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 1.128 by fanzago, Thu Oct 11 16:23:44 2007 UTC vs.
Revision 1.300 by spiga, Wed May 20 13:51:45 2009 UTC

#	Line 2 | Line 2 | from JobType import JobType
2		from crab_logger import Logger
3		from crab_exceptions import *
4		from crab_util import *
5	–	from BlackWhiteListParser import BlackWhiteListParser
5		import common
6		import Scram
7	+	from Splitter import JobSplitter
8
9	+	from IMProv.IMProvNode import IMProvNode
10		import os, string, glob
11
12		class Cmssw(JobType):
13	<	def __init__(self, cfg_params, ncjobs):
13	>	def __init__(self, cfg_params, ncjobs,skip_blocks, isNew):
14		JobType.__init__(self, 'CMSSW')
15		common.logger.debug(3,'CMSSW::__init__')
16	+	self.skip_blocks = skip_blocks
17	+	self.argsList = 1
18
19		self._params = {}
20		self.cfg_params = cfg_params
21
22	<	# init BlackWhiteListParser
23	<	self.blackWhiteListParser = BlackWhiteListParser(cfg_params)
24	<
25	<	try:
26	<	self.MaxTarBallSize = float(self.cfg_params['EDG.maxtarballsize'])
27	<	except KeyError:
25	<	self.MaxTarBallSize = 9.5
22	>	### Temporary patch to automatically skip the ISB size check:
23	>	server=self.cfg_params.get('CRAB.server_name',None)
24	>	size = 9.5
25	>	if server or common.scheduler.name().upper() in ['LSF','CAF']: size = 99999
26	>	### D.S.
27	>	self.MaxTarBallSize = float(self.cfg_params.get('EDG.maxtarballsize',size))
28
29		# number of jobs requested to be created, limit obj splitting
30		self.ncjobs = ncjobs
31
32		log = common.logger
33	<
33	>
34		self.scram = Scram.Scram(cfg_params)
35		self.additional_inbox_files = []
36		self.scriptExe = ''
37		self.executable = ''
38		self.executable_arch = self.scram.getArch()
39	<	self.tgz_name = 'default.tgz'
40	<	self.additional_tgz_name = 'additional.tgz'
39	>	self.tgz_name = 'default.tar.gz'
40	>	self.tar_name = 'default.tar'
41		self.scriptName = 'CMSSW.sh'
42	<	self.pset = ''      #scrip use case Da
43	<	self.datasetPath = '' #scrip use case Da
42	>	self.pset = ''
43	>	self.datasetPath = ''
44
45	+	self.tgzNameWithPath = common.work_space.pathForTgz()+self.tgz_name
46		# set FJR file name
47		self.fjrFileName = 'crab_fjr.xml'
48
49		self.version = self.scram.getSWVersion()
50	<
51	<	#
52	<	# Try to block creation in case of arch/version mismatch
53	<	#
54	<
52	<	a = string.split(self.version, "_")
53	<
54	<	if int(a[1]) == 1 and (int(a[2]) < 5 and self.executable_arch.find('slc4') == 0):
55	<	msg = "Error: CMS does not support %s with %s architecture"%(self.version, self.executable_arch)
50	>	common.logger.write("CMSSW version is: "+str(self.version))
51	>	try:
52	>	type, self.CMSSW_major, self.CMSSW_minor, self.CMSSW_patch = tuple(self.version.split('_'))
53	>	except:
54	>	msg = "Cannot parse CMSSW version string: " + self.version + " for major and minor release number!"
55		raise CrabException(msg)
56	<	if int(a[1]) == 1 and (int(a[2]) >= 5 and self.executable_arch.find('slc3') == 0):
57	<	msg = "Error: CMS does not support %s with %s architecture"%(self.version, self.executable_arch)
56	>
57	>	if self.CMSSW_major < 1 or (self.CMSSW_major == 1 and self.CMSSW_minor < 5):
58	>	msg = "CRAB supports CMSSW >= 1_5_x only. Use an older CRAB version."
59		raise CrabException(msg)
60	<
61	<	common.taskDB.setDict('codeVersion',self.version)
62	<	self.setParam_('application', self.version)
60	>	"""
61	>	As CMSSW versions are dropped we can drop more code:
62	>	1.X dropped: drop support for running .cfg on WN
63	>	2.0 dropped: drop all support for cfg here and in writeCfg
64	>	2.0 dropped: Recheck the random number seed support
65	>	"""
66
67		### collect Data cards
68
66	–	## get DBS mode
67	–	try:
68	–	self.use_dbs_1 = int(self.cfg_params['CMSSW.use_dbs_1'])
69	–	except KeyError:
70	–	self.use_dbs_1 = 0
71	–
72	–	try:
73	–	tmp =  cfg_params['CMSSW.datasetpath']
74	–	log.debug(6, "CMSSW::CMSSW(): datasetPath = "+tmp)
75	–	if string.lower(tmp)=='none':
76	–	self.datasetPath = None
77	–	self.selectNoInput = 1
78	–	else:
79	–	self.datasetPath = tmp
80	–	self.selectNoInput = 0
81	–	except KeyError:
82	–	msg = "Error: datasetpath not defined "
83	–	raise CrabException(msg)
69
70	<	# ML monitoring
71	<	# split dataset path style: /PreProdR3Minbias/SIM/GEN-SIM
72	<	if not self.datasetPath:
73	<	self.setParam_('dataset', 'None')
74	<	self.setParam_('owner', 'None')
70	>	### Temporary: added to remove input file control in the case of PU
71	>	self.dataset_pu = cfg_params.get('CMSSW.dataset_pu', None)
72	>
73	>	tmp =  cfg_params['CMSSW.datasetpath']
74	>	log.debug(6, "CMSSW::CMSSW(): datasetPath = "+tmp)
75	>
76	>	if tmp =='':
77	>	msg = "Error: datasetpath not defined "
78	>	raise CrabException(msg)
79	>	elif string.lower(tmp)=='none':
80	>	self.datasetPath = None
81	>	self.selectNoInput = 1
82		else:
83	<	try:
84	<	datasetpath_split = self.datasetPath.split("/")
93	<	# standard style
94	<	self.setParam_('datasetFull', self.datasetPath)
95	<	if self.use_dbs_1 == 1 :
96	<	self.setParam_('dataset', datasetpath_split[1])
97	<	self.setParam_('owner', datasetpath_split[-1])
98	<	else:
99	<	self.setParam_('dataset', datasetpath_split[1])
100	<	self.setParam_('owner', datasetpath_split[2])
101	<	except:
102	<	self.setParam_('dataset', self.datasetPath)
103	<	self.setParam_('owner', self.datasetPath)
104	<
105	<	self.setTaskid_()
106	<	self.setParam_('taskId', self.cfg_params['taskId'])
83	>	self.datasetPath = tmp
84	>	self.selectNoInput = 0
85
86		self.dataTiers = []
87
88	+	self.debugWrap=''
89	+	self.debug_wrapper = int(cfg_params.get('USER.debug_wrapper',0))
90	+	if self.debug_wrapper == 1: self.debugWrap='--debug'
91	+
92		## now the application
93	<	try:
94	<	self.executable = cfg_params['CMSSW.executable']
95	<	self.setParam_('exe', self.executable)
96	<	log.debug(6, "CMSSW::CMSSW(): executable = "+self.executable)
115	<	msg = "Default executable cmsRun overridden. Switch to " + self.executable
116	<	log.debug(3,msg)
117	<	except KeyError:
118	<	self.executable = 'cmsRun'
119	<	self.setParam_('exe', self.executable)
120	<	msg = "User executable not defined. Use cmsRun"
121	<	log.debug(3,msg)
122	<	pass
93	>	self.managedGenerators = ['madgraph','comphep']
94	>	self.generator = cfg_params.get('CMSSW.generator','pythia').lower()
95	>	self.executable = cfg_params.get('CMSSW.executable','cmsRun')
96	>	log.debug(6, "CMSSW::CMSSW(): executable = "+self.executable)
97
98	<	try:
125	<	self.pset = cfg_params['CMSSW.pset']
126	<	log.debug(6, "Cmssw::Cmssw(): PSet file = "+self.pset)
127	<	if self.pset.lower() != 'none' :
128	<	if (not os.path.exists(self.pset)):
129	<	raise CrabException("User defined PSet file "+self.pset+" does not exist")
130	<	else:
131	<	self.pset = None
132	<	except KeyError:
98	>	if not cfg_params.has_key('CMSSW.pset'):
99		raise CrabException("PSet file missing. Cannot run cmsRun ")
100	+	self.pset = cfg_params['CMSSW.pset']
101	+	log.debug(6, "Cmssw::Cmssw(): PSet file = "+self.pset)
102	+	if self.pset.lower() != 'none' :
103	+	if (not os.path.exists(self.pset)):
104	+	raise CrabException("User defined PSet file "+self.pset+" does not exist")
105	+	else:
106	+	self.pset = None
107
108		# output files
109		## stuff which must be returned always via sandbox
#	Line 140 | Line 113 | class Cmssw(JobType):
113		self.output_file_sandbox.append(self.fjrFileName)
114
115		# other output files to be returned via sandbox or copied to SE
116	<	try:
117	<	self.output_file = []
118	<	tmp = cfg_params['CMSSW.output_file']
119	<	if tmp != '':
120	<	tmpOutFiles = string.split(cfg_params['CMSSW.output_file'],',')
121	<	log.debug(7, 'cmssw::cmssw(): output files '+str(tmpOutFiles))
122	<	for tmp in tmpOutFiles:
123	<	tmp=string.strip(tmp)
151	<	self.output_file.append(tmp)
152	<	pass
153	<	else:
154	<	log.message("No output file defined: only stdout/err and the CRAB Framework Job Report will be available\n")
155	<	pass
156	<	pass
157	<	except KeyError:
158	<	log.message("No output file defined: only stdout/err and the CRAB Framework Job Report will be available\n")
159	<	pass
116	>	outfileflag = False
117	>	self.output_file = []
118	>	tmp = cfg_params.get('CMSSW.output_file',None)
119	>	if tmp :
120	>	self.output_file = [x.strip() for x in tmp.split(',')]
121	>	outfileflag = True #output found
122	>	#else:
123	>	#    log.message("No output file defined: only stdout/err and the CRAB Framework Job Report will be available\n")
124
125		# script_exe file as additional file in inputSandbox
126	<	try:
127	<	self.scriptExe = cfg_params['USER.script_exe']
128	<	if self.scriptExe != '':
129	<	if not os.path.isfile(self.scriptExe):
130	<	msg ="ERROR. file "+self.scriptExe+" not found"
131	<	raise CrabException(msg)
168	<	self.additional_inbox_files.append(string.strip(self.scriptExe))
169	<	except KeyError:
170	<	self.scriptExe = ''
126	>	self.scriptExe = cfg_params.get('USER.script_exe',None)
127	>	if self.scriptExe :
128	>	if not os.path.isfile(self.scriptExe):
129	>	msg ="ERROR. file "+self.scriptExe+" not found"
130	>	raise CrabException(msg)
131	>	self.additional_inbox_files.append(string.strip(self.scriptExe))
132
172	–	#CarlosDaniele
133		if self.datasetPath == None and self.pset == None and self.scriptExe == '' :
134	<	msg ="Error. script_exe  not defined"
135	<	raise CrabException(msg)
134	>	msg ="Error. script_exe  not defined"
135	>	raise CrabException(msg)
136	>
137	>	# use parent files...
138	>	self.useParent = int(self.cfg_params.get('CMSSW.use_parent',0))
139
140		## additional input files
141	<	try:
141	>	if cfg_params.has_key('USER.additional_input_files'):
142		tmpAddFiles = string.split(cfg_params['USER.additional_input_files'],',')
143		for tmp in tmpAddFiles:
144		tmp = string.strip(tmp)
#	Line 192 | Line 155 | class Cmssw(JobType):
155		if not os.path.exists(file):
156		raise CrabException("Additional input file not found: "+file)
157		pass
195	–	# fname = string.split(file, '/')[-1]
196	–	# storedFile = common.work_space.pathForTgz()+'share/'+fname
197	–	# shutil.copyfile(file, storedFile)
158		self.additional_inbox_files.append(string.strip(file))
159		pass
160		pass
161		common.logger.debug(5,"Additional input files: "+str(self.additional_inbox_files))
162	<	except KeyError:
203	<	pass
204	<
205	<	# files per job
206	<	try:
207	<	if (cfg_params['CMSSW.files_per_jobs']):
208	<	raise CrabException("files_per_jobs no longer supported.  Quitting.")
209	<	except KeyError:
210	<	pass
211	<
212	<	## Events per job
213	<	try:
214	<	self.eventsPerJob =int( cfg_params['CMSSW.events_per_job'])
215	<	self.selectEventsPerJob = 1
216	<	except KeyError:
217	<	self.eventsPerJob = -1
218	<	self.selectEventsPerJob = 0
219	<
220	<	## number of jobs
221	<	try:
222	<	self.theNumberOfJobs =int( cfg_params['CMSSW.number_of_jobs'])
223	<	self.selectNumberOfJobs = 1
224	<	except KeyError:
225	<	self.theNumberOfJobs = 0
226	<	self.selectNumberOfJobs = 0
227	<
228	<	try:
229	<	self.total_number_of_events = int(cfg_params['CMSSW.total_number_of_events'])
230	<	self.selectTotalNumberEvents = 1
231	<	except KeyError:
232	<	self.total_number_of_events = 0
233	<	self.selectTotalNumberEvents = 0
234	<
235	<	if self.pset != None: #CarlosDaniele
236	<	if ( (self.selectTotalNumberEvents + self.selectEventsPerJob + self.selectNumberOfJobs) != 2 ):
237	<	msg = 'Must define exactly two of total_number_of_events, events_per_job, or number_of_jobs.'
238	<	raise CrabException(msg)
239	<	else:
240	<	if (self.selectNumberOfJobs == 0):
241	<	msg = 'Must specify  number_of_jobs.'
242	<	raise CrabException(msg)
243	<
244	<	## source seed for pythia
245	<	try:
246	<	self.sourceSeed = int(cfg_params['CMSSW.pythia_seed'])
247	<	except KeyError:
248	<	self.sourceSeed = None
249	<	common.logger.debug(5,"No seed given")
250	<
251	<	try:
252	<	self.sourceSeedVtx = int(cfg_params['CMSSW.vtx_seed'])
253	<	except KeyError:
254	<	self.sourceSeedVtx = None
255	<	common.logger.debug(5,"No vertex seed given")
256	<
257	<	try:
258	<	self.sourceSeedG4 = int(cfg_params['CMSSW.g4_seed'])
259	<	except KeyError:
260	<	self.sourceSeedG4 = None
261	<	common.logger.debug(5,"No g4 sim hits seed given")
262	<
263	<	try:
264	<	self.sourceSeedMix = int(cfg_params['CMSSW.mix_seed'])
265	<	except KeyError:
266	<	self.sourceSeedMix = None
267	<	common.logger.debug(5,"No mix seed given")
162	>	pass
163
269	–	try:
270	–	self.firstRun = int(cfg_params['CMSSW.first_run'])
271	–	except KeyError:
272	–	self.firstRun = None
273	–	common.logger.debug(5,"No first run given")
274	–	if self.pset != None: #CarlosDaniele
275	–	ver = string.split(self.version,"_")
276	–	if (int(ver[1])>=1 and int(ver[2])>=5):
277	–	import PsetManipulator150 as pp
278	–	else:
279	–	import PsetManipulator as pp
280	–	PsetEdit = pp.PsetManipulator(self.pset) #Daniele Pset
164
165	+	## New method of dealing with seeds
166	+	self.incrementSeeds = []
167	+	self.preserveSeeds = []
168	+	if cfg_params.has_key('CMSSW.preserve_seeds'):
169	+	tmpList = cfg_params['CMSSW.preserve_seeds'].split(',')
170	+	for tmp in tmpList:
171	+	tmp.strip()
172	+	self.preserveSeeds.append(tmp)
173	+	if cfg_params.has_key('CMSSW.increment_seeds'):
174	+	tmpList = cfg_params['CMSSW.increment_seeds'].split(',')
175	+	for tmp in tmpList:
176	+	tmp.strip()
177	+	self.incrementSeeds.append(tmp)
178	+
179	+	self.firstRun = cfg_params.get('CMSSW.first_run',None)
180	+
181	+	# Copy/return
182	+	self.copy_data = int(cfg_params.get('USER.copy_data',0))
183	+	self.return_data = int(cfg_params.get('USER.return_data',0))
184	+
185	+	self.conf = {}
186	+	self.conf['pubdata'] = None
187	+	# number of jobs requested to be created, limit obj splitting DD
188		#DBSDLS-start
189	<	## Initialize the variables that are extracted from DBS/DLS and needed in other places of the code
189	>	## Initialize the variables that are extracted from DBS/DLS and needed in other places of the code
190		self.maxEvents=0  # max events available   ( --> check the requested nb. of evts in Creator.py)
191		self.DBSPaths={}  # all dbs paths requested ( --> input to the site local discovery script)
192		self.jobDestination=[]  # Site destination(s) for each job (list of lists)
#	Line 289 | Line 195 | class Cmssw(JobType):
195		blockSites = {}
196		if self.datasetPath:
197		blockSites = self.DataDiscoveryAndLocation(cfg_params)
198	<	#DBSDLS-end
198	>	#DBSDLS-end
199	>	self.conf['blockSites']=blockSites
200
294	–	self.tgzNameWithPath = self.getTarBall(self.executable)
295	–
201		## Select Splitting
202	<	if self.selectNoInput:
203	<	if self.pset == None: #CarlosDaniele
204	<	self.jobSplittingForScript()
202	>	splitByRun = int(cfg_params.get('CMSSW.split_by_run',0))
203	>
204	>	if self.selectNoInput:
205	>	if self.pset == None:
206	>	self.algo = 'ForScript'
207		else:
208	<	self.jobSplittingNoInput()
208	>	self.algo = 'NoInput'
209	>	self.conf['managedGenerators']=self.managedGenerators
210	>	self.conf['generator']=self.generator
211	>	elif splitByRun ==1:
212	>	self.algo = 'RunBased'
213		else:
214	<	self.jobSplittingByBlocks(blockSites)
214	>	self.algo = 'EventBased'
215	>
216	>	#        self.algo = 'LumiBased'
217	>	splitter = JobSplitter(self.cfg_params,self.conf)
218	>	self.dict = splitter.Algos()[self.algo]()
219	>
220	>	self.argsFile= '%s/arguments.xml'%common.work_space.shareDir()
221	>	self.rootArgsFilename= 'arguments'
222	>	# modify Pset only the first time
223	>	if (isNew and self.pset != None): self.ModifyPset()
224	>
225	>	## Prepare inputSandbox TarBall (only the first time)
226	>	self.tarNameWithPath = self.getTarBall(self.executable)
227	>
228	>
229	>	def ModifyPset(self):
230	>	import PsetManipulator as pp
231	>	PsetEdit = pp.PsetManipulator(self.pset)
232	>	try:
233	>	# Add FrameworkJobReport to parameter-set, set max events.
234	>	# Reset later for data jobs by writeCFG which does all modifications
235	>	PsetEdit.maxEvent(1)
236	>	PsetEdit.skipEvent(0)
237	>	PsetEdit.psetWriter(self.configFilename())
238	>	## If present, add TFileService to output files
239	>	if not int(self.cfg_params.get('CMSSW.skip_TFileService_output',0)):
240	>	tfsOutput = PsetEdit.getTFileService()
241	>	if tfsOutput:
242	>	if tfsOutput in self.output_file:
243	>	common.logger.debug(5,"Output from TFileService "+tfsOutput+" already in output files")
244	>	else:
245	>	outfileflag = True #output found
246	>	self.output_file.append(tfsOutput)
247	>	common.logger.message("Adding "+tfsOutput+" (from TFileService) to list of output files")
248	>	pass
249	>	pass
250	>	## If present and requested, add PoolOutputModule to output files
251	>	if int(self.cfg_params.get('CMSSW.get_edm_output',0)):
252	>	edmOutput = PsetEdit.getPoolOutputModule()
253	>	if edmOutput:
254	>	if edmOutput in self.output_file:
255	>	common.logger.debug(5,"Output from PoolOutputModule "+edmOutput+" already in output files")
256	>	else:
257	>	self.output_file.append(edmOutput)
258	>	common.logger.message("Adding "+edmOutput+" (from PoolOutputModule) to list of output files")
259	>	pass
260	>	pass
261	>	# not required: check anyhow if present, to avoid accidental T2 overload
262	>	else:
263	>	edmOutput = PsetEdit.getPoolOutputModule()
264	>	if edmOutput and (edmOutput not in self.output_file):
265	>	msg = "ERROR: a PoolOutputModule is present in your ParameteSet %s \n"%self.pset
266	>	msg +="         but the file produced ( %s ) is not in the list of output files\n"%edmOutput
267	>	msg += "WARNING: please remove it. If you want to keep it, add the file to output_files or use CMSSW.get_edm_output\n"
268	>	raise CrabException(msg)
269	>	pass
270	>	pass
271	>	except CrabException, msg:
272	>	common.logger.message(str(msg))
273	>	msg='Error while manipulating ParameterSet (see previous message, if any): exiting...'
274	>	raise CrabException(msg)
275
305	–	# modify Pset
306	–	if self.pset != None: #CarlosDaniele
307	–	try:
308	–	if (self.datasetPath): # standard job
309	–	# allow to processa a fraction of events in a file
310	–	PsetEdit.inputModule("INPUT")
311	–	PsetEdit.maxEvent("INPUTMAXEVENTS")
312	–	PsetEdit.skipEvent("INPUTSKIPEVENTS")
313	–	else:  # pythia like job
314	–	PsetEdit.maxEvent(self.eventsPerJob)
315	–	if (self.firstRun):
316	–	PsetEdit.pythiaFirstRun("INPUTFIRSTRUN")  #First Run
317	–	if (self.sourceSeed) :
318	–	PsetEdit.pythiaSeed("INPUT")
319	–	if (self.sourceSeedVtx) :
320	–	PsetEdit.vtxSeed("INPUTVTX")
321	–	if (self.sourceSeedG4) :
322	–	PsetEdit.g4Seed("INPUTG4")
323	–	if (self.sourceSeedMix) :
324	–	PsetEdit.mixSeed("INPUTMIX")
325	–	# add FrameworkJobReport to parameter-set
326	–	PsetEdit.addCrabFJR(self.fjrFileName)
327	–	PsetEdit.psetWriter(self.configFilename())
328	–	except:
329	–	msg='Error while manipuliating ParameterSet: exiting...'
330	–	raise CrabException(msg)
276
277		def DataDiscoveryAndLocation(self, cfg_params):
278
279		import DataDiscovery
335	–	import DataDiscovery_DBS2
280		import DataLocation
281		common.logger.debug(10,"CMSSW::DataDiscoveryAndLocation()")
282
#	Line 341 | Line 285 | class Cmssw(JobType):
285		## Contact the DBS
286		common.logger.message("Contacting Data Discovery Services ...")
287		try:
288	<
345	<	if self.use_dbs_1 == 1 :
346	<	self.pubdata=DataDiscovery.DataDiscovery(datasetPath, cfg_params)
347	<	else :
348	<	self.pubdata=DataDiscovery_DBS2.DataDiscovery_DBS2(datasetPath, cfg_params)
288	>	self.pubdata=DataDiscovery.DataDiscovery(datasetPath, cfg_params,self.skip_blocks)
289		self.pubdata.fetchDBSInfo()
290
291		except DataDiscovery.NotExistingDatasetError, ex :
#	Line 357 | Line 297 | class Cmssw(JobType):
297		except DataDiscovery.DataDiscoveryError, ex:
298		msg = 'ERROR ***: failed Data Discovery in DBS :  %s'%ex.getErrorMessage()
299		raise CrabException(msg)
360	–	except DataDiscovery_DBS2.NotExistingDatasetError_DBS2, ex :
361	–	msg = 'ERROR ***: failed Data Discovery in DBS : %s'%ex.getErrorMessage()
362	–	raise CrabException(msg)
363	–	except DataDiscovery_DBS2.NoDataTierinProvenanceError_DBS2, ex :
364	–	msg = 'ERROR ***: failed Data Discovery in DBS : %s'%ex.getErrorMessage()
365	–	raise CrabException(msg)
366	–	except DataDiscovery_DBS2.DataDiscoveryError_DBS2, ex:
367	–	msg = 'ERROR ***: failed Data Discovery in DBS :  %s'%ex.getErrorMessage()
368	–	raise CrabException(msg)
300
301		self.filesbyblock=self.pubdata.getFiles()
302	<	self.eventsbyblock=self.pubdata.getEventsPerBlock()
303	<	self.eventsbyfile=self.pubdata.getEventsPerFile()
302	>	#print self.filesbyblock
303	>	self.conf['pubdata']=self.pubdata
304
305		## get max number of events
306	<	self.maxEvents=self.pubdata.getMaxEvents() ##  self.maxEvents used in Creator.py
306	>	self.maxEvents=self.pubdata.getMaxEvents()
307
308		## Contact the DLS and build a list of sites hosting the fileblocks
309		try:
310		dataloc=DataLocation.DataLocation(self.filesbyblock.keys(),cfg_params)
311		dataloc.fetchDLSInfo()
312	+
313		except DataLocation.DataLocationError , ex:
314		msg = 'ERROR ***: failed Data Location in DLS \n %s '%ex.getErrorMessage()
315		raise CrabException(msg)
384	–
316
317	<	sites = dataloc.getSites()
317	>
318	>	unsorted_sites = dataloc.getSites()
319	>	#print "Unsorted :",unsorted_sites
320	>	sites = self.filesbyblock.fromkeys(self.filesbyblock,'')
321	>	for lfn in self.filesbyblock.keys():
322	>	#print lfn
323	>	if unsorted_sites.has_key(lfn):
324	>	#print "Found ",lfn
325	>	sites[lfn]=unsorted_sites[lfn]
326	>	else:
327	>	#print "Not Found ",lfn
328	>	sites[lfn]=[]
329	>	#print sites
330	>
331	>	#print "Sorted :",sites
332	>	if len(sites)==0:
333	>	msg = 'ERROR ***: no location for any of the blocks of this dataset: \n\t %s \n'%datasetPath
334	>	msg += "\tMaybe the dataset is located only at T1's (or at T0), where analysis jobs are not allowed\n"
335	>	msg += "\tPlease check DataDiscovery page https://cmsweb.cern.ch/dbs_discovery/\n"
336	>	raise CrabException(msg)
337	>
338		allSites = []
339		listSites = sites.values()
340		for listSite in listSites:
341		for oneSite in listSite:
342		allSites.append(oneSite)
343	<	allSites = self.uniquelist(allSites)
343	>	[allSites.append(it) for it in allSites if not allSites.count(it)]
344	>
345
346		# screen output
347		common.logger.message("Requested dataset: " + datasetPath + " has " + str(self.maxEvents) + " events in " + str(len(self.filesbyblock.keys())) + " blocks.\n")
348
349		return sites
398	–
399	–	def jobSplittingByBlocks(self, blockSites):
400	–	"""
401	–	Perform job splitting. Jobs run over an integer number of files
402	–	and no more than one block.
403	–	ARGUMENT: blockSites: dictionary with blocks as keys and list of host sites as values
404	–	REQUIRES: self.selectTotalNumberEvents, self.selectEventsPerJob, self.selectNumberofJobs,
405	–	self.total_number_of_events, self.eventsPerJob, self.theNumberOfJobs,
406	–	self.maxEvents, self.filesbyblock
407	–	SETS: self.jobDestination - Site destination(s) for each job (a list of lists)
408	–	self.total_number_of_jobs - Total # of jobs
409	–	self.list_of_args - File(s) job will run on (a list of lists)
410	–	"""
411	–
412	–	# ---- Handle the possible job splitting configurations ---- #
413	–	if (self.selectTotalNumberEvents):
414	–	totalEventsRequested = self.total_number_of_events
415	–	if (self.selectEventsPerJob):
416	–	eventsPerJobRequested = self.eventsPerJob
417	–	if (self.selectNumberOfJobs):
418	–	totalEventsRequested = self.theNumberOfJobs * self.eventsPerJob
419	–
420	–	# If user requested all the events in the dataset
421	–	if (totalEventsRequested == -1):
422	–	eventsRemaining=self.maxEvents
423	–	# If user requested more events than are in the dataset
424	–	elif (totalEventsRequested > self.maxEvents):
425	–	eventsRemaining = self.maxEvents
426	–	common.logger.message("Requested "+str(self.total_number_of_events)+ " events, but only "+str(self.maxEvents)+" events are available.")
427	–	# If user requested less events than are in the dataset
428	–	else:
429	–	eventsRemaining = totalEventsRequested
350
431	–	# If user requested more events per job than are in the dataset
432	–	if (self.selectEventsPerJob and eventsPerJobRequested > self.maxEvents):
433	–	eventsPerJobRequested = self.maxEvents
434	–
435	–	# For user info at end
436	–	totalEventCount = 0
437	–
438	–	if (self.selectTotalNumberEvents and self.selectNumberOfJobs):
439	–	eventsPerJobRequested = int(eventsRemaining/self.theNumberOfJobs)
440	–
441	–	if (self.selectNumberOfJobs):
442	–	common.logger.message("May not create the exact number_of_jobs requested.")
443	–
444	–	if ( self.ncjobs == 'all' ) :
445	–	totalNumberOfJobs = 999999999
446	–	else :
447	–	totalNumberOfJobs = self.ncjobs
448	–
449	–
450	–	blocks = blockSites.keys()
451	–	blockCount = 0
452	–	# Backup variable in case self.maxEvents counted events in a non-included block
453	–	numBlocksInDataset = len(blocks)
454	–
455	–	jobCount = 0
456	–	list_of_lists = []
457	–
458	–	# list tracking which jobs are in which jobs belong to which block
459	–	jobsOfBlock = {}
460	–
461	–	# ---- Iterate over the blocks in the dataset until ---- #
462	–	# ---- we've met the requested total # of events    ---- #
463	–	while ( (eventsRemaining > 0) and (blockCount < numBlocksInDataset) and (jobCount < totalNumberOfJobs)):
464	–	block = blocks[blockCount]
465	–	blockCount += 1
466	–	if block not in jobsOfBlock.keys() :
467	–	jobsOfBlock[block] = []
468	–
469	–	if self.eventsbyblock.has_key(block) :
470	–	numEventsInBlock = self.eventsbyblock[block]
471	–	common.logger.debug(5,'Events in Block File '+str(numEventsInBlock))
472	–
473	–	files = self.filesbyblock[block]
474	–	numFilesInBlock = len(files)
475	–	if (numFilesInBlock <= 0):
476	–	continue
477	–	fileCount = 0
351
352	<	# ---- New block => New job ---- #
480	<	parString = "\\{"
481	<	# counter for number of events in files currently worked on
482	<	filesEventCount = 0
483	<	# flag if next while loop should touch new file
484	<	newFile = 1
485	<	# job event counter
486	<	jobSkipEventCount = 0
487	<
488	<	# ---- Iterate over the files in the block until we've met the requested ---- #
489	<	# ---- total # of events or we've gone over all the files in this block  ---- #
490	<	while ( (eventsRemaining > 0) and (fileCount < numFilesInBlock) and (jobCount < totalNumberOfJobs) ):
491	<	file = files[fileCount]
492	<	if newFile :
493	<	try:
494	<	numEventsInFile = self.eventsbyfile[file]
495	<	common.logger.debug(6, "File "+str(file)+" has "+str(numEventsInFile)+" events")
496	<	# increase filesEventCount
497	<	filesEventCount += numEventsInFile
498	<	# Add file to current job
499	<	parString += '\\\"' + file + '\\\"\,'
500	<	newFile = 0
501	<	except KeyError:
502	<	common.logger.message("File "+str(file)+" has unknown number of events: skipping")
503	<
504	<
505	<	# if less events in file remain than eventsPerJobRequested
506	<	if ( filesEventCount - jobSkipEventCount < eventsPerJobRequested ) :
507	<	# if last file in block
508	<	if ( fileCount == numFilesInBlock-1 ) :
509	<	# end job using last file, use remaining events in block
510	<	# close job and touch new file
511	<	fullString = parString[:-2]
512	<	fullString += '\\}'
513	<	list_of_lists.append([fullString,str(-1),str(jobSkipEventCount)])
514	<	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(filesEventCount - jobSkipEventCount)+" events (last file in block).")
515	<	self.jobDestination.append(blockSites[block])
516	<	common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
517	<	# fill jobs of block dictionary
518	<	jobsOfBlock[block].append(jobCount+1)
519	<	# reset counter
520	<	jobCount = jobCount + 1
521	<	totalEventCount = totalEventCount + filesEventCount - jobSkipEventCount
522	<	eventsRemaining = eventsRemaining - filesEventCount + jobSkipEventCount
523	<	jobSkipEventCount = 0
524	<	# reset file
525	<	parString = "\\{"
526	<	filesEventCount = 0
527	<	newFile = 1
528	<	fileCount += 1
529	<	else :
530	<	# go to next file
531	<	newFile = 1
532	<	fileCount += 1
533	<	# if events in file equal to eventsPerJobRequested
534	<	elif ( filesEventCount - jobSkipEventCount == eventsPerJobRequested ) :
535	<	# close job and touch new file
536	<	fullString = parString[:-2]
537	<	fullString += '\\}'
538	<	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
539	<	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.")
540	<	self.jobDestination.append(blockSites[block])
541	<	common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
542	<	jobsOfBlock[block].append(jobCount+1)
543	<	# reset counter
544	<	jobCount = jobCount + 1
545	<	totalEventCount = totalEventCount + eventsPerJobRequested
546	<	eventsRemaining = eventsRemaining - eventsPerJobRequested
547	<	jobSkipEventCount = 0
548	<	# reset file
549	<	parString = "\\{"
550	<	filesEventCount = 0
551	<	newFile = 1
552	<	fileCount += 1
553	<
554	<	# if more events in file remain than eventsPerJobRequested
555	<	else :
556	<	# close job but don't touch new file
557	<	fullString = parString[:-2]
558	<	fullString += '\\}'
559	<	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
560	<	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.")
561	<	self.jobDestination.append(blockSites[block])
562	<	common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
563	<	jobsOfBlock[block].append(jobCount+1)
564	<	# increase counter
565	<	jobCount = jobCount + 1
566	<	totalEventCount = totalEventCount + eventsPerJobRequested
567	<	eventsRemaining = eventsRemaining - eventsPerJobRequested
568	<	# calculate skip events for last file
569	<	# use filesEventCount (contains several files), jobSkipEventCount and eventsPerJobRequest
570	<	jobSkipEventCount = eventsPerJobRequested - (filesEventCount - jobSkipEventCount - self.eventsbyfile[file])
571	<	# remove all but the last file
572	<	filesEventCount = self.eventsbyfile[file]
573	<	parString = "\\{"
574	<	parString += '\\\"' + file + '\\\"\,'
575	<	pass # END if
576	<	pass # END while (iterate over files in the block)
577	<	pass # END while (iterate over blocks in the dataset)
578	<	self.ncjobs = self.total_number_of_jobs = jobCount
579	<	if (eventsRemaining > 0 and jobCount < totalNumberOfJobs ):
580	<	common.logger.message("Could not run on all requested events because some blocks not hosted at allowed sites.")
581	<	common.logger.message(str(jobCount)+" job(s) can run on "+str(totalEventCount)+" events.\n")
582	<
583	<	# screen output
584	<	screenOutput = "List of jobs and available destination sites:\n\n"
352	>	def split(self, jobParams,firstJobID):
353
354	<	# keep trace of block with no sites to print a warning at the end
355	<	noSiteBlock = []
356	<	bloskNoSite = []
589	<
590	<	blockCounter = 0
591	<	for block in blocks:
592	<	if block in jobsOfBlock.keys() :
593	<	blockCounter += 1
594	<	screenOutput += "Block %5i: jobs %20s: sites: %s\n" % (blockCounter,spanRanges(jobsOfBlock[block]),','.join(self.blackWhiteListParser.checkWhiteList(self.blackWhiteListParser.checkBlackList(blockSites[block],block),block)))
595	<	if len(self.blackWhiteListParser.checkWhiteList(self.blackWhiteListParser.checkBlackList(blockSites[block],block),block)) == 0:
596	<	noSiteBlock.append( spanRanges(jobsOfBlock[block]) )
597	<	bloskNoSite.append( blockCounter )
598	<
599	<	common.logger.message(screenOutput)
600	<	if len(noSiteBlock) > 0 and len(bloskNoSite) > 0:
601	<	msg = 'WARNING: No sites are hosting any part of data for block:\n                '
602	<	virgola = ""
603	<	if len(bloskNoSite) > 1:
604	<	virgola = ","
605	<	for block in bloskNoSite:
606	<	msg += ' ' + str(block) + virgola
607	<	msg += '\n               Related jobs:\n                 '
608	<	virgola = ""
609	<	if len(noSiteBlock) > 1:
610	<	virgola = ","
611	<	for range_jobs in noSiteBlock:
612	<	msg += str(range_jobs) + virgola
613	<	msg += '\n               will not be submitted and this block of data can not be analyzed!\n'
614	<	common.logger.message(msg)
354	>	jobParams = self.dict['args']
355	>	njobs = self.dict['njobs']
356	>	self.jobDestination = self.dict['jobDestination']
357
358	<	self.list_of_args = list_of_lists
359	<	return
358	>	if njobs==0:
359	>	raise CrabException("Ask to split "+str(njobs)+" jobs: aborting")
360
361	<	def jobSplittingNoInput(self):
362	<	"""
363	<	Perform job splitting based on number of event per job
622	<	"""
623	<	common.logger.debug(5,'Splitting per events')
624	<	common.logger.message('Required '+str(self.eventsPerJob)+' events per job ')
625	<	common.logger.message('Required '+str(self.theNumberOfJobs)+' jobs in total ')
626	<	common.logger.message('Required '+str(self.total_number_of_events)+' events in total ')
627	<
628	<	if (self.total_number_of_events < 0):
629	<	msg='Cannot split jobs per Events with "-1" as total number of events'
630	<	raise CrabException(msg)
631	<
632	<	if (self.selectEventsPerJob):
633	<	if (self.selectTotalNumberEvents):
634	<	self.total_number_of_jobs = int(self.total_number_of_events/self.eventsPerJob)
635	<	elif(self.selectNumberOfJobs) :
636	<	self.total_number_of_jobs =self.theNumberOfJobs
637	<	self.total_number_of_events =int(self.theNumberOfJobs*self.eventsPerJob)
638	<
639	<	elif (self.selectNumberOfJobs) :
640	<	self.total_number_of_jobs = self.theNumberOfJobs
641	<	self.eventsPerJob = int(self.total_number_of_events/self.total_number_of_jobs)
642	<
643	<	common.logger.debug(5,'N jobs  '+str(self.total_number_of_jobs))
644	<
645	<	# is there any remainder?
646	<	check = int(self.total_number_of_events) - (int(self.total_number_of_jobs)*self.eventsPerJob)
647	<
648	<	common.logger.debug(5,'Check  '+str(check))
649	<
650	<	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')
651	<	if check > 0:
652	<	common.logger.message('Warning: asked '+str(self.total_number_of_events)+' but can do only '+str(int(self.total_number_of_jobs)*self.eventsPerJob))
653	<
654	<	# argument is seed number.$i
655	<	self.list_of_args = []
656	<	for i in range(self.total_number_of_jobs):
657	<	## Since there is no input, any site is good
658	<	# self.jobDestination.append(["Any"])
659	<	self.jobDestination.append([""]) #must be empty to write correctly the xml
660	<	args=[]
661	<	if (self.firstRun):
662	<	## pythia first run
663	<	#self.list_of_args.append([(str(self.firstRun)+str(i))])
664	<	args.append(str(self.firstRun)+str(i))
665	<	else:
666	<	## no first run
667	<	#self.list_of_args.append([str(i)])
668	<	args.append(str(i))
669	<	if (self.sourceSeed):
670	<	args.append(str(self.sourceSeed)+str(i))
671	<	if (self.sourceSeedVtx):
672	<	## + vtx random seed
673	<	args.append(str(self.sourceSeedVtx)+str(i))
674	<	if (self.sourceSeedG4):
675	<	## + G4 random seed
676	<	args.append(str(self.sourceSeedG4)+str(i))
677	<	if (self.sourceSeedMix):
678	<	## + Mix random seed
679	<	args.append(str(self.sourceSeedMix)+str(i))
680	<	pass
681	<	pass
682	<	self.list_of_args.append(args)
683	<	pass
684	<
685	<	# print self.list_of_args
361	>	# create the empty structure
362	>	for i in range(njobs):
363	>	jobParams.append("")
364
365	+	listID=[]
366	+	listField=[]
367	+	listDictions=[]
368	+	exist= os.path.exists(self.argsFile)
369	+	for id in range(njobs):
370	+	job = id + int(firstJobID)
371	+	listID.append(job+1)
372	+	job_ToSave ={}
373	+	concString = ' '
374	+	argu=''
375	+	str_argu = str(job+1)
376	+	if len(jobParams[id]):
377	+	argu = {'JobID': job+1}
378	+	for i in range(len(jobParams[id])):
379	+	argu[self.dict['params'][i]]=jobParams[id][i]
380	+	# just for debug
381	+	str_argu += concString.join(jobParams[id])
382	+	listDictions.append(argu)
383	+	job_ToSave['arguments']= str(job+1)
384	+	job_ToSave['dlsDestination']= self.jobDestination[id]
385	+	listField.append(job_ToSave)
386	+	msg="Job  %s  Arguments:  %s\n"%(str(job+1),str_argu)
387	+	msg+="\t  Destination: %s "%(str(self.jobDestination[id]))
388	+	common.logger.debug(5,msg)
389	+	# write xml
390	+	if len(listDictions):
391	+	if exist==False: self.CreateXML()
392	+	self.addEntry(listDictions)
393	+	self.addXMLfile()
394	+	common._db.updateJob_(listID,listField)
395	+	self.zipTarFile()
396		return
397	+
398	+	def addXMLfile(self):
399
400	+	import tarfile
401	+	# try:
402	+	print self.argsFile
403	+	tar = tarfile.open(self.tarNameWithPath, "a")
404	+	tar.add(self.argsFile, os.path.basename(self.argsFile))
405	+	tar.close()
406	+	## except:
407	+	#     pass
408
409	<	def jobSplittingForScript(self):#CarlosDaniele
409	>
410	>	def CreateXML(self):
411		"""
692	–	Perform job splitting based on number of job
412		"""
413	<	common.logger.debug(5,'Splitting per job')
414	<	common.logger.message('Required '+str(self.theNumberOfJobs)+' jobs in total ')
696	<
697	<	self.total_number_of_jobs = self.theNumberOfJobs
698	<
699	<	common.logger.debug(5,'N jobs  '+str(self.total_number_of_jobs))
700	<
701	<	common.logger.message(str(self.total_number_of_jobs)+' jobs can be created')
702	<
703	<	# argument is seed number.$i
704	<	self.list_of_args = []
705	<	for i in range(self.total_number_of_jobs):
706	<	## Since there is no input, any site is good
707	<	# self.jobDestination.append(["Any"])
708	<	self.jobDestination.append([""])
709	<	## no random seed
710	<	self.list_of_args.append([str(i)])
413	>	result = IMProvNode( self.rootArgsFilename )
414	>	outfile = file( self.argsFile, 'w').write(str(result))
415		return
416
417	<	def split(self, jobParams):
418	<
419	<	common.jobDB.load()
716	<	#### Fabio
717	<	njobs = self.total_number_of_jobs
718	<	arglist = self.list_of_args
719	<	# create the empty structure
720	<	for i in range(njobs):
721	<	jobParams.append("")
722	<
723	<	for job in range(njobs):
724	<	jobParams[job] = arglist[job]
725	<	# print str(arglist[job])
726	<	# print jobParams[job]
727	<	common.jobDB.setArguments(job, jobParams[job])
728	<	common.logger.debug(5,"Job "+str(job)+" Destination: "+str(self.jobDestination[job]))
729	<	common.jobDB.setDestination(job, self.jobDestination[job])
417	>	def addEntry(self, listDictions):
418	>	"""
419	>	_addEntry_
420
421	<	common.jobDB.save()
421	>	add an entry to the xml file
422	>	"""
423	>	from IMProv.IMProvLoader import loadIMProvFile
424	>	## load xml
425	>	improvDoc = loadIMProvFile(self.argsFile)
426	>	entrname= 'Job'
427	>	for dictions in listDictions:
428	>	report = IMProvNode(entrname , None, **dictions)
429	>	improvDoc.addNode(report)
430	>	outfile = file( self.argsFile, 'w').write(str(improvDoc))
431		return
432	<
734	<	def getJobTypeArguments(self, nj, sched):
735	<	result = ''
736	<	for i in common.jobDB.arguments(nj):
737	<	result=result+str(i)+" "
738	<	return result
739	<
432	>
433		def numberOfJobs(self):
434	<	# Fabio
742	<	return self.total_number_of_jobs
434	>	return self.dict['njobs']
435
436		def getTarBall(self, exe):
437		"""
438		Return the TarBall with lib and exe
439		"""
440	<
441	<	# if it exist, just return it
442	<	#
751	<	# Marco. Let's start to use relative path for Boss XML files
752	<	#
753	<	self.tgzNameWithPath = common.work_space.pathForTgz()+'share/'+self.tgz_name
754	<	if os.path.exists(self.tgzNameWithPath):
755	<	return self.tgzNameWithPath
440	>	self.tarNameWithPath = common.work_space.pathForTgz()+self.tar_name
441	>	if os.path.exists(self.tarNameWithPath):
442	>	return self.tarNameWithPath
443
444		# Prepare a tar gzipped file with user binaries.
445		self.buildTar_(exe)
446
447	<	return string.strip(self.tgzNameWithPath)
447	>	return string.strip(self.tarNameWithPath)
448
449		def buildTar_(self, executable):
450
451		# First of all declare the user Scram area
452		swArea = self.scram.getSWArea_()
766	–	#print "swArea = ", swArea
767	–	# swVersion = self.scram.getSWVersion()
768	–	# print "swVersion = ", swVersion
453		swReleaseTop = self.scram.getReleaseTop_()
454	<	#print "swReleaseTop = ", swReleaseTop
771	<
454	>
455		## check if working area is release top
456		if swReleaseTop == '' or swArea == swReleaseTop:
457	+	common.logger.debug(3,"swArea = "+swArea+" swReleaseTop ="+swReleaseTop)
458		return
459
460		import tarfile
461		try: # create tar ball
462	<	tar = tarfile.open(self.tgzNameWithPath, "w:gz")
462	>	#tar = tarfile.open(self.tgzNameWithPath, "w:gz")
463	>	tar = tarfile.open(self.tarNameWithPath, "w")
464		## First find the executable
465		if (self.executable != ''):
466		exeWithPath = self.scram.findFile_(executable)
467		if ( not exeWithPath ):
468		raise CrabException('User executable '+executable+' not found')
469	<
469	>
470		## then check if it's private or not
471		if exeWithPath.find(swReleaseTop) == -1:
472		# the exe is private, so we must ship
#	Line 790 | Line 475 | class Cmssw(JobType):
475		# distinguish case when script is in user project area or given by full path somewhere else
476		if exeWithPath.find(path) >= 0 :
477		exe = string.replace(exeWithPath, path,'')
478	<	tar.add(path+exe,os.path.basename(executable))
478	>	tar.add(path+exe,exe)
479		else :
480		tar.add(exeWithPath,os.path.basename(executable))
481		pass
482		else:
483		# the exe is from release, we'll find it on WN
484		pass
485	<
485	>
486		## Now get the libraries: only those in local working area
487	+	tar.dereference=True
488		libDir = 'lib'
489		lib = swArea+'/' +libDir
490		common.logger.debug(5,"lib "+lib+" to be tarred")
491		if os.path.exists(lib):
492		tar.add(lib,libDir)
493	<
493	>
494		## Now check if module dir is present
495		moduleDir = 'module'
496		module = swArea + '/' + moduleDir
497		if os.path.isdir(module):
498		tar.add(module,moduleDir)
499	+	tar.dereference=False
500
501		## Now check if any data dir(s) is present
502	<	swAreaLen=len(swArea)
503	<	for root, dirs, files in os.walk(swArea):
504	<	if "data" in dirs:
505	<	common.logger.debug(5,"data "+root+"/data"+" to be tarred")
506	<	tar.add(root+"/data",root[swAreaLen:]+"/data")
507	<
508	<	## Add ProdAgent dir to tar
509	<	paDir = 'ProdAgentApi'
510	<	pa = os.environ['CRABDIR'] + '/' + 'ProdAgentApi'
511	<	if os.path.isdir(pa):
512	<	tar.add(pa,paDir)
513	<
514	<	### FEDE FOR DBS PUBLICATION
515	<	## Add PRODCOMMON dir to tar
516	<	prodcommonDir = 'ProdCommon'
517	<	prodcommonPath = os.environ['CRABDIR'] + '/' + 'ProdCommon'
518	<	if os.path.isdir(prodcommonPath):
519	<	tar.add(prodcommonPath,prodcommonDir)
520	<	#############################
521	<
522	<	common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
502	>	self.dataExist = False
503	>	todo_list = [(i, i) for i in  os.listdir(swArea+"/src")]
504	>	while len(todo_list):
505	>	entry, name = todo_list.pop()
506	>	if name.startswith('crab_0_') or  name.startswith('.') or name == 'CVS':
507	>	continue
508	>	if os.path.isdir(swArea+"/src/"+entry):
509	>	entryPath = entry + '/'
510	>	todo_list += [(entryPath + i, i) for i in  os.listdir(swArea+"/src/"+entry)]
511	>	if name == 'data':
512	>	self.dataExist=True
513	>	common.logger.debug(5,"data "+entry+" to be tarred")
514	>	tar.add(swArea+"/src/"+entry,"src/"+entry)
515	>	pass
516	>	pass
517	>
518	>	### CMSSW ParameterSet
519	>	if not self.pset is None:
520	>	cfg_file = common.work_space.jobDir()+self.configFilename()
521	>	tar.add(cfg_file,self.configFilename())
522	>
523	>
524	>	## Add ProdCommon dir to tar
525	>	prodcommonDir = './'
526	>	prodcommonPath = os.environ['CRABDIR'] + '/' + 'external/'
527	>	neededStuff = ['ProdCommon/__init__.py','ProdCommon/FwkJobRep', 'ProdCommon/CMSConfigTools', \
528	>	'ProdCommon/Core', 'ProdCommon/MCPayloads', 'IMProv', 'ProdCommon/Storage', \
529	>	'WMCore/__init__.py','WMCore/Algorithms']
530	>	for file in neededStuff:
531	>	tar.add(prodcommonPath+file,prodcommonDir+file)
532	>
533	>	##### ML stuff
534	>	ML_file_list=['report.py', 'DashboardAPI.py', 'Logger.py', 'ProcInfo.py', 'apmon.py']
535	>	path=os.environ['CRABDIR'] + '/python/'
536	>	for file in ML_file_list:
537	>	tar.add(path+file,file)
538	>
539	>	##### Utils
540	>	Utils_file_list=['parseCrabFjr.py','writeCfg.py', 'fillCrabFjr.py','cmscp.py']
541	>	for file in Utils_file_list:
542	>	tar.add(path+file,file)
543	>
544	>	##### AdditionalFiles
545	>	tar.dereference=True
546	>	for file in self.additional_inbox_files:
547	>	tar.add(file,string.split(file,'/')[-1])
548	>	tar.dereference=False
549	>	common.logger.debug(5,"Files in "+self.tarNameWithPath+" : "+str(tar.getnames()))
550	>
551		tar.close()
552	<	except :
553	<	raise CrabException('Could not create tar-ball')
552	>	except IOError, exc:
553	>	common.logger.write(str(exc))
554	>	raise CrabException('Could not create tar-ball '+self.tarNameWithPath)
555	>	except tarfile.TarError, exc:
556	>	common.logger.write(str(exc))
557	>	raise CrabException('Could not create tar-ball '+self.tarNameWithPath)
558	>
559	>	def zipTarFile(self):
560	>
561	>	cmd = "gzip -c %s > %s "%(self.tarNameWithPath,self.tgzNameWithPath)
562	>	res=runCommand(cmd)
563
840	–	## check for tarball size
564		tarballinfo = os.stat(self.tgzNameWithPath)
565		if ( tarballinfo.st_size > self.MaxTarBallSize*1024*1024 ) :
566	<	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.')
566	>	msg  = 'Input sandbox size of ' + str(float(tarballinfo.st_size)/1024.0/1024.0) + ' MB is larger than the allowed ' + str(self.MaxTarBallSize) \
567	>	+'MB input sandbox limit \n'
568	>	msg += '      and not supported by the direct GRID submission system.\n'
569	>	msg += '      Please use the CRAB server mode by setting server_name=<NAME> in section [CRAB] of your crab.cfg.\n'
570	>	msg += '      For further infos please see https://twiki.cern.ch/twiki/bin/view/CMS/CrabServer#CRABSERVER_for_Users'
571	>	raise CrabException(msg)
572
573		## create tar-ball with ML stuff
846	–	self.MLtgzfile =  common.work_space.pathForTgz()+'share/MLfiles.tgz'
847	–	try:
848	–	tar = tarfile.open(self.MLtgzfile, "w:gz")
849	–	path=os.environ['CRABDIR'] + '/python/'
850	–	for file in ['report.py', 'DashboardAPI.py', 'Logger.py', 'ProcInfo.py', 'apmon.py', 'parseCrabFjr.py']:
851	–	tar.add(path+file,file)
852	–	common.logger.debug(5,"Files added to "+self.MLtgzfile+" : "+str(tar.getnames()))
853	–	tar.close()
854	–	except :
855	–	raise CrabException('Could not create ML files tar-ball')
856	–
857	–	return
858	–
859	–	def additionalInputFileTgz(self):
860	–	"""
861	–	Put all additional files into a tar ball and return its name
862	–	"""
863	–	import tarfile
864	–	tarName=  common.work_space.pathForTgz()+'share/'+self.additional_tgz_name
865	–	tar = tarfile.open(tarName, "w:gz")
866	–	for file in self.additional_inbox_files:
867	–	tar.add(file,string.split(file,'/')[-1])
868	–	common.logger.debug(5,"Files added to "+self.additional_tgz_name+" : "+str(tar.getnames()))
869	–	tar.close()
870	–	return tarName
574
575	<	def wsSetupEnvironment(self, nj):
575	>	def wsSetupEnvironment(self, nj=0):
576		"""
577		Returns part of a job script which prepares
578		the execution environment for the job 'nj'.
579		"""
580	+	# FUTURE: Drop support for .cfg when possible
581	+	if (self.CMSSW_major >= 2 and self.CMSSW_minor >= 1) or (self.CMSSW_major >= 3):
582	+	psetName = 'pset.py'
583	+	else:
584	+	psetName = 'pset.cfg'
585		# Prepare JobType-independent part
586	<	txt = ''
587	<
588	<	## OLI_Daniele at this level  middleware already known
881	<
882	<	txt += 'if [ $middleware == LCG ]; then \n'
883	<	txt += '    echo "### First set SCRAM ARCH and BUILD_ARCH to ' + self.executable_arch + ' ###"\n'
884	<	txt += '    export SCRAM_ARCH='+self.executable_arch+'\n'
885	<	txt += '    export BUILD_ARCH='+self.executable_arch+'\n'
586	>	txt = '\n#Written by cms_cmssw::wsSetupEnvironment\n'
587	>	txt += 'echo ">>> setup environment"\n'
588	>	txt += 'if [ $middleware == LCG ] || [ $middleware == CAF ] || [ $middleware == LSF ]; then \n'
589		txt += self.wsSetupCMSLCGEnvironment_()
590		txt += 'elif [ $middleware == OSG ]; then\n'
591		txt += '    WORKING_DIR=`/bin/mktemp  -d $OSG_WN_TMP/cms_XXXXXXXXXXXX`\n'
592	<	txt += '    echo "Created working directory: $WORKING_DIR"\n'
593	<	txt += '    if [ ! -d $WORKING_DIR ] ;then\n'
594	<	txt += '        echo "SET_CMS_ENV 10016 ==> OSG $WORKING_DIR could not be created on WN `hostname`"\n'
595	<	txt += '    echo "JOB_EXIT_STATUS = 10016"\n'
893	<	txt += '    echo "JobExitCode=10016" | tee -a $RUNTIME_AREA/$repo\n'
894	<	txt += '    dumpStatus $RUNTIME_AREA/$repo\n'
895	<	txt += '        rm -f $RUNTIME_AREA/$repo \n'
896	<	txt += '        echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
897	<	txt += '        echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
898	<	txt += '        exit 1\n'
592	>	txt += '    if [ ! $? == 0 ] ;then\n'
593	>	txt += '        echo "ERROR ==> OSG $WORKING_DIR could not be created on WN `hostname`"\n'
594	>	txt += '        job_exit_code=10016\n'
595	>	txt += '        func_exit\n'
596		txt += '    fi\n'
597	+	txt += '    echo ">>> Created working directory: $WORKING_DIR"\n'
598		txt += '\n'
599		txt += '    echo "Change to working directory: $WORKING_DIR"\n'
600		txt += '    cd $WORKING_DIR\n'
601	<	txt += self.wsSetupCMSOSGEnvironment_()
602	<	txt += '    echo "### Set SCRAM ARCH to ' + self.executable_arch + ' ###"\n'
603	<	txt += '    export SCRAM_ARCH='+self.executable_arch+'\n'
601	>	txt += '    echo ">>> current directory (WORKING_DIR): $WORKING_DIR"\n'
602	>	txt += self.wsSetupCMSOSGEnvironment_()
603	>	#Setup SGE Environment
604	>	txt += 'elif [ $middleware == SGE ]; then\n'
605	>	txt += self.wsSetupCMSLCGEnvironment_()
606	>
607	>	txt += 'elif [ $middleware == ARC ]; then\n'
608	>	txt += self.wsSetupCMSLCGEnvironment_()
609	>
610		txt += 'fi\n'
611
612		# Prepare JobType-specific part
613		scram = self.scram.commandName()
614		txt += '\n\n'
615	<	txt += 'echo "### SPECIFIC JOB SETUP ENVIRONMENT ###"\n'
615	>	txt += 'echo ">>> specific cmssw setup environment:"\n'
616	>	txt += 'echo "CMSSW_VERSION =  '+self.version+'"\n'
617		txt += scram+' project CMSSW '+self.version+'\n'
618		txt += 'status=$?\n'
619		txt += 'if [ $status != 0 ] ; then\n'
620	<	txt += '   echo "SET_EXE_ENV 10034 ==>ERROR CMSSW '+self.version+' not found on `hostname`" \n'
621	<	txt += '   echo "JOB_EXIT_STATUS = 10034"\n'
622	<	txt += '   echo "JobExitCode=10034" | tee -a $RUNTIME_AREA/$repo\n'
918	<	txt += '   dumpStatus $RUNTIME_AREA/$repo\n'
919	<	txt += '   rm -f $RUNTIME_AREA/$repo \n'
920	<	txt += '   echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
921	<	txt += '   echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
922	<	## OLI_Daniele
923	<	txt += '    if [ $middleware == OSG ]; then \n'
924	<	txt += '        echo "Remove working directory: $WORKING_DIR"\n'
925	<	txt += '        cd $RUNTIME_AREA\n'
926	<	txt += '        /bin/rm -rf $WORKING_DIR\n'
927	<	txt += '        if [ -d $WORKING_DIR ] ;then\n'
928	<	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'
929	<	txt += '            echo "JOB_EXIT_STATUS = 10018"\n'
930	<	txt += '            echo "JobExitCode=10018" | tee -a $RUNTIME_AREA/$repo\n'
931	<	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
932	<	txt += '            rm -f $RUNTIME_AREA/$repo \n'
933	<	txt += '            echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
934	<	txt += '            echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
935	<	txt += '        fi\n'
936	<	txt += '    fi \n'
937	<	txt += '   exit 1 \n'
620	>	txt += '    echo "ERROR ==> CMSSW '+self.version+' not found on `hostname`" \n'
621	>	txt += '    job_exit_code=10034\n'
622	>	txt += '    func_exit\n'
623		txt += 'fi \n'
939	–	txt += 'echo "CMSSW_VERSION =  '+self.version+'"\n'
624		txt += 'cd '+self.version+'\n'
625	<	########## FEDE FOR DBS2 ######################
626	<	txt += 'SOFTWARE_DIR=`pwd`\n'
943	<	txt += 'echo SOFTWARE_DIR=$SOFTWARE_DIR \n'
944	<	###############################################
945	<	### needed grep for bug in scramv1 ###
946	<	txt += scram+' runtime -sh\n'
625	>	txt += 'SOFTWARE_DIR=`pwd`; export SOFTWARE_DIR\n'
626	>	txt += 'echo ">>> current directory (SOFTWARE_DIR): $SOFTWARE_DIR" \n'
627		txt += 'eval `'+scram+' runtime -sh | grep -v SCRAMRT_LSB_JOBNAME`\n'
628	<	txt += 'echo $PATH\n'
629	<
628	>	txt += 'if [ $? != 0 ] ; then\n'
629	>	txt += '    echo "ERROR ==> Problem with the command: "\n'
630	>	txt += '    echo "eval \`'+scram+' runtime -sh | grep -v SCRAMRT_LSB_JOBNAME \` at `hostname`"\n'
631	>	txt += '    job_exit_code=10034\n'
632	>	txt += '    func_exit\n'
633	>	txt += 'fi \n'
634		# Handle the arguments:
635		txt += "\n"
636		txt += "## number of arguments (first argument always jobnumber)\n"
637		txt += "\n"
638	<	#        txt += "narg=$#\n"
955	<	txt += "if [ $nargs -lt 2 ]\n"
638	>	txt += "if [ $nargs -lt "+str(self.argsList)+" ]\n"
639		txt += "then\n"
640	<	txt += "    echo 'SET_EXE_ENV 1 ==> ERROR Too few arguments' +$nargs+ \n"
641	<	txt += '    echo "JOB_EXIT_STATUS = 50113"\n'
642	<	txt += '    echo "JobExitCode=50113" | tee -a $RUNTIME_AREA/$repo\n'
960	<	txt += '    dumpStatus $RUNTIME_AREA/$repo\n'
961	<	txt += '    rm -f $RUNTIME_AREA/$repo \n'
962	<	txt += '    echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
963	<	txt += '    echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
964	<	## OLI_Daniele
965	<	txt += '    if [ $middleware == OSG ]; then \n'
966	<	txt += '        echo "Remove working directory: $WORKING_DIR"\n'
967	<	txt += '        cd $RUNTIME_AREA\n'
968	<	txt += '        /bin/rm -rf $WORKING_DIR\n'
969	<	txt += '        if [ -d $WORKING_DIR ] ;then\n'
970	<	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'
971	<	txt += '            echo "JOB_EXIT_STATUS = 50114"\n'
972	<	txt += '            echo "JobExitCode=50114" | tee -a $RUNTIME_AREA/$repo\n'
973	<	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
974	<	txt += '            rm -f $RUNTIME_AREA/$repo \n'
975	<	txt += '            echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
976	<	txt += '            echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
977	<	txt += '        fi\n'
978	<	txt += '    fi \n'
979	<	txt += "    exit 1\n"
640	>	txt += "    echo 'ERROR ==> Too few arguments' +$nargs+ \n"
641	>	txt += '    job_exit_code=50113\n'
642	>	txt += "    func_exit\n"
643		txt += "fi\n"
644		txt += "\n"
645
646		# Prepare job-specific part
647		job = common.job_list[nj]
648	<	### FEDE FOR DBS OUTPUT PUBLICATION
649	<	if (self.datasetPath):
648	>	if (self.datasetPath):
649	>	self.primaryDataset = self.datasetPath.split("/")[1]
650	>	DataTier = self.datasetPath.split("/")[2]
651		txt += '\n'
652		txt += 'DatasetPath='+self.datasetPath+'\n'
653
654	<	datasetpath_split = self.datasetPath.split("/")
655	<
992	<	txt += 'PrimaryDataset='+datasetpath_split[1]+'\n'
993	<	txt += 'DataTier='+datasetpath_split[2]+'\n'
994	<	#txt += 'ProcessedDataset='+datasetpath_split[3]+'\n'
654	>	txt += 'PrimaryDataset='+self.primaryDataset +'\n'
655	>	txt += 'DataTier='+DataTier+'\n'
656		txt += 'ApplicationFamily=cmsRun\n'
657
658		else:
659	+	self.primaryDataset = 'null'
660		txt += 'DatasetPath=MCDataTier\n'
661		txt += 'PrimaryDataset=null\n'
662		txt += 'DataTier=null\n'
1001	–	#txt += 'ProcessedDataset=null\n'
663		txt += 'ApplicationFamily=MCDataTier\n'
664	<	if self.pset != None: #CarlosDaniele
664	>	if self.pset != None:
665		pset = os.path.basename(job.configFilename())
666		txt += '\n'
667		txt += 'cp  $RUNTIME_AREA/'+pset+' .\n'
1007	–	if (self.datasetPath): # standard job
1008	–	#txt += 'InputFiles=$2\n'
1009	–	txt += 'InputFiles=${args[1]}\n'
1010	–	txt += 'MaxEvents=${args[2]}\n'
1011	–	txt += 'SkipEvents=${args[3]}\n'
1012	–	txt += 'echo "Inputfiles:<$InputFiles>"\n'
1013	–	txt += 'sed "s#{\'INPUT\'}#$InputFiles#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
1014	–	txt += 'echo "MaxEvents:<$MaxEvents>"\n'
1015	–	txt += 'sed "s#INPUTMAXEVENTS#$MaxEvents#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
1016	–	txt += 'echo "SkipEvents:<$SkipEvents>"\n'
1017	–	txt += 'sed "s#INPUTSKIPEVENTS#$SkipEvents#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
1018	–	else:  # pythia like job
1019	–	seedIndex=1
1020	–	if (self.firstRun):
1021	–	txt += 'FirstRun=${args['+str(seedIndex)+']}\n'
1022	–	txt += 'echo "FirstRun: <$FirstRun>"\n'
1023	–	txt += 'sed "s#\<INPUTFIRSTRUN\>#$FirstRun#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
1024	–	seedIndex=seedIndex+1
1025	–
1026	–	if (self.sourceSeed):
1027	–	txt += 'Seed=${args['+str(seedIndex)+']}\n'
1028	–	txt += 'sed "s#\<INPUT\>#$Seed#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
1029	–	seedIndex=seedIndex+1
1030	–	## the following seeds are not always present
1031	–	if (self.sourceSeedVtx):
1032	–	txt += 'VtxSeed=${args['+str(seedIndex)+']}\n'
1033	–	txt += 'echo "VtxSeed: <$VtxSeed>"\n'
1034	–	txt += 'sed "s#\<INPUTVTX\>#$VtxSeed#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
1035	–	seedIndex += 1
1036	–	if (self.sourceSeedG4):
1037	–	txt += 'G4Seed=${args['+str(seedIndex)+']}\n'
1038	–	txt += 'echo "G4Seed: <$G4Seed>"\n'
1039	–	txt += 'sed "s#\<INPUTG4\>#$G4Seed#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
1040	–	seedIndex += 1
1041	–	if (self.sourceSeedMix):
1042	–	txt += 'mixSeed=${args['+str(seedIndex)+']}\n'
1043	–	txt += 'echo "MixSeed: <$mixSeed>"\n'
1044	–	txt += 'sed "s#\<INPUTMIX\>#$mixSeed#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
1045	–	seedIndex += 1
1046	–	pass
1047	–	pass
1048	–	txt += 'mv -f '+pset+' pset.cfg\n'
668
669	<	if len(self.additional_inbox_files) > 0:
670	<	txt += 'if [ -e $RUNTIME_AREA/'+self.additional_tgz_name+' ] ; then\n'
671	<	txt += '  tar xzvf $RUNTIME_AREA/'+self.additional_tgz_name+'\n'
672	<	txt += 'fi\n'
1054	<	pass
669	>	txt += 'PreserveSeeds='  + ','.join(self.preserveSeeds)  + '; export PreserveSeeds\n'
670	>	txt += 'IncrementSeeds=' + ','.join(self.incrementSeeds) + '; export IncrementSeeds\n'
671	>	txt += 'echo "PreserveSeeds: <$PreserveSeeds>"\n'
672	>	txt += 'echo "IncrementSeeds:<$IncrementSeeds>"\n'
673
674	<	if self.pset != None: #CarlosDaniele
675	<	txt += 'echo "### END JOB SETUP ENVIRONMENT ###"\n\n'
676	<
677	<	txt += '\n'
678	<	txt += 'echo "***** cat pset.cfg *********"\n'
1061	<	txt += 'cat pset.cfg\n'
1062	<	txt += 'echo "****** end pset.cfg ********"\n'
674	>	txt += 'mv -f ' + pset + ' ' + psetName + '\n'
675	>
676	>
677	>	if self.pset != None:
678	>	# FUTURE: Can simply for 2_1_x and higher
679		txt += '\n'
680	<	### FEDE FOR DBS OUTPUT PUBLICATION
681	<	txt += 'PSETHASH=`EdmConfigHash < pset.cfg` \n'
680	>	if self.debug_wrapper == 1:
681	>	txt += 'echo "***** cat ' + psetName + ' *********"\n'
682	>	txt += 'cat ' + psetName + '\n'
683	>	txt += 'echo "****** end ' + psetName + ' ********"\n'
684	>	txt += '\n'
685	>	txt += 'echo "***********************" \n'
686	>	txt += 'which edmConfigHash \n'
687	>	txt += 'echo "***********************" \n'
688	>	if (self.CMSSW_major >= 2 and self.CMSSW_minor >= 1) or (self.CMSSW_major >= 3):
689	>	txt += 'edmConfigHash ' + psetName + ' \n'
690	>	txt += 'PSETHASH=`edmConfigHash ' + psetName + '` \n'
691	>	else:
692	>	txt += 'PSETHASH=`edmConfigHash < ' + psetName + '` \n'
693		txt += 'echo "PSETHASH = $PSETHASH" \n'
694	<	##############
694	>	#### FEDE temporary fix for noEdm files #####
695	>	txt += 'if [ -z "$PSETHASH" ]; then \n'
696	>	txt += '   export PSETHASH=null\n'
697	>	txt += 'fi \n'
698	>	#############################################
699		txt += '\n'
1069	–	# txt += 'echo "***** cat pset1.cfg *********"\n'
1070	–	# txt += 'cat pset1.cfg\n'
1071	–	# txt += 'echo "****** end pset1.cfg ********"\n'
700		return txt
701
702	<	def wsBuildExe(self, nj=0):
702	>	def wsUntarSoftware(self, nj=0):
703		"""
704		Put in the script the commands to build an executable
705		or a library.
706		"""
707
708	<	txt = ""
708	>	txt = '\n#Written by cms_cmssw::wsUntarSoftware\n'
709
710		if os.path.isfile(self.tgzNameWithPath):
711	<	txt += 'echo "tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+'"\n'
712	<	txt += 'tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+'\n'
711	>	txt += 'echo ">>> tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+' :" \n'
712	>	txt += 'tar zxvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+'\n'
713	>	if  self.debug_wrapper==1 :
714	>	txt += 'ls -Al \n'
715		txt += 'untar_status=$? \n'
716		txt += 'if [ $untar_status -ne 0 ]; then \n'
717	<	txt += '   echo "SET_EXE 1 ==> ERROR Untarring .tgz file failed"\n'
718	<	txt += '   echo "JOB_EXIT_STATUS = $untar_status" \n'
719	<	txt += '   echo "JobExitCode=$untar_status" | tee -a $RUNTIME_AREA/$repo\n'
1090	<	txt += '   if [ $middleware == OSG ]; then \n'
1091	<	txt += '       echo "Remove working directory: $WORKING_DIR"\n'
1092	<	txt += '       cd $RUNTIME_AREA\n'
1093	<	txt += '       /bin/rm -rf $WORKING_DIR\n'
1094	<	txt += '       if [ -d $WORKING_DIR ] ;then\n'
1095	<	txt += '           echo "SET_EXE 50999 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after Untarring .tgz file failed"\n'
1096	<	txt += '           echo "JOB_EXIT_STATUS = 50999"\n'
1097	<	txt += '           echo "JobExitCode=50999" | tee -a $RUNTIME_AREA/$repo\n'
1098	<	txt += '           dumpStatus $RUNTIME_AREA/$repo\n'
1099	<	txt += '           rm -f $RUNTIME_AREA/$repo \n'
1100	<	txt += '           echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1101	<	txt += '           echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1102	<	txt += '       fi\n'
1103	<	txt += '   fi \n'
1104	<	txt += '   \n'
1105	<	txt += '   exit 1 \n'
717	>	txt += '   echo "ERROR ==> Untarring .tgz file failed"\n'
718	>	txt += '   job_exit_code=$untar_status\n'
719	>	txt += '   func_exit\n'
720		txt += 'else \n'
721		txt += '   echo "Successful untar" \n'
722		txt += 'fi \n'
723		txt += '\n'
724	<	txt += 'echo "Include ProdAgentApi and PRODCOMMON in PYTHONPATH"\n'
724	>	txt += 'echo ">>> Include $RUNTIME_AREA in PYTHONPATH:"\n'
725		txt += 'if [ -z "$PYTHONPATH" ]; then\n'
726	<	#### FEDE FOR DBS OUTPUT PUBLICATION
1113	<	txt += '   export PYTHONPATH=$SOFTWARE_DIR/ProdAgentApi:$SOFTWARE_DIR/ProdCommon\n'
1114	<	#txt += '   export PYTHONPATH=`pwd`/ProdAgentApi:`pwd`/ProdCommon\n'
1115	<	#txt += '   export PYTHONPATH=ProdAgentApi\n'
726	>	txt += '   export PYTHONPATH=$RUNTIME_AREA/\n'
727		txt += 'else\n'
728	<	txt += '   export PYTHONPATH=$SOFTWARE_DIR/ProdAgentApi:$SOFTWARE_DIR/ProdCommon:${PYTHONPATH}\n'
1118	<	#txt += '   export PYTHONPATH=`pwd`/ProdAgentApi:`pwd`/ProdCommon:${PYTHONPATH}\n'
1119	<	#txt += '   export PYTHONPATH=ProdAgentApi:${PYTHONPATH}\n'
728	>	txt += '   export PYTHONPATH=$RUNTIME_AREA/:${PYTHONPATH}\n'
729		txt += 'echo "PYTHONPATH=$PYTHONPATH"\n'
1121	–	###################
730		txt += 'fi\n'
731		txt += '\n'
732
733		pass
734	<
734	>
735		return txt
736
737	<	def modifySteeringCards(self, nj):
737	>	def wsBuildExe(self, nj=0):
738		"""
739	<	modify the card provided by the user,
740	<	writing a new card into share dir
739	>	Put in the script the commands to build an executable
740	>	or a library.
741		"""
742	<
742	>
743	>	txt = '\n#Written by cms_cmssw::wsBuildExe\n'
744	>	txt += 'echo ">>> moving CMSSW software directories in `pwd`" \n'
745	>
746	>	txt += 'rm -r lib/ module/ \n'
747	>	txt += 'mv $RUNTIME_AREA/lib/ . \n'
748	>	txt += 'mv $RUNTIME_AREA/module/ . \n'
749	>	if self.dataExist == True:
750	>	txt += 'rm -r src/ \n'
751	>	txt += 'mv $RUNTIME_AREA/src/ . \n'
752	>	if len(self.additional_inbox_files)>0:
753	>	for file in self.additional_inbox_files:
754	>	txt += 'mv $RUNTIME_AREA/'+os.path.basename(file)+' . \n'
755	>	# txt += 'mv $RUNTIME_AREA/ProdCommon/ . \n'
756	>	# txt += 'mv $RUNTIME_AREA/IMProv/ . \n'
757	>
758	>	txt += 'echo ">>> Include $RUNTIME_AREA in PYTHONPATH:"\n'
759	>	txt += 'if [ -z "$PYTHONPATH" ]; then\n'
760	>	txt += '   export PYTHONPATH=$RUNTIME_AREA/\n'
761	>	txt += 'else\n'
762	>	txt += '   export PYTHONPATH=$RUNTIME_AREA/:${PYTHONPATH}\n'
763	>	txt += 'echo "PYTHONPATH=$PYTHONPATH"\n'
764	>	txt += 'fi\n'
765	>	txt += '\n'
766	>
767	>	return txt
768	>
769	>
770		def executableName(self):
771	<	if self.scriptExe: #CarlosDaniele
771	>	if self.scriptExe:
772		return "sh "
773		else:
774		return self.executable
775
776		def executableArgs(self):
777	<	if self.scriptExe:#CarlosDaniele
778	<	return   self.scriptExe + " $NJob"
777	>	# FUTURE: This function tests the CMSSW version. Can be simplified as we drop support for old versions
778	>	if self.scriptExe:
779	>	return self.scriptExe + " $NJob"
780		else:
781	<	# if >= CMSSW_1_5_X, add -e
782	<	version_array = self.scram.getSWVersion().split('_')
783	<	major = 0
784	<	minor = 0
785	<	try:
1150	<	major = int(version_array[1])
1151	<	minor = int(version_array[2])
1152	<	except:
1153	<	msg = "Cannot parse CMSSW version string: " + "_".join(version_array) + " for major and minor release number!"
1154	<	raise CrabException(msg)
1155	<	if major >= 1 and minor >= 5 :
1156	<	return " -e -p pset.cfg"
781	>	ex_args = ""
782	>	ex_args += " -j $RUNTIME_AREA/crab_fjr_$NJob.xml"
783	>	# Type of config file depends on CMSSW version
784	>	if self.CMSSW_major >= 2 :
785	>	ex_args += " -p pset.py"
786		else:
787	<	return " -p pset.cfg"
787	>	ex_args += " -p pset.cfg"
788	>	return ex_args
789
790		def inputSandbox(self, nj):
791		"""
792		Returns a list of filenames to be put in JDL input sandbox.
793		"""
794		inp_box = []
1165	–	# # dict added to delete duplicate from input sandbox file list
1166	–	# seen = {}
1167	–	## code
795		if os.path.isfile(self.tgzNameWithPath):
796		inp_box.append(self.tgzNameWithPath)
797	<	if os.path.isfile(self.MLtgzfile):
1171	<	inp_box.append(self.MLtgzfile)
1172	<	## config
1173	<	if not self.pset is None:
1174	<	inp_box.append(common.work_space.pathForTgz() + 'job/' + self.configFilename())
1175	<	## additional input files
1176	<	tgz = self.additionalInputFileTgz()
1177	<	inp_box.append(tgz)
797	>	inp_box.append(common.work_space.jobDir() + self.scriptName)
798		return inp_box
799
800		def outputSandbox(self, nj):
#	Line 1185 | Line 805 | class Cmssw(JobType):
805
806		## User Declared output files
807		for out in (self.output_file+self.output_file_sandbox):
808	<	n_out = nj + 1
809	<	out_box.append(self.numberFile_(out,str(n_out)))
808	>	n_out = nj + 1
809	>	out_box.append(numberFile(out,str(n_out)))
810		return out_box
811
1192	–	def prepareSteeringCards(self):
1193	–	"""
1194	–	Make initial modifications of the user's steering card file.
1195	–	"""
1196	–	return
812
813		def wsRenameOutput(self, nj):
814		"""
815		Returns part of a job script which renames the produced files.
816		"""
817
818	<	txt = '\n'
819	<	txt += '# directory content\n'
820	<	txt += 'ls \n'
821	<
822	<	txt += 'output_exit_status=0\n'
823	<
824	<	for fileWithSuffix in (self.output_file_sandbox):
1210	<	output_file_num = self.numberFile_(fileWithSuffix, '$NJob')
1211	<	txt += '\n'
1212	<	txt += '# check output file\n'
1213	<	txt += 'if [ -e ./'+fileWithSuffix+' ] ; then\n'
1214	<	txt += '    mv '+fileWithSuffix+' $RUNTIME_AREA\n'
1215	<	txt += '    cp $RUNTIME_AREA/'+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n'
1216	<	txt += 'else\n'
1217	<	txt += '    exit_status=60302\n'
1218	<	txt += '    echo "ERROR: Problem with output file '+fileWithSuffix+'"\n'
1219	<	if common.scheduler.boss_scheduler_name == 'condor_g':
1220	<	txt += '    if [ $middleware == OSG ]; then \n'
1221	<	txt += '        echo "prepare dummy output file"\n'
1222	<	txt += '        echo "Processing of job output failed" > $RUNTIME_AREA/'+output_file_num+'\n'
1223	<	txt += '    fi \n'
1224	<	txt += 'fi\n'
1225	<
818	>	txt = '\n#Written by cms_cmssw::wsRenameOutput\n'
819	>	txt += 'echo ">>> current directory (SOFTWARE_DIR): $SOFTWARE_DIR" \n'
820	>	txt += 'echo ">>> current directory content:"\n'
821	>	if self.debug_wrapper==1:
822	>	txt += 'ls -Al\n'
823	>	txt += '\n'
824	>
825		for fileWithSuffix in (self.output_file):
826	<	output_file_num = self.numberFile_(fileWithSuffix, '$NJob')
826	>	output_file_num = numberFile(fileWithSuffix, '$NJob')
827		txt += '\n'
828		txt += '# check output file\n'
829		txt += 'if [ -e ./'+fileWithSuffix+' ] ; then\n'
830	<	txt += '    mv '+fileWithSuffix+' $RUNTIME_AREA\n'
831	<	txt += '    cp $RUNTIME_AREA/'+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n'
830	>	if (self.copy_data == 1):  # For OSG nodes, file is in $WORKING_DIR, should not be moved to $RUNTIME_AREA
831	>	txt += '    mv '+fileWithSuffix+' '+output_file_num+'\n'
832	>	txt += '    ln -s `pwd`/'+output_file_num+' $RUNTIME_AREA/'+fileWithSuffix+'\n'
833	>	else:
834	>	txt += '    mv '+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n'
835	>	txt += '    ln -s $RUNTIME_AREA/'+output_file_num+' $RUNTIME_AREA/'+fileWithSuffix+'\n'
836		txt += 'else\n'
837	<	txt += '    exit_status=60302\n'
838	<	txt += '    echo "ERROR: Problem with output file '+fileWithSuffix+'"\n'
839	<	txt += '    echo "JOB_EXIT_STATUS = $exit_status"\n'
1237	<	txt += '    output_exit_status=$exit_status\n'
1238	<	if common.scheduler.boss_scheduler_name == 'condor_g':
837	>	txt += '    job_exit_code=60302\n'
838	>	txt += '    echo "WARNING: Output file '+fileWithSuffix+' not found"\n'
839	>	if common.scheduler.name().upper() == 'CONDOR_G':
840		txt += '    if [ $middleware == OSG ]; then \n'
841		txt += '        echo "prepare dummy output file"\n'
842		txt += '        echo "Processing of job output failed" > $RUNTIME_AREA/'+output_file_num+'\n'
#	Line 1243 | Line 844 | class Cmssw(JobType):
844		txt += 'fi\n'
845		file_list = []
846		for fileWithSuffix in (self.output_file):
847	<	file_list.append(self.numberFile_(fileWithSuffix, '$NJob'))
848	<
849	<	txt += 'file_list="'+string.join(file_list,' ')+'"\n'
847	>	file_list.append(numberFile('$SOFTWARE_DIR/'+fileWithSuffix, '$NJob'))
848	>
849	>	txt += 'file_list="'+string.join(file_list,',')+'"\n'
850	>	txt += '\n'
851	>	txt += 'echo ">>> current directory (SOFTWARE_DIR): $SOFTWARE_DIR" \n'
852	>	txt += 'echo ">>> current directory content:"\n'
853	>	if self.debug_wrapper==1:
854	>	txt += 'ls -Al\n'
855	>	txt += '\n'
856		txt += 'cd $RUNTIME_AREA\n'
857	+	txt += 'echo ">>> current directory (RUNTIME_AREA):  $RUNTIME_AREA"\n'
858		return txt
859
1252	–	def numberFile_(self, file, txt):
1253	–	"""
1254	–	append _'txt' before last extension of a file
1255	–	"""
1256	–	p = string.split(file,".")
1257	–	# take away last extension
1258	–	name = p[0]
1259	–	for x in p[1:-1]:
1260	–	name=name+"."+x
1261	–	# add "_txt"
1262	–	if len(p)>1:
1263	–	ext = p[len(p)-1]
1264	–	result = name + '_' + txt + "." + ext
1265	–	else:
1266	–	result = name + '_' + txt
1267	–
1268	–	return result
1269	–
860		def getRequirements(self, nj=[]):
861		"""
862	<	return job requirements to add to jdl files
862	>	return job requirements to add to jdl files
863		"""
864		req = ''
865		if self.version:
866		req='Member("VO-cms-' + \
867		self.version + \
868		'", other.GlueHostApplicationSoftwareRunTimeEnvironment)'
869	<	## SL add requirement for OS version only if SL4
1280	<	#reSL4 = re.compile( r'slc4' )
1281	<	if self.executable_arch: # and reSL4.search(self.executable_arch):
869	>	if self.executable_arch:
870		req+=' && Member("VO-cms-' + \
871		self.executable_arch + \
872		'", other.GlueHostApplicationSoftwareRunTimeEnvironment)'
873
874		req = req + ' && (other.GlueHostNetworkAdapterOutboundIP)'
875	+	if ( common.scheduler.name() == "glitecoll" ) or ( common.scheduler.name() == "glite"):
876	+	req += ' && other.GlueCEStateStatus == "Production" '
877
878		return req
879
880		def configFilename(self):
881		""" return the config filename """
882	<	return self.name()+'.cfg'
882	>	# FUTURE: Can remove cfg mode for CMSSW >= 2_1_x
883	>	if (self.CMSSW_major >= 2 and self.CMSSW_minor >= 1) or (self.CMSSW_major >= 3):
884	>	return self.name()+'.py'
885	>	else:
886	>	return self.name()+'.cfg'
887
1294	–	### OLI_DANIELE
888		def wsSetupCMSOSGEnvironment_(self):
889		"""
890		Returns part of a job script which is prepares
891		the execution environment and which is common for all CMS jobs.
892		"""
893	<	txt = '\n'
894	<	txt += '   echo "### SETUP CMS OSG  ENVIRONMENT ###"\n'
895	<	txt += '   if [ -f $GRID3_APP_DIR/cmssoft/cmsset_default.sh ] ;then\n'
896	<	txt += '      # Use $GRID3_APP_DIR/cmssoft/cmsset_default.sh to setup cms software\n'
897	<	txt += '       export SCRAM_ARCH='+self.executable_arch+'\n'
898	<	txt += '       source $GRID3_APP_DIR/cmssoft/cmsset_default.sh '+self.version+'\n'
1306	<	txt += '   elif [ -f $OSG_APP/cmssoft/cms/cmsset_default.sh ] ;then\n'
893	>	txt = '\n#Written by cms_cmssw::wsSetupCMSOSGEnvironment_\n'
894	>	txt += '    echo ">>> setup CMS OSG environment:"\n'
895	>	txt += '    echo "set SCRAM ARCH to ' + self.executable_arch + '"\n'
896	>	txt += '    export SCRAM_ARCH='+self.executable_arch+'\n'
897	>	txt += '    echo "SCRAM_ARCH = $SCRAM_ARCH"\n'
898	>	txt += '    if [ -f $OSG_APP/cmssoft/cms/cmsset_default.sh ] ;then\n'
899		txt += '      # Use $OSG_APP/cmssoft/cms/cmsset_default.sh to setup cms software\n'
900	<	txt += '       export SCRAM_ARCH='+self.executable_arch+'\n'
901	<	txt += '       source $OSG_APP/cmssoft/cms/cmsset_default.sh '+self.version+'\n'
902	<	txt += '   else\n'
903	<	txt += '       echo "SET_CMS_ENV 10020 ==> ERROR $GRID3_APP_DIR/cmssoft/cmsset_default.sh and $OSG_APP/cmssoft/cms/cmsset_default.sh file not found"\n'
904	<	txt += '       echo "JOB_EXIT_STATUS = 10020"\n'
905	<	txt += '       echo "JobExitCode=10020" | tee -a $RUNTIME_AREA/$repo\n'
1314	<	txt += '       dumpStatus $RUNTIME_AREA/$repo\n'
1315	<	txt += '       rm -f $RUNTIME_AREA/$repo \n'
1316	<	txt += '       echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1317	<	txt += '       echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1318	<	txt += '       exit 1\n'
1319	<	txt += '\n'
1320	<	txt += '       echo "Remove working directory: $WORKING_DIR"\n'
1321	<	txt += '       cd $RUNTIME_AREA\n'
1322	<	txt += '       /bin/rm -rf $WORKING_DIR\n'
1323	<	txt += '       if [ -d $WORKING_DIR ] ;then\n'
1324	<	txt += '           echo "SET_CMS_ENV 10017 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after $GRID3_APP_DIR/cmssoft/cmsset_default.sh and $OSG_APP/cmssoft/cms/cmsset_default.sh file not found"\n'
1325	<	txt += '           echo "JOB_EXIT_STATUS = 10017"\n'
1326	<	txt += '           echo "JobExitCode=10017" | tee -a $RUNTIME_AREA/$repo\n'
1327	<	txt += '           dumpStatus $RUNTIME_AREA/$repo\n'
1328	<	txt += '           rm -f $RUNTIME_AREA/$repo \n'
1329	<	txt += '           echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1330	<	txt += '           echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1331	<	txt += '       fi\n'
1332	<	txt += '\n'
1333	<	txt += '       exit 1\n'
1334	<	txt += '   fi\n'
900	>	txt += '        source $OSG_APP/cmssoft/cms/cmsset_default.sh '+self.version+'\n'
901	>	txt += '    else\n'
902	>	txt += '        echo "ERROR ==> $OSG_APP/cmssoft/cms/cmsset_default.sh file not found"\n'
903	>	txt += '        job_exit_code=10020\n'
904	>	txt += '        func_exit\n'
905	>	txt += '    fi\n'
906		txt += '\n'
907	<	txt += '   echo "SET_CMS_ENV 0 ==> setup cms environment ok"\n'
908	<	txt += '   echo " END SETUP CMS OSG  ENVIRONMENT "\n'
907	>	txt += '    echo "==> setup cms environment ok"\n'
908	>	txt += '    echo "SCRAM_ARCH = $SCRAM_ARCH"\n'
909
910		return txt
911	<
1341	<	### OLI_DANIELE
911	>
912		def wsSetupCMSLCGEnvironment_(self):
913		"""
914		Returns part of a job script which is prepares
915		the execution environment and which is common for all CMS jobs.
916		"""
917	<	txt  = '   \n'
918	<	txt += '   echo " ### SETUP CMS LCG  ENVIRONMENT ### "\n'
919	<	txt += '   if [ ! $VO_CMS_SW_DIR ] ;then\n'
920	<	txt += '       echo "SET_CMS_ENV 10031 ==> ERROR CMS software dir not found on WN `hostname`"\n'
921	<	txt += '       echo "JOB_EXIT_STATUS = 10031" \n'
922	<	txt += '       echo "JobExitCode=10031" | tee -a $RUNTIME_AREA/$repo\n'
923	<	txt += '       dumpStatus $RUNTIME_AREA/$repo\n'
924	<	txt += '       rm -f $RUNTIME_AREA/$repo \n'
925	<	txt += '       echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
926	<	txt += '       echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
927	<	txt += '       exit 1\n'
928	<	txt += '   else\n'
929	<	txt += '       echo "Sourcing environment... "\n'
930	<	txt += '       if [ ! -s $VO_CMS_SW_DIR/cmsset_default.sh ] ;then\n'
931	<	txt += '           echo "SET_CMS_ENV 10020 ==> ERROR cmsset_default.sh file not found into dir $VO_CMS_SW_DIR"\n'
932	<	txt += '           echo "JOB_EXIT_STATUS = 10020"\n'
933	<	txt += '           echo "JobExitCode=10020" | tee -a $RUNTIME_AREA/$repo\n'
934	<	txt += '           dumpStatus $RUNTIME_AREA/$repo\n'
935	<	txt += '           rm -f $RUNTIME_AREA/$repo \n'
936	<	txt += '           echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
937	<	txt += '           echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
938	<	txt += '           exit 1\n'
939	<	txt += '       fi\n'
940	<	txt += '       echo "sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
941	<	txt += '       source $VO_CMS_SW_DIR/cmsset_default.sh\n'
942	<	txt += '       result=$?\n'
943	<	txt += '       if [ $result -ne 0 ]; then\n'
1374	<	txt += '           echo "SET_CMS_ENV 10032 ==> ERROR problem sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
1375	<	txt += '           echo "JOB_EXIT_STATUS = 10032"\n'
1376	<	txt += '           echo "JobExitCode=10032" | tee -a $RUNTIME_AREA/$repo\n'
1377	<	txt += '           dumpStatus $RUNTIME_AREA/$repo\n'
1378	<	txt += '           rm -f $RUNTIME_AREA/$repo \n'
1379	<	txt += '           echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1380	<	txt += '           echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1381	<	txt += '           exit 1\n'
1382	<	txt += '       fi\n'
1383	<	txt += '   fi\n'
1384	<	txt += '   \n'
1385	<	txt += '   echo "SET_CMS_ENV 0 ==> setup cms environment ok"\n'
1386	<	txt += '   echo "### END SETUP CMS LCG ENVIRONMENT ###"\n'
917	>	txt = '\n#Written by cms_cmssw::wsSetupCMSLCGEnvironment_\n'
918	>	txt += '    echo ">>> setup CMS LCG environment:"\n'
919	>	txt += '    echo "set SCRAM ARCH and BUILD_ARCH to ' + self.executable_arch + ' ###"\n'
920	>	txt += '    export SCRAM_ARCH='+self.executable_arch+'\n'
921	>	txt += '    export BUILD_ARCH='+self.executable_arch+'\n'
922	>	txt += '    if [ ! $VO_CMS_SW_DIR ] ;then\n'
923	>	txt += '        echo "ERROR ==> CMS software dir not found on WN `hostname`"\n'
924	>	txt += '        job_exit_code=10031\n'
925	>	txt += '        func_exit\n'
926	>	txt += '    else\n'
927	>	txt += '        echo "Sourcing environment... "\n'
928	>	txt += '        if [ ! -s $VO_CMS_SW_DIR/cmsset_default.sh ] ;then\n'
929	>	txt += '            echo "ERROR ==> cmsset_default.sh file not found into dir $VO_CMS_SW_DIR"\n'
930	>	txt += '            job_exit_code=10020\n'
931	>	txt += '            func_exit\n'
932	>	txt += '        fi\n'
933	>	txt += '        echo "sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
934	>	txt += '        source $VO_CMS_SW_DIR/cmsset_default.sh\n'
935	>	txt += '        result=$?\n'
936	>	txt += '        if [ $result -ne 0 ]; then\n'
937	>	txt += '            echo "ERROR ==> problem sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
938	>	txt += '            job_exit_code=10032\n'
939	>	txt += '            func_exit\n'
940	>	txt += '        fi\n'
941	>	txt += '    fi\n'
942	>	txt += '    \n'
943	>	txt += '    echo "==> setup cms environment ok"\n'
944		return txt
945
946	<	### FEDE FOR DBS OUTPUT PUBLICATION
1390	<	def modifyReport(self, nj):
946	>	def wsModifyReport(self, nj):
947		"""
948	<	insert the part of the script that modifies the FrameworkJob Report
948	>	insert the part of the script that modifies the FrameworkJob Report
949		"""
950
951	<	txt = ''
952	<	try:
953	<	publish_data = int(self.cfg_params['USER.publish_data'])
954	<	except KeyError:
955	<	publish_data = 0
956	<	if (publish_data == 1):
1401	<	txt += 'echo "Modify Job Report" \n'
1402	<	#txt += 'chmod a+x $RUNTIME_AREA/'+self.version+'/ProdAgentApi/FwkJobRep/ModifyJobReport.py\n'
1403	<	################ FEDE FOR DBS2 #############################################
1404	<	txt += 'chmod a+x $SOFTWARE_DIR/ProdAgentApi/FwkJobRep/ModifyJobReport.py\n'
1405	<	#############################################################################
1406	<	#try:
1407	<	#    publish_data = int(self.cfg_params['USER.publish_data'])
1408	<	#except KeyError:
1409	<	#    publish_data = 0
1410	<
1411	<	txt += 'if [ -z "$SE" ]; then\n'
1412	<	txt += '    SE="" \n'
1413	<	txt += 'fi \n'
1414	<	txt += 'if [ -z "$SE_PATH" ]; then\n'
1415	<	txt += '    SE_PATH="" \n'
1416	<	txt += 'fi \n'
1417	<	txt += 'echo "SE = $SE"\n'
1418	<	txt += 'echo "SE_PATH = $SE_PATH"\n'
951	>	txt = ''
952	>	publish_data = int(self.cfg_params.get('USER.publish_data',0))
953	>	#if (publish_data == 1):
954	>	if (self.copy_data == 1):
955	>	txt = '\n#Written by cms_cmssw::wsModifyReport\n'
956	>	publish_data = int(self.cfg_params.get('USER.publish_data',0))
957
958	<	#if (publish_data == 1):
959	<	#processedDataset = self.cfg_params['USER.processed_datasetname']
960	<	processedDataset = self.cfg_params['USER.publish_data_name']
961	<	txt += 'ProcessedDataset='+processedDataset+'\n'
1424	<	#### LFN=/store/user/<user>/processedDataset_PSETHASH
1425	<	txt += 'if [ "$SE_PATH" == "" ]; then\n'
1426	<	#### FEDE: added slash in LFN ##############
958	>
959	>	txt += 'if [ $StageOutExitStatus -eq 0 ]; then\n'
960	>	txt += '    FOR_LFN=$LFNBaseName\n'
961	>	txt += 'else\n'
962		txt += '    FOR_LFN=/copy_problems/ \n'
963	<	txt += 'else \n'
964	<	txt += '    tmp=`echo $SE_PATH | awk -F \'store\' \'{print$2}\'` \n'
965	<	#####  FEDE TO BE CHANGED, BECAUSE STORE IS HARDCODED!!!! ########
966	<	txt += '    FOR_LFN=/store$tmp \n'
967	<	txt += 'fi \n'
968	<	txt += 'echo "ProcessedDataset = $ProcessedDataset"\n'
963	>	txt += 'fi\n'
964	>
965	>	txt += 'echo ">>> Modify Job Report:" \n'
966	>	txt += 'chmod a+x $RUNTIME_AREA/ProdCommon/FwkJobRep/ModifyJobReport.py\n'
967	>	txt += 'echo "SE = $SE"\n'
968	>	txt += 'echo "SE_PATH = $SE_PATH"\n'
969		txt += 'echo "FOR_LFN = $FOR_LFN" \n'
970		txt += 'echo "CMSSW_VERSION = $CMSSW_VERSION"\n\n'
971	<	#txt += 'echo "$RUNTIME_AREA/'+self.version+'/ProdAgentApi/FwkJobRep/ModifyJobReport.py crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH"\n'
972	<	txt += 'echo "$SOFTWARE_DIR/ProdAgentApi/FwkJobRep/ModifyJobReport.py crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH"\n'
973	<	txt += '$SOFTWARE_DIR/ProdAgentApi/FwkJobRep/ModifyJobReport.py crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH\n'
974	<	#txt += '$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'
975	<
971	>
972	>
973	>	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'
974	>	if (publish_data == 1):
975	>	processedDataset = self.cfg_params['USER.publish_data_name']
976	>	txt += 'ProcessedDataset='+processedDataset+'\n'
977	>	txt += 'echo "ProcessedDataset = $ProcessedDataset"\n'
978	>	args += ' UserProcessedDataset $USER-$ProcessedDataset-$PSETHASH'
979	>
980	>	txt += 'echo "$RUNTIME_AREA/ProdCommon/FwkJobRep/ModifyJobReport.py '+str(args)+'"\n'
981	>	txt += '$RUNTIME_AREA/ProdCommon/FwkJobRep/ModifyJobReport.py '+str(args)+'\n'
982		txt += 'modifyReport_result=$?\n'
1442	–	txt += 'echo modifyReport_result = $modifyReport_result\n'
983		txt += 'if [ $modifyReport_result -ne 0 ]; then\n'
984	<	txt += '    exit_status=1\n'
985	<	txt += '    echo "ERROR: Problem with ModifyJobReport"\n'
984	>	txt += '    modifyReport_result=70500\n'
985	>	txt += '    job_exit_code=$modifyReport_result\n'
986	>	txt += '    echo "ModifyReportResult=$modifyReport_result" | tee -a $RUNTIME_AREA/$repo\n'
987	>	txt += '    echo "WARNING: Problem with ModifyJobReport"\n'
988		txt += 'else\n'
989	<	txt += '    mv NewFrameworkJobReport.xml crab_fjr_$NJob.xml\n'
989	>	txt += '    mv NewFrameworkJobReport.xml $RUNTIME_AREA/crab_fjr_$NJob.xml\n'
990		txt += 'fi\n'
1449	–	else:
1450	–	txt += 'echo "no data publication required"\n'
1451	–	#txt += 'ProcessedDataset=no_data_to_publish \n'
1452	–	#### FEDE: added slash in LFN ##############
1453	–	#txt += 'FOR_LFN=/local/ \n'
1454	–	#txt += 'echo "ProcessedDataset = $ProcessedDataset"\n'
1455	–	#txt += 'echo "FOR_LFN = $FOR_LFN" \n'
991		return txt
992
993	<	def cleanEnv(self):
994	<	### OLI_DANIELE
995	<	txt = ''
996	<	txt += 'if [ $middleware == OSG ]; then\n'
997	<	txt += '    cd $RUNTIME_AREA\n'
998	<	txt += '    echo "Remove working directory: $WORKING_DIR"\n'
999	<	txt += '    /bin/rm -rf $WORKING_DIR\n'
1000	<	txt += '    if [ -d $WORKING_DIR ] ;then\n'
1001	<	txt += '              echo "SET_EXE 60999 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after cleanup of WN"\n'
1002	<	txt += '              echo "JOB_EXIT_STATUS = 60999"\n'
1003	<	txt += '              echo "JobExitCode=60999" | tee -a $RUNTIME_AREA/$repo\n'
1004	<	txt += '              dumpStatus $RUNTIME_AREA/$repo\n'
1005	<	txt += '        rm -f $RUNTIME_AREA/$repo \n'
1006	<	txt += '        echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1007	<	txt += '        echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
993	>	def wsParseFJR(self):
994	>	"""
995	>	Parse the FrameworkJobReport to obtain useful infos
996	>	"""
997	>	txt = '\n#Written by cms_cmssw::wsParseFJR\n'
998	>	txt += 'echo ">>> Parse FrameworkJobReport crab_fjr.xml"\n'
999	>	txt += 'if [ -s $RUNTIME_AREA/crab_fjr_$NJob.xml ]; then\n'
1000	>	txt += '    if [ -s $RUNTIME_AREA/parseCrabFjr.py ]; then\n'
1001	>	txt += '        cmd_out=`python $RUNTIME_AREA/parseCrabFjr.py --input $RUNTIME_AREA/crab_fjr_$NJob.xml --dashboard $MonitorID,$MonitorJobID '+self.debugWrap+'`\n'
1002	>	if self.debug_wrapper==1 :
1003	>	txt += '        echo "Result of parsing the FrameworkJobReport crab_fjr.xml: $cmd_out"\n'
1004	>	txt += '        executable_exit_status=`python $RUNTIME_AREA/parseCrabFjr.py --input $RUNTIME_AREA/crab_fjr_$NJob.xml --exitcode`\n'
1005	>	txt += '        if [ $executable_exit_status -eq 50115 ];then\n'
1006	>	txt += '            echo ">>> crab_fjr.xml contents: "\n'
1007	>	txt += '            cat $RUNTIME_AREA/crab_fjr_$NJob.xml\n'
1008	>	txt += '            echo "Wrong FrameworkJobReport --> does not contain useful info. ExitStatus: $executable_exit_status"\n'
1009	>	txt += '        elif [ $executable_exit_status -eq -999 ];then\n'
1010	>	txt += '            echo "ExitStatus from FrameworkJobReport not available. not available. Using exit code of executable from command line."\n'
1011	>	txt += '        else\n'
1012	>	txt += '            echo "Extracted ExitStatus from FrameworkJobReport parsing output: $executable_exit_status"\n'
1013	>	txt += '        fi\n'
1014	>	txt += '    else\n'
1015	>	txt += '        echo "CRAB python script to parse CRAB FrameworkJobReport crab_fjr.xml is not available, using exit code of executable from command line."\n'
1016		txt += '    fi\n'
1017	+	#### Patch to check input data reading for CMSSW16x Hopefully we-ll remove it asap
1018	+	txt += '    if [ $executable_exit_status -eq 0 ];then\n'
1019	+	txt += '        echo ">>> Executable succeded  $executable_exit_status"\n'
1020	+	## This cannot more work given the changes on the Job argumentsJob
1021	+	"""
1022	+	if (self.datasetPath and not (self.dataset_pu or self.useParent==1)) :
1023	+	# VERIFY PROCESSED DATA
1024	+	txt += '        echo ">>> Verify list of processed files:"\n'
1025	+	txt += '        echo $InputFiles |tr -d \'\\\\\' |tr \',\' \'\\n\'|tr -d \'"\' > input-files.txt\n'
1026	+	txt += '        python $RUNTIME_AREA/parseCrabFjr.py --input $RUNTIME_AREA/crab_fjr_$NJob.xml --lfn > processed-files.txt\n'
1027	+	txt += '        cat input-files.txt  | sort | uniq > tmp.txt\n'
1028	+	txt += '        mv tmp.txt input-files.txt\n'
1029	+	txt += '        echo "cat input-files.txt"\n'
1030	+	txt += '        echo "----------------------"\n'
1031	+	txt += '        cat input-files.txt\n'
1032	+	txt += '        cat processed-files.txt | sort | uniq > tmp.txt\n'
1033	+	txt += '        mv tmp.txt processed-files.txt\n'
1034	+	txt += '        echo "----------------------"\n'
1035	+	txt += '        echo "cat processed-files.txt"\n'
1036	+	txt += '        echo "----------------------"\n'
1037	+	txt += '        cat processed-files.txt\n'
1038	+	txt += '        echo "----------------------"\n'
1039	+	txt += '        diff -qbB input-files.txt processed-files.txt\n'
1040	+	txt += '        fileverify_status=$?\n'
1041	+	txt += '        if [ $fileverify_status -ne 0 ]; then\n'
1042	+	txt += '            executable_exit_status=30001\n'
1043	+	txt += '            echo "ERROR ==> not all input files processed"\n'
1044	+	txt += '            echo "      ==> list of processed files from crab_fjr.xml differs from list in pset.cfg"\n'
1045	+	txt += '            echo "      ==> diff input-files.txt processed-files.txt"\n'
1046	+	txt += '        fi\n'
1047	+	"""
1048	+	txt += '    fi\n'
1049	+	txt += 'else\n'
1050	+	txt += '    echo "CRAB FrameworkJobReport crab_fjr.xml is not available, using exit code of executable from command line."\n'
1051		txt += 'fi\n'
1052		txt += '\n'
1053	+	txt += 'if [ $executable_exit_status -ne 0 ] && [ $executable_exit_status -ne 50115 ] && [ $executable_exit_status -ne 50117 ] && [ $executable_exit_status -ne 30001 ];then\n'
1054	+	txt += '    echo ">>> Executable failed  $executable_exit_status"\n'
1055	+	txt += '    echo "ExeExitCode=$executable_exit_status" | tee -a $RUNTIME_AREA/$repo\n'
1056	+	txt += '    echo "EXECUTABLE_EXIT_STATUS = $executable_exit_status"\n'
1057	+	txt += '    job_exit_code=$executable_exit_status\n'
1058	+	txt += '    func_exit\n'
1059	+	txt += 'fi\n\n'
1060	+	txt += 'echo "ExeExitCode=$executable_exit_status" | tee -a $RUNTIME_AREA/$repo\n'
1061	+	txt += 'echo "EXECUTABLE_EXIT_STATUS = $executable_exit_status"\n'
1062	+	txt += 'job_exit_code=$executable_exit_status\n'
1063	+
1064		return txt
1065
1066		def setParam_(self, param, value):
#	Line 1481 | Line 1069 | class Cmssw(JobType):
1069		def getParams(self):
1070		return self._params
1071
1072	<	def setTaskid_(self):
1485	<	self._taskId = self.cfg_params['taskId']
1486	<
1487	<	def getTaskid(self):
1488	<	return self._taskId
1489	<
1490	<	def uniquelist(self, old):
1491	<	"""
1492	<	remove duplicates from a list
1493	<	"""
1494	<	nd={}
1495	<	for e in old:
1496	<	nd[e]=0
1497	<	return nd.keys()
1498	<
1499	<
1500	<	def checkOut(self, limit):
1072	>	def outList(self,list=False):
1073		"""
1074		check the dimension of the output files
1075		"""
1076	<	txt = 'echo "*****************************************"\n'
1077	<	txt += 'echo "** Starting output sandbox limit check **"\n'
1506	<	txt += 'echo "*****************************************"\n'
1507	<	allOutFiles = ""
1076	>	txt = ''
1077	>	txt += 'echo ">>> list of expected files on output sandbox"\n'
1078		listOutFiles = []
1079	<	for fileOut in (self.output_file+self.output_file_sandbox):
1080	<	if fileOut.find('crab_fjr') == -1:
1081	<	allOutFiles = allOutFiles + " " + self.numberFile_(fileOut, '$NJob')
1082	<	listOutFiles.append(self.numberFile_(fileOut, '$NJob'))
1083	<	txt += 'echo "OUTPUT files: '+str(allOutFiles)+'";\n'
1084	<	txt += 'ls -gGhrta;\n'
1085	<	txt += 'sum=0;\n'
1086	<	txt += 'for file in '+str(allOutFiles)+' ; do\n'
1087	<	txt += '    if [ -e $file ]; then\n'
1088	<	txt += '        tt=`ls -gGrta $file | awk \'{ print $3 }\'`\n'
1089	<	txt += '        sum=`expr $sum + $tt`\n'
1090	<	txt += '    else\n'
1091	<	txt += '        echo "WARNING: output file $file not found!"\n'
1092	<	txt += '    fi\n'
1093	<	txt += 'done\n'
1094	<	txt += 'echo "Total Output dimension: $sum";\n'
1095	<	txt += 'limit='+str(limit)+';\n'
1096	<	txt += 'echo "OUTPUT FILES LIMIT SET TO: $limit";\n'
1097	<	txt += 'if [ $limit -lt $sum ]; then\n'
1098	<	txt += '    echo "WARNING: output files have to big size - something will be lost;"\n'
1099	<	txt += '    echo "         checking the output file sizes..."\n'
1530	<	"""
1531	<	txt += '    dim=0;\n'
1532	<	txt += '    exclude=0;\n'
1533	<	txt += '    for files in '+str(allOutFiles)+' ; do\n'
1534	<	txt += '        sumTemp=0;\n'
1535	<	txt += '        for file2 in '+str(allOutFiles)+' ; do\n'
1536	<	txt += '            if [ $file != $file2 ]; then\n'
1537	<	txt += '                tt=`ls -gGrta $file2 | awk \'{ print $3 }\';`\n'
1538	<	txt += '                sumTemp=`expr $sumTemp + $tt`;\n'
1539	<	txt += '            fi\n'
1540	<	txt += '        done\n'
1541	<	txt += '        if [ $sumTemp -lt $limit ]; then\n'
1542	<	txt += '            if [ $dim -lt $sumTemp ]; then\n'
1543	<	txt += '                dim=$sumTemp;\n'
1544	<	txt += '                exclude=$file;\n'
1545	<	txt += '            fi\n'
1546	<	txt += '        fi\n'
1547	<	txt += '    done\n'
1548	<	txt += '    echo "Dimension calculated: $dim"; echo "File to exclude: $exclude";\n'
1549	<	"""
1550	<	txt += '    tot=0;\n'
1551	<	txt += '    for file2 in '+str(allOutFiles)+' ; do\n'
1552	<	txt += '        tt=`ls -gGrta $file2 | awk \'{ print $3 }\';`\n'
1553	<	txt += '        tot=`expr $tot + $tt`;\n'
1554	<	txt += '        if [ $limit -lt $tot ]; then\n'
1555	<	txt += '            tot=`expr $tot - $tt`;\n'
1556	<	txt += '            fileLast=$file;\n'
1557	<	txt += '            break;\n'
1558	<	txt += '        fi\n'
1559	<	txt += '    done\n'
1560	<	txt += '    echo "Dimension calculated: $tot"; echo "First file to exclude: $file";\n'
1561	<	txt += '    flag=0;\n'
1562	<	txt += '    for filess in '+str(allOutFiles)+' ; do\n'
1563	<	txt += '        if [ $fileLast = $filess ]; then\n'
1564	<	txt += '            flag=1;\n'
1565	<	txt += '        fi\n'
1566	<	txt += '        if [ $flag -eq 1 ]; then\n'
1567	<	txt += '            rm -f $filess;\n'
1568	<	txt += '        fi\n'
1569	<	txt += '    done\n'
1570	<	txt += '    ls -agGhrt;\n'
1571	<	txt += '    echo "WARNING: output files are too big in dimension: can not put in the output_sandbox.";\n'
1572	<	txt += '    echo "JOB_EXIT_STATUS = 70000";\n'
1573	<	txt += '    exit_status=70000;\n'
1574	<	txt += 'else'
1575	<	txt += '    echo "Total Output dimension $sum is fine.";\n'
1576	<	txt += 'fi\n'
1577	<	txt += 'echo "*****************************************"\n'
1578	<	txt += 'echo "*** Ending output sandbox limit check ***"\n'
1579	<	txt += 'echo "*****************************************"\n'
1079	>	stdout = 'CMSSW_$NJob.stdout'
1080	>	stderr = 'CMSSW_$NJob.stderr'
1081	>	if len(self.output_file) <= 0:
1082	>	msg ="WARNING: no output files name have been defined!!\n"
1083	>	msg+="\tno output files will be reported back/staged\n"
1084	>	common.logger.message(msg)
1085	>	if (self.return_data == 1):
1086	>	for file in (self.output_file+self.output_file_sandbox):
1087	>	listOutFiles.append(numberFile(file, '$NJob'))
1088	>	listOutFiles.append(stdout)
1089	>	listOutFiles.append(stderr)
1090	>	else:
1091	>	for file in (self.output_file_sandbox):
1092	>	listOutFiles.append(numberFile(file, '$NJob'))
1093	>	listOutFiles.append(stdout)
1094	>	listOutFiles.append(stderr)
1095	>	txt += 'echo "output files: '+string.join(listOutFiles,' ')+'"\n'
1096	>	txt += 'filesToCheck="'+string.join(listOutFiles,' ')+'"\n'
1097	>	txt += 'export filesToCheck\n'
1098	>
1099	>	if list : return self.output_file
1100		return txt

Diff Legend

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