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 1.55 by slacapra, Wed Nov 22 14:57:26 2006 UTC vs.
Revision 1.139 by ewv, Tue Nov 20 17:06:50 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	<	import math
5	>	from BlackWhiteListParser import BlackWhiteListParser
6		import common
7	–	import PsetManipulator
8	–
9	–	import DBSInfo
10	–	import DataDiscovery
11	–	import DataLocation
7		import Scram
8
9	<	import glob, os, string, re
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
21	–	# 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)
31	–	scramArea = ''
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.pset = ''      #scrip use case Da
38	>	self.additional_tgz_name = 'additional.tgz'
39	>	self.scriptName = 'CMSSW.sh'
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		try:
67		tmp =  cfg_params['CMSSW.datasetpath']
68		log.debug(6, "CMSSW::CMSSW(): datasetPath = "+tmp)
#	Line 54 | Line 73 | class Cmssw(JobType):
73		self.datasetPath = tmp
74		self.selectNoInput = 0
75		except KeyError:
76	<	msg = "Error: datasetpath not defined "
76	>	msg = "Error: datasetpath not defined "
77		raise CrabException(msg)
78
79		# ML monitoring
#	Line 63 | Line 82 | class Cmssw(JobType):
82		self.setParam_('dataset', 'None')
83		self.setParam_('owner', 'None')
84		else:
85	<	datasetpath_split = self.datasetPath.split("/")
86	<	self.setParam_('dataset', datasetpath_split[1])
87	<	self.setParam_('owner', datasetpath_split[-1])
85	>	try:
86	>	datasetpath_split = self.datasetPath.split("/")
87	>	# standard style
88	>	self.setParam_('datasetFull', self.datasetPath)
89	>	self.setParam_('dataset', datasetpath_split[1])
90	>	self.setParam_('owner', datasetpath_split[2])
91	>	except:
92	>	self.setParam_('dataset', self.datasetPath)
93	>	self.setParam_('owner', self.datasetPath)
94
95		self.setTaskid_()
96		self.setParam_('taskId', self.cfg_params['taskId'])
#	Line 89 | Line 114 | class Cmssw(JobType):
114		try:
115		self.pset = cfg_params['CMSSW.pset']
116		log.debug(6, "Cmssw::Cmssw(): PSet file = "+self.pset)
117	<	if self.pset.lower() != 'none' :
117	>	if self.pset.lower() != 'none' :
118		if (not os.path.exists(self.pset)):
119		raise CrabException("User defined PSet file "+self.pset+" does not exist")
120		else:
#	Line 107 | Line 132 | class Cmssw(JobType):
132		# other output files to be returned via sandbox or copied to SE
133		try:
134		self.output_file = []
110	–
111	–
135		tmp = cfg_params['CMSSW.output_file']
136		if tmp != '':
137		tmpOutFiles = string.split(cfg_params['CMSSW.output_file'],',')
#	Line 118 | Line 141 | class Cmssw(JobType):
141		self.output_file.append(tmp)
142		pass
143		else:
144	<	log.message("No output file defined: only stdout/err and the CRAB Framework Job Report will be available")
144	>	log.message("No output file defined: only stdout/err and the CRAB Framework Job Report will be available\n")
145		pass
146		pass
147		except KeyError:
148	<	log.message("No output file defined: only stdout/err and the CRAB Framework Job Report will be available")
148	>	log.message("No output file defined: only stdout/err and the CRAB Framework Job Report will be available\n")
149		pass
150
151		# script_exe file as additional file in inputSandbox
#	Line 130 | Line 153 | class Cmssw(JobType):
153		self.scriptExe = cfg_params['USER.script_exe']
154		if self.scriptExe != '':
155		if not os.path.isfile(self.scriptExe):
156	<	msg ="WARNING. file "+self.scriptExe+" not found"
156	>	msg ="ERROR. file "+self.scriptExe+" not found"
157		raise CrabException(msg)
158		self.additional_inbox_files.append(string.strip(self.scriptExe))
159		except KeyError:
160		self.scriptExe = ''
161	+
162		#CarlosDaniele
163		if self.datasetPath == None and self.pset == None and self.scriptExe == '' :
164	<	msg ="WARNING. script_exe  not defined"
164	>	msg ="Error. script_exe  not defined"
165		raise CrabException(msg)
166
167		## additional input files
#	Line 147 | Line 171 | class Cmssw(JobType):
171		tmp = string.strip(tmp)
172		dirname = ''
173		if not tmp[0]=="/": dirname = "."
174	<	files = glob.glob(os.path.join(dirname, tmp))
174	>	files = []
175	>	if string.find(tmp,"*")>-1:
176	>	files = glob.glob(os.path.join(dirname, tmp))
177	>	if len(files)==0:
178	>	raise CrabException("No additional input file found with this pattern: "+tmp)
179	>	else:
180	>	files.append(tmp)
181		for file in files:
182		if not os.path.exists(file):
183		raise CrabException("Additional input file not found: "+file)
184		pass
185	+	# fname = string.split(file, '/')[-1]
186	+	# storedFile = common.work_space.pathForTgz()+'share/'+fname
187	+	# shutil.copyfile(file, storedFile)
188		self.additional_inbox_files.append(string.strip(file))
189		pass
190		pass
#	Line 173 | Line 206 | class Cmssw(JobType):
206		except KeyError:
207		self.eventsPerJob = -1
208		self.selectEventsPerJob = 0
209	<
209	>
210		## number of jobs
211		try:
212		self.theNumberOfJobs =int( cfg_params['CMSSW.number_of_jobs'])
#	Line 189 | Line 222 | class Cmssw(JobType):
222		self.total_number_of_events = 0
223		self.selectTotalNumberEvents = 0
224
225	<	if self.pset != None: #CarlosDaniele
225	>	if self.pset != None: #CarlosDaniele
226		if ( (self.selectTotalNumberEvents + self.selectEventsPerJob + self.selectNumberOfJobs) != 2 ):
227		msg = 'Must define exactly two of total_number_of_events, events_per_job, or number_of_jobs.'
228		raise CrabException(msg)
#	Line 210 | Line 243 | class Cmssw(JobType):
243		except KeyError:
244		self.sourceSeedVtx = None
245		common.logger.debug(5,"No vertex seed given")
246	+
247	+	try:
248	+	self.sourceSeedG4 = int(cfg_params['CMSSW.g4_seed'])
249	+	except KeyError:
250	+	self.sourceSeedG4 = None
251	+	common.logger.debug(5,"No g4 sim hits seed given")
252	+
253	+	try:
254	+	self.sourceSeedMix = int(cfg_params['CMSSW.mix_seed'])
255	+	except KeyError:
256	+	self.sourceSeedMix = None
257	+	common.logger.debug(5,"No mix seed given")
258	+
259	+	try:
260	+	self.firstRun = int(cfg_params['CMSSW.first_run'])
261	+	except KeyError:
262	+	self.firstRun = None
263	+	common.logger.debug(5,"No first run given")
264		if self.pset != None: #CarlosDaniele
265	<	self.PsetEdit = PsetManipulator.PsetManipulator(self.pset) #Daniele Pset
265	>	import PsetManipulator as pp
266	>	PsetEdit = pp.PsetManipulator(self.pset) #Daniele Pset
267
268		#DBSDLS-start
269	<	## Initialize the variables that are extracted from DBS/DLS and needed in other places of the code
269	>	## Initialize the variables that are extracted from DBS/DLS and needed in other places of the code
270		self.maxEvents=0  # max events available   ( --> check the requested nb. of evts in Creator.py)
271		self.DBSPaths={}  # all dbs paths requested ( --> input to the site local discovery script)
272		self.jobDestination=[]  # Site destination(s) for each job (list of lists)
#	Line 223 | Line 275 | class Cmssw(JobType):
275		blockSites = {}
276		if self.datasetPath:
277		blockSites = self.DataDiscoveryAndLocation(cfg_params)
278	<	#DBSDLS-end
278	>	#DBSDLS-end
279
280		self.tgzNameWithPath = self.getTarBall(self.executable)
281	<
281	>
282		## Select Splitting
283	<	if self.selectNoInput:
283	>	if self.selectNoInput:
284		if self.pset == None: #CarlosDaniele
285		self.jobSplittingForScript()
286		else:
287		self.jobSplittingNoInput()
288	<	else: self.jobSplittingByBlocks(blockSites)
288	>	else:
289	>	self.jobSplittingByBlocks(blockSites)
290
291		# modify Pset
292		if self.pset != None: #CarlosDaniele
293		try:
294		if (self.datasetPath): # standard job
295		# allow to processa a fraction of events in a file
296	<	self.PsetEdit.inputModule("INPUT")
297	<	self.PsetEdit.maxEvent("INPUTMAXEVENTS")
298	<	self.PsetEdit.skipEvent("INPUTSKIPEVENTS")
296	>	PsetEdit.inputModule("INPUTFILE")
297	>	PsetEdit.maxEvent(0)
298	>	PsetEdit.skipEvent(0)
299		else:  # pythia like job
300	<	self.PsetEdit.maxEvent(self.eventsPerJob)
300	>	PsetEdit.maxEvent(self.eventsPerJob)
301	>	if (self.firstRun):
302	>	PsetEdit.pythiaFirstRun(0)  #First Run
303		if (self.sourceSeed) :
304	<	self.PsetEdit.pythiaSeed("INPUT")
304	>	PsetEdit.pythiaSeed(0)
305		if (self.sourceSeedVtx) :
306	<	self.PsetEdit.pythiaSeedVtx("INPUTVTX")
306	>	PsetEdit.vtxSeed(0)
307	>	if (self.sourceSeedG4) :
308	>	PsetEdit.g4Seed(0)
309	>	if (self.sourceSeedMix) :
310	>	PsetEdit.mixSeed(0)
311		# add FrameworkJobReport to parameter-set
312	<	self.PsetEdit.addCrabFJR(self.fjrFileName)
313	<	self.PsetEdit.psetWriter(self.configFilename())
312	>	PsetEdit.addCrabFJR(self.fjrFileName)
313	>	PsetEdit.psetWriter(self.configFilename())
314		except:
315		msg='Error while manipuliating ParameterSet: exiting...'
316		raise CrabException(msg)
317
318		def DataDiscoveryAndLocation(self, cfg_params):
319
320	+	import DataDiscovery
321	+	import DataLocation
322		common.logger.debug(10,"CMSSW::DataDiscoveryAndLocation()")
323
324		datasetPath=self.datasetPath
325
265	–	## TODO
266	–	dataTiersList = ""
267	–	dataTiers = dataTiersList.split(',')
268	–
326		## Contact the DBS
327	<	common.logger.message("Contacting DBS...")
327	>	common.logger.message("Contacting Data Discovery Services ...")
328		try:
329	<	self.pubdata=DataDiscovery.DataDiscovery(datasetPath, dataTiers, cfg_params)
329	>
330	>	self.pubdata=DataDiscovery.DataDiscovery(datasetPath, cfg_params)
331		self.pubdata.fetchDBSInfo()
332
333		except DataDiscovery.NotExistingDatasetError, ex :
334		msg = 'ERROR ***: failed Data Discovery in DBS : %s'%ex.getErrorMessage()
335		raise CrabException(msg)
278	–
336		except DataDiscovery.NoDataTierinProvenanceError, ex :
337		msg = 'ERROR ***: failed Data Discovery in DBS : %s'%ex.getErrorMessage()
338		raise CrabException(msg)
339		except DataDiscovery.DataDiscoveryError, ex:
340	<	msg = 'ERROR ***: failed Data Discovery in DBS  %s'%ex.getErrorMessage()
340	>	msg = 'ERROR ***: failed Data Discovery in DBS :  %s'%ex.getErrorMessage()
341		raise CrabException(msg)
342
286	–	## get list of all required data in the form of dbs paths  (dbs path = /dataset/datatier/owner)
287	–	## self.DBSPaths=self.pubdata.getDBSPaths()
288	–	common.logger.message("Required data are :"+self.datasetPath)
289	–
343		self.filesbyblock=self.pubdata.getFiles()
344		self.eventsbyblock=self.pubdata.getEventsPerBlock()
345		self.eventsbyfile=self.pubdata.getEventsPerFile()
293	–	# print str(self.filesbyblock)
294	–	# print 'self.eventsbyfile',len(self.eventsbyfile)
295	–	# print str(self.eventsbyfile)
346
347		## get max number of events
348	<	self.maxEvents=self.pubdata.getMaxEvents() ##  self.maxEvents used in Creator.py
299	<	common.logger.message("The number of available events is %s\n"%self.maxEvents)
348	>	self.maxEvents=self.pubdata.getMaxEvents() ##  self.maxEvents used in Creator.py
349
301	–	common.logger.message("Contacting DLS...")
350		## Contact the DLS and build a list of sites hosting the fileblocks
351		try:
352		dataloc=DataLocation.DataLocation(self.filesbyblock.keys(),cfg_params)
#	Line 306 | Line 354 | class Cmssw(JobType):
354		except DataLocation.DataLocationError , ex:
355		msg = 'ERROR ***: failed Data Location in DLS \n %s '%ex.getErrorMessage()
356		raise CrabException(msg)
357	<
357	>
358
359		sites = dataloc.getSites()
360		allSites = []
361		listSites = sites.values()
362	<	for list in listSites:
363	<	for oneSite in list:
362	>	for listSite in listSites:
363	>	for oneSite in listSite:
364		allSites.append(oneSite)
365		allSites = self.uniquelist(allSites)
366
367	<	common.logger.message("Sites ("+str(len(allSites))+") hosting part/all of dataset: "+str(allSites))
368	<	common.logger.debug(6, "List of Sites: "+str(allSites))
367	>	# screen output
368	>	common.logger.message("Requested dataset: " + datasetPath + " has " + str(self.maxEvents) + " events in " + str(len(self.filesbyblock.keys())) + " blocks.\n")
369	>
370		return sites
371	<
371	>
372		def jobSplittingByBlocks(self, blockSites):
373		"""
374		Perform job splitting. Jobs run over an integer number of files
#	Line 369 | Line 418 | class Cmssw(JobType):
418		totalNumberOfJobs = 999999999
419		else :
420		totalNumberOfJobs = self.ncjobs
421	<
421	>
422
423		blocks = blockSites.keys()
424		blockCount = 0
#	Line 379 | Line 428 | class Cmssw(JobType):
428		jobCount = 0
429		list_of_lists = []
430
431	+	# list tracking which jobs are in which jobs belong to which block
432	+	jobsOfBlock = {}
433	+
434		# ---- Iterate over the blocks in the dataset until ---- #
435		# ---- we've met the requested total # of events    ---- #
436		while ( (eventsRemaining > 0) and (blockCount < numBlocksInDataset) and (jobCount < totalNumberOfJobs)):
437		block = blocks[blockCount]
438		blockCount += 1
439	<
439	>	if block not in jobsOfBlock.keys() :
440	>	jobsOfBlock[block] = []
441
442	<	numEventsInBlock = self.eventsbyblock[block]
443	<	common.logger.debug(5,'Events in Block File '+str(numEventsInBlock))
444	<
445	<	files = self.filesbyblock[block]
446	<	numFilesInBlock = len(files)
447	<	if (numFilesInBlock <= 0):
448	<	continue
449	<	fileCount = 0
450	<
451	<	# ---- New block => New job ---- #
452	<	parString = "\\{"
453	<	# counter for number of events in files currently worked on
454	<	filesEventCount = 0
455	<	# flag if next while loop should touch new file
456	<	newFile = 1
457	<	# job event counter
458	<	jobSkipEventCount = 0
459	<
460	<	# ---- Iterate over the files in the block until we've met the requested ---- #
461	<	# ---- total # of events or we've gone over all the files in this block  ---- #
462	<	while ( (eventsRemaining > 0) and (fileCount < numFilesInBlock) and (jobCount < totalNumberOfJobs) ):
463	<	file = files[fileCount]
464	<	if newFile :
465	<	try:
466	<	numEventsInFile = self.eventsbyfile[file]
467	<	common.logger.debug(6, "File "+str(file)+" has "+str(numEventsInFile)+" events")
468	<	# increase filesEventCount
469	<	filesEventCount += numEventsInFile
470	<	# Add file to current job
471	<	parString += '\\\"' + file + '\\\"\,'
472	<	newFile = 0
473	<	except KeyError:
474	<	common.logger.message("File "+str(file)+" has unknown number of events: skipping")
475	<
476	<
477	<	# if less events in file remain than eventsPerJobRequested
478	<	if ( filesEventCount - jobSkipEventCount < eventsPerJobRequested ) :
479	<	# if last file in block
480	<	if ( fileCount == numFilesInBlock-1 ) :
481	<	# end job using last file, use remaining events in block
442	>	if self.eventsbyblock.has_key(block) :
443	>	numEventsInBlock = self.eventsbyblock[block]
444	>	common.logger.debug(5,'Events in Block File '+str(numEventsInBlock))
445	>
446	>	files = self.filesbyblock[block]
447	>	numFilesInBlock = len(files)
448	>	if (numFilesInBlock <= 0):
449	>	continue
450	>	fileCount = 0
451	>
452	>	# ---- New block => New job ---- #
453	>	parString = ""
454	>	# counter for number of events in files currently worked on
455	>	filesEventCount = 0
456	>	# flag if next while loop should touch new file
457	>	newFile = 1
458	>	# job event counter
459	>	jobSkipEventCount = 0
460	>
461	>	# ---- Iterate over the files in the block until we've met the requested ---- #
462	>	# ---- total # of events or we've gone over all the files in this block  ---- #
463	>	while ( (eventsRemaining > 0) and (fileCount < numFilesInBlock) and (jobCount < totalNumberOfJobs) ):
464	>	file = files[fileCount]
465	>	if newFile :
466	>	try:
467	>	numEventsInFile = self.eventsbyfile[file]
468	>	common.logger.debug(6, "File "+str(file)+" has "+str(numEventsInFile)+" events")
469	>	# increase filesEventCount
470	>	filesEventCount += numEventsInFile
471	>	# Add file to current job
472	>	parString += '\\\"' + file + '\\\"\,'
473	>	newFile = 0
474	>	except KeyError:
475	>	common.logger.message("File "+str(file)+" has unknown number of events: skipping")
476	>
477	>
478	>	# if less events in file remain than eventsPerJobRequested
479	>	if ( filesEventCount - jobSkipEventCount < eventsPerJobRequested ) :
480	>	# if last file in block
481	>	if ( fileCount == numFilesInBlock-1 ) :
482	>	# end job using last file, use remaining events in block
483	>	# close job and touch new file
484	>	fullString = parString[:-2]
485	>	list_of_lists.append([fullString,str(-1),str(jobSkipEventCount)])
486	>	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(filesEventCount - jobSkipEventCount)+" events (last file in block).")
487	>	self.jobDestination.append(blockSites[block])
488	>	common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
489	>	# fill jobs of block dictionary
490	>	jobsOfBlock[block].append(jobCount+1)
491	>	# reset counter
492	>	jobCount = jobCount + 1
493	>	totalEventCount = totalEventCount + filesEventCount - jobSkipEventCount
494	>	eventsRemaining = eventsRemaining - filesEventCount + jobSkipEventCount
495	>	jobSkipEventCount = 0
496	>	# reset file
497	>	parString = ""
498	>	filesEventCount = 0
499	>	newFile = 1
500	>	fileCount += 1
501	>	else :
502	>	# go to next file
503	>	newFile = 1
504	>	fileCount += 1
505	>	# if events in file equal to eventsPerJobRequested
506	>	elif ( filesEventCount - jobSkipEventCount == eventsPerJobRequested ) :
507		# close job and touch new file
508		fullString = parString[:-2]
509	<	fullString += '\\}'
510	<	list_of_lists.append([fullString,str(-1),str(jobSkipEventCount)])
433	<	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(filesEventCount - jobSkipEventCount)+" events (last file in block).")
509	>	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
510	>	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.")
511		self.jobDestination.append(blockSites[block])
512		common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
513	+	jobsOfBlock[block].append(jobCount+1)
514		# reset counter
515		jobCount = jobCount + 1
516	<	totalEventCount = totalEventCount + filesEventCount - jobSkipEventCount
517	<	eventsRemaining = eventsRemaining - filesEventCount + jobSkipEventCount
516	>	totalEventCount = totalEventCount + eventsPerJobRequested
517	>	eventsRemaining = eventsRemaining - eventsPerJobRequested
518		jobSkipEventCount = 0
519		# reset file
520	<	parString = "\\{"
520	>	parString = ""
521		filesEventCount = 0
522		newFile = 1
523		fileCount += 1
524	+
525	+	# if more events in file remain than eventsPerJobRequested
526		else :
527	<	# go to next file
528	<	newFile = 1
529	<	fileCount += 1
530	<	# if events in file equal to eventsPerJobRequested
531	<	elif ( filesEventCount - jobSkipEventCount == eventsPerJobRequested ) :
532	<	# close job and touch new file
533	<	fullString = parString[:-2]
534	<	fullString += '\\}'
535	<	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
536	<	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.")
537	<	self.jobDestination.append(blockSites[block])
538	<	common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
539	<	# reset counter
540	<	jobCount = jobCount + 1
541	<	totalEventCount = totalEventCount + eventsPerJobRequested
542	<	eventsRemaining = eventsRemaining - eventsPerJobRequested
543	<	jobSkipEventCount = 0
544	<	# reset file
545	<	parString = "\\{"
546	<	filesEventCount = 0
467	<	newFile = 1
468	<	fileCount += 1
469	<
470	<	# if more events in file remain than eventsPerJobRequested
471	<	else :
472	<	# close job but don't touch new file
473	<	fullString = parString[:-2]
474	<	fullString += '\\}'
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	<	# increase counter
480	<	jobCount = jobCount + 1
481	<	totalEventCount = totalEventCount + eventsPerJobRequested
482	<	eventsRemaining = eventsRemaining - eventsPerJobRequested
483	<	# calculate skip events for last file
484	<	# use filesEventCount (contains several files), jobSkipEventCount and eventsPerJobRequest
485	<	jobSkipEventCount = eventsPerJobRequested - (filesEventCount - jobSkipEventCount - self.eventsbyfile[file])
486	<	# remove all but the last file
487	<	filesEventCount = self.eventsbyfile[file]
488	<	parString = "\\{"
489	<	parString += '\\\"' + file + '\\\"\,'
490	<	pass # END if
491	<	pass # END while (iterate over files in the block)
527	>	# close job but don't touch new file
528	>	fullString = parString[:-2]
529	>	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
530	>	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.")
531	>	self.jobDestination.append(blockSites[block])
532	>	common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
533	>	jobsOfBlock[block].append(jobCount+1)
534	>	# increase counter
535	>	jobCount = jobCount + 1
536	>	totalEventCount = totalEventCount + eventsPerJobRequested
537	>	eventsRemaining = eventsRemaining - eventsPerJobRequested
538	>	# calculate skip events for last file
539	>	# use filesEventCount (contains several files), jobSkipEventCount and eventsPerJobRequest
540	>	jobSkipEventCount = eventsPerJobRequested - (filesEventCount - jobSkipEventCount - self.eventsbyfile[file])
541	>	# remove all but the last file
542	>	filesEventCount = self.eventsbyfile[file]
543	>	parString = ""
544	>	parString += '\\\"' + file + '\\\"\,'
545	>	pass # END if
546	>	pass # END while (iterate over files in the block)
547		pass # END while (iterate over blocks in the dataset)
548		self.ncjobs = self.total_number_of_jobs = jobCount
549		if (eventsRemaining > 0 and jobCount < totalNumberOfJobs ):
550		common.logger.message("Could not run on all requested events because some blocks not hosted at allowed sites.")
551	<	common.logger.message("\n"+str(jobCount)+" job(s) can run on "+str(totalEventCount)+" events.\n")
552	<
551	>	common.logger.message(str(jobCount)+" job(s) can run on "+str(totalEventCount)+" events.\n")
552	>
553	>	# screen output
554	>	screenOutput = "List of jobs and available destination sites:\n\n"
555	>
556	>	# keep trace of block with no sites to print a warning at the end
557	>	noSiteBlock = []
558	>	bloskNoSite = []
559	>
560	>	blockCounter = 0
561	>	for block in blocks:
562	>	if block in jobsOfBlock.keys() :
563	>	blockCounter += 1
564	>	screenOutput += "Block %5i: jobs %20s: sites: %s\n" % (blockCounter,spanRanges(jobsOfBlock[block]),','.join(self.blackWhiteListParser.checkWhiteList(self.blackWhiteListParser.checkBlackList(blockSites[block],block),block)))
565	>	if len(self.blackWhiteListParser.checkWhiteList(self.blackWhiteListParser.checkBlackList(blockSites[block],block),block)) == 0:
566	>	noSiteBlock.append( spanRanges(jobsOfBlock[block]) )
567	>	bloskNoSite.append( blockCounter )
568	>
569	>	common.logger.message(screenOutput)
570	>	if len(noSiteBlock) > 0 and len(bloskNoSite) > 0:
571	>	msg = 'WARNING: No sites are hosting any part of data for block:\n                '
572	>	virgola = ""
573	>	if len(bloskNoSite) > 1:
574	>	virgola = ","
575	>	for block in bloskNoSite:
576	>	msg += ' ' + str(block) + virgola
577	>	msg += '\n               Related jobs:\n                 '
578	>	virgola = ""
579	>	if len(noSiteBlock) > 1:
580	>	virgola = ","
581	>	for range_jobs in noSiteBlock:
582	>	msg += str(range_jobs) + virgola
583	>	msg += '\n               will not be submitted and this block of data can not be analyzed!\n'
584	>	common.logger.message(msg)
585	>
586		self.list_of_args = list_of_lists
587		return
588
#	Line 503 | Line 591 | class Cmssw(JobType):
591		Perform job splitting based on number of event per job
592		"""
593		common.logger.debug(5,'Splitting per events')
594	<	common.logger.message('Required '+str(self.eventsPerJob)+' events per job ')
595	<	common.logger.message('Required '+str(self.theNumberOfJobs)+' jobs in total ')
596	<	common.logger.message('Required '+str(self.total_number_of_events)+' events in total ')
594	>
595	>	if (self.selectEventsPerJob):
596	>	common.logger.message('Required '+str(self.eventsPerJob)+' events per job ')
597	>	if (self.selectNumberOfJobs):
598	>	common.logger.message('Required '+str(self.theNumberOfJobs)+' jobs in total ')
599	>	if (self.selectTotalNumberEvents):
600	>	common.logger.message('Required '+str(self.total_number_of_events)+' events in total ')
601
602		if (self.total_number_of_events < 0):
603		msg='Cannot split jobs per Events with "-1" as total number of events'
604		raise CrabException(msg)
605
606		if (self.selectEventsPerJob):
607	<	self.total_number_of_jobs = int(self.total_number_of_events/self.eventsPerJob)
607	>	if (self.selectTotalNumberEvents):
608	>	self.total_number_of_jobs = int(self.total_number_of_events/self.eventsPerJob)
609	>	elif(self.selectNumberOfJobs) :
610	>	self.total_number_of_jobs =self.theNumberOfJobs
611	>	self.total_number_of_events =int(self.theNumberOfJobs*self.eventsPerJob)
612	>
613		elif (self.selectNumberOfJobs) :
614		self.total_number_of_jobs = self.theNumberOfJobs
615		self.eventsPerJob = int(self.total_number_of_events/self.total_number_of_jobs)
#	Line 532 | Line 629 | class Cmssw(JobType):
629		self.list_of_args = []
630		for i in range(self.total_number_of_jobs):
631		## Since there is no input, any site is good
632	<	# self.jobDestination.append(["Any"])
633	<	self.jobDestination.append([""]) #must be empty to write correctly the xml
632	>	self.jobDestination.append([""]) #must be empty to write correctly the xml
633	>	args=[]
634	>	if (self.firstRun):
635	>	## pythia first run
636	>	args.append(str(self.firstRun)+str(i))
637		if (self.sourceSeed):
638	+	args.append(str(self.sourceSeed)+str(i))
639		if (self.sourceSeedVtx):
640	<	## pythia + vtx random seed
641	<	self.list_of_args.append([
642	<	str(self.sourceSeed)+str(i),
643	<	str(self.sourceSeedVtx)+str(i)
644	<	])
645	<	else:
646	<	## only pythia random seed
647	<	self.list_of_args.append([(str(self.sourceSeed)+str(i))])
648	<	else:
649	<	## no random seed
650	<	self.list_of_args.append([str(i)])
651	<	#print self.list_of_args
640	>	## + vtx random seed
641	>	args.append(str(self.sourceSeedVtx)+str(i))
642	>	if (self.sourceSeedG4):
643	>	## + G4 random seed
644	>	args.append(str(self.sourceSeedG4)+str(i))
645	>	if (self.sourceSeedMix):
646	>	## + Mix random seed
647	>	args.append(str(self.sourceSeedMix)+str(i))
648	>	pass
649	>	pass
650	>	self.list_of_args.append(args)
651	>	pass
652	>
653	>	# print self.list_of_args
654
655		return
656
#	Line 576 | Line 679 | class Cmssw(JobType):
679		return
680
681		def split(self, jobParams):
682	<
682	>
683		common.jobDB.load()
684		#### Fabio
685		njobs = self.total_number_of_jobs
#	Line 584 | Line 687 | class Cmssw(JobType):
687		# create the empty structure
688		for i in range(njobs):
689		jobParams.append("")
690	<
690	>
691		for job in range(njobs):
692		jobParams[job] = arglist[job]
693		# print str(arglist[job])
#	Line 595 | Line 698 | class Cmssw(JobType):
698
699		common.jobDB.save()
700		return
701	<
701	>
702		def getJobTypeArguments(self, nj, sched):
703		result = ''
704		for i in common.jobDB.arguments(nj):
705		result=result+str(i)+" "
706		return result
707	<
707	>
708		def numberOfJobs(self):
709		# Fabio
710		return self.total_number_of_jobs
#	Line 610 | Line 713 | class Cmssw(JobType):
713		"""
714		Return the TarBall with lib and exe
715		"""
716	<
716	>
717		# if it exist, just return it
718	<	self.tgzNameWithPath = common.work_space.shareDir()+self.tgz_name
718	>	#
719	>	# Marco. Let's start to use relative path for Boss XML files
720	>	#
721	>	self.tgzNameWithPath = common.work_space.pathForTgz()+'share/'+self.tgz_name
722		if os.path.exists(self.tgzNameWithPath):
723		return self.tgzNameWithPath
724
#	Line 626 | Line 732 | class Cmssw(JobType):
732		# First of all declare the user Scram area
733		swArea = self.scram.getSWArea_()
734		#print "swArea = ", swArea
735	<	swVersion = self.scram.getSWVersion()
736	<	#print "swVersion = ", swVersion
735	>	# swVersion = self.scram.getSWVersion()
736	>	# print "swVersion = ", swVersion
737		swReleaseTop = self.scram.getReleaseTop_()
738		#print "swReleaseTop = ", swReleaseTop
739	<
739	>
740		## check if working area is release top
741		if swReleaseTop == '' or swArea == swReleaseTop:
742		return
743
744	<	filesToBeTarred = []
745	<	## First find the executable
746	<	if (self.executable != ''):
747	<	exeWithPath = self.scram.findFile_(executable)
748	<	#           print exeWithPath
749	<	if ( not exeWithPath ):
750	<	raise CrabException('User executable '+executable+' not found')
751	<
752	<	## then check if it's private or not
753	<	if exeWithPath.find(swReleaseTop) == -1:
754	<	# the exe is private, so we must ship
755	<	common.logger.debug(5,"Exe "+exeWithPath+" to be tarred")
756	<	path = swArea+'/'
757	<	exe = string.replace(exeWithPath, path,'')
758	<	filesToBeTarred.append(exe)
759	<	pass
760	<	else:
761	<	# the exe is from release, we'll find it on WN
762	<	pass
763	<
764	<	## Now get the libraries: only those in local working area
765	<	libDir = 'lib'
766	<	lib = swArea+'/' +libDir
767	<	common.logger.debug(5,"lib "+lib+" to be tarred")
768	<	if os.path.exists(lib):
769	<	filesToBeTarred.append(libDir)
770	<
771	<	## Now check if module dir is present
772	<	moduleDir = 'module'
773	<	if os.path.isdir(swArea+'/'+moduleDir):
774	<	filesToBeTarred.append(moduleDir)
775	<
776	<	## Now check if the Data dir is present
777	<	dataDir = 'src/Data/'
778	<	if os.path.isdir(swArea+'/'+dataDir):
779	<	filesToBeTarred.append(dataDir)
780	<
781	<	## copy ProdAgent dir to swArea
782	<	cmd = '\cp -rf ' + os.environ['CRABDIR'] + '/ProdAgentApi ' + swArea
783	<	cmd_out = runCommand(cmd)
784	<	if cmd_out != '':
785	<	common.logger.message('ProdAgentApi directory could not be copied to local CMSSW project directory.')
786	<	common.logger.message('No FrameworkJobreport parsing is possible on the WorkerNode.')
787	<
788	<	## Now check if the Data dir is present
789	<	paDir = 'ProdAgentApi'
790	<	if os.path.isdir(swArea+'/'+paDir):
791	<	filesToBeTarred.append(paDir)
792	<
793	<	## Create the tar-ball
794	<	if len(filesToBeTarred)>0:
795	<	cwd = os.getcwd()
796	<	os.chdir(swArea)
797	<	tarcmd = 'tar zcvf ' + self.tgzNameWithPath + ' '
798	<	for line in filesToBeTarred:
799	<	tarcmd = tarcmd + line + ' '
800	<	cout = runCommand(tarcmd)
801	<	if not cout:
802	<	raise CrabException('Could not create tar-ball')
803	<	os.chdir(cwd)
804	<	else:
805	<	common.logger.debug(5,"No files to be to be tarred")
806	<
744	>	import tarfile
745	>	try: # create tar ball
746	>	tar = tarfile.open(self.tgzNameWithPath, "w:gz")
747	>	## First find the executable
748	>	if (self.executable != ''):
749	>	exeWithPath = self.scram.findFile_(executable)
750	>	if ( not exeWithPath ):
751	>	raise CrabException('User executable '+executable+' not found')
752	>
753	>	## then check if it's private or not
754	>	if exeWithPath.find(swReleaseTop) == -1:
755	>	# the exe is private, so we must ship
756	>	common.logger.debug(5,"Exe "+exeWithPath+" to be tarred")
757	>	path = swArea+'/'
758	>	# distinguish case when script is in user project area or given by full path somewhere else
759	>	if exeWithPath.find(path) >= 0 :
760	>	exe = string.replace(exeWithPath, path,'')
761	>	tar.add(path+exe,exe)
762	>	else :
763	>	tar.add(exeWithPath,os.path.basename(executable))
764	>	pass
765	>	else:
766	>	# the exe is from release, we'll find it on WN
767	>	pass
768	>
769	>	## Now get the libraries: only those in local working area
770	>	libDir = 'lib'
771	>	lib = swArea+'/' +libDir
772	>	common.logger.debug(5,"lib "+lib+" to be tarred")
773	>	if os.path.exists(lib):
774	>	tar.add(lib,libDir)
775	>
776	>	## Now check if module dir is present
777	>	moduleDir = 'module'
778	>	module = swArea + '/' + moduleDir
779	>	if os.path.isdir(module):
780	>	tar.add(module,moduleDir)
781	>
782	>	## Now check if any data dir(s) is present
783	>	swAreaLen=len(swArea)
784	>	for root, dirs, files in os.walk(swArea):
785	>	if "data" in dirs:
786	>	common.logger.debug(5,"data "+root+"/data"+" to be tarred")
787	>	tar.add(root+"/data",root[swAreaLen:]+"/data")
788	>
789	>	## Add ProdAgent dir to tar
790	>	paDir = 'ProdAgentApi'
791	>	pa = os.environ['CRABDIR'] + '/' + 'ProdAgentApi'
792	>	if os.path.isdir(pa):
793	>	tar.add(pa,paDir)
794	>
795	>	### FEDE FOR DBS PUBLICATION
796	>	## Add PRODCOMMON dir to tar
797	>	prodcommonDir = 'ProdCommon'
798	>	prodcommonPath = os.environ['CRABDIR'] + '/' + 'ProdCommon'
799	>	if os.path.isdir(prodcommonPath):
800	>	tar.add(prodcommonPath,prodcommonDir)
801	>	#############################
802	>
803	>	common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
804	>	tar.close()
805	>	except :
806	>	raise CrabException('Could not create tar-ball')
807	>
808	>	## check for tarball size
809	>	tarballinfo = os.stat(self.tgzNameWithPath)
810	>	if ( tarballinfo.st_size > self.MaxTarBallSize*1024*1024 ) :
811	>	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.')
812	>
813	>	## create tar-ball with ML stuff
814	>	self.MLtgzfile =  common.work_space.pathForTgz()+'share/MLfiles.tgz'
815	>	try:
816	>	tar = tarfile.open(self.MLtgzfile, "w:gz")
817	>	path=os.environ['CRABDIR'] + '/python/'
818	>	for file in ['report.py', 'DashboardAPI.py', 'Logger.py', 'ProcInfo.py', 'apmon.py', 'parseCrabFjr.py']:
819	>	tar.add(path+file,file)
820	>	common.logger.debug(5,"Files added to "+self.MLtgzfile+" : "+str(tar.getnames()))
821	>	tar.close()
822	>	except :
823	>	raise CrabException('Could not create ML files tar-ball')
824	>
825		return
826	<
826	>
827	>	def additionalInputFileTgz(self):
828	>	"""
829	>	Put all additional files into a tar ball and return its name
830	>	"""
831	>	import tarfile
832	>	tarName=  common.work_space.pathForTgz()+'share/'+self.additional_tgz_name
833	>	tar = tarfile.open(tarName, "w:gz")
834	>	for file in self.additional_inbox_files:
835	>	tar.add(file,string.split(file,'/')[-1])
836	>	common.logger.debug(5,"Files added to "+self.additional_tgz_name+" : "+str(tar.getnames()))
837	>	tar.close()
838	>	return tarName
839	>
840		def wsSetupEnvironment(self, nj):
841		"""
842		Returns part of a job script which prepares
843		the execution environment for the job 'nj'.
844		"""
845		# Prepare JobType-independent part
846	<	txt = ''
847	<
848	<	## OLI_Daniele at this level  middleware already known
712	<
713	<	txt += 'if [ $middleware == LCG ]; then \n'
846	>	txt = ''
847	>	txt += 'echo ">>> setup environment"\n'
848	>	txt += 'if [ $middleware == LCG ]; then \n'
849		txt += self.wsSetupCMSLCGEnvironment_()
850		txt += 'elif [ $middleware == OSG ]; then\n'
851		txt += '    WORKING_DIR=`/bin/mktemp  -d $OSG_WN_TMP/cms_XXXXXXXXXXXX`\n'
852	<	txt += '    echo "Created working directory: $WORKING_DIR"\n'
718	<	txt += '    if [ ! -d $WORKING_DIR ] ;then\n'
852	>	txt += '    if [ ! $? == 0 ] ;then\n'
853		txt += '        echo "SET_CMS_ENV 10016 ==> OSG $WORKING_DIR could not be created on WN `hostname`"\n'
854	<	txt += '        echo "JOB_EXIT_STATUS = 10016"\n'
855	<	txt += '        echo "JobExitCode=10016" | tee -a $RUNTIME_AREA/$repo\n'
856	<	txt += '        dumpStatus $RUNTIME_AREA/$repo\n'
723	<	txt += '        rm -f $RUNTIME_AREA/$repo \n'
724	<	txt += '        echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
725	<	txt += '        echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
854	>	txt += '        echo "JOB_EXIT_STATUS = 10016"\n'
855	>	txt += '        echo "JobExitCode=10016" | tee -a $RUNTIME_AREA/$repo\n'
856	>	txt += '        dumpStatus $RUNTIME_AREA/$repo\n'
857		txt += '        exit 1\n'
858		txt += '    fi\n'
859	+	txt += '    echo ">>> Created working directory: $WORKING_DIR"\n'
860		txt += '\n'
861		txt += '    echo "Change to working directory: $WORKING_DIR"\n'
862		txt += '    cd $WORKING_DIR\n'
863	<	txt += self.wsSetupCMSOSGEnvironment_()
863	>	txt += '    echo ">>> current directory (WORKING_DIR): $WORKING_DIR"\n'
864	>	txt += self.wsSetupCMSOSGEnvironment_()
865	>	#txt += '    echo "### Set SCRAM ARCH to ' + self.executable_arch + ' ###"\n'
866	>	#txt += '    export SCRAM_ARCH='+self.executable_arch+'\n'
867		txt += 'fi\n'
868
869		# Prepare JobType-specific part
870		scram = self.scram.commandName()
871		txt += '\n\n'
872	<	txt += 'echo "### SPECIFIC JOB SETUP ENVIRONMENT ###"\n'
872	>	txt += 'echo ">>> specific cmssw setup environment:"\n'
873	>	txt += 'echo "CMSSW_VERSION =  '+self.version+'"\n'
874		txt += scram+' project CMSSW '+self.version+'\n'
875		txt += 'status=$?\n'
876		txt += 'if [ $status != 0 ] ; then\n'
877	<	txt += '   echo "SET_EXE_ENV 10034 ==>ERROR CMSSW '+self.version+' not found on `hostname`" \n'
878	<	txt += '   echo "JOB_EXIT_STATUS = 10034"\n'
879	<	txt += '   echo "JobExitCode=10034" | tee -a $RUNTIME_AREA/$repo\n'
880	<	txt += '   dumpStatus $RUNTIME_AREA/$repo\n'
745	<	txt += '   rm -f $RUNTIME_AREA/$repo \n'
746	<	txt += '   echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
747	<	txt += '   echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
748	<	## OLI_Daniele
877	>	txt += '    echo "SET_EXE_ENV 10034 ==>ERROR CMSSW '+self.version+' not found on `hostname`" \n'
878	>	txt += '    echo "JOB_EXIT_STATUS = 10034"\n'
879	>	txt += '    echo "JobExitCode=10034" | tee -a $RUNTIME_AREA/$repo\n'
880	>	txt += '    dumpStatus $RUNTIME_AREA/$repo\n'
881		txt += '    if [ $middleware == OSG ]; then \n'
750	–	txt += '        echo "Remove working directory: $WORKING_DIR"\n'
882		txt += '        cd $RUNTIME_AREA\n'
883	+	txt += '        echo ">>> current directory (RUNTIME_AREA): $RUNTIME_AREA"\n'
884	+	txt += '        echo ">>> Remove working directory: $WORKING_DIR"\n'
885		txt += '        /bin/rm -rf $WORKING_DIR\n'
886		txt += '        if [ -d $WORKING_DIR ] ;then\n'
887	<	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'
888	<	txt += '            echo "JOB_EXIT_STATUS = 10018"\n'
889	<	txt += '            echo "JobExitCode=10018" | tee -a $RUNTIME_AREA/$repo\n'
890	<	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
758	<	txt += '            rm -f $RUNTIME_AREA/$repo \n'
759	<	txt += '            echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
760	<	txt += '            echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
887	>	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'
888	>	txt += '            echo "JOB_EXIT_STATUS = 10018"\n'
889	>	txt += '            echo "JobExitCode=10018" | tee -a $RUNTIME_AREA/$repo\n'
890	>	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
891		txt += '        fi\n'
892		txt += '    fi \n'
893	<	txt += '   exit 1 \n'
893	>	txt += '    exit 1 \n'
894		txt += 'fi \n'
765	–	txt += 'echo "CMSSW_VERSION =  '+self.version+'"\n'
895		txt += 'cd '+self.version+'\n'
896	+	########## FEDE FOR DBS2 ######################
897	+	txt += 'SOFTWARE_DIR=`pwd`\n'
898	+	txt += 'echo ">>> current directory (SOFTWARE_DIR): $SOFTWARE_DIR" \n'
899	+	###############################################
900		### needed grep for bug in scramv1 ###
901	+	txt += scram+' runtime -sh\n'
902		txt += 'eval `'+scram+' runtime -sh | grep -v SCRAMRT_LSB_JOBNAME`\n'
903	+	txt += 'echo $PATH\n'
904
905		# Handle the arguments:
906		txt += "\n"
907		txt += "## number of arguments (first argument always jobnumber)\n"
908		txt += "\n"
774	–	#        txt += "narg=$#\n"
909		txt += "if [ $nargs -lt 2 ]\n"
910		txt += "then\n"
911		txt += "    echo 'SET_EXE_ENV 1 ==> ERROR Too few arguments' +$nargs+ \n"
912		txt += '    echo "JOB_EXIT_STATUS = 50113"\n'
913		txt += '    echo "JobExitCode=50113" | tee -a $RUNTIME_AREA/$repo\n'
914		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'
784	–	## OLI_Daniele
915		txt += '    if [ $middleware == OSG ]; then \n'
786	–	txt += '        echo "Remove working directory: $WORKING_DIR"\n'
916		txt += '        cd $RUNTIME_AREA\n'
917	+	txt += '        echo ">>> current directory (RUNTIME_AREA): $RUNTIME_AREA"\n'
918	+	txt += '        echo ">>> Remove working directory: $WORKING_DIR"\n'
919		txt += '        /bin/rm -rf $WORKING_DIR\n'
920		txt += '        if [ -d $WORKING_DIR ] ;then\n'
921	<	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'
922	<	txt += '            echo "JOB_EXIT_STATUS = 50114"\n'
923	<	txt += '            echo "JobExitCode=50114" | tee -a $RUNTIME_AREA/$repo\n'
924	<	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
794	<	txt += '            rm -f $RUNTIME_AREA/$repo \n'
795	<	txt += '            echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
796	<	txt += '            echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
921	>	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'
922	>	txt += '            echo "JOB_EXIT_STATUS = 50114"\n'
923	>	txt += '            echo "JobExitCode=50114" | tee -a $RUNTIME_AREA/$repo\n'
924	>	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
925		txt += '        fi\n'
926		txt += '    fi \n'
927		txt += "    exit 1\n"
#	Line 802 | Line 930 | class Cmssw(JobType):
930
931		# Prepare job-specific part
932		job = common.job_list[nj]
933	+	### FEDE FOR DBS OUTPUT PUBLICATION
934	+	if (self.datasetPath):
935	+	txt += '\n'
936	+	txt += 'DatasetPath='+self.datasetPath+'\n'
937	+
938	+	datasetpath_split = self.datasetPath.split("/")
939	+
940	+	txt += 'PrimaryDataset='+datasetpath_split[1]+'\n'
941	+	txt += 'DataTier='+datasetpath_split[2]+'\n'
942	+	txt += 'ApplicationFamily=cmsRun\n'
943	+
944	+	else:
945	+	txt += 'DatasetPath=MCDataTier\n'
946	+	txt += 'PrimaryDataset=null\n'
947	+	txt += 'DataTier=null\n'
948	+	txt += 'ApplicationFamily=MCDataTier\n'
949		if self.pset != None: #CarlosDaniele
950		pset = os.path.basename(job.configFilename())
951		txt += '\n'
952	+	txt += 'cp  $RUNTIME_AREA/'+pset+' .\n'
953		if (self.datasetPath): # standard job
809	–	#txt += 'InputFiles=$2\n'
954		txt += 'InputFiles=${args[1]}\n'
955		txt += 'MaxEvents=${args[2]}\n'
956		txt += 'SkipEvents=${args[3]}\n'
957		txt += 'echo "Inputfiles:<$InputFiles>"\n'
958	<	txt += 'sed "s#{\'INPUT\'}#$InputFiles#" $RUNTIME_AREA/'+pset+' > pset_tmp_1.cfg\n'
958	>	txt += 'sed "s#\'INPUTFILE\'#$InputFiles#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
959		txt += 'echo "MaxEvents:<$MaxEvents>"\n'
960	<	txt += 'sed "s#INPUTMAXEVENTS#$MaxEvents#" pset_tmp_1.cfg > pset_tmp_2.cfg\n'
960	>	txt += 'sed "s#int32 input = 0#int32 input = $MaxEvents#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
961		txt += 'echo "SkipEvents:<$SkipEvents>"\n'
962	<	txt += 'sed "s#INPUTSKIPEVENTS#$SkipEvents#" pset_tmp_2.cfg > pset.cfg\n'
962	>	txt += 'sed "s#uint32 skipEvents = 0#uint32 skipEvents = $SkipEvents#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
963		else:  # pythia like job
964	+	seedIndex=1
965	+	if (self.firstRun):
966	+	txt += 'FirstRun=${args['+str(seedIndex)+']}\n'
967	+	txt += 'echo "FirstRun: <$FirstRun>"\n'
968	+	txt += 'sed "s#uint32 firstRun = 0#uint32 firstRun = $FirstRun#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
969	+	seedIndex=seedIndex+1
970	+
971		if (self.sourceSeed):
972	<	#                    txt += 'Seed=$2\n'
973	<	txt += 'Seed=${args[1]}\n'
974	<	txt += 'echo "Seed: <$Seed>"\n'
975	<	txt += 'sed "s#\<INPUT\>#$Seed#" $RUNTIME_AREA/'+pset+' > tmp.cfg\n'
972	>	txt += 'Seed=${args['+str(seedIndex)+']}\n'
973	>	txt += 'sed "s#uint32 sourceSeed = 0#uint32 sourceSeed = $Seed#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
974	>	seedIndex=seedIndex+1
975	>	## the following seeds are not always present
976		if (self.sourceSeedVtx):
977	<	#                        txt += 'VtxSeed=$3\n'
827	<	txt += 'VtxSeed=${args[2]}\n'
977	>	txt += 'VtxSeed=${args['+str(seedIndex)+']}\n'
978		txt += 'echo "VtxSeed: <$VtxSeed>"\n'
979	<	txt += 'sed "s#INPUTVTX#$VtxSeed#" tmp.cfg > pset.cfg\n'
980	<	else:
981	<	txt += 'mv tmp.cfg pset.cfg\n'
982	<	else:
983	<	txt += '# Copy untouched pset\n'
984	<	txt += 'cp $RUNTIME_AREA/'+pset+' pset.cfg\n'
985	<
979	>	txt += 'sed "s#uint32 VtxSmeared = 0#uint32 VtxSmeared = $VtxSeed#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
980	>	seedIndex += 1
981	>	if (self.sourceSeedG4):
982	>	txt += 'G4Seed=${args['+str(seedIndex)+']}\n'
983	>	txt += 'echo "G4Seed: <$G4Seed>"\n'
984	>	txt += 'sed "s#uint32 g4SimHits = 0#uint32 g4SimHits = $G4Seed#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
985	>	seedIndex += 1
986	>	if (self.sourceSeedMix):
987	>	txt += 'mixSeed=${args['+str(seedIndex)+']}\n'
988	>	txt += 'echo "MixSeed: <$mixSeed>"\n'
989	>	txt += 'sed "s#uint32 mix = 0#uint32 mix = $mixSeed#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
990	>	seedIndex += 1
991	>	pass
992	>	pass
993	>	txt += 'mv -f '+pset+' pset.cfg\n'
994
995		if len(self.additional_inbox_files) > 0:
996	<	for file in self.additional_inbox_files:
997	<	relFile = file.split("/")[-1]
998	<	txt += 'if [ -e $RUNTIME_AREA/'+relFile+' ] ; then\n'
999	<	txt += '   cp $RUNTIME_AREA/'+relFile+' .\n'
842	<	txt += '   chmod +x '+relFile+'\n'
843	<	txt += 'fi\n'
844	<	pass
996	>	txt += 'if [ -e $RUNTIME_AREA/'+self.additional_tgz_name+' ] ; then\n'
997	>	txt += '  tar xzvf $RUNTIME_AREA/'+self.additional_tgz_name+'\n'
998	>	txt += 'fi\n'
999	>	pass
1000
1001		if self.pset != None: #CarlosDaniele
847	–	txt += 'echo "### END JOB SETUP ENVIRONMENT ###"\n\n'
848	–
1002		txt += '\n'
1003		txt += 'echo "***** cat pset.cfg *********"\n'
1004		txt += 'cat pset.cfg\n'
1005		txt += 'echo "****** end pset.cfg ********"\n'
1006		txt += '\n'
1007	<	# txt += 'echo "***** cat pset1.cfg *********"\n'
1008	<	# txt += 'cat pset1.cfg\n'
1009	<	# txt += 'echo "****** end pset1.cfg ********"\n'
1007	>	### FEDE FOR DBS OUTPUT PUBLICATION
1008	>	txt += 'PSETHASH=`EdmConfigHash < pset.cfg` \n'
1009	>	txt += 'echo "PSETHASH = $PSETHASH" \n'
1010	>	##############
1011	>	txt += '\n'
1012		return txt
1013
1014	<	def wsBuildExe(self, nj):
1014	>	def wsBuildExe(self, nj=0):
1015		"""
1016		Put in the script the commands to build an executable
1017		or a library.
#	Line 865 | Line 1020 | class Cmssw(JobType):
1020		txt = ""
1021
1022		if os.path.isfile(self.tgzNameWithPath):
1023	<	txt += 'echo "tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+'"\n'
1023	>	txt += 'echo ">>> tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+' :" \n'
1024		txt += 'tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+'\n'
1025		txt += 'untar_status=$? \n'
1026		txt += 'if [ $untar_status -ne 0 ]; then \n'
#	Line 873 | Line 1028 | class Cmssw(JobType):
1028		txt += '   echo "JOB_EXIT_STATUS = $untar_status" \n'
1029		txt += '   echo "JobExitCode=$untar_status" | tee -a $RUNTIME_AREA/$repo\n'
1030		txt += '   if [ $middleware == OSG ]; then \n'
876	–	txt += '       echo "Remove working directory: $WORKING_DIR"\n'
1031		txt += '       cd $RUNTIME_AREA\n'
1032	+	txt += '        echo ">>> current directory (RUNTIME_AREA): $RUNTIME_AREA"\n'
1033	+	txt += '        echo ">>> Remove working directory: $WORKING_DIR"\n'
1034		txt += '       /bin/rm -rf $WORKING_DIR\n'
1035		txt += '       if [ -d $WORKING_DIR ] ;then\n'
1036		txt += '           echo "SET_EXE 50999 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after Untarring .tgz file failed"\n'
1037		txt += '           echo "JOB_EXIT_STATUS = 50999"\n'
1038		txt += '           echo "JobExitCode=50999" | tee -a $RUNTIME_AREA/$repo\n'
1039		txt += '           dumpStatus $RUNTIME_AREA/$repo\n'
884	–	txt += '           rm -f $RUNTIME_AREA/$repo \n'
885	–	txt += '           echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
886	–	txt += '           echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1040		txt += '       fi\n'
1041		txt += '   fi \n'
1042		txt += '   \n'
#	Line 892 | Line 1045 | class Cmssw(JobType):
1045		txt += '   echo "Successful untar" \n'
1046		txt += 'fi \n'
1047		txt += '\n'
1048	<	txt += 'echo "Include ProdAgentApi in PYTHONPATH"\n'
1048	>	txt += 'echo ">>> Include ProdAgentApi and PRODCOMMON in PYTHONPATH:"\n'
1049		txt += 'if [ -z "$PYTHONPATH" ]; then\n'
1050	<	txt += '   export PYTHONPATH=ProdAgentApi\n'
1050	>	#### FEDE FOR DBS OUTPUT PUBLICATION
1051	>	txt += '   export PYTHONPATH=$SOFTWARE_DIR/ProdAgentApi:$SOFTWARE_DIR/ProdCommon\n'
1052		txt += 'else\n'
1053	<	txt += '   export PYTHONPATH=ProdAgentApi:${PYTHONPATH}\n'
1053	>	txt += '   export PYTHONPATH=$SOFTWARE_DIR/ProdAgentApi:$SOFTWARE_DIR/ProdCommon:${PYTHONPATH}\n'
1054	>	txt += 'echo "PYTHONPATH=$PYTHONPATH"\n'
1055	>	###################
1056		txt += 'fi\n'
1057		txt += '\n'
1058
1059		pass
1060	<
1060	>
1061		return txt
1062
1063		def modifySteeringCards(self, nj):
1064		"""
1065	<	modify the card provided by the user,
1065	>	modify the card provided by the user,
1066		writing a new card into share dir
1067		"""
1068	<
1068	>
1069		def executableName(self):
1070	<	if self.pset == None: #CarlosDaniele
1070	>	if self.scriptExe: #CarlosDaniele
1071		return "sh "
1072		else:
1073		return self.executable
1074
1075		def executableArgs(self):
1076	<	if self.pset == None:#CarlosDaniele
1076	>	if self.scriptExe:#CarlosDaniele
1077		return   self.scriptExe + " $NJob"
1078	<	else:
1079	<	return " -p pset.cfg"
1078	>	else:
1079	>	# if >= CMSSW_1_5_X, add -j crab_fjr.xml
1080	>	version_array = self.scram.getSWVersion().split('_')
1081	>	major = 0
1082	>	minor = 0
1083	>	try:
1084	>	major = int(version_array[1])
1085	>	minor = int(version_array[2])
1086	>	except:
1087	>	msg = "Cannot parse CMSSW version string: " + "_".join(version_array) + " for major and minor release number!"
1088	>	raise CrabException(msg)
1089	>	if major >= 1 and minor >= 5 :
1090	>	return " -j " + self.fjrFileName + " -p pset.cfg"
1091	>	else:
1092	>	return " -p pset.cfg"
1093
1094		def inputSandbox(self, nj):
1095		"""
#	Line 932 | Line 1101 | class Cmssw(JobType):
1101		## code
1102		if os.path.isfile(self.tgzNameWithPath):
1103		inp_box.append(self.tgzNameWithPath)
1104	+	if os.path.isfile(self.MLtgzfile):
1105	+	inp_box.append(self.MLtgzfile)
1106		## config
1107	<	if not self.pset is None: #CarlosDaniele
1108	<	inp_box.append(common.job_list[nj].configFilename())
1107	>	if not self.pset is None:
1108	>	inp_box.append(common.work_space.pathForTgz() + 'job/' + self.configFilename())
1109		## additional input files
1110	<	#for file in self.additional_inbox_files:
1111	<	#    inp_box.append(common.work_space.cwdDir()+file)
1110	>	tgz = self.additionalInputFileTgz()
1111	>	inp_box.append(tgz)
1112		return inp_box
1113
1114		def outputSandbox(self, nj):
#	Line 948 | Line 1119 | class Cmssw(JobType):
1119
1120		## User Declared output files
1121		for out in (self.output_file+self.output_file_sandbox):
1122	<	n_out = nj + 1
1122	>	n_out = nj + 1
1123		out_box.append(self.numberFile_(out,str(n_out)))
1124		return out_box
1125
#	Line 964 | Line 1135 | class Cmssw(JobType):
1135		"""
1136
1137		txt = '\n'
1138	<	txt += '# directory content\n'
1138	>	txt += 'echo" >>> directory content:"\n'
1139		txt += 'ls \n'
1140	+	txt = '\n'
1141	+
1142	+	txt += 'output_exit_status=0\n'
1143
1144	<	for fileWithSuffix in (self.output_file+self.output_file_sandbox):
1144	>	for fileWithSuffix in (self.output_file_sandbox):
1145		output_file_num = self.numberFile_(fileWithSuffix, '$NJob')
1146		txt += '\n'
1147		txt += '# check output file\n'
1148	<	txt += 'ls '+fileWithSuffix+'\n'
1149	<	txt += 'ls_result=$?\n'
1150	<	txt += 'if [ $ls_result -ne 0 ] ; then\n'
1151	<	txt += '   echo "ERROR: Problem with output file"\n'
1148	>	txt += 'if [ -e ./'+fileWithSuffix+' ] ; then\n'
1149	>	txt += '    mv '+fileWithSuffix+' $RUNTIME_AREA\n'
1150	>	txt += '    cp $RUNTIME_AREA/'+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n'
1151	>	txt += 'else\n'
1152	>	txt += '    exit_status=60302\n'
1153	>	txt += '    echo "ERROR: Problem with output file '+fileWithSuffix+'"\n'
1154		if common.scheduler.boss_scheduler_name == 'condor_g':
1155		txt += '    if [ $middleware == OSG ]; then \n'
1156		txt += '        echo "prepare dummy output file"\n'
1157		txt += '        echo "Processing of job output failed" > $RUNTIME_AREA/'+output_file_num+'\n'
1158		txt += '    fi \n'
983	–	txt += 'else\n'
984	–	txt += '   cp '+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n'
1159		txt += 'fi\n'
986	–
987	–	txt += 'cd $RUNTIME_AREA\n'
988	–	txt += 'cd $RUNTIME_AREA\n'
989	–	### OLI_DANIELE
990	–	txt += 'if [ $middleware == OSG ]; then\n'
991	–	txt += '    cd $RUNTIME_AREA\n'
992	–	txt += '    echo "Remove working directory: $WORKING_DIR"\n'
993	–	txt += '    /bin/rm -rf $WORKING_DIR\n'
994	–	txt += '    if [ -d $WORKING_DIR ] ;then\n'
995	–	txt += '        echo "SET_EXE 60999 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after cleanup of WN"\n'
996	–	txt += '        echo "JOB_EXIT_STATUS = 60999"\n'
997	–	txt += '        echo "JobExitCode=60999" | tee -a $RUNTIME_AREA/$repo\n'
998	–	txt += '        dumpStatus $RUNTIME_AREA/$repo\n'
999	–	txt += '        rm -f $RUNTIME_AREA/$repo \n'
1000	–	txt += '        echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1001	–	txt += '        echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1002	–	txt += '    fi\n'
1003	–	txt += 'fi\n'
1004	–	txt += '\n'
1160
1161	<	file_list = ''
1007	<	## Add to filelist only files to be possibly copied to SE
1008	<	for fileWithSuffix in self.output_file:
1161	>	for fileWithSuffix in (self.output_file):
1162		output_file_num = self.numberFile_(fileWithSuffix, '$NJob')
1163	<	file_list=file_list+output_file_num+' '
1164	<	file_list=file_list[:-1]
1165	<	txt += 'file_list="'+file_list+'"\n'
1163	>	txt += '\n'
1164	>	txt += '# check output file\n'
1165	>	txt += 'if [ -e ./'+fileWithSuffix+' ] ; then\n'
1166	>	txt += '    mv '+fileWithSuffix+' $RUNTIME_AREA\n'
1167	>	txt += '    cp $RUNTIME_AREA/'+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n'
1168	>	txt += 'else\n'
1169	>	txt += '    exit_status=60302\n'
1170	>	txt += '    echo "ERROR: Problem with output file '+fileWithSuffix+'"\n'
1171	>	txt += '    echo "JOB_EXIT_STATUS = $exit_status"\n'
1172	>	txt += '    output_exit_status=$exit_status\n'
1173	>	if common.scheduler.boss_scheduler_name == 'condor_g':
1174	>	txt += '    if [ $middleware == OSG ]; then \n'
1175	>	txt += '        echo "prepare dummy output file"\n'
1176	>	txt += '        echo "Processing of job output failed" > $RUNTIME_AREA/'+output_file_num+'\n'
1177	>	txt += '    fi \n'
1178	>	txt += 'fi\n'
1179	>	file_list = []
1180	>	for fileWithSuffix in (self.output_file):
1181	>	file_list.append(self.numberFile_(fileWithSuffix, '$NJob'))
1182
1183	+	txt += 'file_list="'+string.join(file_list,' ')+'"\n'
1184	+	txt += 'cd $RUNTIME_AREA\n'
1185	+	txt += 'echo ">>> current directory (RUNTIME_AREA):  $RUNTIME_AREA"\n'
1186		return txt
1187
1188		def numberFile_(self, file, txt):
#	Line 1021 | Line 1193 | class Cmssw(JobType):
1193		# take away last extension
1194		name = p[0]
1195		for x in p[1:-1]:
1196	<	name=name+"."+x
1196	>	name=name+"."+x
1197		# add "_txt"
1198		if len(p)>1:
1199	<	ext = p[len(p)-1]
1200	<	result = name + '_' + txt + "." + ext
1199	>	ext = p[len(p)-1]
1200	>	result = name + '_' + txt + "." + ext
1201		else:
1202	<	result = name + '_' + txt
1203	<
1202	>	result = name + '_' + txt
1203	>
1204		return result
1205
1206	<	def getRequirements(self):
1206	>	def getRequirements(self, nj=[]):
1207		"""
1208	<	return job requirements to add to jdl files
1208	>	return job requirements to add to jdl files
1209		"""
1210		req = ''
1211		if self.version:
1212		req='Member("VO-cms-' + \
1213		self.version + \
1214		'", other.GlueHostApplicationSoftwareRunTimeEnvironment)'
1215	+	## SL add requirement for OS version only if SL4
1216	+	#reSL4 = re.compile( r'slc4' )
1217	+	if self.executable_arch: # and reSL4.search(self.executable_arch):
1218	+	req+=' && Member("VO-cms-' + \
1219	+	self.executable_arch + \
1220	+	'", other.GlueHostApplicationSoftwareRunTimeEnvironment)'
1221
1222		req = req + ' && (other.GlueHostNetworkAdapterOutboundIP)'
1223
#	Line 1049 | Line 1227 | class Cmssw(JobType):
1227		""" return the config filename """
1228		return self.name()+'.cfg'
1229
1052	–	### OLI_DANIELE
1230		def wsSetupCMSOSGEnvironment_(self):
1231		"""
1232		Returns part of a job script which is prepares
1233		the execution environment and which is common for all CMS jobs.
1234		"""
1235	<	txt = '\n'
1236	<	txt += '   echo "### SETUP CMS OSG  ENVIRONMENT ###"\n'
1237	<	txt += '   if [ -f $GRID3_APP_DIR/cmssoft/cmsset_default.sh ] ;then\n'
1238	<	txt += '      # Use $GRID3_APP_DIR/cmssoft/cmsset_default.sh to setup cms software\n'
1239	<	txt += '       source $GRID3_APP_DIR/cmssoft/cmsset_default.sh '+self.version+'\n'
1063	<	txt += '   elif [ -f $OSG_APP/cmssoft/cms/cmsset_default.sh ] ;then\n'
1235	>	txt = '    echo ">>> setup CMS OSG environment:"\n'
1236	>	txt += '    echo "set SCRAM ARCH to ' + self.executable_arch + '"\n'
1237	>	txt += '    export SCRAM_ARCH='+self.executable_arch+'\n'
1238	>	txt += '    echo "SCRAM_ARCH = $SCRAM_ARCH"\n'
1239	>	txt += '    if [ -f $OSG_APP/cmssoft/cms/cmsset_default.sh ] ;then\n'
1240		txt += '      # Use $OSG_APP/cmssoft/cms/cmsset_default.sh to setup cms software\n'
1241	<	txt += '       source $OSG_APP/cmssoft/cms/cmsset_default.sh '+self.version+'\n'
1242	<	txt += '   else\n'
1243	<	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'
1244	<	txt += '       echo "JOB_EXIT_STATUS = 10020"\n'
1245	<	txt += '       echo "JobExitCode=10020" | tee -a $RUNTIME_AREA/$repo\n'
1246	<	txt += '       dumpStatus $RUNTIME_AREA/$repo\n'
1071	<	txt += '       rm -f $RUNTIME_AREA/$repo \n'
1072	<	txt += '       echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1073	<	txt += '       echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1074	<	txt += '       exit 1\n'
1241	>	txt += '        source $OSG_APP/cmssoft/cms/cmsset_default.sh '+self.version+'\n'
1242	>	txt += '    else\n'
1243	>	txt += '        echo "SET_CMS_ENV 10020 ==> ERROR $OSG_APP/cmssoft/cms/cmsset_default.sh file not found"\n'
1244	>	txt += '        echo "JOB_EXIT_STATUS = 10020"\n'
1245	>	txt += '        echo "JobExitCode=10020" | tee -a $RUNTIME_AREA/$repo\n'
1246	>	txt += '        dumpStatus $RUNTIME_AREA/$repo\n'
1247		txt += '\n'
1248	<	txt += '       echo "Remove working directory: $WORKING_DIR"\n'
1249	<	txt += '       cd $RUNTIME_AREA\n'
1250	<	txt += '       /bin/rm -rf $WORKING_DIR\n'
1251	<	txt += '       if [ -d $WORKING_DIR ] ;then\n'
1252	<	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'
1253	<	txt += '            echo "JOB_EXIT_STATUS = 10017"\n'
1254	<	txt += '            echo "JobExitCode=10017" | tee -a $RUNTIME_AREA/$repo\n'
1255	<	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
1256	<	txt += '            rm -f $RUNTIME_AREA/$repo \n'
1257	<	txt += '            echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1086	<	txt += '            echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1087	<	txt += '       fi\n'
1248	>	txt += '        cd $RUNTIME_AREA\n'
1249	>	txt += '        echo ">>> current directory (RUNTIME_AREA): $RUNTIME_AREA"\n'
1250	>	txt += '        echo ">>> Remove working directory: $WORKING_DIR"\n'
1251	>	txt += '        /bin/rm -rf $WORKING_DIR\n'
1252	>	txt += '        if [ -d $WORKING_DIR ] ;then\n'
1253	>	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'
1254	>	txt += '            echo "JOB_EXIT_STATUS = 10017"\n'
1255	>	txt += '            echo "JobExitCode=10017" | tee -a $RUNTIME_AREA/$repo\n'
1256	>	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
1257	>	txt += '        fi\n'
1258		txt += '\n'
1259	<	txt += '       exit 1\n'
1260	<	txt += '   fi\n'
1259	>	txt += '        exit 1\n'
1260	>	txt += '    fi\n'
1261		txt += '\n'
1262	<	txt += '   echo "SET_CMS_ENV 0 ==> setup cms environment ok"\n'
1263	<	txt += '   echo " END SETUP CMS OSG  ENVIRONMENT "\n'
1262	>	txt += '    echo "SET_CMS_ENV 0 ==> setup cms environment ok"\n'
1263	>	txt += '    echo "SCRAM_ARCH = $SCRAM_ARCH"\n'
1264
1265		return txt
1266	<
1266	>
1267		### OLI_DANIELE
1268		def wsSetupCMSLCGEnvironment_(self):
1269		"""
1270		Returns part of a job script which is prepares
1271		the execution environment and which is common for all CMS jobs.
1272		"""
1273	<	txt  = '   \n'
1274	<	txt += '   echo " ### SETUP CMS LCG  ENVIRONMENT ### "\n'
1275	<	txt += '   if [ ! $VO_CMS_SW_DIR ] ;then\n'
1276	<	txt += '       echo "SET_CMS_ENV 10031 ==> ERROR CMS software dir not found on WN `hostname`"\n'
1277	<	txt += '       echo "JOB_EXIT_STATUS = 10031" \n'
1278	<	txt += '       echo "JobExitCode=10031" | tee -a $RUNTIME_AREA/$repo\n'
1279	<	txt += '       dumpStatus $RUNTIME_AREA/$repo\n'
1280	<	txt += '       rm -f $RUNTIME_AREA/$repo \n'
1281	<	txt += '       echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1282	<	txt += '       echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1283	<	txt += '       exit 1\n'
1284	<	txt += '   else\n'
1285	<	txt += '       echo "Sourcing environment... "\n'
1286	<	txt += '       if [ ! -s $VO_CMS_SW_DIR/cmsset_default.sh ] ;then\n'
1287	<	txt += '           echo "SET_CMS_ENV 10020 ==> ERROR cmsset_default.sh file not found into dir $VO_CMS_SW_DIR"\n'
1288	<	txt += '           echo "JOB_EXIT_STATUS = 10020"\n'
1289	<	txt += '           echo "JobExitCode=10020" | tee -a $RUNTIME_AREA/$repo\n'
1290	<	txt += '           dumpStatus $RUNTIME_AREA/$repo\n'
1291	<	txt += '           rm -f $RUNTIME_AREA/$repo \n'
1292	<	txt += '           echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1293	<	txt += '           echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1294	<	txt += '           exit 1\n'
1295	<	txt += '       fi\n'
1296	<	txt += '       echo "sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
1297	<	txt += '       source $VO_CMS_SW_DIR/cmsset_default.sh\n'
1298	<	txt += '       result=$?\n'
1299	<	txt += '       if [ $result -ne 0 ]; then\n'
1300	<	txt += '           echo "SET_CMS_ENV 10032 ==> ERROR problem sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
1301	<	txt += '           echo "JOB_EXIT_STATUS = 10032"\n'
1302	<	txt += '           echo "JobExitCode=10032" | tee -a $RUNTIME_AREA/$repo\n'
1303	<	txt += '           dumpStatus $RUNTIME_AREA/$repo\n'
1304	<	txt += '           rm -f $RUNTIME_AREA/$repo \n'
1305	<	txt += '           echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1306	<	txt += '           echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1307	<	txt += '           exit 1\n'
1308	<	txt += '       fi\n'
1309	<	txt += '   fi\n'
1310	<	txt += '   \n'
1311	<	txt += '   string=`cat /etc/redhat-release`\n'
1312	<	txt += '   echo $string\n'
1313	<	txt += '   if [[ $string = *alhalla* ]]; then\n'
1314	<	txt += '       echo "SCRAM_ARCH= $SCRAM_ARCH"\n'
1315	<	txt += '   elif [[ $string = *Enterprise* ]] || [[ $string = *cientific* ]]; then\n'
1316	<	txt += '       export SCRAM_ARCH=slc3_ia32_gcc323\n'
1317	<	txt += '       echo "SCRAM_ARCH= $SCRAM_ARCH"\n'
1318	<	txt += '   else\n'
1319	<	txt += '       echo "SET_CMS_ENV 10033 ==> ERROR OS unknown, LCG environment not initialized"\n'
1320	<	txt += '       echo "JOB_EXIT_STATUS = 10033"\n'
1321	<	txt += '       echo "JobExitCode=10033" | tee -a $RUNTIME_AREA/$repo\n'
1322	<	txt += '       dumpStatus $RUNTIME_AREA/$repo\n'
1323	<	txt += '       rm -f $RUNTIME_AREA/$repo \n'
1324	<	txt += '       echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1325	<	txt += '       echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1326	<	txt += '       exit 1\n'
1327	<	txt += '   fi\n'
1328	<	txt += '   echo "SET_CMS_ENV 0 ==> setup cms environment ok"\n'
1329	<	txt += '   echo "### END SETUP CMS LCG ENVIRONMENT ###"\n'
1273	>	txt = '    echo ">>> setup CMS LCG environment:"\n'
1274	>	txt += '    echo "set SCRAM ARCH and BUILD_ARCH to ' + self.executable_arch + ' ###"\n'
1275	>	txt += '    export SCRAM_ARCH='+self.executable_arch+'\n'
1276	>	txt += '    export BUILD_ARCH='+self.executable_arch+'\n'
1277	>	txt += '    if [ ! $VO_CMS_SW_DIR ] ;then\n'
1278	>	txt += '        echo "SET_CMS_ENV 10031 ==> ERROR CMS software dir not found on WN `hostname`"\n'
1279	>	txt += '        echo "JOB_EXIT_STATUS = 10031" \n'
1280	>	txt += '        echo "JobExitCode=10031" | tee -a $RUNTIME_AREA/$repo\n'
1281	>	txt += '        dumpStatus $RUNTIME_AREA/$repo\n'
1282	>	txt += '        exit 1\n'
1283	>	txt += '    else\n'
1284	>	txt += '        echo "Sourcing environment... "\n'
1285	>	txt += '        if [ ! -s $VO_CMS_SW_DIR/cmsset_default.sh ] ;then\n'
1286	>	txt += '            echo "SET_CMS_ENV 10020 ==> ERROR cmsset_default.sh file not found into dir $VO_CMS_SW_DIR"\n'
1287	>	txt += '            echo "JOB_EXIT_STATUS = 10020"\n'
1288	>	txt += '            echo "JobExitCode=10020" | tee -a $RUNTIME_AREA/$repo\n'
1289	>	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
1290	>	txt += '            exit 1\n'
1291	>	txt += '        fi\n'
1292	>	txt += '        echo "sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
1293	>	txt += '        source $VO_CMS_SW_DIR/cmsset_default.sh\n'
1294	>	txt += '        result=$?\n'
1295	>	txt += '        if [ $result -ne 0 ]; then\n'
1296	>	txt += '            echo "SET_CMS_ENV 10032 ==> ERROR problem sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
1297	>	txt += '            echo "JOB_EXIT_STATUS = 10032"\n'
1298	>	txt += '            echo "JobExitCode=10032" | tee -a $RUNTIME_AREA/$repo\n'
1299	>	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
1300	>	txt += '            exit 1\n'
1301	>	txt += '        fi\n'
1302	>	txt += '    fi\n'
1303	>	txt += '    \n'
1304	>	txt += '    echo "SET_CMS_ENV 0 ==> setup cms environment ok"\n'
1305	>	return txt
1306	>
1307	>	### FEDE FOR DBS OUTPUT PUBLICATION
1308	>	def modifyReport(self, nj):
1309	>	"""
1310	>	insert the part of the script that modifies the FrameworkJob Report
1311	>	"""
1312	>
1313	>	txt = ''
1314	>	try:
1315	>	publish_data = int(self.cfg_params['USER.publish_data'])
1316	>	except KeyError:
1317	>	publish_data = 0
1318	>	if (publish_data == 1):
1319	>	txt += 'echo ">>> Modify Job Report:" \n'
1320	>	################ FEDE FOR DBS2 #############################################
1321	>	txt += 'chmod a+x $SOFTWARE_DIR/ProdAgentApi/FwkJobRep/ModifyJobReport.py\n'
1322	>	#############################################################################
1323	>
1324	>	txt += 'if [ -z "$SE" ]; then\n'
1325	>	txt += '    SE="" \n'
1326	>	txt += 'fi \n'
1327	>	txt += 'if [ -z "$SE_PATH" ]; then\n'
1328	>	txt += '    SE_PATH="" \n'
1329	>	txt += 'fi \n'
1330	>	txt += 'echo "SE = $SE"\n'
1331	>	txt += 'echo "SE_PATH = $SE_PATH"\n'
1332	>
1333	>	processedDataset = self.cfg_params['USER.publish_data_name']
1334	>	txt += 'ProcessedDataset='+processedDataset+'\n'
1335	>	#### LFN=/store/user/<user>/processedDataset_PSETHASH
1336	>	txt += 'if [ "$SE_PATH" == "" ]; then\n'
1337	>	#### FEDE: added slash in LFN ##############
1338	>	txt += '    FOR_LFN=/copy_problems/ \n'
1339	>	txt += 'else \n'
1340	>	txt += '    tmp=`echo $SE_PATH | awk -F \'store\' \'{print$2}\'` \n'
1341	>	#####  FEDE TO BE CHANGED, BECAUSE STORE IS HARDCODED!!!! ########
1342	>	txt += '    FOR_LFN=/store$tmp \n'
1343	>	txt += 'fi \n'
1344	>	txt += 'echo "ProcessedDataset = $ProcessedDataset"\n'
1345	>	txt += 'echo "FOR_LFN = $FOR_LFN" \n'
1346	>	txt += 'echo "CMSSW_VERSION = $CMSSW_VERSION"\n\n'
1347	>	#txt += 'echo "$RUNTIME_AREA/'+self.version+'/ProdAgentApi/FwkJobRep/ModifyJobReport.py crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH"\n'
1348	>	txt += 'echo "$SOFTWARE_DIR/ProdAgentApi/FwkJobRep/ModifyJobReport.py crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH"\n'
1349	>	txt += '$SOFTWARE_DIR/ProdAgentApi/FwkJobRep/ModifyJobReport.py crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH\n'
1350	>	#txt += '$RUNTIME_AREA/'+self.version+'/ProdAgentApi/FwkJobRep/ModifyJobReport.py crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH\n'
1351	>
1352	>	txt += 'modifyReport_result=$?\n'
1353	>	txt += 'echo modifyReport_result = $modifyReport_result\n'
1354	>	txt += 'if [ $modifyReport_result -ne 0 ]; then\n'
1355	>	txt += '    exit_status=1\n'
1356	>	txt += '    echo "ERROR: Problem with ModifyJobReport"\n'
1357	>	txt += 'else\n'
1358	>	txt += '    mv NewFrameworkJobReport.xml crab_fjr_$NJob.xml\n'
1359	>	txt += 'fi\n'
1360	>	else:
1361	>	txt += 'echo "no data publication required"\n'
1362	>	return txt
1363	>
1364	>	def cleanEnv(self):
1365	>	txt = ''
1366	>	txt += 'if [ $middleware == OSG ]; then\n'
1367	>	txt += '    cd $RUNTIME_AREA\n'
1368	>	txt += '    echo ">>> current directory (RUNTIME_AREA): $RUNTIME_AREA"\n'
1369	>	txt += '    echo ">>> Remove working directory: $WORKING_DIR"\n'
1370	>	txt += '    /bin/rm -rf $WORKING_DIR\n'
1371	>	txt += '    if [ -d $WORKING_DIR ] ;then\n'
1372	>	txt += '        echo "SET_EXE 60999 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after cleanup of WN"\n'
1373	>	txt += '        echo "JOB_EXIT_STATUS = 60999"\n'
1374	>	txt += '        echo "JobExitCode=60999" | tee -a $RUNTIME_AREA/$repo\n'
1375	>	txt += '        dumpStatus $RUNTIME_AREA/$repo\n'
1376	>	txt += '    fi\n'
1377	>	txt += 'fi\n'
1378	>	txt += '\n'
1379		return txt
1380
1381		def setParam_(self, param, value):
#	Line 1167 | Line 1386 | class Cmssw(JobType):
1386
1387		def setTaskid_(self):
1388		self._taskId = self.cfg_params['taskId']
1389	<
1389	>
1390		def getTaskid(self):
1391		return self._taskId
1392
1174	–	#######################################################################
1393		def uniquelist(self, old):
1394		"""
1395		remove duplicates from a list
#	Line 1180 | Line 1398 | class Cmssw(JobType):
1398		for e in old:
1399		nd[e]=0
1400		return nd.keys()
1401	+
1402	+
1403	+	def checkOut(self, limit):
1404	+	"""
1405	+	check the dimension of the output files
1406	+	"""
1407	+	txt += 'echo ">>> Starting output sandbox limit check :"\n'
1408	+	allOutFiles = ""
1409	+	listOutFiles = []
1410	+	for fileOut in (self.output_file+self.output_file_sandbox):
1411	+	if fileOut.find('crab_fjr') == -1:
1412	+	allOutFiles = allOutFiles + " " + self.numberFile_(fileOut, '$NJob')
1413	+	listOutFiles.append(self.numberFile_(fileOut, '$NJob'))
1414	+	txt += 'echo "OUTPUT files: '+str(allOutFiles)+'";\n'
1415	+	txt += 'ls -gGhrta;\n'
1416	+	txt += 'sum=0;\n'
1417	+	txt += 'for file in '+str(allOutFiles)+' ; do\n'
1418	+	txt += '    if [ -e $file ]; then\n'
1419	+	txt += '        tt=`ls -gGrta $file | awk \'{ print $3 }\'`\n'
1420	+	txt += '        sum=`expr $sum + $tt`\n'
1421	+	txt += '    else\n'
1422	+	txt += '        echo "WARNING: output file $file not found!"\n'
1423	+	txt += '    fi\n'
1424	+	txt += 'done\n'
1425	+	txt += 'echo "Total Output dimension: $sum";\n'
1426	+	txt += 'limit='+str(limit)+';\n'
1427	+	txt += 'echo "OUTPUT FILES LIMIT SET TO: $limit";\n'
1428	+	txt += 'if [ $limit -lt $sum ]; then\n'
1429	+	txt += '    echo "WARNING: output files have to big size - something will be lost;"\n'
1430	+	txt += '    echo "         checking the output file sizes..."\n'
1431	+	"""
1432	+	txt += '    dim=0;\n'
1433	+	txt += '    exclude=0;\n'
1434	+	txt += '    for files in '+str(allOutFiles)+' ; do\n'
1435	+	txt += '        sumTemp=0;\n'
1436	+	txt += '        for file2 in '+str(allOutFiles)+' ; do\n'
1437	+	txt += '            if [ $file != $file2 ]; then\n'
1438	+	txt += '                tt=`ls -gGrta $file2 | awk \'{ print $3 }\';`\n'
1439	+	txt += '                sumTemp=`expr $sumTemp + $tt`;\n'
1440	+	txt += '            fi\n'
1441	+	txt += '        done\n'
1442	+	txt += '        if [ $sumTemp -lt $limit ]; then\n'
1443	+	txt += '            if [ $dim -lt $sumTemp ]; then\n'
1444	+	txt += '                dim=$sumTemp;\n'
1445	+	txt += '                exclude=$file;\n'
1446	+	txt += '            fi\n'
1447	+	txt += '        fi\n'
1448	+	txt += '    done\n'
1449	+	txt += '    echo "Dimension calculated: $dim"; echo "File to exclude: $exclude";\n'
1450	+	"""
1451	+	txt += '    tot=0;\n'
1452	+	txt += '    for file2 in '+str(allOutFiles)+' ; do\n'
1453	+	txt += '        tt=`ls -gGrta $file2 | awk \'{ print $3 }\';`\n'
1454	+	txt += '        tot=`expr $tot + $tt`;\n'
1455	+	txt += '        if [ $limit -lt $tot ]; then\n'
1456	+	txt += '            tot=`expr $tot - $tt`;\n'
1457	+	txt += '            fileLast=$file;\n'
1458	+	txt += '            break;\n'
1459	+	txt += '        fi\n'
1460	+	txt += '    done\n'
1461	+	txt += '    echo "Dimension calculated: $tot"; echo "First file to exclude: $file";\n'
1462	+	txt += '    flag=0;\n'
1463	+	txt += '    for filess in '+str(allOutFiles)+' ; do\n'
1464	+	txt += '        if [ $fileLast = $filess ]; then\n'
1465	+	txt += '            flag=1;\n'
1466	+	txt += '        fi\n'
1467	+	txt += '        if [ $flag -eq 1 ]; then\n'
1468	+	txt += '            rm -f $filess;\n'
1469	+	txt += '        fi\n'
1470	+	txt += '    done\n'
1471	+	txt += '    ls -agGhrt;\n'
1472	+	txt += '    echo "WARNING: output files are too big in dimension: can not put in the output_sandbox.";\n'
1473	+	txt += '    echo "JOB_EXIT_STATUS = 70000";\n'
1474	+	txt += '    exit_status=70000;\n'
1475	+	txt += 'else'
1476	+	txt += '    echo "Total Output dimension $sum is fine.";\n'
1477	+	txt += 'fi\n'
1478	+	txt += 'echo "Ending output sandbox limit check"\n'
1479	+	return txt

Diff Legend

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