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 1.92 by gutsche, Tue Jun 19 16:21:49 2007 UTC vs.
Revision 1.365 by spiga, Tue Nov 9 21:10:07 2010 UTC

#	Line 1 | Line 1
1	+
2	+	__revision__ = "$Id$"
3	+	__version__ = "$Revision$"
4	+
5		from JobType import JobType
2	–	from crab_logger import Logger
6		from crab_exceptions import *
7		from crab_util import *
8		import common
9	+	import re
10		import Scram
11	<
12	<	import os, string, re, shutil, glob
11	>	from Splitter import JobSplitter
12	>	from Downloader import Downloader
13	>	try:
14	>	import json
15	>	except:
16	>	import simplejson as json
17	>
18	>	from IMProv.IMProvNode import IMProvNode
19	>	from IMProv.IMProvLoader import loadIMProvFile
20	>	import os, string, glob
21	>	from xml.dom import pulldom
22
23		class Cmssw(JobType):
24	<	def __init__(self, cfg_params, ncjobs):
24	>	def __init__(self, cfg_params, ncjobs,skip_blocks, isNew):
25		JobType.__init__(self, 'CMSSW')
26	<	common.logger.debug(3,'CMSSW::__init__')
27	<
28	<	# Marco.
26	>	common.logger.debug('CMSSW::__init__')
27	>	self.skip_blocks = skip_blocks
28	>	self.argsList = 2
29	>	self.NumEvents=0
30		self._params = {}
31		self.cfg_params = cfg_params
32	+	### FEDE FOR MULTI ###
33	+	self.var_filter=''
34
35	<	try:
36	<	self.MaxTarBallSize = float(self.cfg_params['EDG.maxtarballsize'])
37	<	except KeyError:
38	<	self.MaxTarBallSize = 9.5
35	>	### Temporary patch to automatically skip the ISB size check:
36	>	self.server = self.cfg_params.get('CRAB.server_name',None) or \
37	>	self.cfg_params.get('CRAB.use_server',0)
38	>	self.local  = common.scheduler.name().upper() in ['LSF','CAF','CONDOR','SGE','PBS']
39	>	size = 9.5
40	>	if self.server :
41	>	size = 1000
42	>	elif self.local:
43	>	size = 9999999
44	>	self.MaxTarBallSize = float(self.cfg_params.get('GRID.maxtarballsize',size))
45
46		# number of jobs requested to be created, limit obj splitting
47		self.ncjobs = ncjobs
48
27	–	log = common.logger
28	–
49		self.scram = Scram.Scram(cfg_params)
50		self.additional_inbox_files = []
51		self.scriptExe = ''
#	Line 33 | Line 53 | class Cmssw(JobType):
53		self.executable_arch = self.scram.getArch()
54		self.tgz_name = 'default.tgz'
55		self.scriptName = 'CMSSW.sh'
56	<	self.pset = ''      #scrip use case Da
57	<	self.datasetPath = '' #scrip use case Da
56	>	self.pset = ''
57	>	self.datasetPath = ''
58
59	+	self.tgzNameWithPath = common.work_space.pathForTgz()+self.tgz_name
60		# set FJR file name
61		self.fjrFileName = 'crab_fjr.xml'
62
63		self.version = self.scram.getSWVersion()
64	<	common.taskDB.setDict('codeVersion',self.version)
65	<	self.setParam_('application', self.version)
64	>	common.logger.log(10-1,"CMSSW version is: "+str(self.version))
65	>	version_array = self.version.split('_')
66	>	self.CMSSW_major = 0
67	>	self.CMSSW_minor = 0
68	>	self.CMSSW_patch = 0
69	>	try:
70	>	self.CMSSW_major = int(version_array[1])
71	>	self.CMSSW_minor = int(version_array[2])
72	>	self.CMSSW_patch = int(version_array[3])
73	>	except:
74	>	msg = "Cannot parse CMSSW version string: " + self.version + " for major and minor release number!"
75	>	raise CrabException(msg)
76
77	+	if self.CMSSW_major < 2 or (self.CMSSW_major == 2 and self.CMSSW_minor < 1):
78	+	msg = "CRAB supports CMSSW >= 2_1_x only. Use an older CRAB version."
79	+	raise CrabException(msg)
80	+	"""
81	+	As CMSSW versions are dropped we can drop more code:
82	+	2.x dropped: drop check for lumi range setting
83	+	"""
84	+	self.checkCMSSWVersion()
85		### collect Data cards
86
87	<	## get DBS mode
88	<	try:
50	<	self.use_dbs_1 = int(self.cfg_params['CMSSW.use_dbs_1'])
51	<	except KeyError:
52	<	self.use_dbs_1 = 0
53	<
54	<	try:
55	<	tmp =  cfg_params['CMSSW.datasetpath']
56	<	log.debug(6, "CMSSW::CMSSW(): datasetPath = "+tmp)
57	<	if string.lower(tmp)=='none':
58	<	self.datasetPath = None
59	<	self.selectNoInput = 1
60	<	else:
61	<	self.datasetPath = tmp
62	<	self.selectNoInput = 0
63	<	except KeyError:
64	<	msg = "Error: datasetpath not defined "
65	<	raise CrabException(msg)
87	>	### Temporary: added to remove input file control in the case of PU
88	>	self.dataset_pu = cfg_params.get('CMSSW.dataset_pu', None)
89
90	<	# ML monitoring
91	<	# split dataset path style: /PreProdR3Minbias/SIM/GEN-SIM
92	<	if not self.datasetPath:
93	<	self.setParam_('dataset', 'None')
94	<	self.setParam_('owner', 'None')
90	>	tmp =  cfg_params['CMSSW.datasetpath']
91	>	common.logger.log(10-1, "CMSSW::CMSSW(): datasetPath = "+tmp)
92	>
93	>	if tmp =='':
94	>	msg = "Error: datasetpath not defined "
95	>	raise CrabException(msg)
96	>	elif string.lower(tmp)=='none':
97	>	self.datasetPath = None
98	>	self.selectNoInput = 1
99	>	self.primaryDataset = 'null'
100		else:
101	<	try:
102	<	datasetpath_split = self.datasetPath.split("/")
103	<	# standard style
104	<	if self.use_dbs_1 == 1 :
105	<	self.setParam_('dataset', datasetpath_split[1])
106	<	self.setParam_('owner', datasetpath_split[-1])
107	<	else:
108	<	self.setParam_('dataset', datasetpath_split[1])
109	<	self.setParam_('owner', datasetpath_split[2])
82	<	except:
83	<	self.setParam_('dataset', self.datasetPath)
84	<	self.setParam_('owner', self.datasetPath)
85	<
86	<	self.setTaskid_()
87	<	self.setParam_('taskId', self.cfg_params['taskId'])
101	>	self.datasetPath = tmp
102	>	self.selectNoInput = 0
103	>	ll = len(self.datasetPath.split("/"))
104	>	if (ll < 4):
105	>	msg = 'Your datasetpath has a invalid format ' + self.datasetPath + '\n'
106	>	msg += 'Expected a path in format /PRIMARY/PROCESSED/TIER1-TIER2 or /PRIMARY/PROCESSED/TIER/METHOD for ADS'
107	>	raise CrabException(msg)
108	>	self.primaryDataset = self.datasetPath.split("/")[1]
109	>	self.dataTier = self.datasetPath.split("/")[2]
110
111	<	self.dataTiers = []
111	>	# Analysis dataset is primary/processed/tier/definition
112	>	self.ads = False
113	>	if self.datasetPath:
114	>	self.ads = len(self.datasetPath.split("/")) > 4
115	>	self.lumiMask = self.cfg_params.get('CMSSW.lumi_mask',None)
116	>	self.lumiParams = self.cfg_params.get('CMSSW.total_number_of_lumis',None) or \
117	>	self.cfg_params.get('CMSSW.lumis_per_job',None)
118	>
119	>	# FUTURE: Can remove this check
120	>	if self.ads and self.CMSSW_major < 3:
121	>	common.logger.info('Warning: Analysis dataset support is incomplete in CMSSW 2_x.')
122	>	common.logger.info('  Only file level, not lumi level, granularity is supported.')
123	>
124	>	self.debugWrap=''
125	>	self.debug_wrapper = int(cfg_params.get('USER.debug_wrapper',0))
126	>	if self.debug_wrapper == 1: self.debugWrap='--debug'
127
128		## now the application
129	<	try:
130	<	self.executable = cfg_params['CMSSW.executable']
131	<	self.setParam_('exe', self.executable)
132	<	log.debug(6, "CMSSW::CMSSW(): executable = "+self.executable)
96	<	msg = "Default executable cmsRun overridden. Switch to " + self.executable
97	<	log.debug(3,msg)
98	<	except KeyError:
99	<	self.executable = 'cmsRun'
100	<	self.setParam_('exe', self.executable)
101	<	msg = "User executable not defined. Use cmsRun"
102	<	log.debug(3,msg)
103	<	pass
129	>	self.managedGenerators = ['madgraph', 'comphep', 'lhe']
130	>	self.generator = cfg_params.get('CMSSW.generator','pythia').lower()
131	>	self.executable = cfg_params.get('CMSSW.executable','cmsRun')
132	>	common.logger.log(10-1, "CMSSW::CMSSW(): executable = "+self.executable)
133
134	<	try:
106	<	self.pset = cfg_params['CMSSW.pset']
107	<	log.debug(6, "Cmssw::Cmssw(): PSet file = "+self.pset)
108	<	if self.pset.lower() != 'none' :
109	<	if (not os.path.exists(self.pset)):
110	<	raise CrabException("User defined PSet file "+self.pset+" does not exist")
111	<	else:
112	<	self.pset = None
113	<	except KeyError:
134	>	if not cfg_params.has_key('CMSSW.pset'):
135		raise CrabException("PSet file missing. Cannot run cmsRun ")
136	+	self.pset = cfg_params['CMSSW.pset']
137	+	common.logger.log(10-1, "Cmssw::Cmssw(): PSet file = "+self.pset)
138	+	if self.pset.lower() != 'none' :
139	+	if (not os.path.exists(self.pset)):
140	+	raise CrabException("User defined PSet file "+self.pset+" does not exist")
141	+	else:
142	+	self.pset = None
143
144		# output files
145		## stuff which must be returned always via sandbox
#	Line 121 | Line 149 | class Cmssw(JobType):
149		self.output_file_sandbox.append(self.fjrFileName)
150
151		# other output files to be returned via sandbox or copied to SE
152	<	try:
153	<	self.output_file = []
154	<	tmp = cfg_params['CMSSW.output_file']
155	<	if tmp != '':
156	<	tmpOutFiles = string.split(cfg_params['CMSSW.output_file'],',')
157	<	log.debug(7, 'cmssw::cmssw(): output files '+str(tmpOutFiles))
158	<	for tmp in tmpOutFiles:
159	<	tmp=string.strip(tmp)
160	<	self.output_file.append(tmp)
161	<	pass
162	<	else:
163	<	log.message("No output file defined: only stdout/err and the CRAB Framework Job Report will be available\n")
164	<	pass
137	<	pass
138	<	except KeyError:
139	<	log.message("No output file defined: only stdout/err and the CRAB Framework Job Report will be available\n")
140	<	pass
152	>	outfileflag = False
153	>	self.output_file = []
154	>	tmp = cfg_params.get('CMSSW.output_file',None)
155	>	if tmp :
156	>	self.output_file = [x.strip() for x in tmp.split(',')]
157	>	outfileflag = True #output found
158	>
159	>	self.scriptExe = cfg_params.get('USER.script_exe',None)
160	>	if self.scriptExe :
161	>	if not os.path.isfile(self.scriptExe):
162	>	msg ="ERROR. file "+self.scriptExe+" not found"
163	>	raise CrabException(msg)
164	>	self.additional_inbox_files.append(string.strip(self.scriptExe))
165
166	<	# script_exe file as additional file in inputSandbox
167	<	try:
144	<	self.scriptExe = cfg_params['USER.script_exe']
145	<	if self.scriptExe != '':
146	<	if not os.path.isfile(self.scriptExe):
147	<	msg ="ERROR. file "+self.scriptExe+" not found"
148	<	raise CrabException(msg)
149	<	self.additional_inbox_files.append(string.strip(self.scriptExe))
150	<	except KeyError:
151	<	self.scriptExe = ''
166	>	self.AdditionalArgs = cfg_params.get('USER.script_arguments',None)
167	>	if self.AdditionalArgs : self.AdditionalArgs = string.replace(self.AdditionalArgs,',',' ')
168
153	–	#CarlosDaniele
169		if self.datasetPath == None and self.pset == None and self.scriptExe == '' :
170	<	msg ="Error. script_exe  not defined"
171	<	raise CrabException(msg)
170	>	msg ="Error. script_exe  not defined"
171	>	raise CrabException(msg)
172	>
173	>	# use parent files...
174	>	self.useParent = int(self.cfg_params.get('CMSSW.use_parent',0))
175
176		## additional input files
177	<	try:
177	>	if cfg_params.has_key('USER.additional_input_files'):
178		tmpAddFiles = string.split(cfg_params['USER.additional_input_files'],',')
179		for tmp in tmpAddFiles:
180		tmp = string.strip(tmp)
#	Line 173 | Line 191 | class Cmssw(JobType):
191		if not os.path.exists(file):
192		raise CrabException("Additional input file not found: "+file)
193		pass
194	<	fname = string.split(file, '/')[-1]
177	<	storedFile = common.work_space.pathForTgz()+'share/'+fname
178	<	shutil.copyfile(file, storedFile)
179	<	self.additional_inbox_files.append(string.strip(storedFile))
194	>	self.additional_inbox_files.append(string.strip(file))
195		pass
196		pass
197	<	common.logger.debug(5,"Additional input files: "+str(self.additional_inbox_files))
198	<	except KeyError:
184	<	pass
185	<
186	<	# files per job
187	<	try:
188	<	if (cfg_params['CMSSW.files_per_jobs']):
189	<	raise CrabException("files_per_jobs no longer supported.  Quitting.")
190	<	except KeyError:
191	<	pass
192	<
193	<	## Events per job
194	<	try:
195	<	self.eventsPerJob =int( cfg_params['CMSSW.events_per_job'])
196	<	self.selectEventsPerJob = 1
197	<	except KeyError:
198	<	self.eventsPerJob = -1
199	<	self.selectEventsPerJob = 0
200	<
201	<	## number of jobs
202	<	try:
203	<	self.theNumberOfJobs =int( cfg_params['CMSSW.number_of_jobs'])
204	<	self.selectNumberOfJobs = 1
205	<	except KeyError:
206	<	self.theNumberOfJobs = 0
207	<	self.selectNumberOfJobs = 0
208	<
209	<	try:
210	<	self.total_number_of_events = int(cfg_params['CMSSW.total_number_of_events'])
211	<	self.selectTotalNumberEvents = 1
212	<	except KeyError:
213	<	self.total_number_of_events = 0
214	<	self.selectTotalNumberEvents = 0
215	<
216	<	if self.pset != None: #CarlosDaniele
217	<	if ( (self.selectTotalNumberEvents + self.selectEventsPerJob + self.selectNumberOfJobs) != 2 ):
218	<	msg = 'Must define exactly two of total_number_of_events, events_per_job, or number_of_jobs.'
219	<	raise CrabException(msg)
220	<	else:
221	<	if (self.selectNumberOfJobs == 0):
222	<	msg = 'Must specify  number_of_jobs.'
223	<	raise CrabException(msg)
224	<
225	<	## source seed for pythia
226	<	try:
227	<	self.sourceSeed = int(cfg_params['CMSSW.pythia_seed'])
228	<	except KeyError:
229	<	self.sourceSeed = None
230	<	common.logger.debug(5,"No seed given")
231	<
232	<	try:
233	<	self.sourceSeedVtx = int(cfg_params['CMSSW.vtx_seed'])
234	<	except KeyError:
235	<	self.sourceSeedVtx = None
236	<	common.logger.debug(5,"No vertex seed given")
237	<
238	<	try:
239	<	self.sourceSeedG4 = int(cfg_params['CMSSW.g4_seed'])
240	<	except KeyError:
241	<	self.sourceSeedG4 = None
242	<	common.logger.debug(5,"No g4 sim hits seed given")
197	>	common.logger.debug("Additional input files: "+str(self.additional_inbox_files))
198	>	pass
199
244	–	try:
245	–	self.sourceSeedMix = int(cfg_params['CMSSW.mix_seed'])
246	–	except KeyError:
247	–	self.sourceSeedMix = None
248	–	common.logger.debug(5,"No mix seed given")
200
201	<	try:
202	<	self.firstRun = int(cfg_params['CMSSW.first_run'])
203	<	except KeyError:
204	<	self.firstRun = None
205	<	common.logger.debug(5,"No first run given")
206	<	if self.pset != None: #CarlosDaniele
207	<	import PsetManipulator
208	<	PsetEdit = PsetManipulator.PsetManipulator(self.pset) #Daniele Pset
201	>	## New method of dealing with seeds
202	>	self.incrementSeeds = []
203	>	self.preserveSeeds = []
204	>	if cfg_params.has_key('CMSSW.preserve_seeds'):
205	>	tmpList = cfg_params['CMSSW.preserve_seeds'].split(',')
206	>	for tmp in tmpList:
207	>	tmp.strip()
208	>	self.preserveSeeds.append(tmp)
209	>	if cfg_params.has_key('CMSSW.increment_seeds'):
210	>	tmpList = cfg_params['CMSSW.increment_seeds'].split(',')
211	>	for tmp in tmpList:
212	>	tmp.strip()
213	>	self.incrementSeeds.append(tmp)
214	>
215	>	# Copy/return/publish
216	>	self.copy_data = int(cfg_params.get('USER.copy_data',0))
217	>	self.return_data = int(cfg_params.get('USER.return_data',0))
218	>	self.publish_data = int(cfg_params.get('USER.publish_data',0))
219	>	if (self.publish_data == 1):
220	>	if not cfg_params.has_key('USER.publish_data_name'):
221	>	raise CrabException('Cannot publish output data, because you did not specify USER.publish_data_name parameter in the crab.cfg file')
222	>	else:
223	>	self.processedDataset = cfg_params['USER.publish_data_name']
224
225	+	self.conf = {}
226	+	self.conf['pubdata'] = None
227	+	# number of jobs requested to be created, limit obj splitting DD
228		#DBSDLS-start
229	<	## Initialize the variables that are extracted from DBS/DLS and needed in other places of the code
229	>	## Initialize the variables that are extracted from DBS/DLS and needed in other places of the code
230		self.maxEvents=0  # max events available   ( --> check the requested nb. of evts in Creator.py)
231		self.DBSPaths={}  # all dbs paths requested ( --> input to the site local discovery script)
232		self.jobDestination=[]  # Site destination(s) for each job (list of lists)
233		## Perform the data location and discovery (based on DBS/DLS)
234		## SL: Don't if NONE is specified as input (pythia use case)
235		blockSites = {}
236	<	if self.datasetPath:
237	<	blockSites = self.DataDiscoveryAndLocation(cfg_params)
238	<	#DBSDLS-end
236	>	#wmbs
237	>	self.automation = int(self.cfg_params.get('WMBS.automation',0))
238	>	if self.automation == 0:
239	>	if self.datasetPath:
240	>	blockSites = self.DataDiscoveryAndLocation(cfg_params)
241	>	#DBSDLS-end
242	>	self.conf['blockSites']=blockSites
243	>
244	>	## Select Splitting
245	>	splitByRun = int(cfg_params.get('CMSSW.split_by_run',0))
246	>
247	>	if self.selectNoInput:
248	>	if self.pset == None:
249	>	self.algo = 'ForScript'
250	>	else:
251	>	self.algo = 'NoInput'
252	>	self.conf['managedGenerators']=self.managedGenerators
253	>	self.conf['generator']=self.generator
254	>	elif self.ads or self.lumiMask or self.lumiParams:
255	>	self.algo = 'LumiBased'
256	>	if splitByRun:
257	>	msg = "Cannot combine split by run with lumi_mask, ADS, " \
258	>	"or lumis_per_job. Use split by lumi mode instead."
259	>	raise CrabException(msg)
260
261	<	self.tgzNameWithPath = self.getTarBall(self.executable)
262	<
273	<	## Select Splitting
274	<	if self.selectNoInput:
275	<	if self.pset == None: #CarlosDaniele
276	<	self.jobSplittingForScript()
261	>	elif splitByRun ==1:
262	>	self.algo = 'RunBased'
263		else:
264	<	self.jobSplittingNoInput()
265	<	else:
280	<	self.jobSplittingByBlocks(blockSites)
264	>	self.algo = 'EventBased'
265	>	common.logger.debug("Job splitting method: %s" % self.algo)
266
267	<	# modify Pset
268	<	if self.pset != None: #CarlosDaniele
269	<	try:
270	<	if (self.datasetPath): # standard job
271	<	# allow to processa a fraction of events in a file
272	<	PsetEdit.inputModule("INPUT")
273	<	PsetEdit.maxEvent("INPUTMAXEVENTS")
274	<	PsetEdit.skipEvent("INPUTSKIPEVENTS")
275	<	else:  # pythia like job
276	<	PsetEdit.maxEvent(self.eventsPerJob)
277	<	if (self.firstRun):
278	<	PsetEdit.pythiaFirstRun("INPUTFIRSTRUN")  #First Run
279	<	if (self.sourceSeed) :
280	<	PsetEdit.pythiaSeed("INPUT")
281	<	if (self.sourceSeedVtx) :
282	<	PsetEdit.vtxSeed("INPUTVTX")
283	<	if (self.sourceSeedG4) :
284	<	self.PsetEdit.g4Seed("INPUTG4")
285	<	if (self.sourceSeedMix) :
286	<	self.PsetEdit.mixSeed("INPUTMIX")
287	<	# add FrameworkJobReport to parameter-set
288	<	PsetEdit.addCrabFJR(self.fjrFileName)
289	<	PsetEdit.psetWriter(self.configFilename())
290	<	except:
291	<	msg='Error while manipuliating ParameterSet: exiting...'
267	>	splitter = JobSplitter(self.cfg_params,self.conf)
268	>	self.dict = splitter.Algos()[self.algo]()
269	>
270	>	self.argsFile= '%s/arguments.xml'%common.work_space.shareDir()
271	>	self.rootArgsFilename= 'arguments'
272	>	# modify Pset only the first time
273	>	if isNew:
274	>	if self.pset != None: self.ModifyPset()
275	>
276	>	## Prepare inputSandbox TarBall (only the first time)
277	>	self.tarNameWithPath = self.getTarBall(self.executable)
278	>
279	>
280	>	def ModifyPset(self):
281	>	import PsetManipulator as pp
282	>
283	>	# If pycfg_params set, fake out the config script
284	>	# to make it think it was called with those args
285	>	pycfg_params = self.cfg_params.get('CMSSW.pycfg_params',None)
286	>	if pycfg_params:
287	>	trueArgv = sys.argv
288	>	sys.argv = [self.pset]
289	>	sys.argv.extend(pycfg_params.split(' '))
290	>	PsetEdit = pp.PsetManipulator(self.pset)
291	>	if pycfg_params: # Restore original sys.argv
292	>	sys.argv = trueArgv
293	>
294	>	try:
295	>	# Add FrameworkJobReport to parameter-set, set max events.
296	>	# Reset later for data jobs by writeCFG which does all modifications
297	>	PsetEdit.maxEvent(1)
298	>	PsetEdit.skipEvent(0)
299	>	PsetEdit.psetWriter(self.configFilename())
300	>	## If present, add TFileService to output files
301	>	if not int(self.cfg_params.get('CMSSW.skip_tfileservice_output',0)):
302	>	tfsOutput = PsetEdit.getTFileService()
303	>	if tfsOutput:
304	>	if tfsOutput in self.output_file:
305	>	common.logger.debug("Output from TFileService "+tfsOutput+" already in output files")
306	>	else:
307	>	outfileflag = True #output found
308	>	self.output_file.append(tfsOutput)
309	>	common.logger.info("Adding "+tfsOutput+" (from TFileService) to list of output files")
310	>	pass
311	>	pass
312	>
313	>	# If requested, add PoolOutputModule to output files
314	>	### FEDE FOR MULTI ###
315	>	#edmOutput = PsetEdit.getPoolOutputModule()
316	>	edmOutputDict = PsetEdit.getPoolOutputModule()
317	>	common.logger.debug("(test) edmOutputDict = "+str(edmOutputDict))
318	>	filter_dict = {}
319	>	for key in edmOutputDict.keys():
320	>	filter_dict[key]=edmOutputDict[key]['dataset']
321	>	common.logger.debug("(test) filter_dict for multi =  "+str(filter_dict))
322	>
323	>	#### in CMSSW.sh: export var_filter
324	>
325	>	self.var_filter = json.dumps(filter_dict)
326	>	common.logger.debug("(test) var_filter for multi =  "+self.var_filter)
327	>
328	>	edmOutput = edmOutputDict.keys()
329	>	if int(self.cfg_params.get('CMSSW.get_edm_output',0)):
330	>	if edmOutput:
331	>	for outputFile in edmOutput:
332	>	if outputFile in self.output_file:
333	>	common.logger.debug("Output from PoolOutputModule "+outputFile+" already in output files")
334	>	else:
335	>	self.output_file.append(outputFile)
336	>	common.logger.info("Adding "+outputFile+" (from PoolOutputModule) to list of output files")
337	>	# not requested, check anyhow to avoid accidental T2 overload
338	>	else:
339	>	if edmOutput:
340	>	missedFiles = []
341	>	for outputFile in edmOutput:
342	>	if outputFile not in self.output_file:
343	>	missedFiles.append(outputFile)
344	>	if missedFiles:
345	>	msg  = "ERROR: PoolOutputModule(s) are present in your ParameteSet %s \n"%self.pset
346	>	msg += "    but the file(s) produced ( %s ) are not in the list of output files\n" % ', '.join(missedFiles)
347	>	msg += "WARNING: please remove them. If you want to keep them, add the file(s) to output_files or use CMSSW.get_edm_output = 1\n"
348	>	if int(self.cfg_params.get('CMSSW.ignore_edm_output',0)):
349	>	msg += "    CMSSW.ignore_edm_output==1 : Hope you know what you are doing...\n"
350	>	common.logger.info(msg)
351	>	else :
352	>	raise CrabException(msg)
353	>
354	>	if (PsetEdit.getBadFilesSetting()):
355	>	msg = "WARNING: You have set skipBadFiles to True. This will continue processing on some errors and you may not be notified."
356	>	common.logger.info(msg)
357	>
358	>	except CrabException, msg:
359	>	common.logger.info(str(msg))
360	>	msg='Error while manipulating ParameterSet (see previous message, if any): exiting...'
361	>	raise CrabException(msg)
362	>
363	>	valid = re.compile('^[\w\.\-]+$')
364	>	for fileName in self.output_file:
365	>	if not valid.match(fileName):
366	>	msg = "The file %s may only contain alphanumeric characters and -, _, ." % fileName
367		raise CrabException(msg)
368
369	+
370		def DataDiscoveryAndLocation(self, cfg_params):
371
372		import DataDiscovery
312	–	import DataDiscovery_DBS2
373		import DataLocation
374	<	common.logger.debug(10,"CMSSW::DataDiscoveryAndLocation()")
374	>	common.logger.log(10-1,"CMSSW::DataDiscoveryAndLocation()")
375
376		datasetPath=self.datasetPath
377
378		## Contact the DBS
379	<	common.logger.message("Contacting Data Discovery Services ...")
379	>	common.logger.info("Contacting Data Discovery Services ...")
380		try:
381	<
322	<	if self.use_dbs_1 == 1 :
323	<	self.pubdata=DataDiscovery.DataDiscovery(datasetPath, cfg_params)
324	<	else :
325	<	self.pubdata=DataDiscovery_DBS2.DataDiscovery_DBS2(datasetPath, cfg_params)
381	>	self.pubdata=DataDiscovery.DataDiscovery(datasetPath, cfg_params,self.skip_blocks)
382		self.pubdata.fetchDBSInfo()
383
384		except DataDiscovery.NotExistingDatasetError, ex :
#	Line 334 | Line 390 | class Cmssw(JobType):
390		except DataDiscovery.DataDiscoveryError, ex:
391		msg = 'ERROR ***: failed Data Discovery in DBS :  %s'%ex.getErrorMessage()
392		raise CrabException(msg)
337	–	except DataDiscovery_DBS2.NotExistingDatasetError_DBS2, ex :
338	–	msg = 'ERROR ***: failed Data Discovery in DBS : %s'%ex.getErrorMessage()
339	–	raise CrabException(msg)
340	–	except DataDiscovery_DBS2.NoDataTierinProvenanceError_DBS2, ex :
341	–	msg = 'ERROR ***: failed Data Discovery in DBS : %s'%ex.getErrorMessage()
342	–	raise CrabException(msg)
343	–	except DataDiscovery_DBS2.DataDiscoveryError_DBS2, ex:
344	–	msg = 'ERROR ***: failed Data Discovery in DBS :  %s'%ex.getErrorMessage()
345	–	raise CrabException(msg)
393
394		self.filesbyblock=self.pubdata.getFiles()
395	<	self.eventsbyblock=self.pubdata.getEventsPerBlock()
349	<	self.eventsbyfile=self.pubdata.getEventsPerFile()
395	>	self.conf['pubdata']=self.pubdata
396
397		## get max number of events
398	<	self.maxEvents=self.pubdata.getMaxEvents() ##  self.maxEvents used in Creator.py
398	>	self.maxEvents=self.pubdata.getMaxEvents()
399
400		## Contact the DLS and build a list of sites hosting the fileblocks
401		try:
402		dataloc=DataLocation.DataLocation(self.filesbyblock.keys(),cfg_params)
403		dataloc.fetchDLSInfo()
404	+
405		except DataLocation.DataLocationError , ex:
406		msg = 'ERROR ***: failed Data Location in DLS \n %s '%ex.getErrorMessage()
407		raise CrabException(msg)
361	–
408
409	<	sites = dataloc.getSites()
409	>
410	>	unsorted_sites = dataloc.getSites()
411	>	sites = self.filesbyblock.fromkeys(self.filesbyblock,'')
412	>	for lfn in self.filesbyblock.keys():
413	>	if unsorted_sites.has_key(lfn):
414	>	sites[lfn]=unsorted_sites[lfn]
415	>	else:
416	>	sites[lfn]=[]
417	>
418	>	if len(sites)==0:
419	>	msg = 'ERROR ***: no location for any of the blocks of this dataset: \n\t %s \n'%datasetPath
420	>	msg += "\tMaybe the dataset is located only at T1's (or at T0), where analysis jobs are not allowed\n"
421	>	msg += "\tPlease check DataDiscovery page https://cmsweb.cern.ch/dbs_discovery/\n"
422	>	raise CrabException(msg)
423	>
424		allSites = []
425		listSites = sites.values()
426		for listSite in listSites:
427		for oneSite in listSite:
428		allSites.append(oneSite)
429	<	allSites = self.uniquelist(allSites)
429	>	[allSites.append(it) for it in allSites if not allSites.count(it)]
430	>
431
432		# screen output
433	<	common.logger.message("Requested dataset: " + datasetPath + " has " + str(self.maxEvents) + " events in " + str(len(self.filesbyblock.keys())) + " blocks.\n")
433	>	if self.ads or self.lumiMask:
434	>	common.logger.info("Requested (A)DS %s has %s block(s)." %
435	>	(datasetPath, len(self.filesbyblock.keys())))
436	>	else:
437	>	common.logger.info("Requested dataset: " + datasetPath + \
438	>	" has " + str(self.maxEvents) + " events in " + \
439	>	str(len(self.filesbyblock.keys())) + " blocks.\n")
440
441		return sites
375	–
376	–	def jobSplittingByBlocks(self, blockSites):
377	–	"""
378	–	Perform job splitting. Jobs run over an integer number of files
379	–	and no more than one block.
380	–	ARGUMENT: blockSites: dictionary with blocks as keys and list of host sites as values
381	–	REQUIRES: self.selectTotalNumberEvents, self.selectEventsPerJob, self.selectNumberofJobs,
382	–	self.total_number_of_events, self.eventsPerJob, self.theNumberOfJobs,
383	–	self.maxEvents, self.filesbyblock
384	–	SETS: self.jobDestination - Site destination(s) for each job (a list of lists)
385	–	self.total_number_of_jobs - Total # of jobs
386	–	self.list_of_args - File(s) job will run on (a list of lists)
387	–	"""
388	–
389	–	# ---- Handle the possible job splitting configurations ---- #
390	–	if (self.selectTotalNumberEvents):
391	–	totalEventsRequested = self.total_number_of_events
392	–	if (self.selectEventsPerJob):
393	–	eventsPerJobRequested = self.eventsPerJob
394	–	if (self.selectNumberOfJobs):
395	–	totalEventsRequested = self.theNumberOfJobs * self.eventsPerJob
396	–
397	–	# If user requested all the events in the dataset
398	–	if (totalEventsRequested == -1):
399	–	eventsRemaining=self.maxEvents
400	–	# If user requested more events than are in the dataset
401	–	elif (totalEventsRequested > self.maxEvents):
402	–	eventsRemaining = self.maxEvents
403	–	common.logger.message("Requested "+str(self.total_number_of_events)+ " events, but only "+str(self.maxEvents)+" events are available.")
404	–	# If user requested less events than are in the dataset
405	–	else:
406	–	eventsRemaining = totalEventsRequested
442
408	–	# If user requested more events per job than are in the dataset
409	–	if (self.selectEventsPerJob and eventsPerJobRequested > self.maxEvents):
410	–	eventsPerJobRequested = self.maxEvents
411	–
412	–	# For user info at end
413	–	totalEventCount = 0
414	–
415	–	if (self.selectTotalNumberEvents and self.selectNumberOfJobs):
416	–	eventsPerJobRequested = int(eventsRemaining/self.theNumberOfJobs)
417	–
418	–	if (self.selectNumberOfJobs):
419	–	common.logger.message("May not create the exact number_of_jobs requested.")
420	–
421	–	if ( self.ncjobs == 'all' ) :
422	–	totalNumberOfJobs = 999999999
423	–	else :
424	–	totalNumberOfJobs = self.ncjobs
425	–
426	–
427	–	blocks = blockSites.keys()
428	–	blockCount = 0
429	–	# Backup variable in case self.maxEvents counted events in a non-included block
430	–	numBlocksInDataset = len(blocks)
431	–
432	–	jobCount = 0
433	–	list_of_lists = []
434	–
435	–	# list tracking which jobs are in which jobs belong to which block
436	–	jobsOfBlock = {}
437	–
438	–	# ---- Iterate over the blocks in the dataset until ---- #
439	–	# ---- we've met the requested total # of events    ---- #
440	–	while ( (eventsRemaining > 0) and (blockCount < numBlocksInDataset) and (jobCount < totalNumberOfJobs)):
441	–	block = blocks[blockCount]
442	–	blockCount += 1
443	–
444	–	if self.eventsbyblock.has_key(block) :
445	–	numEventsInBlock = self.eventsbyblock[block]
446	–	common.logger.debug(5,'Events in Block File '+str(numEventsInBlock))
447	–
448	–	files = self.filesbyblock[block]
449	–	numFilesInBlock = len(files)
450	–	if (numFilesInBlock <= 0):
451	–	continue
452	–	fileCount = 0
443
444	<	# ---- New block => New job ---- #
455	<	parString = "\\{"
456	<	# counter for number of events in files currently worked on
457	<	filesEventCount = 0
458	<	# flag if next while loop should touch new file
459	<	newFile = 1
460	<	# job event counter
461	<	jobSkipEventCount = 0
462	<
463	<	# ---- Iterate over the files in the block until we've met the requested ---- #
464	<	# ---- total # of events or we've gone over all the files in this block  ---- #
465	<	while ( (eventsRemaining > 0) and (fileCount < numFilesInBlock) and (jobCount < totalNumberOfJobs) ):
466	<	file = files[fileCount]
467	<	if newFile :
468	<	try:
469	<	numEventsInFile = self.eventsbyfile[file]
470	<	common.logger.debug(6, "File "+str(file)+" has "+str(numEventsInFile)+" events")
471	<	# increase filesEventCount
472	<	filesEventCount += numEventsInFile
473	<	# Add file to current job
474	<	parString += '\\\"' + file + '\\\"\,'
475	<	newFile = 0
476	<	except KeyError:
477	<	common.logger.message("File "+str(file)+" has unknown number of events: skipping")
478	<
479	<
480	<	# if less events in file remain than eventsPerJobRequested
481	<	if ( filesEventCount - jobSkipEventCount < eventsPerJobRequested ) :
482	<	# if last file in block
483	<	if ( fileCount == numFilesInBlock-1 ) :
484	<	# end job using last file, use remaining events in block
485	<	# close job and touch new file
486	<	fullString = parString[:-2]
487	<	fullString += '\\}'
488	<	list_of_lists.append([fullString,str(-1),str(jobSkipEventCount)])
489	<	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(filesEventCount - jobSkipEventCount)+" events (last file in block).")
490	<	self.jobDestination.append(blockSites[block])
491	<	common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
492	<	# fill jobs of block dictionary
493	<	if block in jobsOfBlock.keys() :
494	<	jobsOfBlock[block].append(jobCount+1)
495	<	else:
496	<	jobsOfBlock[block] = [jobCount+1]
497	<	# reset counter
498	<	jobCount = jobCount + 1
499	<	totalEventCount = totalEventCount + filesEventCount - jobSkipEventCount
500	<	eventsRemaining = eventsRemaining - filesEventCount + jobSkipEventCount
501	<	jobSkipEventCount = 0
502	<	# reset file
503	<	parString = "\\{"
504	<	filesEventCount = 0
505	<	newFile = 1
506	<	fileCount += 1
507	<	else :
508	<	# go to next file
509	<	newFile = 1
510	<	fileCount += 1
511	<	# if events in file equal to eventsPerJobRequested
512	<	elif ( filesEventCount - jobSkipEventCount == eventsPerJobRequested ) :
513	<	# close job and touch new file
514	<	fullString = parString[:-2]
515	<	fullString += '\\}'
516	<	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
517	<	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.")
518	<	self.jobDestination.append(blockSites[block])
519	<	common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
520	<	if block in jobsOfBlock.keys() :
521	<	jobsOfBlock[block].append(jobCount+1)
522	<	else:
523	<	jobsOfBlock[block] = [jobCount+1]
524	<	# reset counter
525	<	jobCount = jobCount + 1
526	<	totalEventCount = totalEventCount + eventsPerJobRequested
527	<	eventsRemaining = eventsRemaining - eventsPerJobRequested
528	<	jobSkipEventCount = 0
529	<	# reset file
530	<	parString = "\\{"
531	<	filesEventCount = 0
532	<	newFile = 1
533	<	fileCount += 1
534	<
535	<	# if more events in file remain than eventsPerJobRequested
536	<	else :
537	<	# close job but don't touch new file
538	<	fullString = parString[:-2]
539	<	fullString += '\\}'
540	<	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
541	<	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.")
542	<	self.jobDestination.append(blockSites[block])
543	<	common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
544	<	if block in jobsOfBlock.keys() :
545	<	jobsOfBlock[block].append(jobCount+1)
546	<	else:
547	<	jobsOfBlock[block] = [jobCount+1]
548	<	# increase counter
549	<	jobCount = jobCount + 1
550	<	totalEventCount = totalEventCount + eventsPerJobRequested
551	<	eventsRemaining = eventsRemaining - eventsPerJobRequested
552	<	# calculate skip events for last file
553	<	# use filesEventCount (contains several files), jobSkipEventCount and eventsPerJobRequest
554	<	jobSkipEventCount = eventsPerJobRequested - (filesEventCount - jobSkipEventCount - self.eventsbyfile[file])
555	<	# remove all but the last file
556	<	filesEventCount = self.eventsbyfile[file]
557	<	parString = "\\{"
558	<	parString += '\\\"' + file + '\\\"\,'
559	<	pass # END if
560	<	pass # END while (iterate over files in the block)
561	<	pass # END while (iterate over blocks in the dataset)
562	<	self.ncjobs = self.total_number_of_jobs = jobCount
563	<	if (eventsRemaining > 0 and jobCount < totalNumberOfJobs ):
564	<	common.logger.message("Could not run on all requested events because some blocks not hosted at allowed sites.")
565	<	common.logger.message(str(jobCount)+" job(s) can run on "+str(totalEventCount)+" events.\n")
566	<
567	<	# screen output
568	<	screenOutput = "List of jobs and available destination sites:\n\n"
444	>	def split(self, jobParams,firstJobID):
445
446	<	blockCounter = 0
447	<	for block in jobsOfBlock.keys():
448	<	blockCounter += 1
449	<	screenOutput += "Block %5i: jobs %20s: sites: %s\n" % (blockCounter,spanRanges(jobsOfBlock[block]),','.join(blockSites[block]))
446	>	jobParams = self.dict['args']
447	>	njobs = self.dict['njobs']
448	>	self.jobDestination = self.dict['jobDestination']
449	>
450	>	if njobs == 0:
451	>	raise CrabException("Asked to split zero jobs: aborting")
452	>	if not self.server and not self.local and njobs > 500:
453	>	raise CrabException("The CRAB client will not submit more than 500 jobs. You must use the server mode.")
454
455	<	common.logger.message(screenOutput)
455	>	# create the empty structure
456	>	for i in range(njobs):
457	>	jobParams.append("")
458
459	<	self.list_of_args = list_of_lists
459	>	listID=[]
460	>	listField=[]
461	>	listDictions=[]
462	>	exist= os.path.exists(self.argsFile)
463	>	for id in range(njobs):
464	>	job = id + int(firstJobID)
465	>	listID.append(job+1)
466	>	job_ToSave ={}
467	>	concString = ' '
468	>	argu=''
469	>	str_argu = str(job+1)
470	>	if len(jobParams[id]):
471	>	argu = {'JobID': job+1}
472	>	for i in range(len(jobParams[id])):
473	>	argu[self.dict['params'][i]]=jobParams[id][i]
474	>	if len(jobParams[id])==1: self.NumEvents = jobParams[id][i]
475	>	# just for debug
476	>	str_argu += concString.join(jobParams[id])
477	>	if argu != '': listDictions.append(argu)
478	>	job_ToSave['arguments']= '%d %d'%( (job+1), 0)
479	>	job_ToSave['dlsDestination']= self.jobDestination[id]
480	>	listField.append(job_ToSave)
481	>	from ProdCommon.SiteDB.CmsSiteMapper import CmsSEMap
482	>	cms_se = CmsSEMap()
483	>	msg="Job  %s  Arguments:  %s\n"%(str(job+1),str_argu)
484	>	msg+="\t  Destination: %s "%(str(self.jobDestination[id]))
485	>	SEDestination = [cms_se[dest] for dest in self.jobDestination[id]]
486	>	msg+="\t  CMSDestination: %s "%(str(SEDestination))
487	>	common.logger.log(10-1,msg)
488	>	# write xml
489	>	if len(listDictions):
490	>	if exist==False: self.CreateXML()
491	>	self.addEntry(listDictions)
492	>	common._db.updateJob_(listID,listField)
493		return
494
495	<	def jobSplittingNoInput(self):
495	>	def CreateXML(self):
496		"""
582	–	Perform job splitting based on number of event per job
497		"""
498	<	common.logger.debug(5,'Splitting per events')
499	<	common.logger.message('Required '+str(self.eventsPerJob)+' events per job ')
586	<	common.logger.message('Required '+str(self.theNumberOfJobs)+' jobs in total ')
587	<	common.logger.message('Required '+str(self.total_number_of_events)+' events in total ')
588	<
589	<	if (self.total_number_of_events < 0):
590	<	msg='Cannot split jobs per Events with "-1" as total number of events'
591	<	raise CrabException(msg)
592	<
593	<	if (self.selectEventsPerJob):
594	<	if (self.selectTotalNumberEvents):
595	<	self.total_number_of_jobs = int(self.total_number_of_events/self.eventsPerJob)
596	<	elif(self.selectNumberOfJobs) :
597	<	self.total_number_of_jobs =self.theNumberOfJobs
598	<	self.total_number_of_events =int(self.theNumberOfJobs*self.eventsPerJob)
599	<
600	<	elif (self.selectNumberOfJobs) :
601	<	self.total_number_of_jobs = self.theNumberOfJobs
602	<	self.eventsPerJob = int(self.total_number_of_events/self.total_number_of_jobs)
603	<
604	<	common.logger.debug(5,'N jobs  '+str(self.total_number_of_jobs))
605	<
606	<	# is there any remainder?
607	<	check = int(self.total_number_of_events) - (int(self.total_number_of_jobs)*self.eventsPerJob)
608	<
609	<	common.logger.debug(5,'Check  '+str(check))
610	<
611	<	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')
612	<	if check > 0:
613	<	common.logger.message('Warning: asked '+str(self.total_number_of_events)+' but can do only '+str(int(self.total_number_of_jobs)*self.eventsPerJob))
614	<
615	<	# argument is seed number.$i
616	<	self.list_of_args = []
617	<	for i in range(self.total_number_of_jobs):
618	<	## Since there is no input, any site is good
619	<	# self.jobDestination.append(["Any"])
620	<	self.jobDestination.append([""]) #must be empty to write correctly the xml
621	<	args=[]
622	<	if (self.firstRun):
623	<	## pythia first run
624	<	#self.list_of_args.append([(str(self.firstRun)+str(i))])
625	<	args.append(str(self.firstRun)+str(i))
626	<	else:
627	<	## no first run
628	<	#self.list_of_args.append([str(i)])
629	<	args.append(str(i))
630	<	if (self.sourceSeed):
631	<	args.append(str(self.sourceSeed)+str(i))
632	<	if (self.sourceSeedVtx):
633	<	## + vtx random seed
634	<	args.append(str(self.sourceSeedVtx)+str(i))
635	<	if (self.sourceSeedG4):
636	<	## + G4 random seed
637	<	args.append(str(self.sourceSeedG4)+str(i))
638	<	if (self.sourceSeedMix):
639	<	## + Mix random seed
640	<	args.append(str(self.sourceSeedMix)+str(i))
641	<	pass
642	<	pass
643	<	self.list_of_args.append(args)
644	<	pass
645	<
646	<	# print self.list_of_args
647	<
498	>	result = IMProvNode( self.rootArgsFilename )
499	>	outfile = file( self.argsFile, 'w').write(str(result))
500		return
501
502	<
651	<	def jobSplittingForScript(self):#CarlosDaniele
502	>	def addEntry(self, listDictions):
503		"""
504	<	Perform job splitting based on number of job
654	<	"""
655	<	common.logger.debug(5,'Splitting per job')
656	<	common.logger.message('Required '+str(self.theNumberOfJobs)+' jobs in total ')
657	<
658	<	self.total_number_of_jobs = self.theNumberOfJobs
659	<
660	<	common.logger.debug(5,'N jobs  '+str(self.total_number_of_jobs))
661	<
662	<	common.logger.message(str(self.total_number_of_jobs)+' jobs can be created')
504	>	_addEntry_
505
506	<	# argument is seed number.$i
507	<	self.list_of_args = []
508	<	for i in range(self.total_number_of_jobs):
509	<	## Since there is no input, any site is good
510	<	# self.jobDestination.append(["Any"])
511	<	self.jobDestination.append([""])
512	<	## no random seed
513	<	self.list_of_args.append([str(i)])
506	>	add an entry to the xml file
507	>	"""
508	>	## load xml
509	>	improvDoc = loadIMProvFile(self.argsFile)
510	>	entrname= 'Job'
511	>	for dictions in listDictions:
512	>	report = IMProvNode(entrname , None, **dictions)
513	>	improvDoc.addNode(report)
514	>	outfile = file( self.argsFile, 'w').write(str(improvDoc))
515		return
516
674	–	def split(self, jobParams):
675	–
676	–	common.jobDB.load()
677	–	#### Fabio
678	–	njobs = self.total_number_of_jobs
679	–	arglist = self.list_of_args
680	–	# create the empty structure
681	–	for i in range(njobs):
682	–	jobParams.append("")
683	–
684	–	for job in range(njobs):
685	–	jobParams[job] = arglist[job]
686	–	# print str(arglist[job])
687	–	# print jobParams[job]
688	–	common.jobDB.setArguments(job, jobParams[job])
689	–	common.logger.debug(5,"Job "+str(job)+" Destination: "+str(self.jobDestination[job]))
690	–	common.jobDB.setDestination(job, self.jobDestination[job])
691	–
692	–	common.jobDB.save()
693	–	return
694	–
695	–	def getJobTypeArguments(self, nj, sched):
696	–	result = ''
697	–	for i in common.jobDB.arguments(nj):
698	–	result=result+str(i)+" "
699	–	return result
700	–
517		def numberOfJobs(self):
518	<	# Fabio
519	<	return self.total_number_of_jobs
518	>	#wmbs
519	>	if self.automation==0:
520	>	return self.dict['njobs']
521	>	else:
522	>	return None
523
524		def getTarBall(self, exe):
525		"""
526		Return the TarBall with lib and exe
527		"""
528	<
710	<	# if it exist, just return it
711	<	#
712	<	# Marco. Let's start to use relative path for Boss XML files
713	<	#
714	<	self.tgzNameWithPath = common.work_space.pathForTgz()+'share/'+self.tgz_name
528	>	self.tgzNameWithPath = common.work_space.pathForTgz()+self.tgz_name
529		if os.path.exists(self.tgzNameWithPath):
530		return self.tgzNameWithPath
531
#	Line 724 | Line 538 | class Cmssw(JobType):
538
539		# First of all declare the user Scram area
540		swArea = self.scram.getSWArea_()
727	–	#print "swArea = ", swArea
728	–	# swVersion = self.scram.getSWVersion()
729	–	# print "swVersion = ", swVersion
541		swReleaseTop = self.scram.getReleaseTop_()
542	<	#print "swReleaseTop = ", swReleaseTop
732	<
542	>
543		## check if working area is release top
544		if swReleaseTop == '' or swArea == swReleaseTop:
545	+	common.logger.debug("swArea = "+swArea+" swReleaseTop ="+swReleaseTop)
546		return
547
548		import tarfile
#	Line 742 | Line 553 | class Cmssw(JobType):
553		exeWithPath = self.scram.findFile_(executable)
554		if ( not exeWithPath ):
555		raise CrabException('User executable '+executable+' not found')
556	<
556	>
557		## then check if it's private or not
558		if exeWithPath.find(swReleaseTop) == -1:
559		# the exe is private, so we must ship
560	<	common.logger.debug(5,"Exe "+exeWithPath+" to be tarred")
560	>	common.logger.debug("Exe "+exeWithPath+" to be tarred")
561		path = swArea+'/'
562		# distinguish case when script is in user project area or given by full path somewhere else
563		if exeWithPath.find(path) >= 0 :
564		exe = string.replace(exeWithPath, path,'')
565	<	tar.add(path+exe,os.path.basename(executable))
565	>	tar.add(path+exe,exe)
566		else :
567		tar.add(exeWithPath,os.path.basename(executable))
568		pass
569		else:
570		# the exe is from release, we'll find it on WN
571		pass
572	<
572	>
573		## Now get the libraries: only those in local working area
574	+	tar.dereference=True
575		libDir = 'lib'
576		lib = swArea+'/' +libDir
577	<	common.logger.debug(5,"lib "+lib+" to be tarred")
577	>	common.logger.debug("lib "+lib+" to be tarred")
578		if os.path.exists(lib):
579		tar.add(lib,libDir)
580	<
580	>
581		## Now check if module dir is present
582		moduleDir = 'module'
583		module = swArea + '/' + moduleDir
584		if os.path.isdir(module):
585		tar.add(module,moduleDir)
586	+	tar.dereference=False
587
588		## Now check if any data dir(s) is present
589	<	swAreaLen=len(swArea)
590	<	for root, dirs, files in os.walk(swArea):
591	<	if "data" in dirs:
592	<	common.logger.debug(5,"data "+root+"/data"+" to be tarred")
593	<	tar.add(root+"/data",root[swAreaLen:]+"/data")
594	<
595	<	## Add ProdAgent dir to tar
596	<	paDir = 'ProdAgentApi'
597	<	pa = os.environ['CRABDIR'] + '/' + 'ProdAgentApi'
598	<	if os.path.isdir(pa):
599	<	tar.add(pa,paDir)
600	<
601	<	common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
589	>	self.dataExist = False
590	>	todo_list = [(i, i) for i in  os.listdir(swArea+"/src")]
591	>	while len(todo_list):
592	>	entry, name = todo_list.pop()
593	>	if name.startswith('crab_0_') or  name.startswith('.') or name == 'CVS':
594	>	continue
595	>	if os.path.isdir(swArea+"/src/"+entry):
596	>	entryPath = entry + '/'
597	>	todo_list += [(entryPath + i, i) for i in  os.listdir(swArea+"/src/"+entry)]
598	>	if name == 'data':
599	>	self.dataExist=True
600	>	common.logger.debug("data "+entry+" to be tarred")
601	>	tar.add(swArea+"/src/"+entry,"src/"+entry)
602	>	pass
603	>	pass
604	>
605	>	### CMSSW ParameterSet
606	>	if not self.pset is None:
607	>	cfg_file = common.work_space.jobDir()+self.configFilename()
608	>	pickleFile = common.work_space.jobDir()+self.configFilename() + '.pkl'
609	>	tar.add(cfg_file,self.configFilename())
610	>	tar.add(pickleFile,self.configFilename() + '.pkl')
611	>
612	>	try:
613	>	crab_cfg_file = common.work_space.shareDir()+'/crab.cfg'
614	>	tar.add(crab_cfg_file,'crab.cfg')
615	>	except:
616	>	pass
617	>
618	>	## Add ProdCommon dir to tar
619	>	prodcommonDir = './'
620	>	prodcommonPath = os.environ['CRABDIR'] + '/' + 'external/'
621	>	neededStuff = ['ProdCommon/__init__.py','ProdCommon/FwkJobRep', 'ProdCommon/CMSConfigTools', \
622	>	'ProdCommon/Core', 'ProdCommon/MCPayloads', 'IMProv', 'ProdCommon/Storage', \
623	>	'WMCore/__init__.py','WMCore/Algorithms']
624	>	for file in neededStuff:
625	>	tar.add(prodcommonPath+file,prodcommonDir+file)
626	>
627	>	##### ML stuff
628	>	ML_file_list=['report.py', 'DashboardAPI.py', 'Logger.py', 'ProcInfo.py', 'apmon.py']
629	>	path=os.environ['CRABDIR'] + '/python/'
630	>	for file in ML_file_list:
631	>	tar.add(path+file,file)
632	>
633	>	##### Utils
634	>	Utils_file_list=['parseCrabFjr.py','writeCfg.py', 'fillCrabFjr.py','cmscp.py']
635	>	for file in Utils_file_list:
636	>	tar.add(path+file,file)
637	>
638	>	##### AdditionalFiles
639	>	tar.dereference=True
640	>	for file in self.additional_inbox_files:
641	>	tar.add(file,string.split(file,'/')[-1])
642	>	tar.dereference=False
643	>	common.logger.log(10-1,"Files in "+self.tgzNameWithPath+" : "+str(tar.getnames()))
644	>
645		tar.close()
646	<	except :
647	<	raise CrabException('Could not create tar-ball')
646	>	except IOError, exc:
647	>	msg = 'Could not create tar-ball %s \n'%self.tgzNameWithPath
648	>	msg += str(exc)
649	>	raise CrabException(msg)
650	>	except tarfile.TarError, exc:
651	>	msg = 'Could not create tar-ball %s \n'%self.tgzNameWithPath
652	>	msg += str(exc)
653	>	raise CrabException(msg)
654
793	–	## check for tarball size
655		tarballinfo = os.stat(self.tgzNameWithPath)
656		if ( tarballinfo.st_size > self.MaxTarBallSize*1024*1024 ) :
657	<	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.')
657	>	if not self.server:
658	>	msg  = 'Input sandbox size of ' + str(float(tarballinfo.st_size)/1024.0/1024.0) + ' MB is larger than the allowed ' + \
659	>	str(self.MaxTarBallSize) +'MB input sandbox limit \n'
660	>	msg += '      and not supported by the direct GRID submission system.\n'
661	>	msg += '      Please use the CRAB server mode by setting server_name=<NAME> in section [CRAB] of your crab.cfg.\n'
662	>	msg += '      For further infos please see https://twiki.cern.ch/twiki/bin/view/CMS/SWGuideCrabServerForUsers#Server_available_for_users'
663	>	else:
664	>	msg  = 'Input sandbox size of ' + str(float(tarballinfo.st_size)/1024.0/1024.0) + ' MB is larger than the allowed ' +  \
665	>	str(self.MaxTarBallSize) +'MB input sandbox limit in the server.'
666	>	raise CrabException(msg)
667
668		## create tar-ball with ML stuff
669	<	self.MLtgzfile =  common.work_space.pathForTgz()+'share/MLfiles.tgz'
670	<	try:
801	<	tar = tarfile.open(self.MLtgzfile, "w:gz")
802	<	path=os.environ['CRABDIR'] + '/python/'
803	<	for file in ['report.py', 'DashboardAPI.py', 'Logger.py', 'ProcInfo.py', 'apmon.py', 'parseCrabFjr.py']:
804	<	tar.add(path+file,file)
805	<	common.logger.debug(5,"Files added to "+self.MLtgzfile+" : "+str(tar.getnames()))
806	<	tar.close()
807	<	except :
808	<	raise CrabException('Could not create ML files tar-ball')
809	<
810	<	return
811	<
812	<	def wsSetupEnvironment(self, nj):
669	>
670	>	def wsSetupEnvironment(self, nj=0):
671		"""
672		Returns part of a job script which prepares
673		the execution environment for the job 'nj'.
674		"""
675	<	# Prepare JobType-independent part
818	<	txt = ''
819	<
820	<	## OLI_Daniele at this level  middleware already known
675	>	psetName = 'pset.py'
676
677	<	txt += 'if [ $middleware == LCG ]; then \n'
677	>	# Prepare JobType-independent part
678	>	txt = '\n#Written by cms_cmssw::wsSetupEnvironment\n'
679	>	txt += 'echo ">>> setup environment"\n'
680	>	txt += 'echo "set SCRAM ARCH to ' + self.executable_arch + '"\n'
681	>	txt += 'export SCRAM_ARCH=' + self.executable_arch + '\n'
682	>	txt += 'echo "SCRAM_ARCH = $SCRAM_ARCH"\n'
683	>	txt += 'if [ $middleware == LCG ] || [ $middleware == CAF ] || [ $middleware == LSF ]; then \n'
684		txt += self.wsSetupCMSLCGEnvironment_()
685		txt += 'elif [ $middleware == OSG ]; then\n'
686		txt += '    WORKING_DIR=`/bin/mktemp  -d $OSG_WN_TMP/cms_XXXXXXXXXXXX`\n'
687	<	txt += '    echo "Created working directory: $WORKING_DIR"\n'
688	<	txt += '    if [ ! -d $WORKING_DIR ] ;then\n'
689	<	txt += '        echo "SET_CMS_ENV 10016 ==> OSG $WORKING_DIR could not be created on WN `hostname`"\n'
690	<	txt += '    echo "JOB_EXIT_STATUS = 10016"\n'
830	<	txt += '    echo "JobExitCode=10016" | tee -a $RUNTIME_AREA/$repo\n'
831	<	txt += '    dumpStatus $RUNTIME_AREA/$repo\n'
832	<	txt += '        rm -f $RUNTIME_AREA/$repo \n'
833	<	txt += '        echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
834	<	txt += '        echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
835	<	txt += '        exit 1\n'
687	>	txt += '    if [ ! $? == 0 ] ;then\n'
688	>	txt += '        echo "ERROR ==> OSG $WORKING_DIR could not be created on WN `hostname`"\n'
689	>	txt += '        job_exit_code=10016\n'
690	>	txt += '        func_exit\n'
691		txt += '    fi\n'
692	+	txt += '    echo ">>> Created working directory: $WORKING_DIR"\n'
693		txt += '\n'
694		txt += '    echo "Change to working directory: $WORKING_DIR"\n'
695		txt += '    cd $WORKING_DIR\n'
696	<	txt += self.wsSetupCMSOSGEnvironment_()
696	>	txt += '    echo ">>> current directory (WORKING_DIR): $WORKING_DIR"\n'
697	>	txt += self.wsSetupCMSOSGEnvironment_()
698	>	#Setup SGE Environment
699	>	txt += 'elif [ $middleware == SGE ]; then\n'
700	>	txt += self.wsSetupCMSLCGEnvironment_()
701	>
702	>	txt += 'elif [ $middleware == ARC ]; then\n'
703	>	txt += self.wsSetupCMSLCGEnvironment_()
704	>
705	>	#Setup PBS Environment
706	>	txt += 'elif [ $middleware == PBS ]; then\n'
707	>	txt += self.wsSetupCMSLCGEnvironment_()
708	>
709		txt += 'fi\n'
710
711		# Prepare JobType-specific part
712		scram = self.scram.commandName()
713		txt += '\n\n'
714	<	txt += 'echo "### SPECIFIC JOB SETUP ENVIRONMENT ###"\n'
715	<	txt += 'echo "Setting SCRAM_ARCH='+self.executable_arch+'"\n'
848	<	txt += 'export SCRAM_ARCH='+self.executable_arch+'\n'
714	>	txt += 'echo ">>> specific cmssw setup environment:"\n'
715	>	txt += 'echo "CMSSW_VERSION =  '+self.version+'"\n'
716		txt += scram+' project CMSSW '+self.version+'\n'
717		txt += 'status=$?\n'
718		txt += 'if [ $status != 0 ] ; then\n'
719	<	txt += '   echo "SET_EXE_ENV 10034 ==>ERROR CMSSW '+self.version+' not found on `hostname`" \n'
720	<	txt += '   echo "JOB_EXIT_STATUS = 10034"\n'
721	<	txt += '   echo "JobExitCode=10034" | tee -a $RUNTIME_AREA/$repo\n'
855	<	txt += '   dumpStatus $RUNTIME_AREA/$repo\n'
856	<	txt += '   rm -f $RUNTIME_AREA/$repo \n'
857	<	txt += '   echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
858	<	txt += '   echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
859	<	## OLI_Daniele
860	<	txt += '    if [ $middleware == OSG ]; then \n'
861	<	txt += '        echo "Remove working directory: $WORKING_DIR"\n'
862	<	txt += '        cd $RUNTIME_AREA\n'
863	<	txt += '        /bin/rm -rf $WORKING_DIR\n'
864	<	txt += '        if [ -d $WORKING_DIR ] ;then\n'
865	<	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'
866	<	txt += '        echo "JOB_EXIT_STATUS = 10018"\n'
867	<	txt += '        echo "JobExitCode=10018" | tee -a $RUNTIME_AREA/$repo\n'
868	<	txt += '        dumpStatus $RUNTIME_AREA/$repo\n'
869	<	txt += '            rm -f $RUNTIME_AREA/$repo \n'
870	<	txt += '            echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
871	<	txt += '            echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
872	<	txt += '        fi\n'
873	<	txt += '    fi \n'
874	<	txt += '   exit 1 \n'
719	>	txt += '    echo "ERROR ==> CMSSW '+self.version+' not found on `hostname`" \n'
720	>	txt += '    job_exit_code=10034\n'
721	>	txt += '    func_exit\n'
722		txt += 'fi \n'
876	–	txt += 'echo "CMSSW_VERSION =  '+self.version+'"\n'
723		txt += 'cd '+self.version+'\n'
724	<	### needed grep for bug in scramv1 ###
725	<	txt += scram+' runtime -sh\n'
724	>	txt += 'SOFTWARE_DIR=`pwd`; export SOFTWARE_DIR\n'
725	>	txt += 'echo ">>> current directory (SOFTWARE_DIR): $SOFTWARE_DIR" \n'
726		txt += 'eval `'+scram+' runtime -sh | grep -v SCRAMRT_LSB_JOBNAME`\n'
727	<	txt += 'echo $PATH\n'
728	<
727	>	txt += 'if [ $? != 0 ] ; then\n'
728	>	txt += '    echo "ERROR ==> Problem with the command: "\n'
729	>	txt += '    echo "eval \`'+scram+' runtime -sh | grep -v SCRAMRT_LSB_JOBNAME \` at `hostname`"\n'
730	>	txt += '    job_exit_code=10034\n'
731	>	txt += '    func_exit\n'
732	>	txt += 'fi \n'
733		# Handle the arguments:
734		txt += "\n"
735	<	txt += "## number of arguments (first argument always jobnumber)\n"
735	>	txt += "## number of arguments (first argument always jobnumber, the second is the resubmission number)\n"
736		txt += "\n"
737	<	#        txt += "narg=$#\n"
888	<	txt += "if [ $nargs -lt 2 ]\n"
737	>	txt += "if [ $nargs -lt "+str(self.argsList)+" ]\n"
738		txt += "then\n"
739	<	txt += "    echo 'SET_EXE_ENV 1 ==> ERROR Too few arguments' +$nargs+ \n"
740	<	txt += '    echo "JOB_EXIT_STATUS = 50113"\n'
741	<	txt += '    echo "JobExitCode=50113" | tee -a $RUNTIME_AREA/$repo\n'
893	<	txt += '    dumpStatus $RUNTIME_AREA/$repo\n'
894	<	txt += '    rm -f $RUNTIME_AREA/$repo \n'
895	<	txt += '    echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
896	<	txt += '    echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
897	<	## OLI_Daniele
898	<	txt += '    if [ $middleware == OSG ]; then \n'
899	<	txt += '        echo "Remove working directory: $WORKING_DIR"\n'
900	<	txt += '        cd $RUNTIME_AREA\n'
901	<	txt += '        /bin/rm -rf $WORKING_DIR\n'
902	<	txt += '        if [ -d $WORKING_DIR ] ;then\n'
903	<	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'
904	<	txt += '        echo "JOB_EXIT_STATUS = 50114"\n'
905	<	txt += '        echo "JobExitCode=50114" | tee -a $RUNTIME_AREA/$repo\n'
906	<	txt += '        dumpStatus $RUNTIME_AREA/$repo\n'
907	<	txt += '            rm -f $RUNTIME_AREA/$repo \n'
908	<	txt += '            echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
909	<	txt += '            echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
910	<	txt += '        fi\n'
911	<	txt += '    fi \n'
912	<	txt += "    exit 1\n"
739	>	txt += "    echo 'ERROR ==> Too few arguments' +$nargs+ \n"
740	>	txt += '    job_exit_code=50113\n'
741	>	txt += "    func_exit\n"
742		txt += "fi\n"
743		txt += "\n"
744
745		# Prepare job-specific part
746		job = common.job_list[nj]
747	<	if self.pset != None: #CarlosDaniele
919	<	pset = os.path.basename(job.configFilename())
747	>	if (self.datasetPath):
748		txt += '\n'
749	<	if (self.datasetPath): # standard job
922	<	#txt += 'InputFiles=$2\n'
923	<	txt += 'InputFiles=${args[1]}\n'
924	<	txt += 'MaxEvents=${args[2]}\n'
925	<	txt += 'SkipEvents=${args[3]}\n'
926	<	txt += 'echo "Inputfiles:<$InputFiles>"\n'
927	<	txt += 'sed "s#{\'INPUT\'}#$InputFiles#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
928	<	txt += 'echo "MaxEvents:<$MaxEvents>"\n'
929	<	txt += 'sed "s#INPUTMAXEVENTS#$MaxEvents#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
930	<	txt += 'echo "SkipEvents:<$SkipEvents>"\n'
931	<	txt += 'sed "s#INPUTSKIPEVENTS#$SkipEvents#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
932	<	else:  # pythia like job
933	<	seedIndex=1
934	<	if (self.firstRun):
935	<	txt += 'FirstRun=${args['+str(seedIndex)+']}\n'
936	<	txt += 'echo "FirstRun: <$FirstRun>"\n'
937	<	txt += 'sed "s#\<INPUTFIRSTRUN\>#$FirstRun#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
938	<	seedIndex=seedIndex+1
939	<
940	<	if (self.sourceSeed):
941	<	txt += 'Seed=${args['+str(seedIndex)+']}\n'
942	<	txt += 'sed "s#\<INPUT\>#$Seed#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
943	<	seedIndex=seedIndex+1
944	<	## the following seeds are not always present
945	<	if (self.sourceSeedVtx):
946	<	txt += 'VtxSeed=${args['+str(seedIndex)+']}\n'
947	<	txt += 'echo "VtxSeed: <$VtxSeed>"\n'
948	<	txt += 'sed "s#\<INPUTVTX\>#$VtxSeed#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
949	<	seedIndex += 1
950	<	if (self.sourceSeedG4):
951	<	txt += 'G4Seed=${args['+str(seedIndex)+']}\n'
952	<	txt += 'echo "G4Seed: <$G4Seed>"\n'
953	<	txt += 'sed "s#\<INPUTG4\>#$G4Seed#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
954	<	seedIndex += 1
955	<	if (self.sourceSeedMix):
956	<	txt += 'mixSeed=${args['+str(seedIndex)+']}\n'
957	<	txt += 'echo "MixSeed: <$mixSeed>"\n'
958	<	txt += 'sed "s#\<INPUTMIX\>#$mixSeed#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
959	<	seedIndex += 1
960	<	pass
961	<	pass
962	<	txt += 'mv -f '+pset+' pset.cfg\n'
749	>	txt += 'DatasetPath='+self.datasetPath+'\n'
750
751	<	if len(self.additional_inbox_files) > 0:
752	<	for file in self.additional_inbox_files:
753	<	relFile = file.split("/")[-1]
754	<	txt += 'if [ -e $RUNTIME_AREA/'+relFile+' ] ; then\n'
755	<	txt += '   cp $RUNTIME_AREA/'+relFile+' .\n'
756	<	txt += '   chmod +x '+relFile+'\n'
757	<	txt += 'fi\n'
758	<	pass
759	<
760	<	if self.pset != None: #CarlosDaniele
761	<	txt += 'echo "### END JOB SETUP ENVIRONMENT ###"\n\n'
762	<
751	>	txt += 'PrimaryDataset='+self.primaryDataset +'\n'
752	>	txt += 'DataTier='+self.dataTier+'\n'
753	>	txt += 'ApplicationFamily=cmsRun\n'
754	>
755	>	else:
756	>	txt += 'DatasetPath=MCDataTier\n'
757	>	txt += 'PrimaryDataset=null\n'
758	>	txt += 'DataTier=null\n'
759	>	txt += 'ApplicationFamily=MCDataTier\n'
760	>	if self.pset != None:
761	>	pset = os.path.basename(job.configFilename())
762	>	pkl  = os.path.basename(job.configFilename()) + '.pkl'
763		txt += '\n'
764	<	txt += 'echo "***** cat pset.cfg *********"\n'
765	<	txt += 'cat pset.cfg\n'
766	<	txt += 'echo "****** end pset.cfg ********"\n'
764	>	txt += 'cp  $RUNTIME_AREA/'+pset+' .\n'
765	>	txt += 'cp  $RUNTIME_AREA/'+pkl+' .\n'
766	>
767	>	txt += 'PreserveSeeds='  + ','.join(self.preserveSeeds)  + '; export PreserveSeeds\n'
768	>	txt += 'IncrementSeeds=' + ','.join(self.incrementSeeds) + '; export IncrementSeeds\n'
769	>	txt += 'echo "PreserveSeeds: <$PreserveSeeds>"\n'
770	>	txt += 'echo "IncrementSeeds:<$IncrementSeeds>"\n'
771	>
772	>	txt += 'mv -f ' + pset + ' ' + psetName + '\n'
773	>	if self.var_filter:
774	>	#print "self.var_filter = ",self.var_filter
775	>	txt += "export var_filter="+"'"+self.var_filter+"'\n"
776	>	txt += 'echo $var_filter'
777	>	else:
778		txt += '\n'
779	+	if self.AdditionalArgs: txt += 'export AdditionalArgs=\"%s\"\n'%(self.AdditionalArgs)
780	+	if int(self.NumEvents) != 0: txt += 'export MaxEvents=%s\n'%str(self.NumEvents)
781		return txt
782
783	<	def wsBuildExe(self, nj=0):
783	>	def wsUntarSoftware(self, nj=0):
784		"""
785		Put in the script the commands to build an executable
786		or a library.
787		"""
788
789	<	txt = ""
789	>	txt = '\n#Written by cms_cmssw::wsUntarSoftware\n'
790
791		if os.path.isfile(self.tgzNameWithPath):
792	<	txt += 'echo "tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+'"\n'
793	<	txt += 'tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+'\n'
792	>	txt += 'echo ">>> tar xzf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+' :" \n'
793	>	if  self.debug_wrapper==1 :
794	>	txt += 'tar zxvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+'\n'
795	>	txt += 'ls -Al \n'
796	>	else:
797	>	txt += 'tar zxf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+'\n'
798		txt += 'untar_status=$? \n'
799		txt += 'if [ $untar_status -ne 0 ]; then \n'
800	<	txt += '   echo "SET_EXE 1 ==> ERROR Untarring .tgz file failed"\n'
801	<	txt += '   echo "JOB_EXIT_STATUS = $untar_status" \n'
802	<	txt += '   echo "JobExitCode=$untar_status" | tee -a $RUNTIME_AREA/$repo\n'
999	<	txt += '   if [ $middleware == OSG ]; then \n'
1000	<	txt += '       echo "Remove working directory: $WORKING_DIR"\n'
1001	<	txt += '       cd $RUNTIME_AREA\n'
1002	<	txt += '       /bin/rm -rf $WORKING_DIR\n'
1003	<	txt += '       if [ -d $WORKING_DIR ] ;then\n'
1004	<	txt += '           echo "SET_EXE 50999 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after Untarring .tgz file failed"\n'
1005	<	txt += '           echo "JOB_EXIT_STATUS = 50999"\n'
1006	<	txt += '           echo "JobExitCode=50999" | tee -a $RUNTIME_AREA/$repo\n'
1007	<	txt += '           dumpStatus $RUNTIME_AREA/$repo\n'
1008	<	txt += '           rm -f $RUNTIME_AREA/$repo \n'
1009	<	txt += '           echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1010	<	txt += '           echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1011	<	txt += '       fi\n'
1012	<	txt += '   fi \n'
1013	<	txt += '   \n'
1014	<	txt += '   exit 1 \n'
800	>	txt += '   echo "ERROR ==> Untarring .tgz file failed"\n'
801	>	txt += '   job_exit_code=$untar_status\n'
802	>	txt += '   func_exit\n'
803		txt += 'else \n'
804		txt += '   echo "Successful untar" \n'
805		txt += 'fi \n'
806		txt += '\n'
807	<	txt += 'echo "Include ProdAgentApi in PYTHONPATH"\n'
807	>	txt += 'echo ">>> Include $RUNTIME_AREA in PYTHONPATH:"\n'
808		txt += 'if [ -z "$PYTHONPATH" ]; then\n'
809	<	txt += '   export PYTHONPATH=ProdAgentApi\n'
809	>	txt += '   export PYTHONPATH=$RUNTIME_AREA/\n'
810		txt += 'else\n'
811	<	txt += '   export PYTHONPATH=ProdAgentApi:${PYTHONPATH}\n'
811	>	txt += '   export PYTHONPATH=$RUNTIME_AREA/:${PYTHONPATH}\n'
812	>	txt += 'echo "PYTHONPATH=$PYTHONPATH"\n'
813		txt += 'fi\n'
814		txt += '\n'
815
816		pass
817	<
817	>
818		return txt
819
820	<	def modifySteeringCards(self, nj):
820	>	def wsBuildExe(self, nj=0):
821		"""
822	<	modify the card provided by the user,
823	<	writing a new card into share dir
822	>	Put in the script the commands to build an executable
823	>	or a library.
824		"""
825	<
825	>
826	>	txt = '\n#Written by cms_cmssw::wsBuildExe\n'
827	>	txt += 'echo ">>> moving CMSSW software directories in `pwd`" \n'
828	>
829	>	txt += 'rm -r lib/ module/ \n'
830	>	txt += 'mv $RUNTIME_AREA/lib/ . \n'
831	>	txt += 'mv $RUNTIME_AREA/module/ . \n'
832	>	if self.dataExist == True:
833	>	txt += 'rm -r src/ \n'
834	>	txt += 'mv $RUNTIME_AREA/src/ . \n'
835	>	if len(self.additional_inbox_files)>0:
836	>	for file in self.additional_inbox_files:
837	>	txt += 'mv $RUNTIME_AREA/'+os.path.basename(file)+' . \n'
838	>
839	>	txt += 'echo ">>> Include $RUNTIME_AREA in PYTHONPATH:"\n'
840	>	txt += 'if [ -z "$PYTHONPATH" ]; then\n'
841	>	txt += '   export PYTHONPATH=$RUNTIME_AREA/\n'
842	>	txt += 'else\n'
843	>	txt += '   export PYTHONPATH=$RUNTIME_AREA/:${PYTHONPATH}\n'
844	>	txt += 'echo "PYTHONPATH=$PYTHONPATH"\n'
845	>	txt += 'fi\n'
846	>	txt += '\n'
847	>
848	>	if self.pset != None:
849	>	psetName = 'pset.py'
850	>
851	>	txt += '\n'
852	>	if self.debug_wrapper == 1:
853	>	txt += 'echo "***** cat ' + psetName + ' *********"\n'
854	>	txt += 'cat ' + psetName + '\n'
855	>	txt += 'echo "****** end ' + psetName + ' ********"\n'
856	>	txt += '\n'
857	>	txt += 'echo "***********************" \n'
858	>	txt += 'which edmConfigHash \n'
859	>	txt += 'echo "***********************" \n'
860	>	txt += 'edmConfigHash ' + psetName + ' \n'
861	>	txt += 'PSETHASH=`edmConfigHash ' + psetName + '` \n'
862	>	txt += 'echo "PSETHASH = $PSETHASH" \n'
863	>	#### FEDE temporary fix for noEdm files #####
864	>	txt += 'if [ -z "$PSETHASH" ]; then \n'
865	>	txt += '   export PSETHASH=null\n'
866	>	txt += 'fi \n'
867	>	#############################################
868	>	txt += '\n'
869	>	return txt
870	>
871	>
872		def executableName(self):
873	<	if self.scriptExe: #CarlosDaniele
873	>	if self.scriptExe:
874		return "sh "
875		else:
876		return self.executable
877
878		def executableArgs(self):
879	<	if self.scriptExe:#CarlosDaniele
880	<	return   self.scriptExe + " $NJob"
881	<	else:
882	<	return " -p pset.cfg"
879	>	if self.scriptExe:
880	>	return self.scriptExe + " $NJob $AdditionalArgs"
881	>	else:
882	>	return " -j $RUNTIME_AREA/crab_fjr_$NJob.xml -p pset.py"
883
884		def inputSandbox(self, nj):
885		"""
886		Returns a list of filenames to be put in JDL input sandbox.
887		"""
888		inp_box = []
1054	–	# # dict added to delete duplicate from input sandbox file list
1055	–	# seen = {}
1056	–	## code
889		if os.path.isfile(self.tgzNameWithPath):
890		inp_box.append(self.tgzNameWithPath)
891	<	if os.path.isfile(self.MLtgzfile):
892	<	inp_box.append(self.MLtgzfile)
893	<	## config
1062	<	if not self.pset is None:
1063	<	inp_box.append(common.work_space.pathForTgz() + 'job/' + self.configFilename())
1064	<	## additional input files
1065	<	for file in self.additional_inbox_files:
1066	<	inp_box.append(file)
891	>	if os.path.isfile(self.argsFile):
892	>	inp_box.append(self.argsFile)
893	>	inp_box.append(common.work_space.jobDir() + self.scriptName)
894		return inp_box
895
896		def outputSandbox(self, nj):
#	Line 1074 | Line 901 | class Cmssw(JobType):
901
902		## User Declared output files
903		for out in (self.output_file+self.output_file_sandbox):
904	<	n_out = nj + 1
905	<	out_box.append(self.numberFile_(out,str(n_out)))
904	>	n_out = nj + 1
905	>	out_box.append(numberFile(out,str(n_out)))
906		return out_box
907
1081	–	def prepareSteeringCards(self):
1082	–	"""
1083	–	Make initial modifications of the user's steering card file.
1084	–	"""
1085	–	return
908
909		def wsRenameOutput(self, nj):
910		"""
911		Returns part of a job script which renames the produced files.
912		"""
913
914	<	txt = '\n'
915	<	txt += '# directory content\n'
916	<	txt += 'ls \n'
914	>	txt = '\n#Written by cms_cmssw::wsRenameOutput\n'
915	>	txt += 'echo ">>> current directory (SOFTWARE_DIR): $SOFTWARE_DIR" \n'
916	>	txt += 'echo ">>> current directory content:"\n'
917	>	if self.debug_wrapper==1:
918	>	txt += 'ls -Al\n'
919	>	txt += '\n'
920
921	<	for fileWithSuffix in (self.output_file+self.output_file_sandbox):
922	<	output_file_num = self.numberFile_(fileWithSuffix, '$NJob')
921	>	for fileWithSuffix in (self.output_file):
922	>	output_file_num = numberFile(fileWithSuffix, '$OutUniqueID')
923		txt += '\n'
924		txt += '# check output file\n'
925	<	txt += 'ls '+fileWithSuffix+'\n'
926	<	txt += 'ls_result=$?\n'
927	<	txt += 'if [ $ls_result -ne 0 ] ; then\n'
928	<	#txt += '   JOB_EXIT_STATUS=60302\n'
929	<	### FEDE
930	<	txt += '   exit_status=60302\n'
931	<	####
932	<	txt += '   echo "ERROR: Problem with output file"\n'
933	<	if common.scheduler.boss_scheduler_name == 'condor_g':
925	>	txt += 'if [ -e ./'+fileWithSuffix+' ] ; then\n'
926	>	if (self.copy_data == 1):  # For OSG nodes, file is in $WORKING_DIR, should not be moved to $RUNTIME_AREA
927	>	txt += '    mv '+fileWithSuffix+' '+output_file_num+'\n'
928	>	txt += '    ln -s `pwd`/'+output_file_num+' $RUNTIME_AREA/'+fileWithSuffix+'\n'
929	>	else:
930	>	txt += '    mv '+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n'
931	>	txt += '    ln -s $RUNTIME_AREA/'+output_file_num+' $RUNTIME_AREA/'+fileWithSuffix+'\n'
932	>	txt += 'else\n'
933	>	txt += '    job_exit_code=60302\n'
934	>	txt += '    echo "WARNING: Output file '+fileWithSuffix+' not found"\n'
935	>	if common.scheduler.name().upper() == 'CONDOR_G':
936		txt += '    if [ $middleware == OSG ]; then \n'
937		txt += '        echo "prepare dummy output file"\n'
938		txt += '        echo "Processing of job output failed" > $RUNTIME_AREA/'+output_file_num+'\n'
939		txt += '    fi \n'
1113	–	txt += 'else\n'
1114	–	txt += '   cp '+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n'
940		txt += 'fi\n'
941	<
942	<	txt += 'cd $RUNTIME_AREA\n'
943	<	txt += 'cd $RUNTIME_AREA\n'
1119	<	### OLI_DANIELE
1120	<	txt += 'if [ $middleware == OSG ]; then\n'
1121	<	txt += '    cd $RUNTIME_AREA\n'
1122	<	txt += '    echo "Remove working directory: $WORKING_DIR"\n'
1123	<	txt += '    /bin/rm -rf $WORKING_DIR\n'
1124	<	txt += '    if [ -d $WORKING_DIR ] ;then\n'
1125	<	txt += '    echo "SET_EXE 60999 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after cleanup of WN"\n'
1126	<	txt += '    echo "JOB_EXIT_STATUS = 60999"\n'
1127	<	txt += '    echo "JobExitCode=60999" | tee -a $RUNTIME_AREA/$repo\n'
1128	<	txt += '    dumpStatus $RUNTIME_AREA/$repo\n'
1129	<	txt += '        rm -f $RUNTIME_AREA/$repo \n'
1130	<	txt += '        echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1131	<	txt += '        echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1132	<	txt += '    fi\n'
1133	<	txt += 'fi\n'
1134	<	txt += '\n'
1135	<
1136	<	file_list = ''
1137	<	## Add to filelist only files to be possibly copied to SE
1138	<	for fileWithSuffix in self.output_file:
1139	<	output_file_num = self.numberFile_(fileWithSuffix, '$NJob')
1140	<	file_list=file_list+output_file_num+' '
1141	<	file_list=file_list[:-1]
1142	<	txt += 'file_list="'+file_list+'"\n'
941	>	file_list = []
942	>	for fileWithSuffix in (self.output_file):
943	>	file_list.append(numberFile('$SOFTWARE_DIR/'+fileWithSuffix, '$OutUniqueID'))
944
945	+	txt += 'file_list="'+string.join(file_list,',')+'"\n'
946	+	txt += '\n'
947	+	txt += 'echo ">>> current directory (SOFTWARE_DIR): $SOFTWARE_DIR" \n'
948	+	txt += 'echo ">>> current directory content:"\n'
949	+	if self.debug_wrapper==1:
950	+	txt += 'ls -Al\n'
951	+	txt += '\n'
952	+	txt += 'cd $RUNTIME_AREA\n'
953	+	txt += 'echo ">>> current directory (RUNTIME_AREA):  $RUNTIME_AREA"\n'
954		return txt
955
1146	–	def numberFile_(self, file, txt):
1147	–	"""
1148	–	append _'txt' before last extension of a file
1149	–	"""
1150	–	p = string.split(file,".")
1151	–	# take away last extension
1152	–	name = p[0]
1153	–	for x in p[1:-1]:
1154	–	name=name+"."+x
1155	–	# add "_txt"
1156	–	if len(p)>1:
1157	–	ext = p[len(p)-1]
1158	–	result = name + '_' + txt + "." + ext
1159	–	else:
1160	–	result = name + '_' + txt
1161	–
1162	–	return result
1163	–
956		def getRequirements(self, nj=[]):
957		"""
958	<	return job requirements to add to jdl files
958	>	return job requirements to add to jdl files
959		"""
960		req = ''
961		if self.version:
962		req='Member("VO-cms-' + \
963		self.version + \
964		'", other.GlueHostApplicationSoftwareRunTimeEnvironment)'
965	<	# if self.executable_arch:
966	<	#     req='Member("VO-cms-' + \
967	<	#          self.executable_arch + \
968	<	#          '", other.GlueHostApplicationSoftwareRunTimeEnvironment)'
965	>	if self.executable_arch:
966	>	req+=' && Member("VO-cms-' + \
967	>	self.executable_arch + \
968	>	'", other.GlueHostApplicationSoftwareRunTimeEnvironment)'
969
970		req = req + ' && (other.GlueHostNetworkAdapterOutboundIP)'
971	+	if ( common.scheduler.name() in ["glite"] ):
972	+	## 25-Jun-2009 SL: patch to use Cream enabled WMS
973	+	if ( self.cfg_params.get('GRID.use_cream',None) ):
974	+	req += ' && (other.GlueCEStateStatus == "Production" || other.GlueCEStateStatus == "Special")'
975	+	else:
976	+	req += ' && other.GlueCEStateStatus == "Production" '
977
978		return req
979
980		def configFilename(self):
981		""" return the config filename """
982	<	return self.name()+'.cfg'
982	>	return self.name()+'.py'
983
1186	–	### OLI_DANIELE
984		def wsSetupCMSOSGEnvironment_(self):
985		"""
986		Returns part of a job script which is prepares
987		the execution environment and which is common for all CMS jobs.
988		"""
989	<	txt = '\n'
990	<	txt += '   echo "### SETUP CMS OSG  ENVIRONMENT ###"\n'
991	<	txt += '   if [ -f $GRID3_APP_DIR/cmssoft/cmsset_default.sh ] ;then\n'
992	<	txt += '      # Use $GRID3_APP_DIR/cmssoft/cmsset_default.sh to setup cms software\n'
993	<	txt += '       export SCRAM_ARCH='+self.executable_arch+'\n'
994	<	txt += '       source $GRID3_APP_DIR/cmssoft/cmsset_default.sh '+self.version+'\n'
1198	<	txt += '   elif [ -f $OSG_APP/cmssoft/cms/cmsset_default.sh ] ;then\n'
989	>	txt = '\n#Written by cms_cmssw::wsSetupCMSOSGEnvironment_\n'
990	>	txt += '    echo ">>> setup CMS OSG environment:"\n'
991	>	txt += '    echo "set SCRAM ARCH to ' + self.executable_arch + '"\n'
992	>	txt += '    export SCRAM_ARCH='+self.executable_arch+'\n'
993	>	txt += '    echo "SCRAM_ARCH = $SCRAM_ARCH"\n'
994	>	txt += '    if [ -f $OSG_APP/cmssoft/cms/cmsset_default.sh ] ;then\n'
995		txt += '      # Use $OSG_APP/cmssoft/cms/cmsset_default.sh to setup cms software\n'
996	<	txt += '       export SCRAM_ARCH='+self.executable_arch+'\n'
997	<	txt += '       source $OSG_APP/cmssoft/cms/cmsset_default.sh '+self.version+'\n'
998	<	txt += '   else\n'
999	<	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'
1000	<	txt += '       echo "JOB_EXIT_STATUS = 10020"\n'
1001	<	txt += '       echo "JobExitCode=10020" | tee -a $RUNTIME_AREA/$repo\n'
1206	<	txt += '       dumpStatus $RUNTIME_AREA/$repo\n'
1207	<	txt += '       rm -f $RUNTIME_AREA/$repo \n'
1208	<	txt += '       echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1209	<	txt += '       echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1210	<	txt += '       exit 1\n'
1211	<	txt += '\n'
1212	<	txt += '       echo "Remove working directory: $WORKING_DIR"\n'
1213	<	txt += '       cd $RUNTIME_AREA\n'
1214	<	txt += '       /bin/rm -rf $WORKING_DIR\n'
1215	<	txt += '       if [ -d $WORKING_DIR ] ;then\n'
1216	<	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'
1217	<	txt += '        echo "JOB_EXIT_STATUS = 10017"\n'
1218	<	txt += '        echo "JobExitCode=10017" | tee -a $RUNTIME_AREA/$repo\n'
1219	<	txt += '        dumpStatus $RUNTIME_AREA/$repo\n'
1220	<	txt += '            rm -f $RUNTIME_AREA/$repo \n'
1221	<	txt += '            echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1222	<	txt += '            echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1223	<	txt += '       fi\n'
1224	<	txt += '\n'
1225	<	txt += '       exit 1\n'
1226	<	txt += '   fi\n'
996	>	txt += '        source $OSG_APP/cmssoft/cms/cmsset_default.sh '+self.version+'\n'
997	>	txt += '    else\n'
998	>	txt += '        echo "ERROR ==> $OSG_APP/cmssoft/cms/cmsset_default.sh file not found"\n'
999	>	txt += '        job_exit_code=10020\n'
1000	>	txt += '        func_exit\n'
1001	>	txt += '    fi\n'
1002		txt += '\n'
1003	<	txt += '   echo "SET_CMS_ENV 0 ==> setup cms environment ok"\n'
1004	<	txt += '   echo " END SETUP CMS OSG  ENVIRONMENT "\n'
1003	>	txt += '    echo "==> setup cms environment ok"\n'
1004	>	txt += '    echo "SCRAM_ARCH = $SCRAM_ARCH"\n'
1005
1006		return txt
1007	<
1233	<	### OLI_DANIELE
1007	>
1008		def wsSetupCMSLCGEnvironment_(self):
1009		"""
1010		Returns part of a job script which is prepares
1011		the execution environment and which is common for all CMS jobs.
1012		"""
1013	<	txt  = '   \n'
1014	<	txt += '   echo " ### SETUP CMS LCG  ENVIRONMENT ### "\n'
1015	<	txt += '   if [ ! $VO_CMS_SW_DIR ] ;then\n'
1016	<	txt += '       echo "SET_CMS_ENV 10031 ==> ERROR CMS software dir not found on WN `hostname`"\n'
1017	<	txt += '       echo "JOB_EXIT_STATUS = 10031" \n'
1018	<	txt += '       echo "JobExitCode=10031" | tee -a $RUNTIME_AREA/$repo\n'
1019	<	txt += '       dumpStatus $RUNTIME_AREA/$repo\n'
1020	<	txt += '       rm -f $RUNTIME_AREA/$repo \n'
1021	<	txt += '       echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1022	<	txt += '       echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1023	<	txt += '       exit 1\n'
1024	<	txt += '   else\n'
1025	<	txt += '       echo "Sourcing environment... "\n'
1026	<	txt += '       if [ ! -s $VO_CMS_SW_DIR/cmsset_default.sh ] ;then\n'
1027	<	txt += '           echo "SET_CMS_ENV 10020 ==> ERROR cmsset_default.sh file not found into dir $VO_CMS_SW_DIR"\n'
1028	<	txt += '           echo "JOB_EXIT_STATUS = 10020"\n'
1029	<	txt += '           echo "JobExitCode=10020" | tee -a $RUNTIME_AREA/$repo\n'
1030	<	txt += '           dumpStatus $RUNTIME_AREA/$repo\n'
1031	<	txt += '           rm -f $RUNTIME_AREA/$repo \n'
1032	<	txt += '           echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1033	<	txt += '           echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1034	<	txt += '           exit 1\n'
1035	<	txt += '       fi\n'
1036	<	txt += '       echo "sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
1037	<	txt += '       source $VO_CMS_SW_DIR/cmsset_default.sh\n'
1038	<	txt += '       result=$?\n'
1039	<	txt += '       if [ $result -ne 0 ]; then\n'
1040	<	txt += '           echo "SET_CMS_ENV 10032 ==> ERROR problem sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
1041	<	txt += '           echo "JOB_EXIT_STATUS = 10032"\n'
1042	<	txt += '           echo "JobExitCode=10032" | tee -a $RUNTIME_AREA/$repo\n'
1043	<	txt += '           dumpStatus $RUNTIME_AREA/$repo\n'
1044	<	txt += '           rm -f $RUNTIME_AREA/$repo \n'
1045	<	txt += '           echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1046	<	txt += '           echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1047	<	txt += '           exit 1\n'
1048	<	txt += '       fi\n'
1049	<	txt += '   fi\n'
1050	<	txt += '   \n'
1051	<	txt += '   echo "SET_CMS_ENV 0 ==> setup cms environment ok"\n'
1052	<	txt += '   echo "### END SETUP CMS LCG ENVIRONMENT ###"\n'
1013	>	txt = '\n#Written by cms_cmssw::wsSetupCMSLCGEnvironment_\n'
1014	>	txt += '    echo ">>> setup CMS LCG environment:"\n'
1015	>	txt += '    echo "set SCRAM ARCH and BUILD_ARCH to ' + self.executable_arch + ' ###"\n'
1016	>	txt += '    export SCRAM_ARCH='+self.executable_arch+'\n'
1017	>	txt += '    export BUILD_ARCH='+self.executable_arch+'\n'
1018	>	txt += '    if [ ! $VO_CMS_SW_DIR ] ;then\n'
1019	>	txt += '        echo "ERROR ==> CMS software dir not found on WN `hostname`"\n'
1020	>	txt += '        job_exit_code=10031\n'
1021	>	txt += '        func_exit\n'
1022	>	txt += '    else\n'
1023	>	txt += '        echo "Sourcing environment... "\n'
1024	>	txt += '        if [ ! -s $VO_CMS_SW_DIR/cmsset_default.sh ] ;then\n'
1025	>	txt += '            echo "ERROR ==> cmsset_default.sh file not found into dir $VO_CMS_SW_DIR"\n'
1026	>	txt += '            job_exit_code=10020\n'
1027	>	txt += '            func_exit\n'
1028	>	txt += '        fi\n'
1029	>	txt += '        echo "sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
1030	>	txt += '        source $VO_CMS_SW_DIR/cmsset_default.sh\n'
1031	>	txt += '        result=$?\n'
1032	>	txt += '        if [ $result -ne 0 ]; then\n'
1033	>	txt += '            echo "ERROR ==> problem sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
1034	>	txt += '            job_exit_code=10032\n'
1035	>	txt += '            func_exit\n'
1036	>	txt += '        fi\n'
1037	>	txt += '    fi\n'
1038	>	txt += '    \n'
1039	>	txt += '    echo "==> setup cms environment ok"\n'
1040	>	return txt
1041	>
1042	>	def wsModifyReport(self, nj):
1043	>	"""
1044	>	insert the part of the script that modifies the FrameworkJob Report
1045	>	"""
1046	>
1047	>	txt = ''
1048	>	if (self.copy_data == 1):
1049	>	txt = '\n#Written by cms_cmssw::wsModifyReport\n'
1050	>
1051	>	txt += 'echo ">>> Modify Job Report:" \n'
1052	>	txt += 'chmod a+x $RUNTIME_AREA/ProdCommon/FwkJobRep/ModifyJobReport.py\n'
1053	>	txt += 'echo "CMSSW_VERSION = $CMSSW_VERSION"\n\n'
1054	>
1055	>	args = 'fjr $RUNTIME_AREA/crab_fjr_$NJob.xml json $RUNTIME_AREA/resultCopyFile n_job $OutUniqueID PrimaryDataset $PrimaryDataset  ApplicationFamily $ApplicationFamily ApplicationName $executable cmssw_version $CMSSW_VERSION psethash $PSETHASH'
1056	>
1057	>	if (self.publish_data == 1):
1058	>	txt += 'ProcessedDataset='+self.processedDataset+'\n'
1059	>	txt += 'echo "ProcessedDataset = $ProcessedDataset"\n'
1060	>	args += ' UserProcessedDataset $USER-$ProcessedDataset-$PSETHASH'
1061	>
1062	>	txt += 'echo "$RUNTIME_AREA/ProdCommon/FwkJobRep/ModifyJobReport.py '+str(args)+'"\n'
1063	>	txt += '$RUNTIME_AREA/ProdCommon/FwkJobRep/ModifyJobReport.py '+str(args)+'\n'
1064	>	txt += 'modifyReport_result=$?\n'
1065	>	txt += 'if [ $modifyReport_result -ne 0 ]; then\n'
1066	>	txt += '    modifyReport_result=70500\n'
1067	>	txt += '    job_exit_code=$modifyReport_result\n'
1068	>	txt += '    echo "ModifyReportResult=$modifyReport_result" | tee -a $RUNTIME_AREA/$repo\n'
1069	>	txt += '    echo "WARNING: Problem with ModifyJobReport"\n'
1070	>	txt += 'else\n'
1071	>	txt += '    mv NewFrameworkJobReport.xml $RUNTIME_AREA/crab_fjr_$NJob.xml\n'
1072	>	txt += 'fi\n'
1073	>	return txt
1074	>
1075	>	def wsParseFJR(self):
1076	>	"""
1077	>	Parse the FrameworkJobReport to obtain useful infos
1078	>	"""
1079	>	txt = '\n#Written by cms_cmssw::wsParseFJR\n'
1080	>	txt += 'echo ">>> Parse FrameworkJobReport crab_fjr.xml"\n'
1081	>	txt += 'if [ -s $RUNTIME_AREA/crab_fjr_$NJob.xml ]; then\n'
1082	>	txt += '    if [ -s $RUNTIME_AREA/parseCrabFjr.py ]; then\n'
1083	>	txt += '        cmd_out=`python $RUNTIME_AREA/parseCrabFjr.py --input $RUNTIME_AREA/crab_fjr_$NJob.xml --dashboard $MonitorID,$MonitorJobID '+self.debugWrap+'`\n'
1084	>	if self.debug_wrapper==1 :
1085	>	txt += '        echo "Result of parsing the FrameworkJobReport crab_fjr.xml: $cmd_out"\n'
1086	>	txt += '        executable_exit_status=`python $RUNTIME_AREA/parseCrabFjr.py --input $RUNTIME_AREA/crab_fjr_$NJob.xml --exitcode`\n'
1087	>	txt += '        if [ $executable_exit_status -eq 50115 ];then\n'
1088	>	txt += '            echo ">>> crab_fjr.xml contents: "\n'
1089	>	txt += '            cat $RUNTIME_AREA/crab_fjr_$NJob.xml\n'
1090	>	txt += '            echo "Wrong FrameworkJobReport --> does not contain useful info. ExitStatus: $executable_exit_status"\n'
1091	>	txt += '        elif [ $executable_exit_status -eq -999 ];then\n'
1092	>	txt += '            echo "ExitStatus from FrameworkJobReport not available. not available. Using exit code of executable from command line."\n'
1093	>	txt += '        else\n'
1094	>	txt += '            echo "Extracted ExitStatus from FrameworkJobReport parsing output: $executable_exit_status"\n'
1095	>	txt += '        fi\n'
1096	>	txt += '    else\n'
1097	>	txt += '        echo "CRAB python script to parse CRAB FrameworkJobReport crab_fjr.xml is not available, using exit code of executable from command line."\n'
1098	>	txt += '    fi\n'
1099	>	#### Patch to check input data reading for CMSSW16x Hopefully we-ll remove it asap
1100	>	txt += '    if [ $executable_exit_status -eq 0 ];then\n'
1101	>	txt += '        echo ">>> Executable succeded  $executable_exit_status"\n'
1102	>	txt += '    fi\n'
1103	>	txt += 'else\n'
1104	>	txt += '    echo "CRAB FrameworkJobReport crab_fjr.xml is not available, using exit code of executable from command line."\n'
1105	>	txt += 'fi\n'
1106	>	txt += '\n'
1107	>	txt += 'if [ $executable_exit_status -ne 0 ];then\n'
1108	>	txt += '    echo ">>> Executable failed  $executable_exit_status"\n'
1109	>	txt += '    echo "ExeExitCode=$executable_exit_status" | tee -a $RUNTIME_AREA/$repo\n'
1110	>	txt += '    echo "EXECUTABLE_EXIT_STATUS = $executable_exit_status"\n'
1111	>	txt += '    job_exit_code=$executable_exit_status\n'
1112	>	txt += '    func_exit\n'
1113	>	txt += 'fi\n\n'
1114	>	txt += 'echo "ExeExitCode=$executable_exit_status" | tee -a $RUNTIME_AREA/$repo\n'
1115	>	txt += 'echo "EXECUTABLE_EXIT_STATUS = $executable_exit_status"\n'
1116	>	txt += 'job_exit_code=$executable_exit_status\n'
1117	>
1118		return txt
1119
1120		def setParam_(self, param, value):
#	Line 1284 | Line 1123 | class Cmssw(JobType):
1123		def getParams(self):
1124		return self._params
1125
1126	<	def setTaskid_(self):
1127	<	self._taskId = self.cfg_params['taskId']
1128	<
1129	<	def getTaskid(self):
1130	<	return self._taskId
1131	<
1132	<	#######################################################################
1133	<	def uniquelist(self, old):
1134	<	"""
1135	<	remove duplicates from a list
1136	<	"""
1137	<	nd={}
1138	<	for e in old:
1139	<	nd[e]=0
1140	<	return nd.keys()
1126	>	def outList(self,list=False):
1127	>	"""
1128	>	check the dimension of the output files
1129	>	"""
1130	>	txt = ''
1131	>	txt += 'echo ">>> list of expected files on output sandbox"\n'
1132	>	listOutFiles = []
1133	>	stdout = 'CMSSW_$NJob.stdout'
1134	>	stderr = 'CMSSW_$NJob.stderr'
1135	>	if len(self.output_file) <= 0:
1136	>	msg ="WARNING: no output files name have been defined!!\n"
1137	>	msg+="\tno output files will be reported back/staged\n"
1138	>	common.logger.info(msg)
1139	>
1140	>	if (self.return_data == 1):
1141	>	for file in (self.output_file):
1142	>	listOutFiles.append(numberFile(file, '$OutUniqueID'))
1143	>	for file in (self.output_file_sandbox):
1144	>	listOutFiles.append(numberFile(file, '$NJob'))
1145	>	listOutFiles.append(stdout)
1146	>	listOutFiles.append(stderr)
1147	>
1148	>	txt += 'echo "output files: '+string.join(listOutFiles,' ')+'"\n'
1149	>	txt += 'filesToCheck="'+string.join(listOutFiles,' ')+'"\n'
1150	>	txt += 'export filesToCheck\n'
1151	>	taskinfo={}
1152	>	taskinfo['outfileBasename'] = self.output_file
1153	>	common._db.updateTask_(taskinfo)
1154	>
1155	>	if list : return self.output_file
1156	>	return txt
1157	>
1158	>	def checkCMSSWVersion(self, url = "https://cmstags.cern.ch/cgi-bin/CmsTC/", fileName = "ReleasesXML"):
1159	>	"""
1160	>	compare current CMSSW release and arch with allowed releases
1161	>	"""
1162	>
1163	>	downloader = Downloader(url)
1164	>	goodRelease = False
1165	>
1166	>	try:
1167	>	result = downloader.config(fileName)
1168	>	except:
1169	>	common.logger.info("ERROR: Problem reading file of allowed CMSSW releases.")
1170	>
1171	>	try:
1172	>	events = pulldom.parseString(result)
1173	>
1174	>	arch     = None
1175	>	release  = None
1176	>	relType  = None
1177	>	relState = None
1178	>	for (event, node) in events:
1179	>	if event == pulldom.START_ELEMENT:
1180	>	if node.tagName == 'architecture':
1181	>	arch = node.attributes.getNamedItem('name').nodeValue
1182	>	if node.tagName == 'project':
1183	>	relType = node.attributes.getNamedItem('type').nodeValue
1184	>	relState = node.attributes.getNamedItem('state').nodeValue
1185	>	if relType == 'Production' and relState == 'Announced':
1186	>	release = node.attributes.getNamedItem('label').nodeValue
1187	>	if self.executable_arch == arch and self.version == release:
1188	>	goodRelease = True
1189	>	return goodRelease
1190	>
1191	>	if not goodRelease:
1192	>	msg = "WARNING: %s on %s is not a supported release. " % \
1193	>	(self.version, self.executable_arch)
1194	>	msg += "Submission may fail."
1195	>	common.logger.info(msg)
1196	>	except:
1197	>	common.logger.info("Problems parsing file of allowed CMSSW releases.")
1198	>
1199	>	return goodRelease
1200	>

Diff Legend

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