[ViewVC] Diff of: cvsroot/COMP/CRAB/python/cms

Comparing COMP/CRAB/python/cms_cmssw.py (file contents):
Revision 1.164 by fanzago, Tue Mar 4 18:15:38 2008 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	<	self.MaxTarBallSize = float(self.cfg_params.get('EDG.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
28	–	log = common.logger
29	–
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	+	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	<	#
62	<	# Try to block creation in case of arch/version mismatch
63	<	#
64	<
65	<	# a = string.split(self.version, "_")
66	<	#
67	<	# if int(a[1]) == 1 and (int(a[2]) < 5 and self.executable_arch.find('slc4') == 0):
68	<	# 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)
69	<	# common.logger.message(msg)
55	<	# if int(a[1]) == 1 and (int(a[2]) >= 5 and self.executable_arch.find('slc3') == 0):
56	<	# msg = "Error: CMS does not support %s with %s architecture"%(self.version, self.executable_arch)
57	<	# raise CrabException(msg)
58	<	#
59	<
60	<	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	<	if not cfg_params.has_key('CMSSW.datasetpath'):
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	<	tmp = cfg_params['CMSSW.datasetpath']
68	<	log.debug(6, "CMSSW::CMSSW(): datasetPath = "+tmp)
69	<	if string.lower(tmp)=='none':
83	>	elif string.lower(tmp)=='none':
84		self.datasetPath = None
85		self.selectNoInput = 1
86		else:
87		self.datasetPath = tmp
88		self.selectNoInput = 0
89
76	–	# ML monitoring
77	–	# split dataset path style: /PreProdR3Minbias/SIM/GEN-SIM
78	–	if not self.datasetPath:
79	–	self.setParam_('dataset', 'None')
80	–	self.setParam_('owner', 'None')
81	–	else:
82	–	## SL what is supposed to fail here?
83	–	try:
84	–	datasetpath_split = self.datasetPath.split("/")
85	–	# standard style
86	–	self.setParam_('datasetFull', self.datasetPath)
87	–	self.setParam_('dataset', datasetpath_split[1])
88	–	self.setParam_('owner', datasetpath_split[2])
89	–	except:
90	–	self.setParam_('dataset', self.datasetPath)
91	–	self.setParam_('owner', self.datasetPath)
92	–
93	–	self.setParam_('taskId', common._db.queryTask('name')) ## new BL--DS
94	–
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	+	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	<	self.setParam_('exe', self.executable)
100	<	log.debug(6, "CMSSW::CMSSW(): executable = "+self.executable)
100	>	common.logger.log(10-1, "CMSSW::CMSSW(): executable = "+self.executable)
101
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	<	log.debug(6, "Cmssw::Cmssw(): PSet file = "+self.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")
#	Line 117 \| 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	+	outfileflag = False
121		self.output_file = []
122		tmp = cfg_params.get('CMSSW.output_file',None)
123		if tmp :
124	<	tmpOutFiles = string.split(tmp,',')
125	<	log.debug(7, 'cmssw::cmssw(): output files '+str(tmpOutFiles))
126	<	for tmp in tmpOutFiles:
127	<	tmp=string.strip(tmp)
127	<	self.output_file.append(tmp)
128	<	pass
129	<	else:
130	<	log.message("No output file defined: only stdout/err and the CRAB Framework Job Report will be available\n")
131	<	pass
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		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))
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
141	–	#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		if cfg_params.has_key('USER.additional_input_files'):
#	Line 161 \| Line 162 \| class Cmssw(JobType):
162		if not os.path.exists(file):
163		raise CrabException("Additional input file not found: "+file)
164		pass
164	–	# fname = string.split(file, '/')[-1]
165	–	# storedFile = common.work_space.pathForTgz()+'share/'+fname
166	–	# 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))
168	>	common.logger.debug("Additional input files: "+str(self.additional_inbox_files))
169		pass
170
173	–	## Events per job
174	–	if cfg_params.has_key('CMSSW.events_per_job'):
175	–	self.eventsPerJob =int( cfg_params['CMSSW.events_per_job'])
176	–	self.selectEventsPerJob = 1
177	–	else:
178	–	self.eventsPerJob = -1
179	–	self.selectEventsPerJob = 0
180	–
181	–	## number of jobs
182	–	if cfg_params.has_key('CMSSW.number_of_jobs'):
183	–	self.theNumberOfJobs =int( cfg_params['CMSSW.number_of_jobs'])
184	–	self.selectNumberOfJobs = 1
185	–	else:
186	–	self.theNumberOfJobs = 0
187	–	self.selectNumberOfJobs = 0
188	–
189	–	if cfg_params.has_key('CMSSW.total_number_of_events'):
190	–	self.total_number_of_events = int(cfg_params['CMSSW.total_number_of_events'])
191	–	self.selectTotalNumberEvents = 1
192	–	else:
193	–	self.total_number_of_events = 0
194	–	self.selectTotalNumberEvents = 0
195	–
196	–	if self.pset != None: #CarlosDaniele
197	–	if ( (self.selectTotalNumberEvents + self.selectEventsPerJob + self.selectNumberOfJobs) != 2 ):
198	–	msg = 'Must define exactly two of total_number_of_events, events_per_job, or number_of_jobs.'
199	–	raise CrabException(msg)
200	–	else:
201	–	if (self.selectNumberOfJobs == 0):
202	–	msg = 'Must specify number_of_jobs.'
203	–	raise CrabException(msg)
171
172		## New method of dealing with seeds
173		self.incrementSeeds = []
#	Line 216 \| Line 183 \| class Cmssw(JobType):
183		tmp.strip()
184		self.incrementSeeds.append(tmp)
185
219	–	## Old method of dealing with seeds
220	–	## FUTURE: This is for old CMSSW and old CRAB. Can throw exceptions after a couple of CRAB releases and then
221	–	## remove
222	–	self.sourceSeed = cfg_params.get('CMSSW.pythia_seed',None)
223	–	if self.sourceSeed:
224	–	print "pythia_seed is a deprecated parameter. Use preserve_seeds or increment_seeds in the future.\n","Added to increment_seeds."
225	–	self.incrementSeeds.append('sourceSeed')
226	–
227	–	self.sourceSeedVtx = cfg_params.get('CMSSW.vtx_seed',None)
228	–	if self.sourceSeedVtx:
229	–	print "vtx_seed is a deprecated parameter. Use preserve_seeds or increment_seeds in the future.\n","Added to increment_seeds."
230	–	self.incrementSeeds.append('VtxSmeared')
231	–
232	–	self.sourceSeedG4 = cfg_params.get('CMSSW.g4_seed',None)
233	–	if self.sourceSeedG4:
234	–	print "g4_seed is a deprecated parameter. Use preserve_seeds or increment_seeds in the future.\n","Added to increment_seeds."
235	–	self.incrementSeeds.append('g4SimHits')
236	–
237	–	self.sourceSeedMix = cfg_params.get('CMSSW.mix_seed',None)
238	–	if self.sourceSeedMix:
239	–	print "mix_seed is a deprecated parameter. Use preserve_seeds or increment_seeds in the future.\n","Added to increment_seeds."
240	–	self.incrementSeeds.append('mix')
241	–
186		self.firstRun = cfg_params.get('CMSSW.first_run',None)
187
244	–	if self.pset != None: #CarlosDaniele
245	–	import PsetManipulator as pp
246	–	PsetEdit = pp.PsetManipulator(self.pset) #Daniele Pset
247	–
188		# Copy/return
249	–
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 261 \| Line 203 \| class Cmssw(JobType):
203		if self.datasetPath:
204		blockSites = self.DataDiscoveryAndLocation(cfg_params)
205		#DBSDLS-end
206	<
265	<	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
276	–	# modify Pset
277	–	if self.pset != None: #CarlosDaniele
278	–	try:
279	–	# Add FrameworkJobReport to parameter-set, set max events.
280	–	# Reset later for data jobs by writeCFG which does all modifications
281	–	PsetEdit.addCrabFJR(self.fjrFileName)
282	–	PsetEdit.maxEvent(self.eventsPerJob)
283	–	PsetEdit.psetWriter(self.configFilename())
284	–	except:
285	–	msg='Error while manipuliating ParameterSet: exiting...'
286	–	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	<	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 310 \| 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
341	–	def setArgsList(self, argsList):
342	–	self.argsList = argsList
360
361	<	def jobSplittingByBlocks(self, blockSites):
345	<	"""
346	<	Perform job splitting. Jobs run over an integer number of files
347	<	and no more than one block.
348	<	ARGUMENT: blockSites: dictionary with blocks as keys and list of host sites as values
349	<	REQUIRES: self.selectTotalNumberEvents, self.selectEventsPerJob, self.selectNumberofJobs,
350	<	self.total_number_of_events, self.eventsPerJob, self.theNumberOfJobs,
351	<	self.maxEvents, self.filesbyblock
352	<	SETS: self.jobDestination - Site destination(s) for each job (a list of lists)
353	<	self.total_number_of_jobs - Total # of jobs
354	<	self.list_of_args - File(s) job will run on (a list of lists)
355	<	"""
356	<
357	<	# ---- Handle the possible job splitting configurations ---- #
358	<	if (self.selectTotalNumberEvents):
359	<	totalEventsRequested = self.total_number_of_events
360	<	if (self.selectEventsPerJob):
361	<	eventsPerJobRequested = self.eventsPerJob
362	<	if (self.selectNumberOfJobs):
363	<	totalEventsRequested = self.theNumberOfJobs * self.eventsPerJob
364	<
365	<	# If user requested all the events in the dataset
366	<	if (totalEventsRequested == -1):
367	<	eventsRemaining=self.maxEvents
368	<	# If user requested more events than are in the dataset
369	<	elif (totalEventsRequested > self.maxEvents):
370	<	eventsRemaining = self.maxEvents
371	<	common.logger.message("Requested "+str(self.total_number_of_events)+ " events, but only "+str(self.maxEvents)+" events are available.")
372	<	# If user requested less events than are in the dataset
373	<	else:
374	<	eventsRemaining = totalEventsRequested
375	<
376	<	# If user requested more events per job than are in the dataset
377	<	if (self.selectEventsPerJob and eventsPerJobRequested > self.maxEvents):
378	<	eventsPerJobRequested = self.maxEvents
379	<
380	<	# For user info at end
381	<	totalEventCount = 0
382	<
383	<	if (self.selectTotalNumberEvents and self.selectNumberOfJobs):
384	<	eventsPerJobRequested = int(eventsRemaining/self.theNumberOfJobs)
385	<
386	<	if (self.selectNumberOfJobs):
387	<	common.logger.message("May not create the exact number_of_jobs requested.")
388	<
389	<	if ( self.ncjobs == 'all' ) :
390	<	totalNumberOfJobs = 999999999
391	<	else :
392	<	totalNumberOfJobs = self.ncjobs
393	<
394	<
395	<	blocks = blockSites.keys()
396	<	blockCount = 0
397	<	# Backup variable in case self.maxEvents counted events in a non-included block
398	<	numBlocksInDataset = len(blocks)
399	<
400	<	jobCount = 0
401	<	list_of_lists = []
402	<
403	<	# list tracking which jobs are in which jobs belong to which block
404	<	jobsOfBlock = {}
405	<
406	<	# ---- Iterate over the blocks in the dataset until ---- #
407	<	# ---- we've met the requested total # of events ---- #
408	<	while ( (eventsRemaining > 0) and (blockCount < numBlocksInDataset) and (jobCount < totalNumberOfJobs)):
409	<	block = blocks[blockCount]
410	<	blockCount += 1
411	<	if block not in jobsOfBlock.keys() :
412	<	jobsOfBlock[block] = []
413	<
414	<	if self.eventsbyblock.has_key(block) :
415	<	numEventsInBlock = self.eventsbyblock[block]
416	<	common.logger.debug(5,'Events in Block File '+str(numEventsInBlock))
417	<
418	<	files = self.filesbyblock[block]
419	<	numFilesInBlock = len(files)
420	<	if (numFilesInBlock <= 0):
421	<	continue
422	<	fileCount = 0
361	>	def split(self, jobParams,firstJobID):
362
363	<	# ---- New block => New job ---- #
364	<	parString = ""
365	<	# counter for number of events in files currently worked on
427	<	filesEventCount = 0
428	<	# flag if next while loop should touch new file
429	<	newFile = 1
430	<	# job event counter
431	<	jobSkipEventCount = 0
432	<
433	<	# ---- Iterate over the files in the block until we've met the requested ---- #
434	<	# ---- total # of events or we've gone over all the files in this block ---- #
435	<	while ( (eventsRemaining > 0) and (fileCount < numFilesInBlock) and (jobCount < totalNumberOfJobs) ):
436	<	file = files[fileCount]
437	<	if newFile :
438	<	try:
439	<	numEventsInFile = self.eventsbyfile[file]
440	<	common.logger.debug(6, "File "+str(file)+" has "+str(numEventsInFile)+" events")
441	<	# increase filesEventCount
442	<	filesEventCount += numEventsInFile
443	<	# Add file to current job
444	<	parString += '\\\"' + file + '\\\"\,'
445	<	newFile = 0
446	<	except KeyError:
447	<	common.logger.message("File "+str(file)+" has unknown number of events: skipping")
448	<
449	<
450	<	# if less events in file remain than eventsPerJobRequested
451	<	if ( filesEventCount - jobSkipEventCount < eventsPerJobRequested ) :
452	<	# if last file in block
453	<	if ( fileCount == numFilesInBlock-1 ) :
454	<	# end job using last file, use remaining events in block
455	<	# close job and touch new file
456	<	fullString = parString[:-2]
457	<	list_of_lists.append([fullString,str(-1),str(jobSkipEventCount)])
458	<	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(filesEventCount - jobSkipEventCount)+" events (last file in block).")
459	<	self.jobDestination.append(blockSites[block])
460	<	common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
461	<	# fill jobs of block dictionary
462	<	jobsOfBlock[block].append(jobCount+1)
463	<	# reset counter
464	<	jobCount = jobCount + 1
465	<	totalEventCount = totalEventCount + filesEventCount - jobSkipEventCount
466	<	eventsRemaining = eventsRemaining - filesEventCount + jobSkipEventCount
467	<	jobSkipEventCount = 0
468	<	# reset file
469	<	parString = ""
470	<	filesEventCount = 0
471	<	newFile = 1
472	<	fileCount += 1
473	<	else :
474	<	# go to next file
475	<	newFile = 1
476	<	fileCount += 1
477	<	# if events in file equal to eventsPerJobRequested
478	<	elif ( filesEventCount - jobSkipEventCount == eventsPerJobRequested ) :
479	<	# close job and touch new file
480	<	fullString = parString[:-2]
481	<	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
482	<	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.")
483	<	self.jobDestination.append(blockSites[block])
484	<	common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
485	<	jobsOfBlock[block].append(jobCount+1)
486	<	# reset counter
487	<	jobCount = jobCount + 1
488	<	totalEventCount = totalEventCount + eventsPerJobRequested
489	<	eventsRemaining = eventsRemaining - eventsPerJobRequested
490	<	jobSkipEventCount = 0
491	<	# reset file
492	<	parString = ""
493	<	filesEventCount = 0
494	<	newFile = 1
495	<	fileCount += 1
363	>	jobParams = self.dict['args']
364	>	njobs = self.dict['njobs']
365	>	self.jobDestination = self.dict['jobDestination']
366
367	<	# if more events in file remain than eventsPerJobRequested
368	<	else :
499	<	# close job but don't touch new file
500	<	fullString = parString[:-2]
501	<	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
502	<	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.")
503	<	self.jobDestination.append(blockSites[block])
504	<	common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
505	<	jobsOfBlock[block].append(jobCount+1)
506	<	# increase counter
507	<	jobCount = jobCount + 1
508	<	totalEventCount = totalEventCount + eventsPerJobRequested
509	<	eventsRemaining = eventsRemaining - eventsPerJobRequested
510	<	# calculate skip events for last file
511	<	# use filesEventCount (contains several files), jobSkipEventCount and eventsPerJobRequest
512	<	jobSkipEventCount = eventsPerJobRequested - (filesEventCount - jobSkipEventCount - self.eventsbyfile[file])
513	<	# remove all but the last file
514	<	filesEventCount = self.eventsbyfile[file]
515	<	parString = '\\\"' + file + '\\\"\,'
516	<	pass # END if
517	<	pass # END while (iterate over files in the block)
518	<	pass # END while (iterate over blocks in the dataset)
519	<	self.ncjobs = self.total_number_of_jobs = jobCount
520	<	if (eventsRemaining > 0 and jobCount < totalNumberOfJobs ):
521	<	common.logger.message("Could not run on all requested events because some blocks not hosted at allowed sites.")
522	<	common.logger.message(str(jobCount)+" job(s) can run on "+str(totalEventCount)+" events.\n")
523	<
524	<	# screen output
525	<	screenOutput = "List of jobs and available destination sites:\n\n"
526	<
527	<	# keep trace of block with no sites to print a warning at the end
528	<	noSiteBlock = []
529	<	bloskNoSite = []
530	<
531	<	blockCounter = 0
532	<	for block in blocks:
533	<	if block in jobsOfBlock.keys() :
534	<	blockCounter += 1
535	<	screenOutput += "Block %5i: jobs %20s: sites: %s\n" % (blockCounter,spanRanges(jobsOfBlock[block]),','.join(self.blackWhiteListParser.checkWhiteList(self.blackWhiteListParser.checkBlackList(blockSites[block],block),block)))
536	<	if len(self.blackWhiteListParser.checkWhiteList(self.blackWhiteListParser.checkBlackList(blockSites[block],block),block)) == 0:
537	<	noSiteBlock.append( spanRanges(jobsOfBlock[block]) )
538	<	bloskNoSite.append( blockCounter )
539	<
540	<	common.logger.message(screenOutput)
541	<	if len(noSiteBlock) > 0 and len(bloskNoSite) > 0:
542	<	msg = 'WARNING: No sites are hosting any part of data for block:\n '
543	<	virgola = ""
544	<	if len(bloskNoSite) > 1:
545	<	virgola = ","
546	<	for block in bloskNoSite:
547	<	msg += ' ' + str(block) + virgola
548	<	msg += '\n Related jobs:\n '
549	<	virgola = ""
550	<	if len(noSiteBlock) > 1:
551	<	virgola = ","
552	<	for range_jobs in noSiteBlock:
553	<	msg += str(range_jobs) + virgola
554	<	msg += '\n will not be submitted and this block of data can not be analyzed!\n'
555	<	if self.cfg_params.has_key('EDG.se_white_list'):
556	<	msg += 'WARNING: SE White List: '+self.cfg_params['EDG.se_white_list']+'\n'
557	<	msg += '(Hint: By whitelisting you force the job to run at this particular site(s).\n'
558	<	msg += 'Please check if the dataset is available at this site!)\n'
559	<	if self.cfg_params.has_key('EDG.ce_white_list'):
560	<	msg += 'WARNING: CE White List: '+self.cfg_params['EDG.ce_white_list']+'\n'
561	<	msg += '(Hint: By whitelisting you force the job to run at this particular site(s).\n'
562	<	msg += 'Please check if the dataset is available at this site!)\n'
367	>	if njobs==0:
368	>	raise CrabException("Ask to split "+str(njobs)+" jobs: aborting")
369
370	<	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):
570	<	"""
571	<	Perform job splitting based on number of event per job
572	<	"""
573	<	common.logger.debug(5,'Splitting per events')
411	>	def addXMLfile(self):
412
413	<	if (self.selectEventsPerJob):
414	<	common.logger.message('Required '+str(self.eventsPerJob)+' events per job ')
415	<	if (self.selectNumberOfJobs):
416	<	common.logger.message('Required '+str(self.theNumberOfJobs)+' jobs in total ')
417	<	if (self.selectTotalNumberEvents):
418	<	common.logger.message('Required '+str(self.total_number_of_events)+' events in total ')
419	<
420	<	if (self.total_number_of_events < 0):
421	<	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):
587	<	if (self.selectTotalNumberEvents):
588	<	self.total_number_of_jobs = int(self.total_number_of_events/self.eventsPerJob)
589	<	elif(self.selectNumberOfJobs) :
590	<	self.total_number_of_jobs =self.theNumberOfJobs
591	<	self.total_number_of_events =int(self.theNumberOfJobs*self.eventsPerJob)
592	<
593	<	elif (self.selectNumberOfJobs) :
594	<	self.total_number_of_jobs = self.theNumberOfJobs
595	<	self.eventsPerJob = int(self.total_number_of_events/self.total_number_of_jobs)
596	<
597	<	common.logger.debug(5,'N jobs '+str(self.total_number_of_jobs))
598	<
599	<	# is there any remainder?
600	<	check = int(self.total_number_of_events) - (int(self.total_number_of_jobs)*self.eventsPerJob)
601	<
602	<	common.logger.debug(5,'Check '+str(check))
603	<
604	<	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')
605	<	if check > 0:
606	<	common.logger.message('Warning: asked '+str(self.total_number_of_events)+' but can do only '+str(int(self.total_number_of_jobs)*self.eventsPerJob))
607	<
608	<	# argument is seed number.$i
609	<	self.list_of_args = []
610	<	for i in range(self.total_number_of_jobs):
611	<	## Since there is no input, any site is good
612	<	self.jobDestination.append([""]) #must be empty to write correctly the xml
613	<	args=[]
614	<	if (self.firstRun):
615	<	## pythia first run
616	<	args.append(str(self.firstRun)+str(i))
617	<	self.list_of_args.append(args)
618	<
619	<	return
620	<
621	<
622	<	def jobSplittingForScript(self):#CarlosDaniele
427	>	def CreateXML(self):
428		"""
624	–	Perform job splitting based on number of job
429		"""
430	<	common.logger.debug(5,'Splitting per job')
431	<	common.logger.message('Required '+str(self.theNumberOfJobs)+' jobs in total ')
628	<
629	<	self.total_number_of_jobs = self.theNumberOfJobs
630	<
631	<	common.logger.debug(5,'N jobs '+str(self.total_number_of_jobs))
632	<
633	<	common.logger.message(str(self.total_number_of_jobs)+' jobs can be created')
634	<
635	<	# argument is seed number.$i
636	<	self.list_of_args = []
637	<	for i in range(self.total_number_of_jobs):
638	<	## Since there is no input, any site is good
639	<	# self.jobDestination.append(["Any"])
640	<	self.jobDestination.append([""])
641	<	## no random seed
642	<	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	<	#### Fabio
648	<	njobs = self.total_number_of_jobs
649	<	arglist = self.list_of_args
650	<	# create the empty structure
651	<	for i in range(njobs):
652	<	jobParams.append("")
653	<
654	<	for job in range(njobs):
655	<	jobParams[job] = arglist[job]
656	<	job_ToSave ={}
657	<	job_ToSave['arguments']= str(jobParams[job])## new BL--DS
658	<	job_ToSave['dlsDestination']= self.jobDestination[job]## new BL--DS
659	<	common._db.updateJob_(job,job_ToSave)## new BL--DS
660	<	common.logger.debug(5,"Job "+str(job)+" Destination: "+str(self.jobDestination[job]))
434	>	def addEntry(self, listDictions):
435	>	"""
436	>	_addEntry_
437
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
664	–	def getJobTypeArguments(self, nj, sched):
665	–	result = ''
666	–	jobs=[]
667	–	jobs.append(nj)
668	–	for i in common._db.queryJob('arguments',jobs):## BL--DS
669	–	result=result+str(i)+" "
670	–	return result
671	–
450		def numberOfJobs(self):
451	<	# Fabio
674	<	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	<	#
683	<	# Marco. Let's start to use relative path for Boss XML files
684	<	#
685	<	self.tgzNameWithPath = common.work_space.pathForTgz()+'share/'+self.tgz_name
686	<	if os.path.exists(self.tgzNameWithPath):
687	<	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_()
698	–	#print "swArea = ", swArea
699	–	# swVersion = self.scram.getSWVersion()
700	–	# print "swVersion = ", swVersion
470		swReleaseTop = self.scram.getReleaseTop_()
702	–	#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 717 \| 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 731 \| 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 742 \| 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")
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	>	### 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	+	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		## Add ProdCommon dir to tar
547	<	prodcommonDir = 'ProdCommon'
548	<	prodcommonPath = os.environ['CRABDIR'] + '/' + 'ProdCommon'
549	<	if os.path.isdir(prodcommonPath):
550	<	tar.add(prodcommonPath,prodcommonDir)
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 ML_file_list:
559	>	tar.add(path+file,file)
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
760	–	common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
573		tar.close()
574	<	except :
575	<	raise CrabException('Could not create 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	>	def zipTarFile(self):
584	>
585	>	cmd = "gzip -c %s > %s "%(self.tarNameWithPath,self.tgzNameWithPath)
586	>	res=runCommand(cmd)
587
765	–	## check for tarball size
588		tarballinfo = os.stat(self.tgzNameWithPath)
589		if ( tarballinfo.st_size > self.MaxTarBallSize10241024 ) :
590	<	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.')
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
771	–	self.MLtgzfile = common.work_space.pathForTgz()+'share/MLfiles.tgz'
772	–	try:
773	–	tar = tarfile.open(self.MLtgzfile, "w:gz")
774	–	path=os.environ['CRABDIR'] + '/python/'
775	–	for file in ['report.py', 'DashboardAPI.py', 'Logger.py', 'ProcInfo.py', 'apmon.py', 'parseCrabFjr.py','writeCfg.py']:
776	–	tar.add(path+file,file)
777	–	common.logger.debug(5,"Files added to "+self.MLtgzfile+" : "+str(tar.getnames()))
778	–	tar.close()
779	–	except :
780	–	raise CrabException('Could not create ML files tar-ball')
781	–
782	–	return
783	–
784	–	def additionalInputFileTgz(self):
785	–	"""
786	–	Put all additional files into a tar ball and return its name
787	–	"""
788	–	import tarfile
789	–	tarName= common.work_space.pathForTgz()+'share/'+self.additional_tgz_name
790	–	tar = tarfile.open(tarName, "w:gz")
791	–	for file in self.additional_inbox_files:
792	–	tar.add(file,string.split(file,'/')[-1])
793	–	common.logger.debug(5,"Files added to "+self.additional_tgz_name+" : "+str(tar.getnames()))
794	–	tar.close()
795	–	return tarName
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 = '\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'
810	–	#txt += ' echo "SET_CMS_ENV 10016 ==> OSG $WORKING_DIR could not be created on WN `hostname`"\n'
811	–	#txt += ' echo "JOB_EXIT_STATUS = 10016"\n'
812	–	#txt += ' echo "JobExitCode=10016" \| tee -a $RUNTIME_AREA/$repo\n'
813	–	#txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
814	–	#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'
#	Line 822 \| 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 [ $middleware == ARC ]; then\n'
632	>	txt += self.wsSetupCMSLCGEnvironment_()
633	>
634		txt += 'fi\n'
635
636		# Prepare JobType-specific part
#	Line 834 \| Line 641 \| class Cmssw(JobType):
641		txt += scram+' project CMSSW '+self.version+'\n'
642		txt += 'status=$?\n'
643		txt += 'if [ $status != 0 ] ; then\n'
837	–	#txt += ' echo "SET_EXE_ENV 10034 ==>ERROR CMSSW '+self.version+' not found on `hostname`" \n'
838	–	#txt += ' echo "JOB_EXIT_STATUS = 10034"\n'
839	–	#txt += ' echo "JobExitCode=10034" \| tee -a $RUNTIME_AREA/$repo\n'
840	–	#txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
644		txt += ' echo "ERROR ==> CMSSW '+self.version+' not found on `hostname`" \n'
645		txt += ' job_exit_code=10034\n'
843	–	txt += ' if [ $middleware == OSG ]; then \n'
844	–	txt += ' cd $RUNTIME_AREA\n'
845	–	txt += ' echo ">>> current directory (RUNTIME_AREA): $RUNTIME_AREA"\n'
846	–	txt += ' echo ">>> Remove working directory: $WORKING_DIR"\n'
847	–	txt += ' /bin/rm -rf $WORKING_DIR\n'
848	–	txt += ' if [ -d $WORKING_DIR ] ;then\n'
849	–	#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'
850	–	#txt += ' echo "JOB_EXIT_STATUS = 10018"\n'
851	–	#txt += ' echo "JobExitCode=10018" \| tee -a $RUNTIME_AREA/$repo\n'
852	–	#txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
853	–	txt += ' echo "ERROR ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after CMSSW CMSSW_0_6_1 not found on `hostname`"\n'
854	–	txt += ' job_exit_code=10017\n'
855	–	txt += ' fi\n'
856	–	txt += ' fi \n'
857	–	#txt += ' exit 1 \n'
646		txt += ' func_exit\n'
647		txt += 'fi \n'
648		txt += 'cd '+self.version+'\n'
649	<	########## FEDE FOR DBS2 ######################
862	<	txt += 'SOFTWARE_DIR=`pwd`\n'
649	>	txt += 'SOFTWARE_DIR=`pwd`; export SOFTWARE_DIR\n'
650		txt += 'echo ">>> current directory (SOFTWARE_DIR): $SOFTWARE_DIR" \n'
864	–	###############################################
865	–	### needed grep for bug in scramv1 ###
651		txt += 'eval `'+scram+' runtime -sh \| grep -v SCRAMRT_LSB_JOBNAME`\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 [ $nargs -lt "+str(self.argsList)+" ]\n"
663		txt += "then\n"
873	–	#txt += " echo 'SET_EXE_ENV 1 ==> ERROR Too few arguments' +$nargs+ \n"
874	–	#txt += ' echo "JOB_EXIT_STATUS = 50113"\n'
875	–	#txt += ' echo "JobExitCode=50113" \| tee -a $RUNTIME_AREA/$repo\n'
876	–	#txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
664		txt += " echo 'ERROR ==> Too few arguments' +$nargs+ \n"
665		txt += ' job_exit_code=50113\n'
879	–	txt += ' if [ $middleware == OSG ]; then \n'
880	–	txt += ' cd $RUNTIME_AREA\n'
881	–	txt += ' echo ">>> current directory (RUNTIME_AREA): $RUNTIME_AREA"\n'
882	–	txt += ' echo ">>> Remove working directory: $WORKING_DIR"\n'
883	–	txt += ' /bin/rm -rf $WORKING_DIR\n'
884	–	txt += ' if [ -d $WORKING_DIR ] ;then\n'
885	–	#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'
886	–	#txt += ' echo "JOB_EXIT_STATUS = 50114"\n'
887	–	#txt += ' echo "JobExitCode=50114" \| tee -a $RUNTIME_AREA/$repo\n'
888	–	#txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
889	–	txt += ' echo "ERROR ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after Too few arguments for CRAB job wrapper"\n'
890	–	txt += ' job_exit_code=10017\n'
891	–	txt += ' fi\n'
892	–	txt += ' fi\n'
893	–	#txt += " exit 1\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]
900	–	### 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	<
907	<	txt += 'PrimaryDataset='+datasetpath_split[1]+'\n'
908	<	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:
688	>	if self.pset != None:
689		pset = os.path.basename(job.configFilename())
690		txt += '\n'
691		txt += 'cp $RUNTIME_AREA/'+pset+' .\n'
920	–	if (self.datasetPath): # standard job
921	–	txt += 'InputFiles=${args[1]}; export InputFiles\n'
922	–	txt += 'MaxEvents=${args[2]}; export MaxEvents\n'
923	–	txt += 'SkipEvents=${args[3]}; export SkipEvents\n'
924	–	txt += 'echo "Inputfiles:<$InputFiles>"\n'
925	–	txt += 'echo "MaxEvents:<$MaxEvents>"\n'
926	–	txt += 'echo "SkipEvents:<$SkipEvents>"\n'
927	–	else: # pythia like job
928	–	txt += 'PreserveSeeds=' + ','.join(self.preserveSeeds) + '; export PreserveSeeds\n'
929	–	txt += 'IncrementSeeds=' + ','.join(self.incrementSeeds) + '; export IncrementSeeds\n'
930	–	txt += 'echo "PreserveSeeds: <$PreserveSeeds>"\n'
931	–	txt += 'echo "IncrementSeeds:<$IncrementSeeds>"\n'
932	–	if (self.firstRun):
933	–	txt += 'FirstRun=${args[1]}; export FirstRun\n'
934	–	txt += 'echo "FirstRun: <$FirstRun>"\n'
935	–
936	–	txt += 'mv -f '+pset+' pset.cfg\n'
937	–
938	–	if len(self.additional_inbox_files) > 0:
939	–	txt += 'if [ -e $RUNTIME_AREA/'+self.additional_tgz_name+' ] ; then\n'
940	–	txt += ' tar xzvf $RUNTIME_AREA/'+self.additional_tgz_name+'\n'
941	–	txt += 'fi\n'
942	–	pass
692
693	<	if self.pset != None:
694	<	txt += '\n'
695	<	txt += 'echo "*** cat pset.cfg *******"\n'
696	<	txt += 'cat pset.cfg\n'
697	<	txt += 'echo "**** end pset.cfg ******"\n'
698	<	txt += '\n'
699	<	txt += 'PSETHASH=`EdmConfigHash < pset.cfg` \n'
951	<	txt += 'echo "PSETHASH = $PSETHASH" \n'
693	>	txt += 'PreserveSeeds=' + ','.join(self.preserveSeeds) + '; export PreserveSeeds\n'
694	>	txt += 'IncrementSeeds=' + ','.join(self.incrementSeeds) + '; export IncrementSeeds\n'
695	>	txt += 'echo "PreserveSeeds: <$PreserveSeeds>"\n'
696	>	txt += 'echo "IncrementSeeds:<$IncrementSeeds>"\n'
697	>
698	>	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 = '\n#Written by cms_cmssw::wsBuildExe\n'
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'
968	–	#txt += ' echo "SET_EXE 1 ==> ERROR Untarring .tgz file failed"\n'
969	–	#txt += ' echo "JOB_EXIT_STATUS = $untar_status" \n'
970	–	#txt += ' echo "JobExitCode=$untar_status" \| tee -a $RUNTIME_AREA/$repo\n'
720		txt += ' echo "ERROR ==> Untarring .tgz file failed"\n'
721		txt += ' job_exit_code=$untar_status\n'
973	–	txt += ' echo "JobExitCode=$untar_status" >> $RUNTIME_AREA/$repo\n'
974	–	txt += ' if [ $middleware == OSG ]; then \n'
975	–	txt += ' cd $RUNTIME_AREA\n'
976	–	txt += ' echo ">>> current directory (RUNTIME_AREA): $RUNTIME_AREA"\n'
977	–	txt += ' echo ">>> Remove working directory: $WORKING_DIR"\n'
978	–	txt += ' /bin/rm -rf $WORKING_DIR\n'
979	–	txt += ' if [ -d $WORKING_DIR ] ;then\n'
980	–	#txt += ' echo "SET_EXE 50999 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after Untarring .tgz file failed"\n'
981	–	#txt += ' echo "JOB_EXIT_STATUS = 50999"\n'
982	–	#txt += ' echo "JobExitCode=50999" \| tee -a $RUNTIME_AREA/$repo\n'
983	–	#txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
984	–	txt += ' echo "ERROR ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after Untarring .tgz file failed"\n'
985	–	txt += ' job_exit_code=10017\n'
986	–	txt += ' fi\n'
987	–	txt += ' fi \n'
988	–	txt += ' \n'
989	–	#txt += ' exit 1 \n'
722		txt += ' func_exit\n'
723		txt += 'else \n'
724		txt += ' echo "Successful untar" \n'
725		txt += 'fi \n'
726		txt += '\n'
727	<	#### Removed ProdAgent API dependencies
996	<	txt += 'echo ">>> Include 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
999	<	txt += ' export PYTHONPATH=$SOFTWARE_DIR/ProdCommon\n'
729	>	txt += ' export PYTHONPATH=$RUNTIME_AREA/\n'
730		txt += 'else\n'
731	<	txt += ' export PYTHONPATH=$SOFTWARE_DIR/ProdCommon:${PYTHONPATH}\n'
731	>	txt += ' export PYTHONPATH=$RUNTIME_AREA/:${PYTHONPATH}\n'
732		txt += 'echo "PYTHONPATH=$PYTHONPATH"\n'
1003	–	###################
733		txt += 'fi\n'
734		txt += '\n'
735
#	Line 1008 \| 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		# FUTURE: This function tests the CMSSW version. Can be simplified as we drop support for old versions
809	<	if self.scriptExe:#CarlosDaniele
810	<	return self.scriptExe + " $NJob"
809	>	if self.scriptExe:
810	>	return self.scriptExe + " $NJob $AdditionalArgs"
811		else:
1028	–	version_array = self.scram.getSWVersion().split('_')
1029	–	major = 0
1030	–	minor = 0
1031	–	try:
1032	–	major = int(version_array[1])
1033	–	minor = int(version_array[2])
1034	–	except:
1035	–	msg = "Cannot parse CMSSW version string: " + "_".join(version_array) + " for major and minor release number!"
1036	–	raise CrabException(msg)
1037	–
812		ex_args = ""
813	<
814	<	# Framework job report
815	<	if major >= 1 and minor >= 5 :
816	<	ex_args += " -j " + self.fjrFileName
1043	<
1044	<	# Type of cfg file
1045	<	if major >= 2 :
1046	<	ex_args += " -p pset.pycfg"
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		ex_args += " -p pset.cfg"
819		return ex_args
#	Line 1053 \| Line 823 \| class Cmssw(JobType):
823		Returns a list of filenames to be put in JDL input sandbox.
824		"""
825		inp_box = []
1056	–	# # dict added to delete duplicate from input sandbox file list
1057	–	# seen = {}
1058	–	## code
826		if os.path.isfile(self.tgzNameWithPath):
827		inp_box.append(self.tgzNameWithPath)
828	<	if os.path.isfile(self.MLtgzfile):
1062	<	inp_box.append(self.MLtgzfile)
1063	<	## config
1064	<	if not self.pset is None:
1065	<	inp_box.append(common.work_space.pathForTgz() + 'job/' + self.configFilename())
1066	<	## additional input files
1067	<	tgz = self.additionalInputFileTgz()
1068	<	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 1077 \| 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
1083	–	def prepareSteeringCards(self):
1084	–	"""
1085	–	Make initial modifications of the user's steering card file.
1086	–	"""
1087	–	return
843
844		def wsRenameOutput(self, nj):
845		"""
#	Line 1094 \| Line 849 \| class Cmssw(JobType):
849		txt = '\n#Written by cms_cmssw::wsRenameOutput\n'
850		txt += 'echo ">>> current directory (SOFTWARE_DIR): $SOFTWARE_DIR" \n'
851		txt += 'echo ">>> current directory content:"\n'
852	<	txt += 'ls \n'
852	>	if self.debug_wrapper==1:
853	>	txt += 'ls -Al\n'
854		txt += '\n'
855
1100	–	#txt += 'output_exit_status=0\n'
1101	–
1102	–	for fileWithSuffix in (self.output_file_sandbox):
1103	–	output_file_num = self.numberFile_(fileWithSuffix, '$NJob')
1104	–	txt += '\n'
1105	–	txt += '# check output file\n'
1106	–	txt += 'if [ -e ./'+fileWithSuffix+' ] ; then\n'
1107	–	txt += ' mv '+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n'
1108	–	txt += ' ln -s $RUNTIME_AREA/'+output_file_num+' $RUNTIME_AREA/'+fileWithSuffix+'\n'
1109	–	txt += 'else\n'
1110	–	#txt += ' exit_status=60302\n'
1111	–	#txt += ' echo "ERROR: Output file '+fileWithSuffix+' not found"\n'
1112	–	txt += ' echo "WARNING: Output file '+fileWithSuffix+' not found"\n'
1113	–	txt += ' job_exit_code=60302\n'
1114	–	if common.scheduler.name().upper() == 'CONDOR_G':
1115	–	txt += ' if [ $middleware == OSG ]; then \n'
1116	–	txt += ' echo "prepare dummy output file"\n'
1117	–	txt += ' echo "Processing of job output failed" > $RUNTIME_AREA/'+output_file_num+'\n'
1118	–	txt += ' fi \n'
1119	–	txt += 'fi\n'
1120	–
856		for fileWithSuffix in (self.output_file):
857	<	output_file_num = self.numberFile_(fileWithSuffix, '$NJob')
857	>	output_file_num = numberFile(fileWithSuffix, '$NJob')
858		txt += '\n'
859		txt += '# check output file\n'
860		txt += 'if [ -e ./'+fileWithSuffix+' ] ; then\n'
#	Line 1130 \| Line 865 \| class Cmssw(JobType):
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'
1133	–	#txt += ' exit_status=60302\n'
1134	–	#txt += ' echo "ERROR: Output file '+fileWithSuffix+' not found"\n'
1135	–	#txt += ' echo "JOB_EXIT_STATUS = $exit_status"\n'
1136	–	#txt += ' output_exit_status=$exit_status\n'
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':
#	Line 1144 \| 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	<	txt += 'ls \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
1159	–	def numberFile_(self, file, txt):
1160	–	"""
1161	–	append _'txt' before last extension of a file
1162	–	"""
1163	–	p = string.split(file,".")
1164	–	# take away last extension
1165	–	name = p[0]
1166	–	for x in p[1:-1]:
1167	–	name=name+"."+x
1168	–	# add "_txt"
1169	–	if len(p)>1:
1170	–	ext = p[len(p)-1]
1171	–	result = name + '_' + txt + "." + ext
1172	–	else:
1173	–	result = name + '_' + txt
1174	–
1175	–	return result
1176	–
891		def getRequirements(self, nj=[]):
892		"""
893		return job requirements to add to jdl files
#	Line 1183 \| 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
1187	<	#reSL4 = re.compile( r'slc4' )
1188	<	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":
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		"""
#	Line 1214 \| 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'
1217	–	#txt += ' echo "SET_CMS_ENV 10020 ==> ERROR $OSG_APP/cmssoft/cms/cmsset_default.sh file not found"\n'
1218	–	#txt += ' echo "JOB_EXIT_STATUS = 10020"\n'
1219	–	#txt += ' echo "JobExitCode=10020" \| tee -a $RUNTIME_AREA/$repo\n'
1220	–	#txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
933		txt += ' echo "ERROR ==> $OSG_APP/cmssoft/cms/cmsset_default.sh file not found"\n'
934		txt += ' job_exit_code=10020\n'
1223	–	txt += ' cd $RUNTIME_AREA\n'
1224	–	txt += ' echo ">>> current directory (RUNTIME_AREA): $RUNTIME_AREA"\n'
1225	–	txt += ' echo ">>> Remove working directory: $WORKING_DIR"\n'
1226	–	txt += ' /bin/rm -rf $WORKING_DIR\n'
1227	–	txt += ' if [ -d $WORKING_DIR ] ;then\n'
1228	–	#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'
1229	–	#txt += ' echo "JOB_EXIT_STATUS = 10017"\n'
1230	–	#txt += ' echo "JobExitCode=10017" \| tee -a $RUNTIME_AREA/$repo\n'
1231	–	#txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
1232	–	txt += ' echo "ERROR ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after $OSG_APP/cmssoft/cms/cmsset_default.sh file not found"\n'
1233	–	txt += ' job_exit_code=10017\n'
1234	–	txt += ' fi\n'
1235	–	txt += '\n'
1236	–	#txt += ' exit 1\n'
935		txt += ' func_exit\n'
936		txt += ' fi\n'
937		txt += '\n'
#	Line 1242 \| Line 940 \| class Cmssw(JobType):
940
941		return txt
942
1245	–	### OLI_DANIELE
943		def wsSetupCMSLCGEnvironment_(self):
944		"""
945		Returns part of a job script which is prepares
#	Line 1254 \| Line 951 \| class Cmssw(JobType):
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'
1257	–	#txt += ' echo "SET_CMS_ENV 10031 ==> ERROR CMS software dir not found on WN `hostname`"\n'
1258	–	#txt += ' echo "JOB_EXIT_STATUS = 10031" \n'
1259	–	#txt += ' echo "JobExitCode=10031" \| tee -a $RUNTIME_AREA/$repo\n'
1260	–	#txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
1261	–	#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'
1268	–	#txt += ' echo "SET_CMS_ENV 10020 ==> ERROR cmsset_default.sh file not found into dir $VO_CMS_SW_DIR"\n'
1269	–	#txt += ' echo "JOB_EXIT_STATUS = 10020"\n'
1270	–	#txt += ' echo "JobExitCode=10020" \| tee -a $RUNTIME_AREA/$repo\n'
1271	–	#txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
1272	–	#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'
#	Line 1278 \| Line 965 \| class Cmssw(JobType):
965		txt += ' source $VO_CMS_SW_DIR/cmsset_default.sh\n'
966		txt += ' result=$?\n'
967		txt += ' if [ $result -ne 0 ]; then\n'
1281	–	#txt += ' echo "SET_CMS_ENV 10032 ==> ERROR problem sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
1282	–	#txt += ' echo "JOB_EXIT_STATUS = 10032"\n'
1283	–	#txt += ' echo "JobExitCode=10032" \| tee -a $RUNTIME_AREA/$repo\n'
1284	–	#txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
1285	–	#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'
#	Line 1292 \| Line 974 \| class Cmssw(JobType):
974		txt += ' echo "==> setup cms environment ok"\n'
975		return txt
976
977	<	### FEDE FOR DBS OUTPUT PUBLICATION
1296	<	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 = '\n#Written by cms_cmssw::modifyReport\n'
983	<	try:
984	<	publish_data = int(self.cfg_params['USER.publish_data'])
985	<	except KeyError:
986	<	publish_data = 0
987	<	if (publish_data == 1):
988	<
989	<	txt += 'if [ $copy_exit_status -eq 0 ]; then\n'
990	<	txt += ' echo ">>> Modify Job Report:" \n'
991	<	txt += ' chmod a+x $SOFTWARE_DIR/ProdCommon/ProdCommon/FwkJobRep/ModifyJobReport.py\n'
992	<	#txt += ' if [ -z "$SE" ]; then\n'
993	<	#txt += ' SE="" \n'
994	<	#txt += ' fi \n'
995	<	#txt += ' if [ -z "$SE_PATH" ]; then\n'
996	<	#txt += ' SE_PATH="" \n'
997	<	#txt += ' fi \n'
998	<	txt += ' echo "SE = $SE"\n'
999	<	txt += ' echo "SE_PATH = $SE_PATH"\n'
1000	<
1001	<	processedDataset = self.cfg_params['USER.publish_data_name']
1002	<	txt += ' ProcessedDataset='+processedDataset+'\n'
1003	<	#txt += ' if [ "$SE_PATH" == "" ]; then\n'
1004	<	#txt += ' FOR_LFN=/copy_problems/ \n'
1005	<	#txt += ' else \n'
1006	<	#txt += ' tmp=`echo $SE_PATH \| awk -F \'store\' \'{print$2}\'` \n'
1007	<	#txt += ' FOR_LFN=/store$tmp \n'
1008	<	#txt += ' fi \n'
1009	<	txt += ' tmp=`echo $SE_PATH \| awk -F \'store\' \'{print$2}\'` \n'
1010	<	txt += ' FOR_LFN=/store$tmp \n'
1011	<	txt += ' echo "ProcessedDataset = $ProcessedDataset"\n'
1012	<	txt += ' echo "FOR_LFN = $FOR_LFN" \n'
1013	<	txt += ' echo "CMSSW_VERSION = $CMSSW_VERSION"\n\n'
1014	<	txt += ' echo "$SOFTWARE_DIR/ProdCommon/ProdCommon/FwkJobRep/ModifyJobReport.py crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH"\n'
1015	<	txt += ' $SOFTWARE_DIR/ProdCommon/ProdCommon/FwkJobRep/ModifyJobReport.py crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH\n'
1016	<
1017	<	txt += ' modifyReport_result=$?\n'
1018	<	txt += ' if [ $modifyReport_result -ne 0 ]; then\n'
1019	<	txt += ' modifyReport_result=70500\n'
1020	<	txt += ' job_exit_code=$modifyReport_result\n'
1340	<	txt += ' echo "ModifyReportResult=$modifyReport_result" \| tee -a $RUNTIME_AREA/$repo\n'
1341	<	txt += ' echo "WARNING: Problem with ModifyJobReport"\n'
1342	<	txt += ' else\n'
1343	<	txt += ' mv NewFrameworkJobReport.xml crab_fjr_$NJob.xml\n'
1344	<	txt += ' fi\n'
982	>	txt = ''
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	>
989	>
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 += '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'
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'
1014	>	txt += 'if [ $modifyReport_result -ne 0 ]; then\n'
1015	>	txt += ' modifyReport_result=70500\n'
1016	>	txt += ' job_exit_code=$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 $RUNTIME_AREA/crab_fjr_$NJob.xml\n'
1021		txt += 'fi\n'
1022		return txt
1023
1024	<	def cleanEnv(self):
1025	<	txt = '\n#Written by cms_cmssw::cleanEnv\n'
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'
1036	<	txt += ' echo "ERROR ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after cleanup of WN"\n'
1037	<	txt += ' job_exit_code=10017\n'
1038	<	txt += ' func_exit\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 1371 \| Line 1100 \| class Cmssw(JobType):
1100		def getParams(self):
1101		return self._params
1102
1103	<	def uniquelist(self, old):
1375	<	"""
1376	<	remove duplicates from a list
1377	<	"""
1378	<	nd={}
1379	<	for e in old:
1380	<	nd[e]=0
1381	<	return nd.keys()
1382	<
1383	<
1384	<	def checkOut(self, limit):
1103	>	def outList(self,list=False):
1104		"""
1105		check the dimension of the output files
1106		"""
1107	<	txt = '\n#Written by cms_cmssw::checkOut\n'
1108	<	txt += 'echo ">>> Starting output sandbox limit check :"\n'
1107	>	txt = ''
1108	>	txt += 'echo ">>> list of expected files on output sandbox"\n'
1109		listOutFiles = []
1110	<	txt += 'stdoutFile=`ls *stdout` \n'
1111	<	txt += 'stderrFile=`ls *stderr` \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(self.numberFile_(file, '$NJob'))
1119	<	listOutFiles.append('$stdoutFile')
1120	<	listOutFiles.append('$stderrFile')
1118	>	listOutFiles.append(numberFile(file, '$NJob'))
1119	>	listOutFiles.append(stdout)
1120	>	listOutFiles.append(stderr)
1121		else:
1122		for file in (self.output_file_sandbox):
1123	<	listOutFiles.append(self.numberFile_(file, '$NJob'))
1124	<	listOutFiles.append('$stdoutFile')
1125	<	listOutFiles.append('$stderrFile')
1403	<
1123	>	listOutFiles.append(numberFile(file, '$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 += 'ls -gGhrta;\n'
1407	<	txt += 'sum=0;\n'
1408	<	txt += 'for file in $filesToCheck ; do\n'
1409	<	txt += ' if [ -e $file ]; then\n'
1410	<	txt += ' tt=`ls -gGrta $file \| awk \'{ print $3 }\'`\n'
1411	<	txt += ' sum=`expr $sum + $tt`\n'
1412	<	txt += ' else\n'
1413	<	txt += ' echo "WARNING: output file $file not found!"\n'
1414	<	txt += ' fi\n'
1415	<	txt += 'done\n'
1416	<	txt += 'echo "Total Output dimension: $sum";\n'
1417	<	txt += 'limit='+str(limit)+';\n'
1418	<	txt += 'echo "WARNING: output files size limit is set to: $limit";\n'
1419	<	txt += 'if [ $limit -lt $sum ]; then\n'
1420	<	txt += ' echo "WARNING: output files have to big size - something will be lost;"\n'
1421	<	txt += ' echo " checking the output file sizes..."\n'
1422	<	txt += ' tot=0;\n'
1423	<	txt += ' for filefile in $filesToCheck ; do\n'
1424	<	txt += ' dimFile=`ls -gGrta $filefile \| awk \'{ print $3 }\';`\n'
1425	<	txt += ' tot=`expr $tot + $tt`;\n'
1426	<	txt += ' if [ $limit -lt $dimFile ]; then\n'
1427	<	txt += ' echo "deleting file: $filefile";\n'
1428	<	txt += ' rm -f $filefile\n'
1429	<	txt += ' elif [ $limit -lt $tot ]; then\n'
1430	<	txt += ' echo "deleting file: $filefile";\n'
1431	<	txt += ' rm -f $filefile\n'
1432	<	txt += ' else\n'
1433	<	txt += ' echo "saving file: $filefile"\n'
1434	<	txt += ' fi\n'
1435	<	txt += ' done\n'
1128	>	txt += 'export filesToCheck\n'
1129
1130	<	txt += ' ls -agGhrt\n'
1438	<	txt += ' echo "WARNING: output files are too big in dimension: can not put in the output_sandbox."\n'
1439	<	#txt += ' echo "JOB_EXIT_STATUS = 70000"\n'
1440	<	#txt += ' exit_status=70000\n'
1441	<	txt += ' job_exit_code=70000\n'
1442	<	txt += 'else\n'
1443	<	txt += ' echo "Total Output dimension $sum is fine."\n'
1444	<	txt += 'fi\n'
1445	<	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

Comparing COMP/CRAB/python/cms_cmssw.py (file contents): Revision 1.164 by fanzago, Tue Mar 4 18:15:38 2008 UTC vs. Revision 1.314 by spiga, Thu Jun 18 17:23:02 2009 UTC

Diff Legend

Comparing COMP/CRAB/python/cms_cmssw.py (file contents):
Revision 1.164 by fanzago, Tue Mar 4 18:15:38 2008 UTC vs.
Revision 1.314 by spiga, Thu Jun 18 17:23:02 2009 UTC