFile $stderrFile"
1115	<	listOutFiles.append(self.numberFile_(fileOut, '$NJob'))
1116	<	txt += 'echo "OUTPUT files: '+str(allOutFiles)+'";\n'
1117	<	txt += 'ls -gGhrta;\n'
1118	<	txt += 'sum=0;\n'
1119	<	txt += 'for file in '+str(allOutFiles)+' ; do\n'
1120	<	txt += '    if [ -e $file ]; then\n'
1121	<	txt += '        tt=`ls -gGrta $file | awk \'{ print $3 }\'`\n'
1122	<	txt += '        sum=`expr $sum + $tt`\n'
1123	<	txt += '    else\n'
1124	<	txt += '        echo "WARNING: output file $file not found!"\n'
1125	<	txt += '    fi\n'
1126	<	txt += 'done\n'
1127	<	txt += 'echo "Total Output dimension: $sum";\n'
1128	<	txt += 'limit='+str(limit)+';\n'
1434	<	txt += 'echo "OUTPUT FILES LIMIT SET TO: $limit";\n'
1435	<	txt += 'if [ $limit -lt $sum ]; then\n'
1436	<	txt += '    echo "WARNING: output files have to big size - something will be lost;"\n'
1437	<	txt += '    echo "         checking the output file sizes..."\n'
1438	<	txt += '    tot=0;\n'
1439	<	txt += '    for filefile in '+str(allOutFiles)+' ; do\n'
1440	<	txt += '        dimFile=`ls -gGrta $filefile | awk \'{ print $3 }\';`\n'
1441	<	txt += '        tot=`expr $tot + $tt`;\n'
1442	<	txt += '        if [ $limit -lt $dimFile ]; then\n'
1443	<	txt += '            echo "deleting file: $filefile";\n'
1444	<	txt += '            rm -f $filefile\n'
1445	<	txt += '        elif [ $limit -lt $tot ]; then\n'
1446	<	txt += '            echo "deleting file: $filefile";\n'
1447	<	txt += '            rm -f $filefile\n'
1448	<	txt += '        else\n'
1449	<	txt += '            echo "saving file: $filefile"\n'
1450	<	txt += '        fi\n'
1451	<	txt += '    done\n'
1110	>	stdout = 'CMSSW_$NJob.stdout'
1111	>	stderr = 'CMSSW_$NJob.stderr'
1112	>	if len(self.output_file) <= 0:
1113	>	msg ="WARNING: no output files name have been defined!!\n"
1114	>	msg+="\tno output files will be reported back/staged\n"
1115	>	common.logger.info(msg)
1116	>	if (self.return_data == 1):
1117	>	for file in (self.output_file+self.output_file_sandbox):
1118	>	listOutFiles.append(numberFile(file, '$NJob'))
1119	>	listOutFiles.append(stdout)
1120	>	listOutFiles.append(stderr)
1121	>	else:
1122	>	for file in (self.output_file_sandbox):
1123	>	listOutFiles.append(numberFile(file, '$NJob'))
1124	>	listOutFiles.append(stdout)
1125	>	listOutFiles.append(stderr)
1126	>	txt += 'echo "output files: '+string.join(listOutFiles,' ')+'"\n'
1127	>	txt += 'filesToCheck="'+string.join(listOutFiles,' ')+'"\n'
1128	>	txt += 'export filesToCheck\n'
1129
1130	<	txt += '    ls -agGhrt;\n'
1454	<	txt += '    echo "WARNING: output files are too big in dimension: can not put in the output_sandbox.";\n'
1455	<	txt += '    echo "JOB_EXIT_STATUS = 70000";\n'
1456	<	txt += '    exit_status=70000;\n'
1457	<	txt += 'else'
1458	<	txt += '    echo "Total Output dimension $sum is fine.";\n'
1459	<	txt += 'fi\n'
1460	<	txt += 'echo "Ending output sandbox limit check"\n'
1130	>	if list : return self.output_file
1131		return txt

Diff Legend

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