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 1.64 by slacapra, Thu Jan 18 18:29:51 2007 UTC vs.
Revision 1.136 by fanzago, Wed Nov 14 17:53:21 2007 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	–	import PsetManipulator
7	–
8	–	import DataDiscovery
9	–	import DataLocation
7		import Scram
8
9	<	import os, string, re, shutil
9	>	import os, string, glob
10
11		class Cmssw(JobType):
12		def __init__(self, cfg_params, ncjobs):
13		JobType.__init__(self, 'CMSSW')
14		common.logger.debug(3,'CMSSW::__init__')
15
19	–	# Marco.
16		self._params = {}
17		self.cfg_params = cfg_params
18
19	+	# init BlackWhiteListParser
20	+	self.blackWhiteListParser = BlackWhiteListParser(cfg_params)
21	+
22	+	try:
23	+	self.MaxTarBallSize = float(self.cfg_params['EDG.maxtarballsize'])
24	+	except KeyError:
25	+	self.MaxTarBallSize = 9.5
26	+
27		# number of jobs requested to be created, limit obj splitting
28		self.ncjobs = ncjobs
29
30		log = common.logger
31	<
31	>
32		self.scram = Scram.Scram(cfg_params)
33		self.additional_inbox_files = []
34		self.scriptExe = ''
35		self.executable = ''
36	+	self.executable_arch = self.scram.getArch()
37		self.tgz_name = 'default.tgz'
38	+	self.additional_tgz_name = 'additional.tgz'
39		self.scriptName = 'CMSSW.sh'
40	<	self.pset = ''      #scrip use case Da
40	>	self.pset = ''      #scrip use case Da
41		self.datasetPath = '' #scrip use case Da
42
43		# set FJR file name
44		self.fjrFileName = 'crab_fjr.xml'
45
46		self.version = self.scram.getSWVersion()
47	+
48	+	#
49	+	# Try to block creation in case of arch/version mismatch
50	+	#
51	+
52	+	a = string.split(self.version, "_")
53	+
54	+	if int(a[1]) == 1 and (int(a[2]) < 5 and self.executable_arch.find('slc4') == 0):
55	+	msg = "Warning: You are using %s version of CMSSW  with %s architecture. \n--> Did you compile your libraries with SLC3? Otherwise you can find some problems running on SLC4 Grid nodes.\n"%(self.version, self.executable_arch)
56	+	common.logger.message(msg)
57	+	if int(a[1]) == 1 and (int(a[2]) >= 5 and self.executable_arch.find('slc3') == 0):
58	+	msg = "Error: CMS does not support %s with %s architecture"%(self.version, self.executable_arch)
59	+	raise CrabException(msg)
60	+
61		common.taskDB.setDict('codeVersion',self.version)
62		self.setParam_('application', self.version)
63
64		### collect Data cards
65	+
66	+	## get DBS mode
67	+	try:
68	+	self.use_dbs_1 = int(self.cfg_params['CMSSW.use_dbs_1'])
69	+	except KeyError:
70	+	self.use_dbs_1 = 0
71	+
72		try:
73		tmp =  cfg_params['CMSSW.datasetpath']
74		log.debug(6, "CMSSW::CMSSW(): datasetPath = "+tmp)
#	Line 52 | Line 79 | class Cmssw(JobType):
79		self.datasetPath = tmp
80		self.selectNoInput = 0
81		except KeyError:
82	<	msg = "Error: datasetpath not defined "
82	>	msg = "Error: datasetpath not defined "
83		raise CrabException(msg)
84
85		# ML monitoring
#	Line 61 | Line 88 | class Cmssw(JobType):
88		self.setParam_('dataset', 'None')
89		self.setParam_('owner', 'None')
90		else:
91	<	datasetpath_split = self.datasetPath.split("/")
92	<	self.setParam_('dataset', datasetpath_split[1])
93	<	self.setParam_('owner', datasetpath_split[-1])
91	>	try:
92	>	datasetpath_split = self.datasetPath.split("/")
93	>	# standard style
94	>	self.setParam_('datasetFull', self.datasetPath)
95	>	if self.use_dbs_1 == 1 :
96	>	self.setParam_('dataset', datasetpath_split[1])
97	>	self.setParam_('owner', datasetpath_split[-1])
98	>	else:
99	>	self.setParam_('dataset', datasetpath_split[1])
100	>	self.setParam_('owner', datasetpath_split[2])
101	>	except:
102	>	self.setParam_('dataset', self.datasetPath)
103	>	self.setParam_('owner', self.datasetPath)
104
105		self.setTaskid_()
106		self.setParam_('taskId', self.cfg_params['taskId'])
#	Line 87 | Line 124 | class Cmssw(JobType):
124		try:
125		self.pset = cfg_params['CMSSW.pset']
126		log.debug(6, "Cmssw::Cmssw(): PSet file = "+self.pset)
127	<	if self.pset.lower() != 'none' :
127	>	if self.pset.lower() != 'none' :
128		if (not os.path.exists(self.pset)):
129		raise CrabException("User defined PSet file "+self.pset+" does not exist")
130		else:
#	Line 114 | Line 151 | class Cmssw(JobType):
151		self.output_file.append(tmp)
152		pass
153		else:
154	<	log.message("No output file defined: only stdout/err and the CRAB Framework Job Report will be available")
154	>	log.message("No output file defined: only stdout/err and the CRAB Framework Job Report will be available\n")
155		pass
156		pass
157		except KeyError:
158	<	log.message("No output file defined: only stdout/err and the CRAB Framework Job Report will be available")
158	>	log.message("No output file defined: only stdout/err and the CRAB Framework Job Report will be available\n")
159		pass
160
161		# script_exe file as additional file in inputSandbox
#	Line 131 | Line 168 | class Cmssw(JobType):
168		self.additional_inbox_files.append(string.strip(self.scriptExe))
169		except KeyError:
170		self.scriptExe = ''
171	+
172		#CarlosDaniele
173		if self.datasetPath == None and self.pset == None and self.scriptExe == '' :
174	<	msg ="WARNING. script_exe  not defined"
174	>	msg ="Error. script_exe  not defined"
175		raise CrabException(msg)
176
177		## additional input files
178		try:
179		tmpAddFiles = string.split(cfg_params['USER.additional_input_files'],',')
180	<	common.logger.debug(5,"Additional input files: "+str(tmpAddFiles))
181	<	for tmpFile in tmpAddFiles:
182	<	tmpFile = string.strip(tmpFile)
183	<	if not os.path.exists(tmpFile):
184	<	raise CrabException("Additional input file not found: "+tmpFile)
180	>	for tmp in tmpAddFiles:
181	>	tmp = string.strip(tmp)
182	>	dirname = ''
183	>	if not tmp[0]=="/": dirname = "."
184	>	files = []
185	>	if string.find(tmp,"*")>-1:
186	>	files = glob.glob(os.path.join(dirname, tmp))
187	>	if len(files)==0:
188	>	raise CrabException("No additional input file found with this pattern: "+tmp)
189	>	else:
190	>	files.append(tmp)
191	>	for file in files:
192	>	if not os.path.exists(file):
193	>	raise CrabException("Additional input file not found: "+file)
194		pass
195	<	storedFile = common.work_space.shareDir()+ tmpFile
196	<	shutil.copyfile(tmpFile, storedFile)
197	<	self.additional_inbox_files.append(string.strip(storedFile))
195	>	# fname = string.split(file, '/')[-1]
196	>	# storedFile = common.work_space.pathForTgz()+'share/'+fname
197	>	# shutil.copyfile(file, storedFile)
198	>	self.additional_inbox_files.append(string.strip(file))
199		pass
152	–	common.logger.debug(5,"Inbox files so far : "+str(self.additional_inbox_files))
200		pass
201	+	common.logger.debug(5,"Additional input files: "+str(self.additional_inbox_files))
202		except KeyError:
203		pass
204
#	Line 168 | Line 216 | class Cmssw(JobType):
216		except KeyError:
217		self.eventsPerJob = -1
218		self.selectEventsPerJob = 0
219	<
219	>
220		## number of jobs
221		try:
222		self.theNumberOfJobs =int( cfg_params['CMSSW.number_of_jobs'])
#	Line 184 | Line 232 | class Cmssw(JobType):
232		self.total_number_of_events = 0
233		self.selectTotalNumberEvents = 0
234
235	<	if self.pset != None: #CarlosDaniele
235	>	if self.pset != None: #CarlosDaniele
236		if ( (self.selectTotalNumberEvents + self.selectEventsPerJob + self.selectNumberOfJobs) != 2 ):
237		msg = 'Must define exactly two of total_number_of_events, events_per_job, or number_of_jobs.'
238		raise CrabException(msg)
#	Line 205 | Line 253 | class Cmssw(JobType):
253		except KeyError:
254		self.sourceSeedVtx = None
255		common.logger.debug(5,"No vertex seed given")
256	+
257	+	try:
258	+	self.sourceSeedG4 = int(cfg_params['CMSSW.g4_seed'])
259	+	except KeyError:
260	+	self.sourceSeedG4 = None
261	+	common.logger.debug(5,"No g4 sim hits seed given")
262	+
263	+	try:
264	+	self.sourceSeedMix = int(cfg_params['CMSSW.mix_seed'])
265	+	except KeyError:
266	+	self.sourceSeedMix = None
267	+	common.logger.debug(5,"No mix seed given")
268	+
269		try:
270		self.firstRun = int(cfg_params['CMSSW.first_run'])
271		except KeyError:
272		self.firstRun = None
273		common.logger.debug(5,"No first run given")
274		if self.pset != None: #CarlosDaniele
275	<	self.PsetEdit = PsetManipulator.PsetManipulator(self.pset) #Daniele Pset
275	>	import PsetManipulator as pp
276	>	PsetEdit = pp.PsetManipulator(self.pset) #Daniele Pset
277
278		#DBSDLS-start
279	<	## Initialize the variables that are extracted from DBS/DLS and needed in other places of the code
279	>	## Initialize the variables that are extracted from DBS/DLS and needed in other places of the code
280		self.maxEvents=0  # max events available   ( --> check the requested nb. of evts in Creator.py)
281		self.DBSPaths={}  # all dbs paths requested ( --> input to the site local discovery script)
282		self.jobDestination=[]  # Site destination(s) for each job (list of lists)
#	Line 223 | Line 285 | class Cmssw(JobType):
285		blockSites = {}
286		if self.datasetPath:
287		blockSites = self.DataDiscoveryAndLocation(cfg_params)
288	<	#DBSDLS-end
288	>	#DBSDLS-end
289
290		self.tgzNameWithPath = self.getTarBall(self.executable)
291	<
291	>
292		## Select Splitting
293	<	if self.selectNoInput:
293	>	if self.selectNoInput:
294		if self.pset == None: #CarlosDaniele
295		self.jobSplittingForScript()
296		else:
297		self.jobSplittingNoInput()
298	<	else:
298	>	else:
299		self.jobSplittingByBlocks(blockSites)
300
301		# modify Pset
#	Line 241 | Line 303 | class Cmssw(JobType):
303		try:
304		if (self.datasetPath): # standard job
305		# allow to processa a fraction of events in a file
306	<	self.PsetEdit.inputModule("INPUT")
307	<	self.PsetEdit.maxEvent("INPUTMAXEVENTS")
308	<	self.PsetEdit.skipEvent("INPUTSKIPEVENTS")
306	>	PsetEdit.inputModule("INPUTFILE")
307	>	PsetEdit.maxEvent(0)
308	>	PsetEdit.skipEvent(0)
309		else:  # pythia like job
310	<	self.PsetEdit.maxEvent(self.eventsPerJob)
310	>	PsetEdit.maxEvent(self.eventsPerJob)
311		if (self.firstRun):
312	<	self.PsetEdit.pythiaFirstRun("INPUTFIRSTRUN")  #First Run
312	>	PsetEdit.pythiaFirstRun(0)  #First Run
313		if (self.sourceSeed) :
314	<	self.PsetEdit.pythiaSeed("INPUT")
314	>	PsetEdit.pythiaSeed(0)
315		if (self.sourceSeedVtx) :
316	<	self.PsetEdit.pythiaSeedVtx("INPUTVTX")
316	>	PsetEdit.vtxSeed(0)
317	>	if (self.sourceSeedG4) :
318	>	PsetEdit.g4Seed(0)
319	>	if (self.sourceSeedMix) :
320	>	PsetEdit.mixSeed(0)
321		# add FrameworkJobReport to parameter-set
322	<	self.PsetEdit.addCrabFJR(self.fjrFileName)
323	<	self.PsetEdit.psetWriter(self.configFilename())
322	>	PsetEdit.addCrabFJR(self.fjrFileName)
323	>	PsetEdit.psetWriter(self.configFilename())
324		except:
325		msg='Error while manipuliating ParameterSet: exiting...'
326		raise CrabException(msg)
327
328		def DataDiscoveryAndLocation(self, cfg_params):
329
330	+	import DataDiscovery
331	+	import DataDiscovery_DBS2
332	+	import DataLocation
333		common.logger.debug(10,"CMSSW::DataDiscoveryAndLocation()")
334
335		datasetPath=self.datasetPath
336
268	–	## TODO
269	–	dataTiersList = ""
270	–	dataTiers = dataTiersList.split(',')
271	–
337		## Contact the DBS
338	<	common.logger.message("Contacting DBS...")
338	>	common.logger.message("Contacting Data Discovery Services ...")
339		try:
340	<	self.pubdata=DataDiscovery.DataDiscovery(datasetPath, dataTiers, cfg_params)
340	>
341	>	if self.use_dbs_1 == 1 :
342	>	self.pubdata=DataDiscovery.DataDiscovery(datasetPath, cfg_params)
343	>	else :
344	>	self.pubdata=DataDiscovery_DBS2.DataDiscovery_DBS2(datasetPath, cfg_params)
345		self.pubdata.fetchDBSInfo()
346
347		except DataDiscovery.NotExistingDatasetError, ex :
348		msg = 'ERROR ***: failed Data Discovery in DBS : %s'%ex.getErrorMessage()
349		raise CrabException(msg)
281	–
350		except DataDiscovery.NoDataTierinProvenanceError, ex :
351		msg = 'ERROR ***: failed Data Discovery in DBS : %s'%ex.getErrorMessage()
352		raise CrabException(msg)
353		except DataDiscovery.DataDiscoveryError, ex:
354	<	msg = 'ERROR ***: failed Data Discovery in DBS  %s'%ex.getErrorMessage()
354	>	msg = 'ERROR ***: failed Data Discovery in DBS :  %s'%ex.getErrorMessage()
355	>	raise CrabException(msg)
356	>	except DataDiscovery_DBS2.NotExistingDatasetError_DBS2, ex :
357	>	msg = 'ERROR ***: failed Data Discovery in DBS : %s'%ex.getErrorMessage()
358	>	raise CrabException(msg)
359	>	except DataDiscovery_DBS2.NoDataTierinProvenanceError_DBS2, ex :
360	>	msg = 'ERROR ***: failed Data Discovery in DBS : %s'%ex.getErrorMessage()
361	>	raise CrabException(msg)
362	>	except DataDiscovery_DBS2.DataDiscoveryError_DBS2, ex:
363	>	msg = 'ERROR ***: failed Data Discovery in DBS :  %s'%ex.getErrorMessage()
364		raise CrabException(msg)
288	–
289	–	## get list of all required data in the form of dbs paths  (dbs path = /dataset/datatier/owner)
290	–	## self.DBSPaths=self.pubdata.getDBSPaths()
291	–	common.logger.message("Required data are :"+self.datasetPath)
365
366		self.filesbyblock=self.pubdata.getFiles()
367		self.eventsbyblock=self.pubdata.getEventsPerBlock()
368		self.eventsbyfile=self.pubdata.getEventsPerFile()
296	–	# print str(self.filesbyblock)
297	–	# print 'self.eventsbyfile',len(self.eventsbyfile)
298	–	# print str(self.eventsbyfile)
369
370		## get max number of events
371	<	self.maxEvents=self.pubdata.getMaxEvents() ##  self.maxEvents used in Creator.py
302	<	common.logger.message("The number of available events is %s\n"%self.maxEvents)
371	>	self.maxEvents=self.pubdata.getMaxEvents() ##  self.maxEvents used in Creator.py
372
304	–	common.logger.message("Contacting DLS...")
373		## Contact the DLS and build a list of sites hosting the fileblocks
374		try:
375		dataloc=DataLocation.DataLocation(self.filesbyblock.keys(),cfg_params)
#	Line 309 | Line 377 | class Cmssw(JobType):
377		except DataLocation.DataLocationError , ex:
378		msg = 'ERROR ***: failed Data Location in DLS \n %s '%ex.getErrorMessage()
379		raise CrabException(msg)
380	<
380	>
381
382		sites = dataloc.getSites()
383		allSites = []
#	Line 319 | Line 387 | class Cmssw(JobType):
387		allSites.append(oneSite)
388		allSites = self.uniquelist(allSites)
389
390	<	common.logger.message("Sites ("+str(len(allSites))+") hosting part/all of dataset: "+str(allSites))
391	<	common.logger.debug(6, "List of Sites: "+str(allSites))
390	>	# screen output
391	>	common.logger.message("Requested dataset: " + datasetPath + " has " + str(self.maxEvents) + " events in " + str(len(self.filesbyblock.keys())) + " blocks.\n")
392	>
393		return sites
394	<
394	>
395		def jobSplittingByBlocks(self, blockSites):
396		"""
397		Perform job splitting. Jobs run over an integer number of files
#	Line 372 | Line 441 | class Cmssw(JobType):
441		totalNumberOfJobs = 999999999
442		else :
443		totalNumberOfJobs = self.ncjobs
444	<
444	>
445
446		blocks = blockSites.keys()
447		blockCount = 0
#	Line 382 | Line 451 | class Cmssw(JobType):
451		jobCount = 0
452		list_of_lists = []
453
454	+	# list tracking which jobs are in which jobs belong to which block
455	+	jobsOfBlock = {}
456	+
457		# ---- Iterate over the blocks in the dataset until ---- #
458		# ---- we've met the requested total # of events    ---- #
459		while ( (eventsRemaining > 0) and (blockCount < numBlocksInDataset) and (jobCount < totalNumberOfJobs)):
460		block = blocks[blockCount]
461		blockCount += 1
462	<
462	>	if block not in jobsOfBlock.keys() :
463	>	jobsOfBlock[block] = []
464
465	<	numEventsInBlock = self.eventsbyblock[block]
466	<	common.logger.debug(5,'Events in Block File '+str(numEventsInBlock))
467	<
468	<	files = self.filesbyblock[block]
469	<	numFilesInBlock = len(files)
470	<	if (numFilesInBlock <= 0):
471	<	continue
472	<	fileCount = 0
473	<
474	<	# ---- New block => New job ---- #
475	<	parString = "\\{"
476	<	# counter for number of events in files currently worked on
477	<	filesEventCount = 0
478	<	# flag if next while loop should touch new file
479	<	newFile = 1
480	<	# job event counter
481	<	jobSkipEventCount = 0
482	<
483	<	# ---- Iterate over the files in the block until we've met the requested ---- #
484	<	# ---- total # of events or we've gone over all the files in this block  ---- #
485	<	while ( (eventsRemaining > 0) and (fileCount < numFilesInBlock) and (jobCount < totalNumberOfJobs) ):
486	<	file = files[fileCount]
487	<	if newFile :
488	<	try:
489	<	numEventsInFile = self.eventsbyfile[file]
490	<	common.logger.debug(6, "File "+str(file)+" has "+str(numEventsInFile)+" events")
491	<	# increase filesEventCount
492	<	filesEventCount += numEventsInFile
493	<	# Add file to current job
494	<	parString += '\\\"' + file + '\\\"\,'
495	<	newFile = 0
496	<	except KeyError:
497	<	common.logger.message("File "+str(file)+" has unknown number of events: skipping")
498	<
499	<
500	<	# if less events in file remain than eventsPerJobRequested
501	<	if ( filesEventCount - jobSkipEventCount < eventsPerJobRequested ) :
502	<	# if last file in block
503	<	if ( fileCount == numFilesInBlock-1 ) :
504	<	# end job using last file, use remaining events in block
465	>	if self.eventsbyblock.has_key(block) :
466	>	numEventsInBlock = self.eventsbyblock[block]
467	>	common.logger.debug(5,'Events in Block File '+str(numEventsInBlock))
468	>
469	>	files = self.filesbyblock[block]
470	>	numFilesInBlock = len(files)
471	>	if (numFilesInBlock <= 0):
472	>	continue
473	>	fileCount = 0
474	>
475	>	# ---- New block => New job ---- #
476	>	parString = ""
477	>	# counter for number of events in files currently worked on
478	>	filesEventCount = 0
479	>	# flag if next while loop should touch new file
480	>	newFile = 1
481	>	# job event counter
482	>	jobSkipEventCount = 0
483	>
484	>	# ---- Iterate over the files in the block until we've met the requested ---- #
485	>	# ---- total # of events or we've gone over all the files in this block  ---- #
486	>	while ( (eventsRemaining > 0) and (fileCount < numFilesInBlock) and (jobCount < totalNumberOfJobs) ):
487	>	file = files[fileCount]
488	>	if newFile :
489	>	try:
490	>	numEventsInFile = self.eventsbyfile[file]
491	>	common.logger.debug(6, "File "+str(file)+" has "+str(numEventsInFile)+" events")
492	>	# increase filesEventCount
493	>	filesEventCount += numEventsInFile
494	>	# Add file to current job
495	>	parString += '\\\"' + file + '\\\"\,'
496	>	newFile = 0
497	>	except KeyError:
498	>	common.logger.message("File "+str(file)+" has unknown number of events: skipping")
499	>
500	>
501	>	# if less events in file remain than eventsPerJobRequested
502	>	if ( filesEventCount - jobSkipEventCount < eventsPerJobRequested ) :
503	>	# if last file in block
504	>	if ( fileCount == numFilesInBlock-1 ) :
505	>	# end job using last file, use remaining events in block
506	>	# close job and touch new file
507	>	fullString = parString[:-2]
508	>	list_of_lists.append([fullString,str(-1),str(jobSkipEventCount)])
509	>	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(filesEventCount - jobSkipEventCount)+" events (last file in block).")
510	>	self.jobDestination.append(blockSites[block])
511	>	common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
512	>	# fill jobs of block dictionary
513	>	jobsOfBlock[block].append(jobCount+1)
514	>	# reset counter
515	>	jobCount = jobCount + 1
516	>	totalEventCount = totalEventCount + filesEventCount - jobSkipEventCount
517	>	eventsRemaining = eventsRemaining - filesEventCount + jobSkipEventCount
518	>	jobSkipEventCount = 0
519	>	# reset file
520	>	parString = ""
521	>	filesEventCount = 0
522	>	newFile = 1
523	>	fileCount += 1
524	>	else :
525	>	# go to next file
526	>	newFile = 1
527	>	fileCount += 1
528	>	# if events in file equal to eventsPerJobRequested
529	>	elif ( filesEventCount - jobSkipEventCount == eventsPerJobRequested ) :
530		# close job and touch new file
531		fullString = parString[:-2]
532	<	fullString += '\\}'
533	<	list_of_lists.append([fullString,str(-1),str(jobSkipEventCount)])
436	<	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(filesEventCount - jobSkipEventCount)+" events (last file in block).")
532	>	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
533	>	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.")
534		self.jobDestination.append(blockSites[block])
535		common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
536	+	jobsOfBlock[block].append(jobCount+1)
537		# reset counter
538		jobCount = jobCount + 1
539	<	totalEventCount = totalEventCount + filesEventCount - jobSkipEventCount
540	<	eventsRemaining = eventsRemaining - filesEventCount + jobSkipEventCount
539	>	totalEventCount = totalEventCount + eventsPerJobRequested
540	>	eventsRemaining = eventsRemaining - eventsPerJobRequested
541		jobSkipEventCount = 0
542		# reset file
543	<	parString = "\\{"
543	>	parString = ""
544		filesEventCount = 0
545		newFile = 1
546		fileCount += 1
547	+
548	+	# if more events in file remain than eventsPerJobRequested
549		else :
550	<	# go to next file
551	<	newFile = 1
552	<	fileCount += 1
553	<	# if events in file equal to eventsPerJobRequested
554	<	elif ( filesEventCount - jobSkipEventCount == eventsPerJobRequested ) :
555	<	# close job and touch new file
556	<	fullString = parString[:-2]
557	<	fullString += '\\}'
558	<	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
559	<	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.")
560	<	self.jobDestination.append(blockSites[block])
561	<	common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
562	<	# reset counter
563	<	jobCount = jobCount + 1
564	<	totalEventCount = totalEventCount + eventsPerJobRequested
565	<	eventsRemaining = eventsRemaining - eventsPerJobRequested
566	<	jobSkipEventCount = 0
567	<	# reset file
568	<	parString = "\\{"
569	<	filesEventCount = 0
470	<	newFile = 1
471	<	fileCount += 1
472	<
473	<	# if more events in file remain than eventsPerJobRequested
474	<	else :
475	<	# close job but don't touch new file
476	<	fullString = parString[:-2]
477	<	fullString += '\\}'
478	<	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
479	<	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.")
480	<	self.jobDestination.append(blockSites[block])
481	<	common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
482	<	# increase counter
483	<	jobCount = jobCount + 1
484	<	totalEventCount = totalEventCount + eventsPerJobRequested
485	<	eventsRemaining = eventsRemaining - eventsPerJobRequested
486	<	# calculate skip events for last file
487	<	# use filesEventCount (contains several files), jobSkipEventCount and eventsPerJobRequest
488	<	jobSkipEventCount = eventsPerJobRequested - (filesEventCount - jobSkipEventCount - self.eventsbyfile[file])
489	<	# remove all but the last file
490	<	filesEventCount = self.eventsbyfile[file]
491	<	parString = "\\{"
492	<	parString += '\\\"' + file + '\\\"\,'
493	<	pass # END if
494	<	pass # END while (iterate over files in the block)
550	>	# close job but don't touch new file
551	>	fullString = parString[:-2]
552	>	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
553	>	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.")
554	>	self.jobDestination.append(blockSites[block])
555	>	common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
556	>	jobsOfBlock[block].append(jobCount+1)
557	>	# increase counter
558	>	jobCount = jobCount + 1
559	>	totalEventCount = totalEventCount + eventsPerJobRequested
560	>	eventsRemaining = eventsRemaining - eventsPerJobRequested
561	>	# calculate skip events for last file
562	>	# use filesEventCount (contains several files), jobSkipEventCount and eventsPerJobRequest
563	>	jobSkipEventCount = eventsPerJobRequested - (filesEventCount - jobSkipEventCount - self.eventsbyfile[file])
564	>	# remove all but the last file
565	>	filesEventCount = self.eventsbyfile[file]
566	>	parString = ""
567	>	parString += '\\\"' + file + '\\\"\,'
568	>	pass # END if
569	>	pass # END while (iterate over files in the block)
570		pass # END while (iterate over blocks in the dataset)
571		self.ncjobs = self.total_number_of_jobs = jobCount
572		if (eventsRemaining > 0 and jobCount < totalNumberOfJobs ):
573		common.logger.message("Could not run on all requested events because some blocks not hosted at allowed sites.")
574	<	common.logger.message("\n"+str(jobCount)+" job(s) can run on "+str(totalEventCount)+" events.\n")
575	<
574	>	common.logger.message(str(jobCount)+" job(s) can run on "+str(totalEventCount)+" events.\n")
575	>
576	>	# screen output
577	>	screenOutput = "List of jobs and available destination sites:\n\n"
578	>
579	>	# keep trace of block with no sites to print a warning at the end
580	>	noSiteBlock = []
581	>	bloskNoSite = []
582	>
583	>	blockCounter = 0
584	>	for block in blocks:
585	>	if block in jobsOfBlock.keys() :
586	>	blockCounter += 1
587	>	screenOutput += "Block %5i: jobs %20s: sites: %s\n" % (blockCounter,spanRanges(jobsOfBlock[block]),','.join(self.blackWhiteListParser.checkWhiteList(self.blackWhiteListParser.checkBlackList(blockSites[block],block),block)))
588	>	if len(self.blackWhiteListParser.checkWhiteList(self.blackWhiteListParser.checkBlackList(blockSites[block],block),block)) == 0:
589	>	noSiteBlock.append( spanRanges(jobsOfBlock[block]) )
590	>	bloskNoSite.append( blockCounter )
591	>
592	>	common.logger.message(screenOutput)
593	>	if len(noSiteBlock) > 0 and len(bloskNoSite) > 0:
594	>	msg = 'WARNING: No sites are hosting any part of data for block:\n                '
595	>	virgola = ""
596	>	if len(bloskNoSite) > 1:
597	>	virgola = ","
598	>	for block in bloskNoSite:
599	>	msg += ' ' + str(block) + virgola
600	>	msg += '\n               Related jobs:\n                 '
601	>	virgola = ""
602	>	if len(noSiteBlock) > 1:
603	>	virgola = ","
604	>	for range_jobs in noSiteBlock:
605	>	msg += str(range_jobs) + virgola
606	>	msg += '\n               will not be submitted and this block of data can not be analyzed!\n'
607	>	common.logger.message(msg)
608	>
609		self.list_of_args = list_of_lists
610		return
611
#	Line 506 | Line 614 | class Cmssw(JobType):
614		Perform job splitting based on number of event per job
615		"""
616		common.logger.debug(5,'Splitting per events')
617	<	common.logger.message('Required '+str(self.eventsPerJob)+' events per job ')
618	<	common.logger.message('Required '+str(self.theNumberOfJobs)+' jobs in total ')
619	<	common.logger.message('Required '+str(self.total_number_of_events)+' events in total ')
617	>
618	>	if (self.selectEventsPerJob):
619	>	common.logger.message('Required '+str(self.eventsPerJob)+' events per job ')
620	>	if (self.selectNumberOfJobs):
621	>	common.logger.message('Required '+str(self.theNumberOfJobs)+' jobs in total ')
622	>	if (self.selectTotalNumberEvents):
623	>	common.logger.message('Required '+str(self.total_number_of_events)+' events in total ')
624
625		if (self.total_number_of_events < 0):
626		msg='Cannot split jobs per Events with "-1" as total number of events'
627		raise CrabException(msg)
628
629		if (self.selectEventsPerJob):
630	<	self.total_number_of_jobs = int(self.total_number_of_events/self.eventsPerJob)
630	>	if (self.selectTotalNumberEvents):
631	>	self.total_number_of_jobs = int(self.total_number_of_events/self.eventsPerJob)
632	>	elif(self.selectNumberOfJobs) :
633	>	self.total_number_of_jobs =self.theNumberOfJobs
634	>	self.total_number_of_events =int(self.theNumberOfJobs*self.eventsPerJob)
635	>
636		elif (self.selectNumberOfJobs) :
637		self.total_number_of_jobs = self.theNumberOfJobs
638		self.eventsPerJob = int(self.total_number_of_events/self.total_number_of_jobs)
#	Line 536 | Line 653 | class Cmssw(JobType):
653		for i in range(self.total_number_of_jobs):
654		## Since there is no input, any site is good
655		# self.jobDestination.append(["Any"])
656	<	self.jobDestination.append([""]) #must be empty to write correctly the xml
657	<	args=''
656	>	self.jobDestination.append([""]) #must be empty to write correctly the xml
657	>	args=[]
658		if (self.firstRun):
659		## pythia first run
660		#self.list_of_args.append([(str(self.firstRun)+str(i))])
661	<	args=args+(str(self.firstRun)+str(i))
661	>	args.append(str(self.firstRun)+str(i))
662		else:
663		## no first run
664		#self.list_of_args.append([str(i)])
665	<	args=args+str(i)
665	>	args.append(str(i))
666		if (self.sourceSeed):
667	+	args.append(str(self.sourceSeed)+str(i))
668		if (self.sourceSeedVtx):
669	<	## pythia + vtx random seed
670	<	#self.list_of_args.append([
671	<	#                          str(self.sourceSeed)+str(i),
672	<	#                          str(self.sourceSeedVtx)+str(i)
673	<	#                          ])
674	<	args=args+str(',')+str(self.sourceSeed)+str(i)+str(',')+str(self.sourceSeedVtx)+str(i)
675	<	else:
676	<	## only pythia random seed
677	<	#self.list_of_args.append([(str(self.sourceSeed)+str(i))])
678	<	args=args +str(',')+str(self.sourceSeed)+str(i)
679	<	else:
680	<	## no random seed
681	<	if str(args)=='': args=args+(str(self.firstRun)+str(i))
682	<	arguments=args.split(',')
565	<	if len(arguments)==3:self.list_of_args.append([str(arguments[0]),str(arguments[1]),str(arguments[2])])
566	<	elif len(arguments)==2:self.list_of_args.append([str(arguments[0]),str(arguments[1])])
567	<	else :self.list_of_args.append([str(arguments[0])])
568	<
569	<	#   print self.list_of_args
669	>	## + vtx random seed
670	>	args.append(str(self.sourceSeedVtx)+str(i))
671	>	if (self.sourceSeedG4):
672	>	## + G4 random seed
673	>	args.append(str(self.sourceSeedG4)+str(i))
674	>	if (self.sourceSeedMix):
675	>	## + Mix random seed
676	>	args.append(str(self.sourceSeedMix)+str(i))
677	>	pass
678	>	pass
679	>	self.list_of_args.append(args)
680	>	pass
681	>
682	>	# print self.list_of_args
683
684		return
685
#	Line 595 | Line 708 | class Cmssw(JobType):
708		return
709
710		def split(self, jobParams):
711	<
711	>
712		common.jobDB.load()
713		#### Fabio
714		njobs = self.total_number_of_jobs
#	Line 603 | Line 716 | class Cmssw(JobType):
716		# create the empty structure
717		for i in range(njobs):
718		jobParams.append("")
719	<
719	>
720		for job in range(njobs):
721		jobParams[job] = arglist[job]
722		# print str(arglist[job])
#	Line 614 | Line 727 | class Cmssw(JobType):
727
728		common.jobDB.save()
729		return
730	<
730	>
731		def getJobTypeArguments(self, nj, sched):
732		result = ''
733		for i in common.jobDB.arguments(nj):
734		result=result+str(i)+" "
735		return result
736	<
736	>
737		def numberOfJobs(self):
738		# Fabio
739		return self.total_number_of_jobs
#	Line 629 | Line 742 | class Cmssw(JobType):
742		"""
743		Return the TarBall with lib and exe
744		"""
745	<
745	>
746		# if it exist, just return it
747		#
748		# Marco. Let's start to use relative path for Boss XML files
#	Line 652 | Line 765 | class Cmssw(JobType):
765		# print "swVersion = ", swVersion
766		swReleaseTop = self.scram.getReleaseTop_()
767		#print "swReleaseTop = ", swReleaseTop
768	<
768	>
769		## check if working area is release top
770		if swReleaseTop == '' or swArea == swReleaseTop:
771		return
#	Line 665 | Line 778 | class Cmssw(JobType):
778		exeWithPath = self.scram.findFile_(executable)
779		if ( not exeWithPath ):
780		raise CrabException('User executable '+executable+' not found')
781	<
781	>
782		## then check if it's private or not
783		if exeWithPath.find(swReleaseTop) == -1:
784		# the exe is private, so we must ship
785		common.logger.debug(5,"Exe "+exeWithPath+" to be tarred")
786		path = swArea+'/'
787	<	exe = string.replace(exeWithPath, path,'')
788	<	tar.add(path+exe,exe)
787	>	# distinguish case when script is in user project area or given by full path somewhere else
788	>	if exeWithPath.find(path) >= 0 :
789	>	exe = string.replace(exeWithPath, path,'')
790	>	tar.add(path+exe,exe)
791	>	else :
792	>	tar.add(exeWithPath,os.path.basename(executable))
793		pass
794		else:
795		# the exe is from release, we'll find it on WN
796		pass
797	<
797	>
798		## Now get the libraries: only those in local working area
799		libDir = 'lib'
800		lib = swArea+'/' +libDir
801		common.logger.debug(5,"lib "+lib+" to be tarred")
802		if os.path.exists(lib):
803		tar.add(lib,libDir)
804	<
804	>
805		## Now check if module dir is present
806		moduleDir = 'module'
807		module = swArea + '/' + moduleDir
#	Line 703 | Line 820 | class Cmssw(JobType):
820		pa = os.environ['CRABDIR'] + '/' + 'ProdAgentApi'
821		if os.path.isdir(pa):
822		tar.add(pa,paDir)
823	<
823	>
824	>	### FEDE FOR DBS PUBLICATION
825	>	## Add PRODCOMMON dir to tar
826	>	prodcommonDir = 'ProdCommon'
827	>	prodcommonPath = os.environ['CRABDIR'] + '/' + 'ProdCommon'
828	>	if os.path.isdir(prodcommonPath):
829	>	tar.add(prodcommonPath,prodcommonDir)
830	>	#############################
831	>
832		common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
833		tar.close()
834		except :
835		raise CrabException('Could not create tar-ball')
836	<
836	>
837	>	## check for tarball size
838	>	tarballinfo = os.stat(self.tgzNameWithPath)
839	>	if ( tarballinfo.st_size > self.MaxTarBallSize*1024*1024 ) :
840	>	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.')
841	>
842		## create tar-ball with ML stuff
843	<	self.MLtgzfile =  common.work_space.pathForTgz()+'share/MLfiles.tgz'
843	>	self.MLtgzfile =  common.work_space.pathForTgz()+'share/MLfiles.tgz'
844		try:
845		tar = tarfile.open(self.MLtgzfile, "w:gz")
846		path=os.environ['CRABDIR'] + '/python/'
#	Line 720 | Line 850 | class Cmssw(JobType):
850		tar.close()
851		except :
852		raise CrabException('Could not create ML files tar-ball')
853	<
853	>
854		return
855	<
855	>
856	>	def additionalInputFileTgz(self):
857	>	"""
858	>	Put all additional files into a tar ball and return its name
859	>	"""
860	>	import tarfile
861	>	tarName=  common.work_space.pathForTgz()+'share/'+self.additional_tgz_name
862	>	tar = tarfile.open(tarName, "w:gz")
863	>	for file in self.additional_inbox_files:
864	>	tar.add(file,string.split(file,'/')[-1])
865	>	common.logger.debug(5,"Files added to "+self.additional_tgz_name+" : "+str(tar.getnames()))
866	>	tar.close()
867	>	return tarName
868	>
869		def wsSetupEnvironment(self, nj):
870		"""
871		Returns part of a job script which prepares
872		the execution environment for the job 'nj'.
873		"""
874		# Prepare JobType-independent part
875	<	txt = ''
876	<
877	<	## OLI_Daniele at this level  middleware already known
735	<
736	<	txt += 'if [ $middleware == LCG ]; then \n'
875	>	txt = ''
876	>	txt += 'echo ">>> setup environment"\n'
877	>	txt += 'if [ $middleware == LCG ]; then \n'
878		txt += self.wsSetupCMSLCGEnvironment_()
879		txt += 'elif [ $middleware == OSG ]; then\n'
880		txt += '    WORKING_DIR=`/bin/mktemp  -d $OSG_WN_TMP/cms_XXXXXXXXXXXX`\n'
881	<	txt += '    echo "Created working directory: $WORKING_DIR"\n'
741	<	txt += '    if [ ! -d $WORKING_DIR ] ;then\n'
881	>	txt += '    if [ ! $? == 0 ] ;then\n'
882		txt += '        echo "SET_CMS_ENV 10016 ==> OSG $WORKING_DIR could not be created on WN `hostname`"\n'
883	<	txt += '        echo "JOB_EXIT_STATUS = 10016"\n'
884	<	txt += '        echo "JobExitCode=10016" | tee -a $RUNTIME_AREA/$repo\n'
885	<	txt += '        dumpStatus $RUNTIME_AREA/$repo\n'
746	<	txt += '        rm -f $RUNTIME_AREA/$repo \n'
747	<	txt += '        echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
748	<	txt += '        echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
883	>	txt += '        echo "JOB_EXIT_STATUS = 10016"\n'
884	>	txt += '        echo "JobExitCode=10016" | tee -a $RUNTIME_AREA/$repo\n'
885	>	txt += '        dumpStatus $RUNTIME_AREA/$repo\n'
886		txt += '        exit 1\n'
887		txt += '    fi\n'
888	+	txt += '    echo ">>> Created working directory: $WORKING_DIR"\n'
889		txt += '\n'
890		txt += '    echo "Change to working directory: $WORKING_DIR"\n'
891		txt += '    cd $WORKING_DIR\n'
892	<	txt += self.wsSetupCMSOSGEnvironment_()
892	>	txt += '    echo ">>> current directory (WORKING_DIR): $WORKING_DIR"\n'
893	>	txt += self.wsSetupCMSOSGEnvironment_()
894	>	#txt += '    echo "### Set SCRAM ARCH to ' + self.executable_arch + ' ###"\n'
895	>	#txt += '    export SCRAM_ARCH='+self.executable_arch+'\n'
896		txt += 'fi\n'
897
898		# Prepare JobType-specific part
899		scram = self.scram.commandName()
900		txt += '\n\n'
901	<	txt += 'echo "### SPECIFIC JOB SETUP ENVIRONMENT ###"\n'
901	>	txt += 'echo ">>> specific cmssw setup environment:"\n'
902	>	txt += 'echo "CMSSW_VERSION =  '+self.version+'"\n'
903		txt += scram+' project CMSSW '+self.version+'\n'
904		txt += 'status=$?\n'
905		txt += 'if [ $status != 0 ] ; then\n'
906	<	txt += '   echo "SET_EXE_ENV 10034 ==>ERROR CMSSW '+self.version+' not found on `hostname`" \n'
907	<	txt += '   echo "JOB_EXIT_STATUS = 10034"\n'
908	<	txt += '   echo "JobExitCode=10034" | tee -a $RUNTIME_AREA/$repo\n'
909	<	txt += '   dumpStatus $RUNTIME_AREA/$repo\n'
768	<	txt += '   rm -f $RUNTIME_AREA/$repo \n'
769	<	txt += '   echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
770	<	txt += '   echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
771	<	## OLI_Daniele
906	>	txt += '    echo "SET_EXE_ENV 10034 ==>ERROR CMSSW '+self.version+' not found on `hostname`" \n'
907	>	txt += '    echo "JOB_EXIT_STATUS = 10034"\n'
908	>	txt += '    echo "JobExitCode=10034" | tee -a $RUNTIME_AREA/$repo\n'
909	>	txt += '    dumpStatus $RUNTIME_AREA/$repo\n'
910		txt += '    if [ $middleware == OSG ]; then \n'
773	–	txt += '        echo "Remove working directory: $WORKING_DIR"\n'
911		txt += '        cd $RUNTIME_AREA\n'
912	+	txt += '        echo ">>> current directory (RUNTIME_AREA): $RUNTIME_AREA"\n'
913	+	txt += '        echo ">>> Remove working directory: $WORKING_DIR"\n'
914		txt += '        /bin/rm -rf $WORKING_DIR\n'
915		txt += '        if [ -d $WORKING_DIR ] ;then\n'
916	<	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'
917	<	txt += '            echo "JOB_EXIT_STATUS = 10018"\n'
918	<	txt += '            echo "JobExitCode=10018" | tee -a $RUNTIME_AREA/$repo\n'
919	<	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
781	<	txt += '            rm -f $RUNTIME_AREA/$repo \n'
782	<	txt += '            echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
783	<	txt += '            echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
916	>	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'
917	>	txt += '            echo "JOB_EXIT_STATUS = 10018"\n'
918	>	txt += '            echo "JobExitCode=10018" | tee -a $RUNTIME_AREA/$repo\n'
919	>	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
920		txt += '        fi\n'
921		txt += '    fi \n'
922	<	txt += '   exit 1 \n'
922	>	txt += '    exit 1 \n'
923		txt += 'fi \n'
788	–	txt += 'echo "CMSSW_VERSION =  '+self.version+'"\n'
924		txt += 'cd '+self.version+'\n'
925	+	########## FEDE FOR DBS2 ######################
926	+	txt += 'SOFTWARE_DIR=`pwd`\n'
927	+	txt += 'echo ">>> current directory (SOFTWARE_DIR): $SOFTWARE_DIR" \n'
928	+	###############################################
929		### needed grep for bug in scramv1 ###
930		txt += scram+' runtime -sh\n'
931		txt += 'eval `'+scram+' runtime -sh | grep -v SCRAMRT_LSB_JOBNAME`\n'
#	Line 796 | Line 935 | class Cmssw(JobType):
935		txt += "\n"
936		txt += "## number of arguments (first argument always jobnumber)\n"
937		txt += "\n"
799	–	#        txt += "narg=$#\n"
938		txt += "if [ $nargs -lt 2 ]\n"
939		txt += "then\n"
940		txt += "    echo 'SET_EXE_ENV 1 ==> ERROR Too few arguments' +$nargs+ \n"
941		txt += '    echo "JOB_EXIT_STATUS = 50113"\n'
942		txt += '    echo "JobExitCode=50113" | tee -a $RUNTIME_AREA/$repo\n'
943		txt += '    dumpStatus $RUNTIME_AREA/$repo\n'
806	–	txt += '    rm -f $RUNTIME_AREA/$repo \n'
807	–	txt += '    echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
808	–	txt += '    echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
809	–	## OLI_Daniele
944		txt += '    if [ $middleware == OSG ]; then \n'
811	–	txt += '        echo "Remove working directory: $WORKING_DIR"\n'
945		txt += '        cd $RUNTIME_AREA\n'
946	+	txt += '        echo ">>> current directory (RUNTIME_AREA): $RUNTIME_AREA"\n'
947	+	txt += '        echo ">>> Remove working directory: $WORKING_DIR"\n'
948		txt += '        /bin/rm -rf $WORKING_DIR\n'
949		txt += '        if [ -d $WORKING_DIR ] ;then\n'
950	<	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'
951	<	txt += '            echo "JOB_EXIT_STATUS = 50114"\n'
952	<	txt += '            echo "JobExitCode=50114" | tee -a $RUNTIME_AREA/$repo\n'
953	<	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
819	<	txt += '            rm -f $RUNTIME_AREA/$repo \n'
820	<	txt += '            echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
821	<	txt += '            echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
950	>	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'
951	>	txt += '            echo "JOB_EXIT_STATUS = 50114"\n'
952	>	txt += '            echo "JobExitCode=50114" | tee -a $RUNTIME_AREA/$repo\n'
953	>	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
954		txt += '        fi\n'
955		txt += '    fi \n'
956		txt += "    exit 1\n"
#	Line 827 | Line 959 | class Cmssw(JobType):
959
960		# Prepare job-specific part
961		job = common.job_list[nj]
962	+	### FEDE FOR DBS OUTPUT PUBLICATION
963	+	if (self.datasetPath):
964	+	txt += '\n'
965	+	txt += 'DatasetPath='+self.datasetPath+'\n'
966	+
967	+	datasetpath_split = self.datasetPath.split("/")
968	+
969	+	txt += 'PrimaryDataset='+datasetpath_split[1]+'\n'
970	+	txt += 'DataTier='+datasetpath_split[2]+'\n'
971	+	txt += 'ApplicationFamily=cmsRun\n'
972	+
973	+	else:
974	+	txt += 'DatasetPath=MCDataTier\n'
975	+	txt += 'PrimaryDataset=null\n'
976	+	txt += 'DataTier=null\n'
977	+	txt += 'ApplicationFamily=MCDataTier\n'
978		if self.pset != None: #CarlosDaniele
979		pset = os.path.basename(job.configFilename())
980		txt += '\n'
981	+	txt += 'cp  $RUNTIME_AREA/'+pset+' .\n'
982		if (self.datasetPath): # standard job
834	–	#txt += 'InputFiles=$2\n'
983		txt += 'InputFiles=${args[1]}\n'
984		txt += 'MaxEvents=${args[2]}\n'
985		txt += 'SkipEvents=${args[3]}\n'
986		txt += 'echo "Inputfiles:<$InputFiles>"\n'
987	<	txt += 'sed "s#{\'INPUT\'}#$InputFiles#" $RUNTIME_AREA/'+pset+' > pset_tmp_1.cfg\n'
987	>	txt += 'sed "s#\'INPUTFILE\'#$InputFiles#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
988		txt += 'echo "MaxEvents:<$MaxEvents>"\n'
989	<	txt += 'sed "s#INPUTMAXEVENTS#$MaxEvents#" pset_tmp_1.cfg > pset_tmp_2.cfg\n'
989	>	txt += 'sed "s#int32 input = 0#int32 input = $MaxEvents#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
990		txt += 'echo "SkipEvents:<$SkipEvents>"\n'
991	<	txt += 'sed "s#INPUTSKIPEVENTS#$SkipEvents#" pset_tmp_2.cfg > pset.cfg\n'
991	>	txt += 'sed "s#uint32 skipEvents = 0#uint32 skipEvents = $SkipEvents#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
992		else:  # pythia like job
993	<	if (self.sourceSeed):
994	<	txt += 'FirstRun=${args[1]}\n'
993	>	seedIndex=1
994	>	if (self.firstRun):
995	>	txt += 'FirstRun=${args['+str(seedIndex)+']}\n'
996		txt += 'echo "FirstRun: <$FirstRun>"\n'
997	<	txt += 'sed "s#\<INPUTFIRSTRUN\>#$FirstRun#" $RUNTIME_AREA/'+pset+' > tmp_1.cfg\n'
998	<	else:
999	<	txt += '# Copy untouched pset\n'
851	<	txt += 'cp $RUNTIME_AREA/'+pset+' tmp_1.cfg\n'
997	>	txt += 'sed "s#uint32 firstRun = 0#uint32 firstRun = $FirstRun#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
998	>	seedIndex=seedIndex+1
999	>
1000		if (self.sourceSeed):
1001	<	#                    txt += 'Seed=$2\n'
1002	<	txt += 'Seed=${args[2]}\n'
1003	<	txt += 'echo "Seed: <$Seed>"\n'
1004	<	txt += 'sed "s#\<INPUT\>#$Seed#" tmp_1.cfg > tmp_2.cfg\n'
1001	>	txt += 'Seed=${args['+str(seedIndex)+']}\n'
1002	>	txt += 'sed "s#uint32 sourceSeed = 0#uint32 sourceSeed = $Seed#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
1003	>	seedIndex=seedIndex+1
1004	>	## the following seeds are not always present
1005		if (self.sourceSeedVtx):
1006	<	#                        txt += 'VtxSeed=$3\n'
859	<	txt += 'VtxSeed=${args[3]}\n'
1006	>	txt += 'VtxSeed=${args['+str(seedIndex)+']}\n'
1007		txt += 'echo "VtxSeed: <$VtxSeed>"\n'
1008	<	txt += 'sed "s#INPUTVTX#$VtxSeed#" tmp_2.cfg > pset.cfg\n'
1009	<	else:
1010	<	txt += 'mv tmp_2.cfg pset.cfg\n'
1011	<	else:
1012	<	txt += 'mv tmp_1.cfg pset.cfg\n'
1013	<	# txt += '# Copy untouched pset\n'
1014	<	# txt += 'cp $RUNTIME_AREA/'+pset+' pset.cfg\n'
1015	<
1008	>	txt += 'sed "s#uint32 VtxSmeared = 0#uint32 VtxSmeared = $VtxSeed#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
1009	>	seedIndex += 1
1010	>	if (self.sourceSeedG4):
1011	>	txt += 'G4Seed=${args['+str(seedIndex)+']}\n'
1012	>	txt += 'echo "G4Seed: <$G4Seed>"\n'
1013	>	txt += 'sed "s#uint32 g4SimHits = 0#uint32 g4SimHits = $G4Seed#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
1014	>	seedIndex += 1
1015	>	if (self.sourceSeedMix):
1016	>	txt += 'mixSeed=${args['+str(seedIndex)+']}\n'
1017	>	txt += 'echo "MixSeed: <$mixSeed>"\n'
1018	>	txt += 'sed "s#uint32 mix = 0#uint32 mix = $mixSeed#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
1019	>	seedIndex += 1
1020	>	pass
1021	>	pass
1022	>	txt += 'mv -f '+pset+' pset.cfg\n'
1023
1024		if len(self.additional_inbox_files) > 0:
1025	<	for file in self.additional_inbox_files:
1026	<	relFile = file.split("/")[-1]
1027	<	txt += 'if [ -e $RUNTIME_AREA/'+relFile+' ] ; then\n'
1028	<	txt += '   cp $RUNTIME_AREA/'+relFile+' .\n'
875	<	txt += '   chmod +x '+relFile+'\n'
876	<	txt += 'fi\n'
877	<	pass
1025	>	txt += 'if [ -e $RUNTIME_AREA/'+self.additional_tgz_name+' ] ; then\n'
1026	>	txt += '  tar xzvf $RUNTIME_AREA/'+self.additional_tgz_name+'\n'
1027	>	txt += 'fi\n'
1028	>	pass
1029
1030		if self.pset != None: #CarlosDaniele
880	–	txt += 'echo "### END JOB SETUP ENVIRONMENT ###"\n\n'
881	–
1031		txt += '\n'
1032		txt += 'echo "***** cat pset.cfg *********"\n'
1033		txt += 'cat pset.cfg\n'
1034		txt += 'echo "****** end pset.cfg ********"\n'
1035		txt += '\n'
1036	<	# txt += 'echo "***** cat pset1.cfg *********"\n'
1037	<	# txt += 'cat pset1.cfg\n'
1038	<	# txt += 'echo "****** end pset1.cfg ********"\n'
1036	>	### FEDE FOR DBS OUTPUT PUBLICATION
1037	>	txt += 'PSETHASH=`EdmConfigHash < pset.cfg` \n'
1038	>	txt += 'echo "PSETHASH = $PSETHASH" \n'
1039	>	##############
1040	>	txt += '\n'
1041		return txt
1042
1043		def wsBuildExe(self, nj=0):
#	Line 898 | Line 1049 | class Cmssw(JobType):
1049		txt = ""
1050
1051		if os.path.isfile(self.tgzNameWithPath):
1052	<	txt += 'echo "tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+'"\n'
1052	>	txt += 'echo ">>> tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+' :" \n'
1053		txt += 'tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+'\n'
1054		txt += 'untar_status=$? \n'
1055		txt += 'if [ $untar_status -ne 0 ]; then \n'
#	Line 906 | Line 1057 | class Cmssw(JobType):
1057		txt += '   echo "JOB_EXIT_STATUS = $untar_status" \n'
1058		txt += '   echo "JobExitCode=$untar_status" | tee -a $RUNTIME_AREA/$repo\n'
1059		txt += '   if [ $middleware == OSG ]; then \n'
909	–	txt += '       echo "Remove working directory: $WORKING_DIR"\n'
1060		txt += '       cd $RUNTIME_AREA\n'
1061	+	txt += '        echo ">>> current directory (RUNTIME_AREA): $RUNTIME_AREA"\n'
1062	+	txt += '        echo ">>> Remove working directory: $WORKING_DIR"\n'
1063		txt += '       /bin/rm -rf $WORKING_DIR\n'
1064		txt += '       if [ -d $WORKING_DIR ] ;then\n'
1065		txt += '           echo "SET_EXE 50999 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after Untarring .tgz file failed"\n'
1066		txt += '           echo "JOB_EXIT_STATUS = 50999"\n'
1067		txt += '           echo "JobExitCode=50999" | tee -a $RUNTIME_AREA/$repo\n'
1068		txt += '           dumpStatus $RUNTIME_AREA/$repo\n'
917	–	txt += '           rm -f $RUNTIME_AREA/$repo \n'
918	–	txt += '           echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
919	–	txt += '           echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1069		txt += '       fi\n'
1070		txt += '   fi \n'
1071		txt += '   \n'
#	Line 925 | Line 1074 | class Cmssw(JobType):
1074		txt += '   echo "Successful untar" \n'
1075		txt += 'fi \n'
1076		txt += '\n'
1077	<	txt += 'echo "Include ProdAgentApi in PYTHONPATH"\n'
1077	>	txt += 'echo ">>> Include ProdAgentApi and PRODCOMMON in PYTHONPATH:"\n'
1078		txt += 'if [ -z "$PYTHONPATH" ]; then\n'
1079	<	txt += '   export PYTHONPATH=ProdAgentApi\n'
1079	>	#### FEDE FOR DBS OUTPUT PUBLICATION
1080	>	txt += '   export PYTHONPATH=$SOFTWARE_DIR/ProdAgentApi:$SOFTWARE_DIR/ProdCommon\n'
1081		txt += 'else\n'
1082	<	txt += '   export PYTHONPATH=ProdAgentApi:${PYTHONPATH}\n'
1082	>	txt += '   export PYTHONPATH=$SOFTWARE_DIR/ProdAgentApi:$SOFTWARE_DIR/ProdCommon:${PYTHONPATH}\n'
1083	>	txt += 'echo "PYTHONPATH=$PYTHONPATH"\n'
1084	>	###################
1085		txt += 'fi\n'
1086		txt += '\n'
1087
1088		pass
1089	<
1089	>
1090		return txt
1091
1092		def modifySteeringCards(self, nj):
1093		"""
1094	<	modify the card provided by the user,
1094	>	modify the card provided by the user,
1095		writing a new card into share dir
1096		"""
1097	<
1097	>
1098		def executableName(self):
1099	<	if self.pset == None: #CarlosDaniele
1099	>	if self.scriptExe: #CarlosDaniele
1100		return "sh "
1101		else:
1102		return self.executable
1103
1104		def executableArgs(self):
1105	<	if self.pset == None:#CarlosDaniele
1105	>	if self.scriptExe:#CarlosDaniele
1106		return   self.scriptExe + " $NJob"
1107	<	else:
1108	<	return " -p pset.cfg"
1107	>	else:
1108	>	# if >= CMSSW_1_5_X, add -e
1109	>	version_array = self.scram.getSWVersion().split('_')
1110	>	major = 0
1111	>	minor = 0
1112	>	try:
1113	>	major = int(version_array[1])
1114	>	minor = int(version_array[2])
1115	>	except:
1116	>	msg = "Cannot parse CMSSW version string: " + "_".join(version_array) + " for major and minor release number!"
1117	>	raise CrabException(msg)
1118	>	if major >= 1 and minor >= 5 :
1119	>	return " -e -p pset.cfg"
1120	>	else:
1121	>	return " -p pset.cfg"
1122
1123		def inputSandbox(self, nj):
1124		"""
#	Line 968 | Line 1133 | class Cmssw(JobType):
1133		if os.path.isfile(self.MLtgzfile):
1134		inp_box.append(self.MLtgzfile)
1135		## config
1136	<	if not self.pset is None: #CarlosDaniele
1136	>	if not self.pset is None:
1137		inp_box.append(common.work_space.pathForTgz() + 'job/' + self.configFilename())
1138		## additional input files
1139	<	#for file in self.additional_inbox_files:
1140	<	#    inp_box.append(common.work_space.cwdDir()+file)
1139	>	tgz = self.additionalInputFileTgz()
1140	>	inp_box.append(tgz)
1141		return inp_box
1142
1143		def outputSandbox(self, nj):
#	Line 983 | Line 1148 | class Cmssw(JobType):
1148
1149		## User Declared output files
1150		for out in (self.output_file+self.output_file_sandbox):
1151	<	n_out = nj + 1
1151	>	n_out = nj + 1
1152		out_box.append(self.numberFile_(out,str(n_out)))
1153		return out_box
1154
#	Line 999 | Line 1164 | class Cmssw(JobType):
1164		"""
1165
1166		txt = '\n'
1167	<	txt += '# directory content\n'
1167	>	txt += 'echo" >>> directory content:"\n'
1168		txt += 'ls \n'
1169	+	txt = '\n'
1170	+
1171	+	txt += 'output_exit_status=0\n'
1172
1173	<	for fileWithSuffix in (self.output_file+self.output_file_sandbox):
1173	>	for fileWithSuffix in (self.output_file_sandbox):
1174		output_file_num = self.numberFile_(fileWithSuffix, '$NJob')
1175		txt += '\n'
1176		txt += '# check output file\n'
1177	<	txt += 'ls '+fileWithSuffix+'\n'
1178	<	txt += 'ls_result=$?\n'
1179	<	txt += 'if [ $ls_result -ne 0 ] ; then\n'
1180	<	txt += '   echo "ERROR: Problem with output file"\n'
1177	>	txt += 'if [ -e ./'+fileWithSuffix+' ] ; then\n'
1178	>	txt += '    mv '+fileWithSuffix+' $RUNTIME_AREA\n'
1179	>	txt += '    cp $RUNTIME_AREA/'+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n'
1180	>	txt += 'else\n'
1181	>	txt += '    exit_status=60302\n'
1182	>	txt += '    echo "ERROR: Problem with output file '+fileWithSuffix+'"\n'
1183		if common.scheduler.boss_scheduler_name == 'condor_g':
1184		txt += '    if [ $middleware == OSG ]; then \n'
1185		txt += '        echo "prepare dummy output file"\n'
1186		txt += '        echo "Processing of job output failed" > $RUNTIME_AREA/'+output_file_num+'\n'
1187		txt += '    fi \n'
1018	–	txt += 'else\n'
1019	–	txt += '   cp '+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n'
1188		txt += 'fi\n'
1021	–
1022	–	txt += 'cd $RUNTIME_AREA\n'
1023	–	txt += 'cd $RUNTIME_AREA\n'
1024	–	### OLI_DANIELE
1025	–	txt += 'if [ $middleware == OSG ]; then\n'
1026	–	txt += '    cd $RUNTIME_AREA\n'
1027	–	txt += '    echo "Remove working directory: $WORKING_DIR"\n'
1028	–	txt += '    /bin/rm -rf $WORKING_DIR\n'
1029	–	txt += '    if [ -d $WORKING_DIR ] ;then\n'
1030	–	txt += '        echo "SET_EXE 60999 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after cleanup of WN"\n'
1031	–	txt += '        echo "JOB_EXIT_STATUS = 60999"\n'
1032	–	txt += '        echo "JobExitCode=60999" | tee -a $RUNTIME_AREA/$repo\n'
1033	–	txt += '        dumpStatus $RUNTIME_AREA/$repo\n'
1034	–	txt += '        rm -f $RUNTIME_AREA/$repo \n'
1035	–	txt += '        echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1036	–	txt += '        echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1037	–	txt += '    fi\n'
1038	–	txt += 'fi\n'
1039	–	txt += '\n'
1189
1190	<	file_list = ''
1042	<	## Add to filelist only files to be possibly copied to SE
1043	<	for fileWithSuffix in self.output_file:
1190	>	for fileWithSuffix in (self.output_file):
1191		output_file_num = self.numberFile_(fileWithSuffix, '$NJob')
1192	<	file_list=file_list+output_file_num+' '
1193	<	file_list=file_list[:-1]
1194	<	txt += 'file_list="'+file_list+'"\n'
1192	>	txt += '\n'
1193	>	txt += '# check output file\n'
1194	>	txt += 'if [ -e ./'+fileWithSuffix+' ] ; then\n'
1195	>	txt += '    mv '+fileWithSuffix+' $RUNTIME_AREA\n'
1196	>	txt += '    cp $RUNTIME_AREA/'+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n'
1197	>	txt += 'else\n'
1198	>	txt += '    exit_status=60302\n'
1199	>	txt += '    echo "ERROR: Problem with output file '+fileWithSuffix+'"\n'
1200	>	txt += '    echo "JOB_EXIT_STATUS = $exit_status"\n'
1201	>	txt += '    output_exit_status=$exit_status\n'
1202	>	if common.scheduler.boss_scheduler_name == 'condor_g':
1203	>	txt += '    if [ $middleware == OSG ]; then \n'
1204	>	txt += '        echo "prepare dummy output file"\n'
1205	>	txt += '        echo "Processing of job output failed" > $RUNTIME_AREA/'+output_file_num+'\n'
1206	>	txt += '    fi \n'
1207	>	txt += 'fi\n'
1208	>	file_list = []
1209	>	for fileWithSuffix in (self.output_file):
1210	>	file_list.append(self.numberFile_(fileWithSuffix, '$NJob'))
1211
1212	+	txt += 'file_list="'+string.join(file_list,' ')+'"\n'
1213	+	txt += 'cd $RUNTIME_AREA\n'
1214	+	txt += 'echo ">>> current directory (RUNTIME_AREA):  $RUNTIME_AREA"\n'
1215		return txt
1216
1217		def numberFile_(self, file, txt):
#	Line 1056 | Line 1222 | class Cmssw(JobType):
1222		# take away last extension
1223		name = p[0]
1224		for x in p[1:-1]:
1225	<	name=name+"."+x
1225	>	name=name+"."+x
1226		# add "_txt"
1227		if len(p)>1:
1228	<	ext = p[len(p)-1]
1229	<	result = name + '_' + txt + "." + ext
1228	>	ext = p[len(p)-1]
1229	>	result = name + '_' + txt + "." + ext
1230		else:
1231	<	result = name + '_' + txt
1232	<
1231	>	result = name + '_' + txt
1232	>
1233		return result
1234
1235		def getRequirements(self, nj=[]):
1236		"""
1237	<	return job requirements to add to jdl files
1237	>	return job requirements to add to jdl files
1238		"""
1239		req = ''
1240		if self.version:
1241		req='Member("VO-cms-' + \
1242		self.version + \
1243		'", other.GlueHostApplicationSoftwareRunTimeEnvironment)'
1244	+	## SL add requirement for OS version only if SL4
1245	+	#reSL4 = re.compile( r'slc4' )
1246	+	if self.executable_arch: # and reSL4.search(self.executable_arch):
1247	+	req+=' && Member("VO-cms-' + \
1248	+	self.executable_arch + \
1249	+	'", other.GlueHostApplicationSoftwareRunTimeEnvironment)'
1250
1251		req = req + ' && (other.GlueHostNetworkAdapterOutboundIP)'
1252
#	Line 1084 | Line 1256 | class Cmssw(JobType):
1256		""" return the config filename """
1257		return self.name()+'.cfg'
1258
1087	–	### OLI_DANIELE
1259		def wsSetupCMSOSGEnvironment_(self):
1260		"""
1261		Returns part of a job script which is prepares
1262		the execution environment and which is common for all CMS jobs.
1263		"""
1264	<	txt = '\n'
1265	<	txt += '   echo "### SETUP CMS OSG  ENVIRONMENT ###"\n'
1266	<	txt += '   if [ -f $GRID3_APP_DIR/cmssoft/cmsset_default.sh ] ;then\n'
1267	<	txt += '      # Use $GRID3_APP_DIR/cmssoft/cmsset_default.sh to setup cms software\n'
1268	<	txt += '       source $GRID3_APP_DIR/cmssoft/cmsset_default.sh '+self.version+'\n'
1098	<	txt += '   elif [ -f $OSG_APP/cmssoft/cms/cmsset_default.sh ] ;then\n'
1264	>	txt = '    echo ">>> setup CMS OSG environment:"\n'
1265	>	txt += '    echo "set SCRAM ARCH to ' + self.executable_arch + '"\n'
1266	>	txt += '    export SCRAM_ARCH='+self.executable_arch+'\n'
1267	>	txt += '    echo "SCRAM_ARCH = $SCRAM_ARCH"\n'
1268	>	txt += '    if [ -f $OSG_APP/cmssoft/cms/cmsset_default.sh ] ;then\n'
1269		txt += '      # Use $OSG_APP/cmssoft/cms/cmsset_default.sh to setup cms software\n'
1270	<	txt += '       source $OSG_APP/cmssoft/cms/cmsset_default.sh '+self.version+'\n'
1271	<	txt += '   else\n'
1272	<	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'
1273	<	txt += '       echo "JOB_EXIT_STATUS = 10020"\n'
1274	<	txt += '       echo "JobExitCode=10020" | tee -a $RUNTIME_AREA/$repo\n'
1275	<	txt += '       dumpStatus $RUNTIME_AREA/$repo\n'
1106	<	txt += '       rm -f $RUNTIME_AREA/$repo \n'
1107	<	txt += '       echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1108	<	txt += '       echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1109	<	txt += '       exit 1\n'
1270	>	txt += '        source $OSG_APP/cmssoft/cms/cmsset_default.sh '+self.version+'\n'
1271	>	txt += '    else\n'
1272	>	txt += '        echo "SET_CMS_ENV 10020 ==> ERROR $OSG_APP/cmssoft/cms/cmsset_default.sh file not found"\n'
1273	>	txt += '        echo "JOB_EXIT_STATUS = 10020"\n'
1274	>	txt += '        echo "JobExitCode=10020" | tee -a $RUNTIME_AREA/$repo\n'
1275	>	txt += '        dumpStatus $RUNTIME_AREA/$repo\n'
1276		txt += '\n'
1277	<	txt += '       echo "Remove working directory: $WORKING_DIR"\n'
1278	<	txt += '       cd $RUNTIME_AREA\n'
1279	<	txt += '       /bin/rm -rf $WORKING_DIR\n'
1280	<	txt += '       if [ -d $WORKING_DIR ] ;then\n'
1281	<	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'
1282	<	txt += '            echo "JOB_EXIT_STATUS = 10017"\n'
1283	<	txt += '            echo "JobExitCode=10017" | tee -a $RUNTIME_AREA/$repo\n'
1284	<	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
1285	<	txt += '            rm -f $RUNTIME_AREA/$repo \n'
1286	<	txt += '            echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1121	<	txt += '            echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1122	<	txt += '       fi\n'
1277	>	txt += '        cd $RUNTIME_AREA\n'
1278	>	txt += '        echo ">>> current directory (RUNTIME_AREA): $RUNTIME_AREA"\n'
1279	>	txt += '        echo ">>> Remove working directory: $WORKING_DIR"\n'
1280	>	txt += '        /bin/rm -rf $WORKING_DIR\n'
1281	>	txt += '        if [ -d $WORKING_DIR ] ;then\n'
1282	>	txt += '            echo "SET_CMS_ENV 10017 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after $OSG_APP/cmssoft/cms/cmsset_default.sh file not found"\n'
1283	>	txt += '            echo "JOB_EXIT_STATUS = 10017"\n'
1284	>	txt += '            echo "JobExitCode=10017" | tee -a $RUNTIME_AREA/$repo\n'
1285	>	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
1286	>	txt += '        fi\n'
1287		txt += '\n'
1288	<	txt += '       exit 1\n'
1289	<	txt += '   fi\n'
1288	>	txt += '        exit 1\n'
1289	>	txt += '    fi\n'
1290		txt += '\n'
1291	<	txt += '   echo "SET_CMS_ENV 0 ==> setup cms environment ok"\n'
1292	<	txt += '   echo " END SETUP CMS OSG  ENVIRONMENT "\n'
1291	>	txt += '    echo "SET_CMS_ENV 0 ==> setup cms environment ok"\n'
1292	>	txt += '    echo "SCRAM_ARCH = $SCRAM_ARCH"\n'
1293
1294		return txt
1295	<
1295	>
1296		### OLI_DANIELE
1297		def wsSetupCMSLCGEnvironment_(self):
1298		"""
1299		Returns part of a job script which is prepares
1300		the execution environment and which is common for all CMS jobs.
1301		"""
1302	<	txt  = '   \n'
1303	<	txt += '   echo " ### SETUP CMS LCG  ENVIRONMENT ### "\n'
1304	<	txt += '   if [ ! $VO_CMS_SW_DIR ] ;then\n'
1305	<	txt += '       echo "SET_CMS_ENV 10031 ==> ERROR CMS software dir not found on WN `hostname`"\n'
1306	<	txt += '       echo "JOB_EXIT_STATUS = 10031" \n'
1307	<	txt += '       echo "JobExitCode=10031" | tee -a $RUNTIME_AREA/$repo\n'
1308	<	txt += '       dumpStatus $RUNTIME_AREA/$repo\n'
1309	<	txt += '       rm -f $RUNTIME_AREA/$repo \n'
1310	<	txt += '       echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1311	<	txt += '       echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1312	<	txt += '       exit 1\n'
1313	<	txt += '   else\n'
1314	<	txt += '       echo "Sourcing environment... "\n'
1315	<	txt += '       if [ ! -s $VO_CMS_SW_DIR/cmsset_default.sh ] ;then\n'
1316	<	txt += '           echo "SET_CMS_ENV 10020 ==> ERROR cmsset_default.sh file not found into dir $VO_CMS_SW_DIR"\n'
1317	<	txt += '           echo "JOB_EXIT_STATUS = 10020"\n'
1318	<	txt += '           echo "JobExitCode=10020" | tee -a $RUNTIME_AREA/$repo\n'
1319	<	txt += '           dumpStatus $RUNTIME_AREA/$repo\n'
1320	<	txt += '           rm -f $RUNTIME_AREA/$repo \n'
1321	<	txt += '           echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1322	<	txt += '           echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1323	<	txt += '           exit 1\n'
1324	<	txt += '       fi\n'
1325	<	txt += '       echo "sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
1326	<	txt += '       source $VO_CMS_SW_DIR/cmsset_default.sh\n'
1327	<	txt += '       result=$?\n'
1328	<	txt += '       if [ $result -ne 0 ]; then\n'
1329	<	txt += '           echo "SET_CMS_ENV 10032 ==> ERROR problem sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
1330	<	txt += '           echo "JOB_EXIT_STATUS = 10032"\n'
1331	<	txt += '           echo "JobExitCode=10032" | tee -a $RUNTIME_AREA/$repo\n'
1332	<	txt += '           dumpStatus $RUNTIME_AREA/$repo\n'
1333	<	txt += '           rm -f $RUNTIME_AREA/$repo \n'
1334	<	txt += '           echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1335	<	txt += '           echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1336	<	txt += '           exit 1\n'
1337	<	txt += '       fi\n'
1338	<	txt += '   fi\n'
1339	<	txt += '   \n'
1340	<	txt += '   string=`cat /etc/redhat-release`\n'
1341	<	txt += '   echo $string\n'
1342	<	txt += '   if [[ $string = *alhalla* ]]; then\n'
1343	<	txt += '       echo "SCRAM_ARCH= $SCRAM_ARCH"\n'
1344	<	txt += '   elif [[ $string = *Enterprise* ]] || [[ $string = *cientific* ]]; then\n'
1345	<	txt += '       export SCRAM_ARCH=slc3_ia32_gcc323\n'
1346	<	txt += '       echo "SCRAM_ARCH= $SCRAM_ARCH"\n'
1347	<	txt += '   else\n'
1348	<	txt += '       echo "SET_CMS_ENV 10033 ==> ERROR OS unknown, LCG environment not initialized"\n'
1349	<	txt += '       echo "JOB_EXIT_STATUS = 10033"\n'
1350	<	txt += '       echo "JobExitCode=10033" | tee -a $RUNTIME_AREA/$repo\n'
1351	<	txt += '       dumpStatus $RUNTIME_AREA/$repo\n'
1352	<	txt += '       rm -f $RUNTIME_AREA/$repo \n'
1353	<	txt += '       echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1354	<	txt += '       echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1355	<	txt += '       exit 1\n'
1356	<	txt += '   fi\n'
1357	<	txt += '   echo "SET_CMS_ENV 0 ==> setup cms environment ok"\n'
1358	<	txt += '   echo "### END SETUP CMS LCG ENVIRONMENT ###"\n'
1302	>	txt = '    echo ">>> setup CMS LCG environment:"\n'
1303	>	txt += '    echo "set SCRAM ARCH and BUILD_ARCH to ' + self.executable_arch + ' ###"\n'
1304	>	txt += '    export SCRAM_ARCH='+self.executable_arch+'\n'
1305	>	txt += '    export BUILD_ARCH='+self.executable_arch+'\n'
1306	>	txt += '    if [ ! $VO_CMS_SW_DIR ] ;then\n'
1307	>	txt += '        echo "SET_CMS_ENV 10031 ==> ERROR CMS software dir not found on WN `hostname`"\n'
1308	>	txt += '        echo "JOB_EXIT_STATUS = 10031" \n'
1309	>	txt += '        echo "JobExitCode=10031" | tee -a $RUNTIME_AREA/$repo\n'
1310	>	txt += '        dumpStatus $RUNTIME_AREA/$repo\n'
1311	>	txt += '        exit 1\n'
1312	>	txt += '    else\n'
1313	>	txt += '        echo "Sourcing environment... "\n'
1314	>	txt += '        if [ ! -s $VO_CMS_SW_DIR/cmsset_default.sh ] ;then\n'
1315	>	txt += '            echo "SET_CMS_ENV 10020 ==> ERROR cmsset_default.sh file not found into dir $VO_CMS_SW_DIR"\n'
1316	>	txt += '            echo "JOB_EXIT_STATUS = 10020"\n'
1317	>	txt += '            echo "JobExitCode=10020" | tee -a $RUNTIME_AREA/$repo\n'
1318	>	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
1319	>	txt += '            exit 1\n'
1320	>	txt += '        fi\n'
1321	>	txt += '        echo "sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
1322	>	txt += '        source $VO_CMS_SW_DIR/cmsset_default.sh\n'
1323	>	txt += '        result=$?\n'
1324	>	txt += '        if [ $result -ne 0 ]; then\n'
1325	>	txt += '            echo "SET_CMS_ENV 10032 ==> ERROR problem sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
1326	>	txt += '            echo "JOB_EXIT_STATUS = 10032"\n'
1327	>	txt += '            echo "JobExitCode=10032" | tee -a $RUNTIME_AREA/$repo\n'
1328	>	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
1329	>	txt += '            exit 1\n'
1330	>	txt += '        fi\n'
1331	>	txt += '    fi\n'
1332	>	txt += '    \n'
1333	>	txt += '    echo "SET_CMS_ENV 0 ==> setup cms environment ok"\n'
1334	>	return txt
1335	>
1336	>	### FEDE FOR DBS OUTPUT PUBLICATION
1337	>	def modifyReport(self, nj):
1338	>	"""
1339	>	insert the part of the script that modifies the FrameworkJob Report
1340	>	"""
1341	>
1342	>	txt = ''
1343	>	try:
1344	>	publish_data = int(self.cfg_params['USER.publish_data'])
1345	>	except KeyError:
1346	>	publish_data = 0
1347	>	if (publish_data == 1):
1348	>	txt += 'echo ">>> Modify Job Report:" \n'
1349	>	################ FEDE FOR DBS2 #############################################
1350	>	txt += 'chmod a+x $SOFTWARE_DIR/ProdAgentApi/FwkJobRep/ModifyJobReport.py\n'
1351	>	#############################################################################
1352	>
1353	>	txt += 'if [ -z "$SE" ]; then\n'
1354	>	txt += '    SE="" \n'
1355	>	txt += 'fi \n'
1356	>	txt += 'if [ -z "$SE_PATH" ]; then\n'
1357	>	txt += '    SE_PATH="" \n'
1358	>	txt += 'fi \n'
1359	>	txt += 'echo "SE = $SE"\n'
1360	>	txt += 'echo "SE_PATH = $SE_PATH"\n'
1361	>
1362	>	processedDataset = self.cfg_params['USER.publish_data_name']
1363	>	txt += 'ProcessedDataset='+processedDataset+'\n'
1364	>	#### LFN=/store/user/<user>/processedDataset_PSETHASH
1365	>	txt += 'if [ "$SE_PATH" == "" ]; then\n'
1366	>	#### FEDE: added slash in LFN ##############
1367	>	txt += '    FOR_LFN=/copy_problems/ \n'
1368	>	txt += 'else \n'
1369	>	txt += '    tmp=`echo $SE_PATH | awk -F \'store\' \'{print$2}\'` \n'
1370	>	#####  FEDE TO BE CHANGED, BECAUSE STORE IS HARDCODED!!!! ########
1371	>	txt += '    FOR_LFN=/store$tmp \n'
1372	>	txt += 'fi \n'
1373	>	txt += 'echo "ProcessedDataset = $ProcessedDataset"\n'
1374	>	txt += 'echo "FOR_LFN = $FOR_LFN" \n'
1375	>	txt += 'echo "CMSSW_VERSION = $CMSSW_VERSION"\n\n'
1376	>	#txt += 'echo "$RUNTIME_AREA/'+self.version+'/ProdAgentApi/FwkJobRep/ModifyJobReport.py crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH"\n'
1377	>	txt += 'echo "$SOFTWARE_DIR/ProdAgentApi/FwkJobRep/ModifyJobReport.py crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH"\n'
1378	>	txt += '$SOFTWARE_DIR/ProdAgentApi/FwkJobRep/ModifyJobReport.py crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH\n'
1379	>	#txt += '$RUNTIME_AREA/'+self.version+'/ProdAgentApi/FwkJobRep/ModifyJobReport.py crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH\n'
1380	>
1381	>	txt += 'modifyReport_result=$?\n'
1382	>	txt += 'echo modifyReport_result = $modifyReport_result\n'
1383	>	txt += 'if [ $modifyReport_result -ne 0 ]; then\n'
1384	>	txt += '    exit_status=1\n'
1385	>	txt += '    echo "ERROR: Problem with ModifyJobReport"\n'
1386	>	txt += 'else\n'
1387	>	txt += '    mv NewFrameworkJobReport.xml crab_fjr_$NJob.xml\n'
1388	>	txt += 'fi\n'
1389	>	else:
1390	>	txt += 'echo "no data publication required"\n'
1391	>	return txt
1392	>
1393	>	def cleanEnv(self):
1394	>	txt = ''
1395	>	txt += 'if [ $middleware == OSG ]; then\n'
1396	>	txt += '    cd $RUNTIME_AREA\n'
1397	>	txt += '    echo ">>> current directory (RUNTIME_AREA): $RUNTIME_AREA"\n'
1398	>	txt += '    echo ">>> Remove working directory: $WORKING_DIR"\n'
1399	>	txt += '    /bin/rm -rf $WORKING_DIR\n'
1400	>	txt += '    if [ -d $WORKING_DIR ] ;then\n'
1401	>	txt += '        echo "SET_EXE 60999 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after cleanup of WN"\n'
1402	>	txt += '        echo "JOB_EXIT_STATUS = 60999"\n'
1403	>	txt += '        echo "JobExitCode=60999" | tee -a $RUNTIME_AREA/$repo\n'
1404	>	txt += '        dumpStatus $RUNTIME_AREA/$repo\n'
1405	>	txt += '    fi\n'
1406	>	txt += 'fi\n'
1407	>	txt += '\n'
1408		return txt
1409
1410		def setParam_(self, param, value):
#	Line 1202 | Line 1415 | class Cmssw(JobType):
1415
1416		def setTaskid_(self):
1417		self._taskId = self.cfg_params['taskId']
1418	<
1418	>
1419		def getTaskid(self):
1420		return self._taskId
1421
1209	–	#######################################################################
1422		def uniquelist(self, old):
1423		"""
1424		remove duplicates from a list
#	Line 1215 | Line 1427 | class Cmssw(JobType):
1427		for e in old:
1428		nd[e]=0
1429		return nd.keys()
1430	+
1431	+
1432	+	def checkOut(self, limit):
1433	+	"""
1434	+	check the dimension of the output files
1435	+	"""
1436	+	txt += 'echo ">>> Starting output sandbox limit check :"\n'
1437	+	allOutFiles = ""
1438	+	listOutFiles = []
1439	+	for fileOut in (self.output_file+self.output_file_sandbox):
1440	+	if fileOut.find('crab_fjr') == -1:
1441	+	allOutFiles = allOutFiles + " " + self.numberFile_(fileOut, '$NJob')
1442	+	listOutFiles.append(self.numberFile_(fileOut, '$NJob'))
1443	+	txt += 'echo "OUTPUT files: '+str(allOutFiles)+'";\n'
1444	+	txt += 'ls -gGhrta;\n'
1445	+	txt += 'sum=0;\n'
1446	+	txt += 'for file in '+str(allOutFiles)+' ; do\n'
1447	+	txt += '    if [ -e $file ]; then\n'
1448	+	txt += '        tt=`ls -gGrta $file | awk \'{ print $3 }\'`\n'
1449	+	txt += '        sum=`expr $sum + $tt`\n'
1450	+	txt += '    else\n'
1451	+	txt += '        echo "WARNING: output file $file not found!"\n'
1452	+	txt += '    fi\n'
1453	+	txt += 'done\n'
1454	+	txt += 'echo "Total Output dimension: $sum";\n'
1455	+	txt += 'limit='+str(limit)+';\n'
1456	+	txt += 'echo "OUTPUT FILES LIMIT SET TO: $limit";\n'
1457	+	txt += 'if [ $limit -lt $sum ]; then\n'
1458	+	txt += '    echo "WARNING: output files have to big size - something will be lost;"\n'
1459	+	txt += '    echo "         checking the output file sizes..."\n'
1460	+	"""
1461	+	txt += '    dim=0;\n'
1462	+	txt += '    exclude=0;\n'
1463	+	txt += '    for files in '+str(allOutFiles)+' ; do\n'
1464	+	txt += '        sumTemp=0;\n'
1465	+	txt += '        for file2 in '+str(allOutFiles)+' ; do\n'
1466	+	txt += '            if [ $file != $file2 ]; then\n'
1467	+	txt += '                tt=`ls -gGrta $file2 | awk \'{ print $3 }\';`\n'
1468	+	txt += '                sumTemp=`expr $sumTemp + $tt`;\n'
1469	+	txt += '            fi\n'
1470	+	txt += '        done\n'
1471	+	txt += '        if [ $sumTemp -lt $limit ]; then\n'
1472	+	txt += '            if [ $dim -lt $sumTemp ]; then\n'
1473	+	txt += '                dim=$sumTemp;\n'
1474	+	txt += '                exclude=$file;\n'
1475	+	txt += '            fi\n'
1476	+	txt += '        fi\n'
1477	+	txt += '    done\n'
1478	+	txt += '    echo "Dimension calculated: $dim"; echo "File to exclude: $exclude";\n'
1479	+	"""
1480	+	txt += '    tot=0;\n'
1481	+	txt += '    for file2 in '+str(allOutFiles)+' ; do\n'
1482	+	txt += '        tt=`ls -gGrta $file2 | awk \'{ print $3 }\';`\n'
1483	+	txt += '        tot=`expr $tot + $tt`;\n'
1484	+	txt += '        if [ $limit -lt $tot ]; then\n'
1485	+	txt += '            tot=`expr $tot - $tt`;\n'
1486	+	txt += '            fileLast=$file;\n'
1487	+	txt += '            break;\n'
1488	+	txt += '        fi\n'
1489	+	txt += '    done\n'
1490	+	txt += '    echo "Dimension calculated: $tot"; echo "First file to exclude: $file";\n'
1491	+	txt += '    flag=0;\n'
1492	+	txt += '    for filess in '+str(allOutFiles)+' ; do\n'
1493	+	txt += '        if [ $fileLast = $filess ]; then\n'
1494	+	txt += '            flag=1;\n'
1495	+	txt += '        fi\n'
1496	+	txt += '        if [ $flag -eq 1 ]; then\n'
1497	+	txt += '            rm -f $filess;\n'
1498	+	txt += '        fi\n'
1499	+	txt += '    done\n'
1500	+	txt += '    ls -agGhrt;\n'
1501	+	txt += '    echo "WARNING: output files are too big in dimension: can not put in the output_sandbox.";\n'
1502	+	txt += '    echo "JOB_EXIT_STATUS = 70000";\n'
1503	+	txt += '    exit_status=70000;\n'
1504	+	txt += 'else'
1505	+	txt += '    echo "Total Output dimension $sum is fine.";\n'
1506	+	txt += 'fi\n'
1507	+	txt += 'echo "Ending output sandbox limit check"\n'
1508	+	return txt

Diff Legend

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