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 1.1 by slacapra, Thu Apr 6 16:18:17 2006 UTC vs.
Revision 1.195 by slacapra, Wed May 28 15:42:01 2008 UTC

#	Line 2 | Line 2 | from JobType import JobType
2		from crab_logger import Logger
3		from crab_exceptions import *
4		from crab_util import *
5	+	from BlackWhiteListParser import BlackWhiteListParser
6		import common
6	–
7	–	import DataDiscovery
8	–	import DataLocation
7		import Scram
8	+	from LFNBaseName import *
9
10	<	import os, string, re
10	>	import os, string, glob
11
12		class Cmssw(JobType):
13	<	def __init__(self, cfg_params):
13	>	def __init__(self, cfg_params, ncjobs):
14		JobType.__init__(self, 'CMSSW')
15		common.logger.debug(3,'CMSSW::__init__')
16
17	<	self.analisys_common_info = {}
17	>	self.argsList = []
18	>
19	>	self._params = {}
20	>	self.cfg_params = cfg_params
21	>	# init BlackWhiteListParser
22	>	self.blackWhiteListParser = BlackWhiteListParser(cfg_params)
23	>
24	>	self.MaxTarBallSize = float(self.cfg_params.get('EDG.maxtarballsize',9.5))
25	>
26	>	# number of jobs requested to be created, limit obj splitting
27	>	self.ncjobs = ncjobs
28
29		log = common.logger
30	<
30	>
31		self.scram = Scram.Scram(cfg_params)
23	–	scramArea = ''
32		self.additional_inbox_files = []
33		self.scriptExe = ''
34		self.executable = ''
35	+	self.executable_arch = self.scram.getArch()
36		self.tgz_name = 'default.tgz'
37	+	self.scriptName = 'CMSSW.sh'
38	+	self.pset = ''
39	+	self.datasetPath = ''
40	+
41	+	# set FJR file name
42	+	self.fjrFileName = 'crab_fjr.xml'
43
44		self.version = self.scram.getSWVersion()
45	<	common.analisys_common_info['sw_version'] = self.version
45	>	version_array = self.version.split('_')
46	>	self.CMSSW_major = 0
47	>	self.CMSSW_minor = 0
48	>	self.CMSSW_patch = 0
49	>	try:
50	>	self.CMSSW_major = int(version_array[1])
51	>	self.CMSSW_minor = int(version_array[2])
52	>	self.CMSSW_patch = int(version_array[3])
53	>	except:
54	>	msg = "Cannot parse CMSSW version string: " + self.version + " for major and minor release number!"
55	>	raise CrabException(msg)
56
57		### collect Data cards
58	<	try:
59	<	self.owner = cfg_params['CMSSW.owner']
60	<	log.debug(6, "CMSSW::CMSSW(): owner = "+self.owner)
36	<	self.dataset = cfg_params['CMSSW.dataset']
37	<	log.debug(6, "CMSSW::CMSSW(): dataset = "+self.dataset)
38	<	except KeyError:
39	<	msg = "Error: owner and/or dataset not defined "
58	>
59	>	if not cfg_params.has_key('CMSSW.datasetpath'):
60	>	msg = "Error: datasetpath not defined "
61		raise CrabException(msg)
62	+	tmp =  cfg_params['CMSSW.datasetpath']
63	+	log.debug(6, "CMSSW::CMSSW(): datasetPath = "+tmp)
64	+	if string.lower(tmp)=='none':
65	+	self.datasetPath = None
66	+	self.selectNoInput = 1
67	+	else:
68	+	self.datasetPath = tmp
69	+	self.selectNoInput = 0
70
71		self.dataTiers = []
72	<	try:
73	<	tmpDataTiers = string.split(cfg_params['CMSSW.data_tier'],',')
45	<	for tmp in tmpDataTiers:
46	<	tmp=string.strip(tmp)
47	<	self.dataTiers.append(tmp)
48	<	pass
49	<	pass
50	<	except KeyError:
51	<	pass
52	<	log.debug(6, "Cmssw::Cmssw(): dataTiers = "+str(self.dataTiers))
72	>
73	>	self.debug_pset = cfg_params.get('USER.debug_pset',False)
74
75		## now the application
76	<	try:
77	<	self.executable = cfg_params['CMSSW.executable']
57	<	log.debug(6, "CMSSW::CMSSW(): executable = "+self.executable)
58	<	msg = "Default executable cmsRun overridden. Switch to " + self.executable
59	<	log.debug(3,msg)
60	<	except KeyError:
61	<	self.executable = 'cmsRun'
62	<	msg = "User executable not defined. Use cmsRun"
63	<	log.debug(3,msg)
64	<	pass
76	>	self.executable = cfg_params.get('CMSSW.executable','cmsRun')
77	>	log.debug(6, "CMSSW::CMSSW(): executable = "+self.executable)
78
79	<	try:
80	<	self.pset = cfg_params['CMSSW.pset']
81	<	log.debug(6, "Cmssw::Cmssw(): PSet file = "+self.pset)
79	>	if not cfg_params.has_key('CMSSW.pset'):
80	>	raise CrabException("PSet file missing. Cannot run cmsRun ")
81	>	self.pset = cfg_params['CMSSW.pset']
82	>	log.debug(6, "Cmssw::Cmssw(): PSet file = "+self.pset)
83	>	if self.pset.lower() != 'none' :
84		if (not os.path.exists(self.pset)):
85		raise CrabException("User defined PSet file "+self.pset+" does not exist")
86	<	except KeyError:
87	<	raise CrabException("PSet file missing. Cannot run cmsRun ")
86	>	else:
87	>	self.pset = None
88
89		# output files
90	<	try:
91	<	self.output_file = []
90	>	## stuff which must be returned always via sandbox
91	>	self.output_file_sandbox = []
92
93	<	tmp = cfg_params['CMSSW.output_file']
94	<	if tmp != '':
80	<	tmpOutFiles = string.split(cfg_params['CMSSW.output_file'],',')
81	<	log.debug(7, 'cmssw::cmssw(): output files '+str(tmpOutFiles))
82	<	for tmp in tmpOutFiles:
83	<	tmp=string.strip(tmp)
84	<	self.output_file.append(tmp)
85	<	pass
93	>	# add fjr report by default via sandbox
94	>	self.output_file_sandbox.append(self.fjrFileName)
95
96	<	else:
97	<	log.message("No output file defined: only stdout/err will be available")
96	>	# other output files to be returned via sandbox or copied to SE
97	>	self.output_file = []
98	>	tmp = cfg_params.get('CMSSW.output_file',None)
99	>	if tmp :
100	>	tmpOutFiles = string.split(tmp,',')
101	>	log.debug(7, 'cmssw::cmssw(): output files '+str(tmpOutFiles))
102	>	for tmp in tmpOutFiles:
103	>	tmp=string.strip(tmp)
104	>	self.output_file.append(tmp)
105		pass
106	<	pass
107	<	except KeyError:
108	<	log.message("No output file defined: only stdout/err will be available")
93	<	pass
106	>	else:
107	>	log.message("No output file defined: only stdout/err and the CRAB Framework Job Report will be available\n")
108	>	pass
109
110		# script_exe file as additional file in inputSandbox
111	<	try:
112	<	self.scriptExe = cfg_params['CMSSW.script_exe']
113	<	self.additional_inbox_files.append(self.scriptExe)
114	<	except KeyError:
115	<	pass
116	<	if self.scriptExe != '':
117	<	if os.path.isfile(self.scriptExe):
118	<	pass
119	<	else:
120	<	log.message("WARNING. file "+self.scriptExe+" not found")
121	<	sys.exit()
107	<
111	>	self.scriptExe = cfg_params.get('USER.script_exe',None)
112	>	if self.scriptExe :
113	>	if not os.path.isfile(self.scriptExe):
114	>	msg ="ERROR. file "+self.scriptExe+" not found"
115	>	raise CrabException(msg)
116	>	self.additional_inbox_files.append(string.strip(self.scriptExe))
117	>
118	>	if self.datasetPath == None and self.pset == None and self.scriptExe == '' :
119	>	msg ="Error. script_exe  not defined"
120	>	raise CrabException(msg)
121	>
122		## additional input files
123	<	try:
124	<	tmpAddFiles = string.split(cfg_params['CMSSW.additional_input_files'],',')
123	>	if cfg_params.has_key('USER.additional_input_files'):
124	>	tmpAddFiles = string.split(cfg_params['USER.additional_input_files'],',')
125		for tmp in tmpAddFiles:
126	<	tmp=string.strip(tmp)
127	<	self.additional_inbox_files.append(tmp)
126	>	tmp = string.strip(tmp)
127	>	dirname = ''
128	>	if not tmp[0]=="/": dirname = "."
129	>	files = []
130	>	if string.find(tmp,"*")>-1:
131	>	files = glob.glob(os.path.join(dirname, tmp))
132	>	if len(files)==0:
133	>	raise CrabException("No additional input file found with this pattern: "+tmp)
134	>	else:
135	>	files.append(tmp)
136	>	for file in files:
137	>	if not os.path.exists(file):
138	>	raise CrabException("Additional input file not found: "+file)
139	>	pass
140	>	self.additional_inbox_files.append(string.strip(file))
141		pass
142		pass
143	<	except KeyError:
144	<	pass
143	>	common.logger.debug(5,"Additional input files: "+str(self.additional_inbox_files))
144	>	pass
145
146	<	try:
147	<	self.total_number_of_events = int(cfg_params['CMSSW.total_number_of_events'])
148	<	except KeyError:
149	<	msg = 'Must define total_number_of_events and job_number_of_events'
150	<	raise CrabException(msg)
151	<
152	<	#Marco: FirstEvent is nolonger used inside PSet
126	<	#        try:
127	<	#            self.first = int(cfg_params['CMSSW.first_event'])
128	<	#        except KeyError:
129	<	#            self.first = 0
130	<	#            pass
131	<	#        log.debug(6, "Orca::Orca(): total number of events = "+`self.total_number_of_events`)
132	<	#log.debug(6, "Orca::Orca(): events per job = "+`self.job_number_of_events`)
133	<	#        log.debug(6, "Orca::Orca(): first event = "+`self.first`)
134	<
135	<	CEBlackList = []
136	<	try:
137	<	tmpBad = string.split(cfg_params['EDG.ce_black_list'],',')
138	<	for tmp in tmpBad:
139	<	tmp=string.strip(tmp)
140	<	CEBlackList.append(tmp)
141	<	except KeyError:
142	<	pass
146	>	## Events per job
147	>	if cfg_params.has_key('CMSSW.events_per_job'):
148	>	self.eventsPerJob =int( cfg_params['CMSSW.events_per_job'])
149	>	self.selectEventsPerJob = 1
150	>	else:
151	>	self.eventsPerJob = -1
152	>	self.selectEventsPerJob = 0
153
154	<	self.reCEBlackList=[]
155	<	for bad in CEBlackList:
156	<	self.reCEBlackList.append(re.compile( bad ))
154	>	## number of jobs
155	>	if cfg_params.has_key('CMSSW.number_of_jobs'):
156	>	self.theNumberOfJobs =int( cfg_params['CMSSW.number_of_jobs'])
157	>	self.selectNumberOfJobs = 1
158	>	else:
159	>	self.theNumberOfJobs = 0
160	>	self.selectNumberOfJobs = 0
161
162	<	common.logger.debug(5,'CEBlackList: '+str(CEBlackList))
162	>	if cfg_params.has_key('CMSSW.total_number_of_events'):
163	>	self.total_number_of_events = int(cfg_params['CMSSW.total_number_of_events'])
164	>	self.selectTotalNumberEvents = 1
165	>	if self.selectNumberOfJobs  == 1:
166	>	if int(self.total_number_of_events) < int(self.theNumberOfJobs):
167	>	msg = 'Must specify at least one event per job. total_number_of_events > number_of_jobs '
168	>	raise CrabException(msg)
169	>	else:
170	>	self.total_number_of_events = 0
171	>	self.selectTotalNumberEvents = 0
172
173	<	CEWhiteList = []
174	<	try:
175	<	tmpGood = string.split(cfg_params['EDG.ce_white_list'],',')
176	<	#tmpGood = ['cern']
177	<	for tmp in tmpGood:
178	<	tmp=string.strip(tmp)
179	<	#if (tmp == 'cnaf'): tmp = 'webserver' ########## warning: temp. patch
180	<	CEWhiteList.append(tmp)
181	<	except KeyError:
182	<	pass
173	>	if self.pset != None:
174	>	if ( (self.selectTotalNumberEvents + self.selectEventsPerJob + self.selectNumberOfJobs) != 2 ):
175	>	msg = 'Must define exactly two of total_number_of_events, events_per_job, or number_of_jobs.'
176	>	raise CrabException(msg)
177	>	else:
178	>	if (self.selectNumberOfJobs == 0):
179	>	msg = 'Must specify  number_of_jobs.'
180	>	raise CrabException(msg)
181	>
182	>	## New method of dealing with seeds
183	>	self.incrementSeeds = []
184	>	self.preserveSeeds = []
185	>	if cfg_params.has_key('CMSSW.preserve_seeds'):
186	>	tmpList = cfg_params['CMSSW.preserve_seeds'].split(',')
187	>	for tmp in tmpList:
188	>	tmp.strip()
189	>	self.preserveSeeds.append(tmp)
190	>	if cfg_params.has_key('CMSSW.increment_seeds'):
191	>	tmpList = cfg_params['CMSSW.increment_seeds'].split(',')
192	>	for tmp in tmpList:
193	>	tmp.strip()
194	>	self.incrementSeeds.append(tmp)
195	>
196	>	## Old method of dealing with seeds
197	>	## FUTURE: This is for old CMSSW and old CRAB. Can throw exceptions after a couple of CRAB releases and then
198	>	## remove
199	>	self.sourceSeed = cfg_params.get('CMSSW.pythia_seed',None)
200	>	if self.sourceSeed:
201	>	print "pythia_seed is a deprecated parameter. Use preserve_seeds or increment_seeds in the future.\n","Added to increment_seeds."
202	>	self.incrementSeeds.append('sourceSeed')
203	>	self.incrementSeeds.append('theSource')
204	>
205	>	self.sourceSeedVtx = cfg_params.get('CMSSW.vtx_seed',None)
206	>	if self.sourceSeedVtx:
207	>	print "vtx_seed is a deprecated parameter. Use preserve_seeds or increment_seeds in the future.\n","Added to increment_seeds."
208	>	self.incrementSeeds.append('VtxSmeared')
209	>
210	>	self.sourceSeedG4 = cfg_params.get('CMSSW.g4_seed',None)
211	>	if self.sourceSeedG4:
212	>	print "g4_seed is a deprecated parameter. Use preserve_seeds or increment_seeds in the future.\n","Added to increment_seeds."
213	>	self.incrementSeeds.append('g4SimHits')
214	>
215	>	self.sourceSeedMix = cfg_params.get('CMSSW.mix_seed',None)
216	>	if self.sourceSeedMix:
217	>	print "mix_seed is a deprecated parameter. Use preserve_seeds or increment_seeds in the future.\n","Added to increment_seeds."
218	>	self.incrementSeeds.append('mix')
219	>
220	>	self.firstRun = cfg_params.get('CMSSW.first_run',None)
221	>
222	>	if self.pset != None: #CarlosDaniele
223	>	import PsetManipulator as pp
224	>	PsetEdit = pp.PsetManipulator(self.pset) #Daniele Pset
225
226	<	#print 'CEWhiteList: ',CEWhiteList
162	<	self.reCEWhiteList=[]
163	<	for Good in CEWhiteList:
164	<	self.reCEWhiteList.append(re.compile( Good ))
226	>	# Copy/return
227
228	<	common.logger.debug(5,'CEWhiteList: '+str(CEWhiteList))
228	>	self.copy_data = int(cfg_params.get('USER.copy_data',0))
229	>	self.return_data = int(cfg_params.get('USER.return_data',0))
230
231		#DBSDLS-start
232	<	## Initialize the variables that are extracted from DBS/DLS and needed in other places of the code
232	>	## Initialize the variables that are extracted from DBS/DLS and needed in other places of the code
233		self.maxEvents=0  # max events available   ( --> check the requested nb. of evts in Creator.py)
234		self.DBSPaths={}  # all dbs paths requested ( --> input to the site local discovery script)
235	+	self.jobDestination=[]  # Site destination(s) for each job (list of lists)
236		## Perform the data location and discovery (based on DBS/DLS)
237	<	self.DataDiscoveryAndLocation(cfg_params)
238	<	#DBSDLS-end
237	>	## SL: Don't if NONE is specified as input (pythia use case)
238	>	blockSites = {}
239	>	if self.datasetPath:
240	>	blockSites = self.DataDiscoveryAndLocation(cfg_params)
241	>	#DBSDLS-end
242
243	+
244	+	## Select Splitting
245	+	if self.selectNoInput:
246	+	if self.pset == None:
247	+	self.jobSplittingForScript()
248	+	else:
249	+	self.jobSplittingNoInput()
250	+	else:
251	+	self.jobSplittingByBlocks(blockSites)
252	+
253	+	# modify Pset
254	+	if self.pset != None:
255	+	try:
256	+	# Add FrameworkJobReport to parameter-set, set max events.
257	+	# Reset later for data jobs by writeCFG which does all modifications
258	+	PsetEdit.addCrabFJR(self.fjrFileName) # FUTURE: Job report addition not needed by CMSSW>1.5
259	+	PsetEdit.maxEvent(self.eventsPerJob)
260	+	PsetEdit.psetWriter(self.configFilename())
261	+	except:
262	+	msg='Error while manipulating ParameterSet: exiting...'
263	+	raise CrabException(msg)
264		self.tgzNameWithPath = self.getTarBall(self.executable)
265
266		def DataDiscoveryAndLocation(self, cfg_params):
267
268	<	fun = "CMSSW::DataDiscoveryAndLocation()"
268	>	import DataDiscovery
269	>	import DataLocation
270	>	common.logger.debug(10,"CMSSW::DataDiscoveryAndLocation()")
271	>
272	>	datasetPath=self.datasetPath
273
274		## Contact the DBS
275	+	common.logger.message("Contacting Data Discovery Services ...")
276		try:
277	<	self.pubdata=DataDiscovery.DataDiscovery(self.owner,
185	<	self.dataset,
186	<	self.dataTiers,
187	<	cfg_params)
277	>	self.pubdata=DataDiscovery.DataDiscovery(datasetPath, cfg_params)
278		self.pubdata.fetchDBSInfo()
279
280		except DataDiscovery.NotExistingDatasetError, ex :
281		msg = 'ERROR ***: failed Data Discovery in DBS : %s'%ex.getErrorMessage()
282		raise CrabException(msg)
193	–
283		except DataDiscovery.NoDataTierinProvenanceError, ex :
284		msg = 'ERROR ***: failed Data Discovery in DBS : %s'%ex.getErrorMessage()
285		raise CrabException(msg)
286		except DataDiscovery.DataDiscoveryError, ex:
287	<	msg = 'ERROR ***: failed Data Discovery in DBS  %s'%ex.getErrorMessage()
287	>	msg = 'ERROR ***: failed Data Discovery in DBS :  %s'%ex.getErrorMessage()
288		raise CrabException(msg)
289
290	<	## get list of all required data in the form of dbs paths  (dbs path = /dataset/datatier/owner)
291	<	self.DBSPaths=self.pubdata.getDBSPaths()
292	<	common.logger.message("Required data are : ")
204	<	for path in self.DBSPaths:
205	<	common.logger.message(" --> "+path )
290	>	self.filesbyblock=self.pubdata.getFiles()
291	>	self.eventsbyblock=self.pubdata.getEventsPerBlock()
292	>	self.eventsbyfile=self.pubdata.getEventsPerFile()
293
294		## get max number of events
295	<	common.logger.debug(10,"number of events for primary fileblocks %i"%self.pubdata.getMaxEvents())
209	<	self.maxEvents=self.pubdata.getMaxEvents() ##  self.maxEvents used in Creator.py
210	<	common.logger.message("\nThe number of available events is %s"%self.maxEvents)
211	<
212	<	## get fileblocks corresponding to the required data
213	<	fb=self.pubdata.getFileBlocks()
214	<	common.logger.debug(5,"fileblocks are %s"%fb)
295	>	self.maxEvents=self.pubdata.getMaxEvents()
296
297		## Contact the DLS and build a list of sites hosting the fileblocks
298		try:
299	<	dataloc=DataLocation.DataLocation(self.pubdata.getFileBlocks(),cfg_params)
299	>	dataloc=DataLocation.DataLocation(self.filesbyblock.keys(),cfg_params)
300		dataloc.fetchDLSInfo()
301		except DataLocation.DataLocationError , ex:
302		msg = 'ERROR ***: failed Data Location in DLS \n %s '%ex.getErrorMessage()
303		raise CrabException(msg)
223	–
224	–	allsites=dataloc.getSites()
225	–	common.logger.debug(5,"sites are %s"%allsites)
226	–	sites=self.checkBlackList(allsites)
227	–	common.logger.debug(5,"sites are (after black list) %s"%sites)
228	–	sites=self.checkWhiteList(sites)
229	–	common.logger.debug(5,"sites are (after white list) %s"%sites)
304
305	<	if len(sites)==0:
306	<	msg = 'No sites hosting all the needed data! Exiting... '
307	<	raise CrabException(msg)
308	<	common.logger.message("List of Sites hosting the data : "+str(sites))
309	<	common.logger.debug(6, "List of Sites: "+str(sites))
310	<	common.analisys_common_info['sites']=sites    ## used in SchedulerEdg.py in createSchScript
311	<	return
312	<
313	<	def checkBlackList(self, allSites):
314	<	if len(self.reCEBlackList)==0: return allSites
315	<	sites = []
316	<	for site in allSites:
243	<	common.logger.debug(10,'Site '+site)
244	<	good=1
245	<	for re in self.reCEBlackList:
246	<	if re.search(site):
247	<	common.logger.message('CE in black list, skipping site '+site)
248	<	good=0
249	<	pass
250	<	if good: sites.append(site)
251	<	if len(sites) == 0:
252	<	common.logger.debug(3,"No sites found after BlackList")
305	>
306	>	sites = dataloc.getSites()
307	>	allSites = []
308	>	listSites = sites.values()
309	>	for listSite in listSites:
310	>	for oneSite in listSite:
311	>	allSites.append(oneSite)
312	>	allSites = self.uniquelist(allSites)
313	>
314	>	# screen output
315	>	common.logger.message("Requested dataset: " + datasetPath + " has " + str(self.maxEvents) + " events in " + str(len(self.filesbyblock.keys())) + " blocks.\n")
316	>
317		return sites
318
319	<	def checkWhiteList(self, allsites):
319	>	def jobSplittingByBlocks(self, blockSites):
320	>	"""
321	>	Perform job splitting. Jobs run over an integer number of files
322	>	and no more than one block.
323	>	ARGUMENT: blockSites: dictionary with blocks as keys and list of host sites as values
324	>	REQUIRES: self.selectTotalNumberEvents, self.selectEventsPerJob, self.selectNumberofJobs,
325	>	self.total_number_of_events, self.eventsPerJob, self.theNumberOfJobs,
326	>	self.maxEvents, self.filesbyblock
327	>	SETS: self.jobDestination - Site destination(s) for each job (a list of lists)
328	>	self.total_number_of_jobs - Total # of jobs
329	>	self.list_of_args - File(s) job will run on (a list of lists)
330	>	"""
331
332	<	if len(self.reCEWhiteList)==0: return pubDBUrls
333	<	sites = []
334	<	for site in allsites:
335	<	#print 'connecting to the URL ',url
336	<	good=0
337	<	for re in self.reCEWhiteList:
338	<	if re.search(site):
339	<	common.logger.debug(5,'CE in white list, adding site '+site)
340	<	good=1
341	<	if not good: continue
342	<	sites.append(site)
343	<	if len(sites) == 0:
344	<	common.logger.message("No sites found after WhiteList\n")
332	>	# ---- Handle the possible job splitting configurations ---- #
333	>	if (self.selectTotalNumberEvents):
334	>	totalEventsRequested = self.total_number_of_events
335	>	if (self.selectEventsPerJob):
336	>	eventsPerJobRequested = self.eventsPerJob
337	>	if (self.selectNumberOfJobs):
338	>	totalEventsRequested = self.theNumberOfJobs * self.eventsPerJob
339	>
340	>	# If user requested all the events in the dataset
341	>	if (totalEventsRequested == -1):
342	>	eventsRemaining=self.maxEvents
343	>	# If user requested more events than are in the dataset
344	>	elif (totalEventsRequested > self.maxEvents):
345	>	eventsRemaining = self.maxEvents
346	>	common.logger.message("Requested "+str(self.total_number_of_events)+ " events, but only "+str(self.maxEvents)+" events are available.")
347	>	# If user requested less events than are in the dataset
348		else:
349	<	common.logger.debug(5,"Selected sites via WhiteList are "+str(sites)+"\n")
350	<	return sites
349	>	eventsRemaining = totalEventsRequested
350	>
351	>	# If user requested more events per job than are in the dataset
352	>	if (self.selectEventsPerJob and eventsPerJobRequested > self.maxEvents):
353	>	eventsPerJobRequested = self.maxEvents
354	>
355	>	# For user info at end
356	>	totalEventCount = 0
357	>
358	>	if (self.selectTotalNumberEvents and self.selectNumberOfJobs):
359	>	eventsPerJobRequested = int(eventsRemaining/self.theNumberOfJobs)
360	>
361	>	if (self.selectNumberOfJobs):
362	>	common.logger.message("May not create the exact number_of_jobs requested.")
363	>
364	>	if ( self.ncjobs == 'all' ) :
365	>	totalNumberOfJobs = 999999999
366	>	else :
367	>	totalNumberOfJobs = self.ncjobs
368	>
369	>	blocks = blockSites.keys()
370	>	blockCount = 0
371	>	# Backup variable in case self.maxEvents counted events in a non-included block
372	>	numBlocksInDataset = len(blocks)
373	>
374	>	jobCount = 0
375	>	list_of_lists = []
376	>
377	>	# list tracking which jobs are in which jobs belong to which block
378	>	jobsOfBlock = {}
379	>
380	>	# ---- Iterate over the blocks in the dataset until ---- #
381	>	# ---- we've met the requested total # of events    ---- #
382	>	while ( (eventsRemaining > 0) and (blockCount < numBlocksInDataset) and (jobCount < totalNumberOfJobs)):
383	>	block = blocks[blockCount]
384	>	blockCount += 1
385	>	if block not in jobsOfBlock.keys() :
386	>	jobsOfBlock[block] = []
387	>
388	>	if self.eventsbyblock.has_key(block) :
389	>	numEventsInBlock = self.eventsbyblock[block]
390	>	common.logger.debug(5,'Events in Block File '+str(numEventsInBlock))
391	>
392	>	files = self.filesbyblock[block]
393	>	numFilesInBlock = len(files)
394	>	if (numFilesInBlock <= 0):
395	>	continue
396	>	fileCount = 0
397	>
398	>	# ---- New block => New job ---- #
399	>	parString = ""
400	>	# counter for number of events in files currently worked on
401	>	filesEventCount = 0
402	>	# flag if next while loop should touch new file
403	>	newFile = 1
404	>	# job event counter
405	>	jobSkipEventCount = 0
406	>
407	>	# ---- Iterate over the files in the block until we've met the requested ---- #
408	>	# ---- total # of events or we've gone over all the files in this block  ---- #
409	>	while ( (eventsRemaining > 0) and (fileCount < numFilesInBlock) and (jobCount < totalNumberOfJobs) ):
410	>	file = files[fileCount]
411	>	if newFile :
412	>	try:
413	>	numEventsInFile = self.eventsbyfile[file]
414	>	common.logger.debug(6, "File "+str(file)+" has "+str(numEventsInFile)+" events")
415	>	# increase filesEventCount
416	>	filesEventCount += numEventsInFile
417	>	# Add file to current job
418	>	parString += '\\\"' + file + '\\\"\,'
419	>	newFile = 0
420	>	except KeyError:
421	>	common.logger.message("File "+str(file)+" has unknown number of events: skipping")
422	>
423	>	eventsPerJobRequested = min(eventsPerJobRequested, eventsRemaining)
424	>	# if less events in file remain than eventsPerJobRequested
425	>	if ( filesEventCount - jobSkipEventCount < eventsPerJobRequested):
426	>	# if last file in block
427	>	if ( fileCount == numFilesInBlock-1 ) :
428	>	# end job using last file, use remaining events in block
429	>	# close job and touch new file
430	>	fullString = parString[:-2]
431	>	list_of_lists.append([fullString,str(-1),str(jobSkipEventCount)])
432	>	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(filesEventCount - jobSkipEventCount)+" events (last file in block).")
433	>	self.jobDestination.append(blockSites[block])
434	>	common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
435	>	# fill jobs of block dictionary
436	>	jobsOfBlock[block].append(jobCount+1)
437	>	# reset counter
438	>	jobCount = jobCount + 1
439	>	totalEventCount = totalEventCount + filesEventCount - jobSkipEventCount
440	>	eventsRemaining = eventsRemaining - filesEventCount + jobSkipEventCount
441	>	jobSkipEventCount = 0
442	>	# reset file
443	>	parString = ""
444	>	filesEventCount = 0
445	>	newFile = 1
446	>	fileCount += 1
447	>	else :
448	>	# go to next file
449	>	newFile = 1
450	>	fileCount += 1
451	>	# if events in file equal to eventsPerJobRequested
452	>	elif ( filesEventCount - jobSkipEventCount == eventsPerJobRequested ) :
453	>	# close job and touch new file
454	>	fullString = parString[:-2]
455	>	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
456	>	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.")
457	>	self.jobDestination.append(blockSites[block])
458	>	common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
459	>	jobsOfBlock[block].append(jobCount+1)
460	>	# reset counter
461	>	jobCount = jobCount + 1
462	>	totalEventCount = totalEventCount + eventsPerJobRequested
463	>	eventsRemaining = eventsRemaining - eventsPerJobRequested
464	>	jobSkipEventCount = 0
465	>	# reset file
466	>	parString = ""
467	>	filesEventCount = 0
468	>	newFile = 1
469	>	fileCount += 1
470	>
471	>	# if more events in file remain than eventsPerJobRequested
472	>	else :
473	>	# close job but don't touch new file
474	>	fullString = parString[:-2]
475	>	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
476	>	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.")
477	>	self.jobDestination.append(blockSites[block])
478	>	common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
479	>	jobsOfBlock[block].append(jobCount+1)
480	>	# increase counter
481	>	jobCount = jobCount + 1
482	>	totalEventCount = totalEventCount + eventsPerJobRequested
483	>	eventsRemaining = eventsRemaining - eventsPerJobRequested
484	>	# calculate skip events for last file
485	>	# use filesEventCount (contains several files), jobSkipEventCount and eventsPerJobRequest
486	>	jobSkipEventCount = eventsPerJobRequested - (filesEventCount - jobSkipEventCount - self.eventsbyfile[file])
487	>	# remove all but the last file
488	>	filesEventCount = self.eventsbyfile[file]
489	>	parString = '\\\"' + file + '\\\"\,'
490	>	pass # END if
491	>	pass # END while (iterate over files in the block)
492	>	pass # END while (iterate over blocks in the dataset)
493	>	self.ncjobs = self.total_number_of_jobs = jobCount
494	>	if (eventsRemaining > 0 and jobCount < totalNumberOfJobs ):
495	>	common.logger.message("Could not run on all requested events because some blocks not hosted at allowed sites.")
496	>	common.logger.message(str(jobCount)+" job(s) can run on "+str(totalEventCount)+" events.\n")
497	>
498	>	# screen output
499	>	screenOutput = "List of jobs and available destination sites:\n\n"
500	>
501	>	# keep trace of block with no sites to print a warning at the end
502	>	noSiteBlock = []
503	>	bloskNoSite = []
504	>
505	>	blockCounter = 0
506	>	for block in blocks:
507	>	if block in jobsOfBlock.keys() :
508	>	blockCounter += 1
509	>	screenOutput += "Block %5i: jobs %20s: sites: %s\n" % (blockCounter,spanRanges(jobsOfBlock[block]),
510	>	','.join(self.blackWhiteListParser.checkWhiteList(self.blackWhiteListParser.checkBlackList(blockSites[block],block),block)))
511	>	if len(self.blackWhiteListParser.checkWhiteList(self.blackWhiteListParser.checkBlackList(blockSites[block],block),block)) == 0:
512	>	noSiteBlock.append( spanRanges(jobsOfBlock[block]) )
513	>	bloskNoSite.append( blockCounter )
514	>
515	>	common.logger.message(screenOutput)
516	>	if len(noSiteBlock) > 0 and len(bloskNoSite) > 0:
517	>	msg = 'WARNING: No sites are hosting any part of data for block:\n                '
518	>	virgola = ""
519	>	if len(bloskNoSite) > 1:
520	>	virgola = ","
521	>	for block in bloskNoSite:
522	>	msg += ' ' + str(block) + virgola
523	>	msg += '\n               Related jobs:\n                 '
524	>	virgola = ""
525	>	if len(noSiteBlock) > 1:
526	>	virgola = ","
527	>	for range_jobs in noSiteBlock:
528	>	msg += str(range_jobs) + virgola
529	>	msg += '\n               will not be submitted and this block of data can not be analyzed!\n'
530	>	if self.cfg_params.has_key('EDG.se_white_list'):
531	>	msg += 'WARNING: SE White List: '+self.cfg_params['EDG.se_white_list']+'\n'
532	>	msg += '(Hint: By whitelisting you force the job to run at this particular site(s).\n'
533	>	msg += 'Please check if the dataset is available at this site!)\n'
534	>	if self.cfg_params.has_key('EDG.ce_white_list'):
535	>	msg += 'WARNING: CE White List: '+self.cfg_params['EDG.ce_white_list']+'\n'
536	>	msg += '(Hint: By whitelisting you force the job to run at this particular site(s).\n'
537	>	msg += 'Please check if the dataset is available at this site!)\n'
538	>
539	>	common.logger.message(msg)
540	>
541	>	self.list_of_args = list_of_lists
542	>	return
543	>
544	>	def jobSplittingNoInput(self):
545	>	"""
546	>	Perform job splitting based on number of event per job
547	>	"""
548	>	common.logger.debug(5,'Splitting per events')
549	>
550	>	if (self.selectEventsPerJob):
551	>	common.logger.message('Required '+str(self.eventsPerJob)+' events per job ')
552	>	if (self.selectNumberOfJobs):
553	>	common.logger.message('Required '+str(self.theNumberOfJobs)+' jobs in total ')
554	>	if (self.selectTotalNumberEvents):
555	>	common.logger.message('Required '+str(self.total_number_of_events)+' events in total ')
556	>
557	>	if (self.total_number_of_events < 0):
558	>	msg='Cannot split jobs per Events with "-1" as total number of events'
559	>	raise CrabException(msg)
560	>
561	>	if (self.selectEventsPerJob):
562	>	if (self.selectTotalNumberEvents):
563	>	self.total_number_of_jobs = int(self.total_number_of_events/self.eventsPerJob)
564	>	elif(self.selectNumberOfJobs) :
565	>	self.total_number_of_jobs =self.theNumberOfJobs
566	>	self.total_number_of_events =int(self.theNumberOfJobs*self.eventsPerJob)
567	>
568	>	elif (self.selectNumberOfJobs) :
569	>	self.total_number_of_jobs = self.theNumberOfJobs
570	>	self.eventsPerJob = int(self.total_number_of_events/self.total_number_of_jobs)
571	>
572	>	common.logger.debug(5,'N jobs  '+str(self.total_number_of_jobs))
573	>
574	>	# is there any remainder?
575	>	check = int(self.total_number_of_events) - (int(self.total_number_of_jobs)*self.eventsPerJob)
576	>
577	>	common.logger.debug(5,'Check  '+str(check))
578	>
579	>	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')
580	>	if check > 0:
581	>	common.logger.message('Warning: asked '+str(self.total_number_of_events)+' but can do only '+str(int(self.total_number_of_jobs)*self.eventsPerJob))
582	>
583	>	# argument is seed number.$i
584	>	self.list_of_args = []
585	>	for i in range(self.total_number_of_jobs):
586	>	## Since there is no input, any site is good
587	>	self.jobDestination.append([""]) #must be empty to write correctly the xml
588	>	args=[]
589	>	if (self.firstRun):
590	>	## pythia first run
591	>	args.append(str(self.firstRun)+str(i))
592	>	self.list_of_args.append(args)
593	>
594	>	return
595	>
596	>
597	>	def jobSplittingForScript(self):
598	>	"""
599	>	Perform job splitting based on number of job
600	>	"""
601	>	common.logger.debug(5,'Splitting per job')
602	>	common.logger.message('Required '+str(self.theNumberOfJobs)+' jobs in total ')
603	>
604	>	self.total_number_of_jobs = self.theNumberOfJobs
605	>
606	>	common.logger.debug(5,'N jobs  '+str(self.total_number_of_jobs))
607	>
608	>	common.logger.message(str(self.total_number_of_jobs)+' jobs can be created')
609	>
610	>	# argument is seed number.$i
611	>	self.list_of_args = []
612	>	for i in range(self.total_number_of_jobs):
613	>	self.jobDestination.append([""])
614	>	self.list_of_args.append([str(i)])
615	>	return
616	>
617	>	def split(self, jobParams):
618	>
619	>	njobs = self.total_number_of_jobs
620	>	arglist = self.list_of_args
621	>	# create the empty structure
622	>	for i in range(njobs):
623	>	jobParams.append("")
624	>
625	>	listID=[]
626	>	listField=[]
627	>	for job in range(njobs):
628	>	jobParams[job] = arglist[job]
629	>	listID.append(job+1)
630	>	job_ToSave ={}
631	>	concString = ' '
632	>	argu=''
633	>	if len(jobParams[job]):
634	>	argu +=   concString.join(jobParams[job] )
635	>	job_ToSave['arguments']= str(job+1)+' '+argu
636	>	job_ToSave['dlsDestination']= self.jobDestination[job]
637	>	listField.append(job_ToSave)
638	>	msg="Job "+str(job)+" Arguments:   "+str(job+1)+" "+argu+"\n"  \
639	>	+"                     Destination: "+str(self.jobDestination[job])
640	>	common.logger.debug(5,msg)
641	>	common._db.updateJob_(listID,listField)
642	>	self.argsList = (len(jobParams[0])+1)
643	>
644	>	return
645	>
646	>	def numberOfJobs(self):
647	>	return self.total_number_of_jobs
648
649		def getTarBall(self, exe):
650		"""
651		Return the TarBall with lib and exe
652		"""
653	<
279	<	# if it exist, just return it
280	<	self.tgzNameWithPath = common.work_space.shareDir()+self.tgz_name
653	>	self.tgzNameWithPath = common.work_space.pathForTgz()+'share/'+self.tgz_name
654		if os.path.exists(self.tgzNameWithPath):
655		return self.tgzNameWithPath
656
#	Line 290 | Line 663 | class Cmssw(JobType):
663
664		# First of all declare the user Scram area
665		swArea = self.scram.getSWArea_()
293	–	#print "swArea = ", swArea
294	–	swVersion = self.scram.getSWVersion()
295	–	#print "swVersion = ", swVersion
666		swReleaseTop = self.scram.getReleaseTop_()
667	<	#print "swReleaseTop = ", swReleaseTop
298	<
667	>
668		## check if working area is release top
669		if swReleaseTop == '' or swArea == swReleaseTop:
670	+	common.logger.debug(3,"swArea = "+swArea+" swReleaseTop ="+swReleaseTop)
671		return
672
673	<	filesToBeTarred = []
674	<	## First find the executable
675	<	if (self.executable != ''):
676	<	exeWithPath = self.scram.findFile_(executable)
677	<	#           print exeWithPath
678	<	if ( not exeWithPath ):
679	<	raise CrabException('User executable '+executable+' not found')
680	<
681	<	## then check if it's private or not
682	<	if exeWithPath.find(swReleaseTop) == -1:
683	<	# the exe is private, so we must ship
684	<	common.logger.debug(5,"Exe "+exeWithPath+" to be tarred")
685	<	path = swArea+'/'
686	<	exe = string.replace(exeWithPath, path,'')
687	<	filesToBeTarred.append(exe)
688	<	pass
689	<	else:
690	<	# the exe is from release, we'll find it on WN
691	<	pass
692	<
693	<	## Now get the libraries: only those in local working area
694	<	libDir = 'lib'
695	<	lib = swArea+'/' +libDir
696	<	common.logger.debug(5,"lib "+lib+" to be tarred")
697	<	if os.path.exists(lib):
698	<	filesToBeTarred.append(libDir)
699	<
700	<	## Now check if the Data dir is present
701	<	dataDir = 'src/Data/'
702	<	if os.path.isdir(swArea+'/'+dataDir):
703	<	filesToBeTarred.append(dataDir)
704	<
705	<	## Create the tar-ball
706	<	if len(filesToBeTarred)>0:
707	<	cwd = os.getcwd()
708	<	os.chdir(swArea)
709	<	tarcmd = 'tar zcvf ' + self.tgzNameWithPath + ' '
710	<	for line in filesToBeTarred:
711	<	tarcmd = tarcmd + line + ' '
712	<	cout = runCommand(tarcmd)
713	<	if not cout:
714	<	raise CrabException('Could not create tar-ball')
715	<	os.chdir(cwd)
716	<	else:
717	<	common.logger.debug(5,"No files to be to be tarred")
718	<
719	<	return
720	<
721	<	def wsSetupEnvironment(self, nj):
673	>	import tarfile
674	>	try: # create tar ball
675	>	tar = tarfile.open(self.tgzNameWithPath, "w:gz")
676	>	## First find the executable
677	>	if (self.executable != ''):
678	>	exeWithPath = self.scram.findFile_(executable)
679	>	if ( not exeWithPath ):
680	>	raise CrabException('User executable '+executable+' not found')
681	>
682	>	## then check if it's private or not
683	>	if exeWithPath.find(swReleaseTop) == -1:
684	>	# the exe is private, so we must ship
685	>	common.logger.debug(5,"Exe "+exeWithPath+" to be tarred")
686	>	path = swArea+'/'
687	>	# distinguish case when script is in user project area or given by full path somewhere else
688	>	if exeWithPath.find(path) >= 0 :
689	>	exe = string.replace(exeWithPath, path,'')
690	>	tar.add(path+exe,exe)
691	>	else :
692	>	tar.add(exeWithPath,os.path.basename(executable))
693	>	pass
694	>	else:
695	>	# the exe is from release, we'll find it on WN
696	>	pass
697	>
698	>	## Now get the libraries: only those in local working area
699	>	libDir = 'lib'
700	>	lib = swArea+'/' +libDir
701	>	common.logger.debug(5,"lib "+lib+" to be tarred")
702	>	if os.path.exists(lib):
703	>	tar.add(lib,libDir)
704	>
705	>	## Now check if module dir is present
706	>	moduleDir = 'module'
707	>	module = swArea + '/' + moduleDir
708	>	if os.path.isdir(module):
709	>	tar.add(module,moduleDir)
710	>
711	>	## Now check if any data dir(s) is present
712	>	swAreaLen=len(swArea)
713	>	self.dataExist = False
714	>	for root, dirs, files in os.walk(swArea):
715	>	if "data" in dirs:
716	>	self.dataExist=True
717	>	common.logger.debug(5,"data "+root+"/data"+" to be tarred")
718	>	tar.add(root+"/data",root[swAreaLen:]+"/data")
719	>
720	>	### CMSSW ParameterSet
721	>	if not self.pset is None:
722	>	cfg_file = common.work_space.jobDir()+self.configFilename()
723	>	tar.add(cfg_file,self.configFilename())
724	>	common.logger.debug(5,"File added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
725	>
726	>
727	>	## Add ProdCommon dir to tar
728	>	prodcommonDir = 'ProdCommon'
729	>	prodcommonPath = os.environ['CRABDIR'] + '/' + 'ProdCommon'
730	>	if os.path.isdir(prodcommonPath):
731	>	tar.add(prodcommonPath,prodcommonDir)
732	>	common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
733	>
734	>	##### ML stuff
735	>	ML_file_list=['report.py', 'DashboardAPI.py', 'Logger.py', 'ProcInfo.py', 'apmon.py']
736	>	path=os.environ['CRABDIR'] + '/python/'
737	>	for file in ML_file_list:
738	>	tar.add(path+file,file)
739	>	common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
740	>
741	>	##### Utils
742	>	Utils_file_list=['parseCrabFjr.py','writeCfg.py', 'JobReportErrorCode.py']
743	>	for file in Utils_file_list:
744	>	tar.add(path+file,file)
745	>	common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
746	>
747	>	##### AdditionalFiles
748	>	for file in self.additional_inbox_files:
749	>	tar.add(file,string.split(file,'/')[-1])
750	>	common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
751	>
752	>	tar.close()
753	>	except :
754	>	raise CrabException('Could not create tar-ball')
755	>
756	>	## check for tarball size
757	>	tarballinfo = os.stat(self.tgzNameWithPath)
758	>	if ( tarballinfo.st_size > self.MaxTarBallSize*1024*1024 ) :
759	>	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.')
760	>
761	>	## create tar-ball with ML stuff
762	>
763	>	def wsSetupEnvironment(self, nj=0):
764		"""
765		Returns part of a job script which prepares
766		the execution environment for the job 'nj'.
767		"""
768	+	if (self.CMSSW_major >= 2 and self.CMSSW_minor >= 1) or (self.CMSSW_major >= 3):
769	+	psetName = 'pset.py'
770	+	else:
771	+	psetName = 'pset.cfg'
772		# Prepare JobType-independent part
773	<	txt = self.wsSetupCMSEnvironment_()
773	>	txt = '\n#Written by cms_cmssw::wsSetupEnvironment\n'
774	>	txt += 'echo ">>> setup environment"\n'
775	>	txt += 'if [ $middleware == LCG ]; then \n'
776	>	txt += self.wsSetupCMSLCGEnvironment_()
777	>	txt += 'elif [ $middleware == OSG ]; then\n'
778	>	txt += '    WORKING_DIR=`/bin/mktemp  -d $OSG_WN_TMP/cms_XXXXXXXXXXXX`\n'
779	>	txt += '    if [ ! $? == 0 ] ;then\n'
780	>	txt += '        echo "ERROR ==> OSG $WORKING_DIR could not be created on WN `hostname`"\n'
781	>	txt += '        job_exit_code=10016\n'
782	>	txt += '        func_exit\n'
783	>	txt += '    fi\n'
784	>	txt += '    echo ">>> Created working directory: $WORKING_DIR"\n'
785	>	txt += '\n'
786	>	txt += '    echo "Change to working directory: $WORKING_DIR"\n'
787	>	txt += '    cd $WORKING_DIR\n'
788	>	txt += '    echo ">>> current directory (WORKING_DIR): $WORKING_DIR"\n'
789	>	txt += self.wsSetupCMSOSGEnvironment_()
790	>	txt += 'fi\n'
791
792		# Prepare JobType-specific part
793		scram = self.scram.commandName()
794		txt += '\n\n'
795	<	txt += 'echo "### SPECIFIC JOB SETUP ENVIRONMENT ###"\n'
795	>	txt += 'echo ">>> specific cmssw setup environment:"\n'
796	>	txt += 'echo "CMSSW_VERSION =  '+self.version+'"\n'
797		txt += scram+' project CMSSW '+self.version+'\n'
798		txt += 'status=$?\n'
799		txt += 'if [ $status != 0 ] ; then\n'
800	<	txt += '   echo "SET_EXE_ENV 1 ==>ERROR CMSSW '+self.version+' not found on `hostname`" \n'
801	<	txt += '   echo "JOB_EXIT_STATUS = 5"\n'
802	<	txt += '   echo "SanityCheckCode = 5" | tee -a $RUNTIME_AREA/$repo\n'
369	<	txt += '   dumpStatus $RUNTIME_AREA/$repo\n'
370	<	txt += '   exit 5 \n'
800	>	txt += '    echo "ERROR ==> CMSSW '+self.version+' not found on `hostname`" \n'
801	>	txt += '    job_exit_code=10034\n'
802	>	txt += '    func_exit\n'
803		txt += 'fi \n'
372	–	txt += 'echo "CMSSW_VERSION =  '+self.version+'"\n'
804		txt += 'cd '+self.version+'\n'
805	<	### needed grep for bug in scramv1 ###
805	>	txt += 'SOFTWARE_DIR=`pwd`\n'
806	>	txt += 'echo ">>> current directory (SOFTWARE_DIR): $SOFTWARE_DIR" \n'
807		txt += 'eval `'+scram+' runtime -sh | grep -v SCRAMRT_LSB_JOBNAME`\n'
808	<
808	>	txt += 'if [ $? != 0 ] ; then\n'
809	>	txt += '    echo "ERROR ==> Problem with the command: "\n'
810	>	txt += '    echo "eval \`'+scram+' runtime -sh | grep -v SCRAMRT_LSB_JOBNAME \` at `hostname`"\n'
811	>	txt += '    job_exit_code=10034\n'
812	>	txt += '    func_exit\n'
813	>	txt += 'fi \n'
814		# Handle the arguments:
815		txt += "\n"
816	<	txt += "## ARGUMNETS: $1 Job Number\n"
380	<	# txt += "## ARGUMNETS: $2 First Event for this job\n"
381	<	# txt += "## ARGUMNETS: $3 Max Event for this job\n"
816	>	txt += "## number of arguments (first argument always jobnumber)\n"
817		txt += "\n"
818	<	txt += "narg=$#\n"
384	<	txt += "if [ $narg -lt 1 ]\n"
818	>	txt += "if [ $nargs -lt "+str(self.argsList)+" ]\n"
819		txt += "then\n"
820	<	txt += "    echo 'SET_EXE_ENV 1 ==> ERROR Too few arguments' +$narg+ \n"
821	<	txt += '    echo "JOB_EXIT_STATUS = 1"\n'
822	<	txt += '    echo "SanityCheckCode = 1" | tee -a $RUNTIME_AREA/$repo\n'
389	<	txt += '    dumpStatus $RUNTIME_AREA/$repo\n'
390	<	txt += "    exit 1\n"
820	>	txt += "    echo 'ERROR ==> Too few arguments' +$nargs+ \n"
821	>	txt += '    job_exit_code=50113\n'
822	>	txt += "    func_exit\n"
823		txt += "fi\n"
824		txt += "\n"
393	–	txt += "NJob=$1\n"
394	–	# txt += "FirstEvent=$2\n"
395	–	# txt += "MaxEvents=$3\n"
825
826		# Prepare job-specific part
827		job = common.job_list[nj]
828	<	pset = os.path.basename(job.configFilename())
829	<	txt += '\n'
830	<	txt += 'cp $RUNTIME_AREA/'+pset+' pset.cfg\n'
831	<	# txt += 'if [ -e $RUNTIME_AREA/orcarc_$CE ] ; then\n'
832	<	# txt += '  cat $RUNTIME_AREA/orcarc_$CE .orcarc >> .orcarc_tmp\n'
833	<	# txt += '  mv .orcarc_tmp .orcarc\n'
834	<	# txt += 'fi\n'
835	<	# txt += 'if [ -e $RUNTIME_AREA/init_$CE.sh ] ; then\n'
836	<	# txt += '  cp $RUNTIME_AREA/init_$CE.sh init.sh\n'
837	<	# txt += 'fi\n'
828	>	if (self.datasetPath):
829	>	txt += '\n'
830	>	txt += 'DatasetPath='+self.datasetPath+'\n'
831	>
832	>	datasetpath_split = self.datasetPath.split("/")
833	>
834	>	txt += 'PrimaryDataset='+datasetpath_split[1]+'\n'
835	>	txt += 'DataTier='+datasetpath_split[2]+'\n'
836	>	txt += 'ApplicationFamily=cmsRun\n'
837	>
838	>	else:
839	>	txt += 'DatasetPath=MCDataTier\n'
840	>	txt += 'PrimaryDataset=null\n'
841	>	txt += 'DataTier=null\n'
842	>	txt += 'ApplicationFamily=MCDataTier\n'
843	>	if self.pset != None:
844	>	pset = os.path.basename(job.configFilename())
845	>	txt += '\n'
846	>	txt += 'cp  $RUNTIME_AREA/'+pset+' .\n'
847	>	if (self.datasetPath): # standard job
848	>	txt += 'InputFiles=${args[1]}; export InputFiles\n'
849	>	txt += 'MaxEvents=${args[2]}; export MaxEvents\n'
850	>	txt += 'SkipEvents=${args[3]}; export SkipEvents\n'
851	>	txt += 'echo "Inputfiles:<$InputFiles>"\n'
852	>	txt += 'echo "MaxEvents:<$MaxEvents>"\n'
853	>	txt += 'echo "SkipEvents:<$SkipEvents>"\n'
854	>	else:  # pythia like job
855	>	txt += 'PreserveSeeds='  + ','.join(self.preserveSeeds)  + '; export PreserveSeeds\n'
856	>	txt += 'IncrementSeeds=' + ','.join(self.incrementSeeds) + '; export IncrementSeeds\n'
857	>	txt += 'echo "PreserveSeeds: <$PreserveSeeds>"\n'
858	>	txt += 'echo "IncrementSeeds:<$IncrementSeeds>"\n'
859	>	if (self.firstRun):
860	>	txt += 'FirstRun=${args[1]}; export FirstRun\n'
861	>	txt += 'echo "FirstRun: <$FirstRun>"\n'
862	>
863	>	txt += 'mv -f ' + pset + ' ' + psetName + '\n'
864
865	<	if len(self.additional_inbox_files) > 0:
865	>
866	>	if self.pset != None:
867	>	# FUTURE: Can simply for 2_1_x and higher
868	>	txt += '\n'
869	>	if self.debug_pset==True:
870	>	txt += 'echo "***** cat ' + psetName + ' *********"\n'
871	>	txt += 'cat ' + psetName + '\n'
872	>	txt += 'echo "****** end ' + psetName + ' ********"\n'
873	>	txt += '\n'
874	>	txt += 'PSETHASH=`edmConfigHash < ' + psetName + '` \n'
875	>	txt += 'echo "PSETHASH = $PSETHASH" \n'
876	>	txt += '\n'
877	>	return txt
878	>
879	>	def wsUntarSoftware(self, nj=0):
880	>	"""
881	>	Put in the script the commands to build an executable
882	>	or a library.
883	>	"""
884	>
885	>	txt = '\n#Written by cms_cmssw::wsUntarSoftware\n'
886	>
887	>	if os.path.isfile(self.tgzNameWithPath):
888	>	txt += 'echo ">>> tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+' :" \n'
889	>	txt += 'tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+'\n'
890	>	txt += 'ls -Al \n'
891	>	txt += 'untar_status=$? \n'
892	>	txt += 'if [ $untar_status -ne 0 ]; then \n'
893	>	txt += '   echo "ERROR ==> Untarring .tgz file failed"\n'
894	>	txt += '   job_exit_code=$untar_status\n'
895	>	txt += '   func_exit\n'
896	>	txt += 'else \n'
897	>	txt += '   echo "Successful untar" \n'
898	>	txt += 'fi \n'
899	>	txt += '\n'
900	>	txt += 'echo ">>> Include ProdCommon in PYTHONPATH:"\n'
901	>	txt += 'if [ -z "$PYTHONPATH" ]; then\n'
902	>	txt += '   export PYTHONPATH=$RUNTIME_AREA/ProdCommon\n'
903	>	txt += 'else\n'
904	>	txt += '   export PYTHONPATH=$RUNTIME_AREA/ProdCommon:${PYTHONPATH}\n'
905	>	txt += 'echo "PYTHONPATH=$PYTHONPATH"\n'
906	>	txt += 'fi\n'
907	>	txt += '\n'
908	>
909	>	pass
910	>
911	>	return txt
912	>
913	>	def wsBuildExe(self, nj=0):
914	>	"""
915	>	Put in the script the commands to build an executable
916	>	or a library.
917	>	"""
918	>
919	>	txt = '\n#Written by cms_cmssw::wsBuildExe\n'
920	>	txt += 'echo ">>> moving CMSSW software directories in `pwd`" \n'
921	>
922	>	txt += 'rm -r lib/ module/ \n'
923	>	txt += 'mv $RUNTIME_AREA/lib/ . \n'
924	>	txt += 'mv $RUNTIME_AREA/module/ . \n'
925	>	if self.dataExist == True:
926	>	txt += 'rm -r src/ \n'
927	>	txt += 'mv $RUNTIME_AREA/src/ . \n'
928	>	if len(self.additional_inbox_files)>0:
929		for file in self.additional_inbox_files:
930	<	txt += 'if [ -e $RUNTIME_AREA/'+file+' ] ; then\n'
931	<	txt += '   cp $RUNTIME_AREA/'+file+' .\n'
414	<	txt += '   chmod +x '+file+'\n'
415	<	txt += 'fi\n'
416	<	pass
417	<
418	<	# txt += '\n'
419	<	# txt += 'chmod +x ./init.sh\n'
420	<	# txt += './init.sh\n'
421	<	# txt += 'exitStatus=$?\n'
422	<	# txt += 'if [ $exitStatus != 0 ] ; then\n'
423	<	# txt += '  echo "SET_EXE_ENV 1 ==> ERROR StageIn init script failed"\n'
424	<	# txt += '  echo "JOB_EXIT_STATUS = $exitStatus" \n'
425	<	# txt += '  echo "SanityCheckCode = $exitStatus" | tee -a $RUNTIME_AREA/$repo\n'
426	<	# txt += '  dumpStatus $RUNTIME_AREA/$repo\n'
427	<	# txt += '  exit $exitStatus\n'
428	<	# txt += 'fi\n'
429	<	# txt += "echo 'SET_EXE_ENV 0 ==> job setup ok'\n"
430	<	txt += 'echo "### END JOB SETUP ENVIRONMENT ###"\n\n'
431	<
432	<	# txt += 'echo "FirstEvent=$FirstEvent" >> .orcarc\n'
433	<	# txt += 'echo "MaxEvents=$MaxEvents" >> .orcarc\n'
434	<	# if self.ML:
435	<	#     txt += 'echo "MonalisaJobId=$NJob" >> .orcarc\n'
930	>	txt += 'mv $RUNTIME_AREA/'+os.path.basename(file)+' . \n'
931	>	txt += 'mv $RUNTIME_AREA/ProdCommon/ . \n'
932
933	+	txt += 'if [ -z "$PYTHONPATH" ]; then\n'
934	+	txt += '   export PYTHONPATH=$SOFTWARE_DIR/ProdCommon\n'
935	+	txt += 'else\n'
936	+	txt += '   export PYTHONPATH=$SOFTWARE_DIR/ProdCommon:${PYTHONPATH}\n'
937	+	txt += 'echo "PYTHONPATH=$PYTHONPATH"\n'
938	+	txt += 'fi\n'
939		txt += '\n'
940	<	txt += 'echo "***** cat pset.cfg *********"\n'
439	<	txt += 'cat pset.cfg\n'
440	<	txt += 'echo "****** end pset.cfg ********"\n'
940	>
941		return txt
942
943		def modifySteeringCards(self, nj):
944		"""
945	<	modify the card provided by the user,
945	>	modify the card provided by the user,
946		writing a new card into share dir
947		"""
948	<
948	>
949		def executableName(self):
950	<	return self.executable
950	>	if self.scriptExe:
951	>	return "sh "
952	>	else:
953	>	return self.executable
954
955		def executableArgs(self):
956	<	return "-p pset.cfg"
956	>	# FUTURE: This function tests the CMSSW version. Can be simplified as we drop support for old versions
957	>	if self.scriptExe:#CarlosDaniele
958	>	return   self.scriptExe + " $NJob"
959	>	else:
960	>	ex_args = ""
961	>	# FUTURE: This tests the CMSSW version. Can remove code as versions deprecated
962	>	# Framework job report
963	>	if (self.CMSSW_major >= 1 and self.CMSSW_minor >= 5) or (self.CMSSW_major >= 2):
964	>	ex_args += " -j $RUNTIME_AREA/crab_fjr_$NJob.xml"
965	>	# Type of config file
966	>	if self.CMSSW_major >= 2 :
967	>	ex_args += " -p pset.py"
968	>	else:
969	>	ex_args += " -p pset.cfg"
970	>	return ex_args
971
972		def inputSandbox(self, nj):
973		"""
974		Returns a list of filenames to be put in JDL input sandbox.
975		"""
976		inp_box = []
460	–	# dict added to delete duplicate from input sandbox file list
461	–	seen = {}
462	–	## code
977		if os.path.isfile(self.tgzNameWithPath):
978		inp_box.append(self.tgzNameWithPath)
979	<	## config
980	<	inp_box.append(common.job_list[nj].configFilename())
467	<	## additional input files
468	<	for file in self.additional_inbox_files:
469	<	inp_box.append(common.work_space.cwdDir()+file)
470	<	#print "sono inputSandbox, inp_box = ", inp_box
979	>	wrapper = os.path.basename(str(common._db.queryTask('scriptName')))
980	>	inp_box.append(common.work_space.pathForTgz() +'job/'+ wrapper)
981		return inp_box
982
983		def outputSandbox(self, nj):
#	Line 476 | Line 986 | class Cmssw(JobType):
986		"""
987		out_box = []
988
479	–	stdout=common.job_list[nj].stdout()
480	–	stderr=common.job_list[nj].stderr()
481	–
989		## User Declared output files
990	<	for out in self.output_file:
991	<	n_out = nj + 1
990	>	for out in (self.output_file+self.output_file_sandbox):
991	>	n_out = nj + 1
992		out_box.append(self.numberFile_(out,str(n_out)))
993		return out_box
487	–	return []
994
995		def prepareSteeringCards(self):
996		"""
997		Make initial modifications of the user's steering card file.
998		"""
493	–	infile = open(self.pset,'r')
494	–
495	–	outfile = open(common.work_space.jobDir()+self.name()+'.cfg', 'w')
496	–
497	–	outfile.write('\n\n##### The following cards have been created by CRAB: DO NOT TOUCH #####\n')
498	–
499	–	outfile.write('InputCollections=/System/'+self.owner+'/'+self.dataset+'/'+self.dataset+'\n')
500	–
501	–	infile.close()
502	–	outfile.close()
999		return
1000
1001		def wsRenameOutput(self, nj):
#	Line 507 | Line 1003 | class Cmssw(JobType):
1003		Returns part of a job script which renames the produced files.
1004		"""
1005
1006	<	txt = '\n'
1007	<	file_list = ''
1008	<	for fileWithSuffix in self.output_file:
1006	>	txt = '\n#Written by cms_cmssw::wsRenameOutput\n'
1007	>	txt += 'echo ">>> current directory (SOFTWARE_DIR): $SOFTWARE_DIR" \n'
1008	>	txt += 'echo ">>> current directory content:"\n'
1009	>	txt += 'ls \n'
1010	>	txt += '\n'
1011	>
1012	>	for fileWithSuffix in (self.output_file):
1013		output_file_num = self.numberFile_(fileWithSuffix, '$NJob')
514	–	file_list=file_list+output_file_num+','
515	–	txt += '\n'
516	–	txt += 'ls \n'
1014		txt += '\n'
1015	<	txt += 'ls '+fileWithSuffix+'\n'
1016	<	txt += 'exe_result=$?\n'
1017	<	txt += 'if [ $exe_result -ne 0 ] ; then\n'
1018	<	txt += '   echo "ERROR: No output file to manage"\n'
1019	<	txt += '   echo "JOB_EXIT_STATUS = $exe_result"\n'
1020	<	txt += '   echo "SanityCheckCode = $exe_result" | tee -a $RUNTIME_AREA/$repo\n'
1021	<	txt += '   dumpStatus $RUNTIME_AREA/$repo\n'
1022	<	txt += '   exit $exe_result \n'
1015	>	txt += '# check output file\n'
1016	>	txt += 'if [ -e ./'+fileWithSuffix+' ] ; then\n'
1017	>	if (self.copy_data == 1):  # For OSG nodes, file is in $WORKING_DIR, should not be moved to $RUNTIME_AREA
1018	>	txt += '    mv '+fileWithSuffix+' '+output_file_num+'\n'
1019	>	txt += '    ln -s `pwd`/'+output_file_num+' $RUNTIME_AREA/'+fileWithSuffix+'\n'
1020	>	else:
1021	>	txt += '    mv '+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n'
1022	>	txt += '    ln -s $RUNTIME_AREA/'+output_file_num+' $RUNTIME_AREA/'+fileWithSuffix+'\n'
1023		txt += 'else\n'
1024	<	txt += '   cp '+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n'
1024	>	txt += '    job_exit_code=60302\n'
1025	>	txt += '    echo "WARNING: Output file '+fileWithSuffix+' not found"\n'
1026	>	if common.scheduler.name().upper() == 'CONDOR_G':
1027	>	txt += '    if [ $middleware == OSG ]; then \n'
1028	>	txt += '        echo "prepare dummy output file"\n'
1029	>	txt += '        echo "Processing of job output failed" > $RUNTIME_AREA/'+output_file_num+'\n'
1030	>	txt += '    fi \n'
1031		txt += 'fi\n'
1032	<	txt += 'cd $RUNTIME_AREA\n'
1033	<
1034	<	pass
1035	<
1036	<	file_list=file_list[:-1]
1037	<	txt += 'file_list='+file_list+'\n'
1032	>	file_list = []
1033	>	for fileWithSuffix in (self.output_file):
1034	>	file_list.append(self.numberFile_(fileWithSuffix, '$NJob'))
1035	>
1036	>	txt += 'file_list="'+string.join(file_list,' ')+'"\n'
1037	>	txt += '\n'
1038	>	txt += 'echo ">>> current directory (SOFTWARE_DIR): $SOFTWARE_DIR" \n'
1039	>	txt += 'echo ">>> current directory content:"\n'
1040	>	txt += 'ls \n'
1041	>	txt += '\n'
1042	>	txt += 'cd $RUNTIME_AREA\n'
1043	>	txt += 'echo ">>> current directory (RUNTIME_AREA):  $RUNTIME_AREA"\n'
1044		return txt
1045
1046		def numberFile_(self, file, txt):
#	Line 542 | Line 1051 | class Cmssw(JobType):
1051		# take away last extension
1052		name = p[0]
1053		for x in p[1:-1]:
1054	<	name=name+"."+x
1054	>	name=name+"."+x
1055		# add "_txt"
1056		if len(p)>1:
1057	<	ext = p[len(p)-1]
1058	<	#result = name + '_' + str(txt) + "." + ext
1059	<	result = name + '_' + txt + "." + ext
1060	<	else:
1061	<	#result = name + '_' + str(txt)
553	<	result = name + '_' + txt
554	<
1057	>	ext = p[len(p)-1]
1058	>	result = name + '_' + txt + "." + ext
1059	>	else:
1060	>	result = name + '_' + txt
1061	>
1062		return result
1063
1064	<	def getRequirements(self):
1064	>	def getRequirements(self, nj=[]):
1065		"""
1066	<	return job requirements to add to jdl files
1066	>	return job requirements to add to jdl files
1067		"""
1068		req = ''
1069	<	if common.analisys_common_info['sites']:
1070	<	if common.analisys_common_info['sw_version']:
1071	<	req='Member("VO-cms-' + \
1072	<	common.analisys_common_info['sw_version'] + \
1073	<	'", other.GlueHostApplicationSoftwareRunTimeEnvironment)'
1074	<	if len(common.analisys_common_info['sites'])>0:
1075	<	req = req + ' && ('
1076	<	for i in range(len(common.analisys_common_info['sites'])):
1077	<	req = req + 'other.GlueCEInfoHostName == "' \
1078	<	+ common.analisys_common_info['sites'][i] + '"'
1079	<	if ( i < (int(len(common.analisys_common_info['sites']) - 1)) ):
1080	<	req = req + ' || '
1081	<	req = req + ')'
575	<	#print "req = ", req
1069	>	if self.version:
1070	>	req='Member("VO-cms-' + \
1071	>	self.version + \
1072	>	'", other.GlueHostApplicationSoftwareRunTimeEnvironment)'
1073	>	if self.executable_arch:
1074	>	req+=' && Member("VO-cms-' + \
1075	>	self.executable_arch + \
1076	>	'", other.GlueHostApplicationSoftwareRunTimeEnvironment)'
1077	>
1078	>	req = req + ' && (other.GlueHostNetworkAdapterOutboundIP)'
1079	>	if common.scheduler.name() == "glitecoll":
1080	>	req += ' && other.GlueCEStateStatus == "Production" '
1081	>
1082		return req
1083	+
1084	+	def configFilename(self):
1085	+	""" return the config filename """
1086	+	# FUTURE: Can remove cfg mode for CMSSW >= 2_1_x
1087	+	if (self.CMSSW_major >= 2 and self.CMSSW_minor >= 1) or (self.CMSSW_major >= 3):
1088	+	return self.name()+'.py'
1089	+	else:
1090	+	return self.name()+'.cfg'
1091	+
1092	+	def wsSetupCMSOSGEnvironment_(self):
1093	+	"""
1094	+	Returns part of a job script which is prepares
1095	+	the execution environment and which is common for all CMS jobs.
1096	+	"""
1097	+	txt = '\n#Written by cms_cmssw::wsSetupCMSOSGEnvironment_\n'
1098	+	txt += '    echo ">>> setup CMS OSG environment:"\n'
1099	+	txt += '    echo "set SCRAM ARCH to ' + self.executable_arch + '"\n'
1100	+	txt += '    export SCRAM_ARCH='+self.executable_arch+'\n'
1101	+	txt += '    echo "SCRAM_ARCH = $SCRAM_ARCH"\n'
1102	+	txt += '    if [ -f $OSG_APP/cmssoft/cms/cmsset_default.sh ] ;then\n'
1103	+	txt += '      # Use $OSG_APP/cmssoft/cms/cmsset_default.sh to setup cms software\n'
1104	+	txt += '        source $OSG_APP/cmssoft/cms/cmsset_default.sh '+self.version+'\n'
1105	+	txt += '    else\n'
1106	+	txt += '        echo "ERROR ==> $OSG_APP/cmssoft/cms/cmsset_default.sh file not found"\n'
1107	+	txt += '        job_exit_code=10020\n'
1108	+	txt += '        func_exit\n'
1109	+	txt += '    fi\n'
1110	+	txt += '\n'
1111	+	txt += '    echo "==> setup cms environment ok"\n'
1112	+	txt += '    echo "SCRAM_ARCH = $SCRAM_ARCH"\n'
1113	+
1114	+	return txt
1115	+
1116	+	def wsSetupCMSLCGEnvironment_(self):
1117	+	"""
1118	+	Returns part of a job script which is prepares
1119	+	the execution environment and which is common for all CMS jobs.
1120	+	"""
1121	+	txt = '\n#Written by cms_cmssw::wsSetupCMSLCGEnvironment_\n'
1122	+	txt += '    echo ">>> setup CMS LCG environment:"\n'
1123	+	txt += '    echo "set SCRAM ARCH and BUILD_ARCH to ' + self.executable_arch + ' ###"\n'
1124	+	txt += '    export SCRAM_ARCH='+self.executable_arch+'\n'
1125	+	txt += '    export BUILD_ARCH='+self.executable_arch+'\n'
1126	+	txt += '    if [ ! $VO_CMS_SW_DIR ] ;then\n'
1127	+	txt += '        echo "ERROR ==> CMS software dir not found on WN `hostname`"\n'
1128	+	txt += '        job_exit_code=10031\n'
1129	+	txt += '        func_exit\n'
1130	+	txt += '    else\n'
1131	+	txt += '        echo "Sourcing environment... "\n'
1132	+	txt += '        if [ ! -s $VO_CMS_SW_DIR/cmsset_default.sh ] ;then\n'
1133	+	txt += '            echo "ERROR ==> cmsset_default.sh file not found into dir $VO_CMS_SW_DIR"\n'
1134	+	txt += '            job_exit_code=10020\n'
1135	+	txt += '            func_exit\n'
1136	+	txt += '        fi\n'
1137	+	txt += '        echo "sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
1138	+	txt += '        source $VO_CMS_SW_DIR/cmsset_default.sh\n'
1139	+	txt += '        result=$?\n'
1140	+	txt += '        if [ $result -ne 0 ]; then\n'
1141	+	txt += '            echo "ERROR ==> problem sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
1142	+	txt += '            job_exit_code=10032\n'
1143	+	txt += '            func_exit\n'
1144	+	txt += '        fi\n'
1145	+	txt += '    fi\n'
1146	+	txt += '    \n'
1147	+	txt += '    echo "==> setup cms environment ok"\n'
1148	+	return txt
1149	+
1150	+	def modifyReport(self, nj):
1151	+	"""
1152	+	insert the part of the script that modifies the FrameworkJob Report
1153	+	"""
1154	+	txt = '\n#Written by cms_cmssw::modifyReport\n'
1155	+	publish_data = int(self.cfg_params.get('USER.publish_data',0))
1156	+	if (publish_data == 1):
1157	+	processedDataset = self.cfg_params['USER.publish_data_name']
1158	+	LFNBaseName = LFNBase(processedDataset)
1159	+
1160	+	txt += 'if [ $copy_exit_status -eq 0 ]; then\n'
1161	+	txt += '    FOR_LFN=%s_${PSETHASH}/\n'%(LFNBaseName)
1162	+	txt += 'else\n'
1163	+	txt += '    FOR_LFN=/copy_problems/ \n'
1164	+	txt += '    SE=""\n'
1165	+	txt += '    SE_PATH=""\n'
1166	+	txt += 'fi\n'
1167	+
1168	+	txt += 'echo ">>> Modify Job Report:" \n'
1169	+	txt += 'chmod a+x $SOFTWARE_DIR/ProdCommon/ProdCommon/FwkJobRep/ModifyJobReport.py\n'
1170	+	txt += 'ProcessedDataset='+processedDataset+'\n'
1171	+	txt += 'echo "ProcessedDataset = $ProcessedDataset"\n'
1172	+	txt += 'echo "SE = $SE"\n'
1173	+	txt += 'echo "SE_PATH = $SE_PATH"\n'
1174	+	txt += 'echo "FOR_LFN = $FOR_LFN" \n'
1175	+	txt += 'echo "CMSSW_VERSION = $CMSSW_VERSION"\n\n'
1176	+	txt += 'echo "$SOFTWARE_DIR/ProdCommon/ProdCommon/FwkJobRep/ModifyJobReport.py $RUNTIME_AREA/crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH"\n'
1177	+	txt += '$SOFTWARE_DIR/ProdCommon/ProdCommon/FwkJobRep/ModifyJobReport.py $RUNTIME_AREA/crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH\n'
1178	+	txt += 'modifyReport_result=$?\n'
1179	+	txt += 'if [ $modifyReport_result -ne 0 ]; then\n'
1180	+	txt += '    modifyReport_result=70500\n'
1181	+	txt += '    job_exit_code=$modifyReport_result\n'
1182	+	txt += '    echo "ModifyReportResult=$modifyReport_result" | tee -a $RUNTIME_AREA/$repo\n'
1183	+	txt += '    echo "WARNING: Problem with ModifyJobReport"\n'
1184	+	txt += 'else\n'
1185	+	txt += '    mv NewFrameworkJobReport.xml $RUNTIME_AREA/crab_fjr_$NJob.xml\n'
1186	+	txt += 'fi\n'
1187	+	return txt
1188	+
1189	+	def wsParseFJR(self):
1190	+	"""
1191	+	Parse the FrameworkJobReport to obtain useful infos
1192	+	"""
1193	+	txt = '\n#Written by cms_cmssw::wsParseFJR\n'
1194	+	txt += 'echo ">>> Parse FrameworkJobReport crab_fjr.xml"\n'
1195	+	txt += 'if [ -s $RUNTIME_AREA/crab_fjr_$NJob.xml ]; then\n'
1196	+	txt += '    if [ -s $RUNTIME_AREA/parseCrabFjr.py ]; then\n'
1197	+	txt += '        cmd_out=`python $RUNTIME_AREA/parseCrabFjr.py --input $RUNTIME_AREA/crab_fjr_$NJob.xml --MonitorID $MonitorID --MonitorJobID $MonitorJobID`\n'
1198	+	txt += '        echo "Result of parsing the FrameworkJobReport crab_fjr.xml: $cmd_out"\n'
1199	+	txt += '        tmp_executable_exit_status=`echo $cmd_out | awk -F\; \'{print $1}\' | awk -F \' \' \'{print $NF}\'`\n'
1200	+	txt += '        if [ -n $tmp_executable_exit_status ];then\n'
1201	+	txt += '            executable_exit_status=$tmp_executable_exit_status\n'
1202	+	txt += '        fi\n'
1203	+	txt += '        if [ $executable_exit_status -eq 50115 ];then\n'
1204	+	txt += '            echo ">>> crab_fjr.xml contents: "\n'
1205	+	txt += '            cat $RUNTIME_AREA/crab_fjr_NJob.xml\n'
1206	+	txt += '            echo "Wrong FrameworkJobReport --> does not contain useful info. ExitStatus: $executable_exit_status"\n'
1207	+	txt += '        else\n'
1208	+	txt += '            echo "Extracted ExitStatus from FrameworkJobReport parsing output: $executable_exit_status"\n'
1209	+	txt += '        fi\n'
1210	+	txt += '    else\n'
1211	+	txt += '        echo "CRAB python script to parse CRAB FrameworkJobReport crab_fjr.xml is not available, using exit code of executable from command line."\n'
1212	+	txt += '    fi\n'
1213	+	#### Patch to check input data reading for CMSSW16x Hopefully we-ll remove it asap
1214	+
1215	+	if self.datasetPath:
1216	+	# VERIFY PROCESSED DATA
1217	+	txt += '    if [ $executable_exit_status -eq 0 ];then\n'
1218	+	txt += '      echo ">>> Verify list of processed files:"\n'
1219	+	txt += '      echo $InputFiles |tr -d "\\\\" |tr "," "\\n"|tr -d "\\"" > input-files.txt\n'
1220	+	txt += '      grep LFN $RUNTIME_AREA/crab_fjr_$NJob.xml |cut -d">" -f2|cut -d"<" -f1|grep "/" > processed-files.txt\n'
1221	+	txt += '      cat input-files.txt  | sort | uniq > tmp.txt\n'
1222	+	txt += '      mv tmp.txt input-files.txt\n'
1223	+	txt += '      echo "cat input-files.txt"\n'
1224	+	txt += '      echo "----------------------"\n'
1225	+	txt += '      cat input-files.txt\n'
1226	+	txt += '      cat processed-files.txt | sort | uniq > tmp.txt\n'
1227	+	txt += '      mv tmp.txt processed-files.txt\n'
1228	+	txt += '      echo "----------------------"\n'
1229	+	txt += '      echo "cat processed-files.txt"\n'
1230	+	txt += '      echo "----------------------"\n'
1231	+	txt += '      cat processed-files.txt\n'
1232	+	txt += '      echo "----------------------"\n'
1233	+	txt += '      diff -q input-files.txt processed-files.txt\n'
1234	+	txt += '      fileverify_status=$?\n'
1235	+	txt += '      if [ $fileverify_status -ne 0 ]; then\n'
1236	+	txt += '         executable_exit_status=30001\n'
1237	+	txt += '         echo "ERROR ==> not all input files processed"\n'
1238	+	txt += '         echo "      ==> list of processed files from crab_fjr.xml differs from list in pset.cfg"\n'
1239	+	txt += '         echo "      ==> diff input-files.txt processed-files.txt"\n'
1240	+	txt += '      fi\n'
1241	+	txt += '    fi\n'
1242	+	txt += '\n'
1243	+	txt += 'else\n'
1244	+	txt += '    echo "CRAB FrameworkJobReport crab_fjr.xml is not available, using exit code of executable from command line."\n'
1245	+	txt += 'fi\n'
1246	+	txt += '\n'
1247	+	txt += 'echo "ExeExitCode=$executable_exit_status" | tee -a $RUNTIME_AREA/$repo\n'
1248	+	txt += 'echo "EXECUTABLE_EXIT_STATUS = $executable_exit_status"\n'
1249	+	txt += 'job_exit_code=$executable_exit_status\n'
1250	+
1251	+	return txt
1252	+
1253	+	def setParam_(self, param, value):
1254	+	self._params[param] = value
1255	+
1256	+	def getParams(self):
1257	+	return self._params
1258	+
1259	+	def uniquelist(self, old):
1260	+	"""
1261	+	remove duplicates from a list
1262	+	"""
1263	+	nd={}
1264	+	for e in old:
1265	+	nd[e]=0
1266	+	return nd.keys()
1267	+
1268	+	def outList(self):
1269	+	"""
1270	+	check the dimension of the output files
1271	+	"""
1272	+	txt = ''
1273	+	txt += 'echo ">>> list of expected files on output sandbox"\n'
1274	+	listOutFiles = []
1275	+	stdout = 'CMSSW_$NJob.stdout'
1276	+	stderr = 'CMSSW_$NJob.stderr'
1277	+	if (self.return_data == 1):
1278	+	for file in (self.output_file+self.output_file_sandbox):
1279	+	listOutFiles.append(self.numberFile_(file, '$NJob'))
1280	+	listOutFiles.append(stdout)
1281	+	listOutFiles.append(stderr)
1282	+	else:
1283	+	for file in (self.output_file_sandbox):
1284	+	listOutFiles.append(self.numberFile_(file, '$NJob'))
1285	+	listOutFiles.append(stdout)
1286	+	listOutFiles.append(stderr)
1287	+	txt += 'echo "output files: '+string.join(listOutFiles,' ')+'"\n'
1288	+	txt += 'filesToCheck="'+string.join(listOutFiles,' ')+'"\n'
1289	+	txt += 'export filesToCheck\n'
1290	+	return txt

Diff Legend

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