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 1.107 by gutsche, Tue Jul 24 19:49:35 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		import common
7	–	import PsetManipulator
8	–
9	–	import DBSInfo
10	–	import DataDiscovery
11	–	import DataLocation
6		import Scram
7
8	<	import glob, os, string, re
8	>	import os, string, glob
9
10		class Cmssw(JobType):
11		def __init__(self, cfg_params, ncjobs):
12		JobType.__init__(self, 'CMSSW')
13		common.logger.debug(3,'CMSSW::__init__')
14
21	–	# Marco.
15		self._params = {}
16		self.cfg_params = cfg_params
17
18	+	try:
19	+	self.MaxTarBallSize = float(self.cfg_params['EDG.maxtarballsize'])
20	+	except KeyError:
21	+	self.MaxTarBallSize = 9.5
22	+
23		# number of jobs requested to be created, limit obj splitting
24		self.ncjobs = ncjobs
25
26		log = common.logger
27
28		self.scram = Scram.Scram(cfg_params)
31	–	scramArea = ''
29		self.additional_inbox_files = []
30		self.scriptExe = ''
31		self.executable = ''
32	+	self.executable_arch = self.scram.getArch()
33		self.tgz_name = 'default.tgz'
34	+	self.additional_tgz_name = 'additional.tgz'
35	+	self.scriptName = 'CMSSW.sh'
36		self.pset = ''      #scrip use case Da
37		self.datasetPath = '' #scrip use case Da
38
#	Line 44 | Line 44 | class Cmssw(JobType):
44		self.setParam_('application', self.version)
45
46		### collect Data cards
47	+
48	+	## get DBS mode
49	+	try:
50	+	self.use_dbs_1 = int(self.cfg_params['CMSSW.use_dbs_1'])
51	+	except KeyError:
52	+	self.use_dbs_1 = 0
53	+
54		try:
55		tmp =  cfg_params['CMSSW.datasetpath']
56		log.debug(6, "CMSSW::CMSSW(): datasetPath = "+tmp)
#	Line 63 | Line 70 | class Cmssw(JobType):
70		self.setParam_('dataset', 'None')
71		self.setParam_('owner', 'None')
72		else:
73	<	datasetpath_split = self.datasetPath.split("/")
74	<	self.setParam_('dataset', datasetpath_split[1])
75	<	self.setParam_('owner', datasetpath_split[-1])
76	<
73	>	try:
74	>	datasetpath_split = self.datasetPath.split("/")
75	>	# standard style
76	>	if self.use_dbs_1 == 1 :
77	>	self.setParam_('dataset', datasetpath_split[1])
78	>	self.setParam_('owner', datasetpath_split[-1])
79	>	else:
80	>	self.setParam_('dataset', datasetpath_split[1])
81	>	self.setParam_('owner', datasetpath_split[2])
82	>	except:
83	>	self.setParam_('dataset', self.datasetPath)
84	>	self.setParam_('owner', self.datasetPath)
85	>
86		self.setTaskid_()
87		self.setParam_('taskId', self.cfg_params['taskId'])
88
#	Line 107 | Line 123 | class Cmssw(JobType):
123		# other output files to be returned via sandbox or copied to SE
124		try:
125		self.output_file = []
110	–
111	–
126		tmp = cfg_params['CMSSW.output_file']
127		if tmp != '':
128		tmpOutFiles = string.split(cfg_params['CMSSW.output_file'],',')
#	Line 118 | Line 132 | class Cmssw(JobType):
132		self.output_file.append(tmp)
133		pass
134		else:
135	<	log.message("No output file defined: only stdout/err and the CRAB Framework Job Report will be available")
135	>	log.message("No output file defined: only stdout/err and the CRAB Framework Job Report will be available\n")
136		pass
137		pass
138		except KeyError:
139	<	log.message("No output file defined: only stdout/err and the CRAB Framework Job Report will be available")
139	>	log.message("No output file defined: only stdout/err and the CRAB Framework Job Report will be available\n")
140		pass
141
142		# script_exe file as additional file in inputSandbox
#	Line 130 | Line 144 | class Cmssw(JobType):
144		self.scriptExe = cfg_params['USER.script_exe']
145		if self.scriptExe != '':
146		if not os.path.isfile(self.scriptExe):
147	<	msg ="WARNING. file "+self.scriptExe+" not found"
147	>	msg ="ERROR. file "+self.scriptExe+" not found"
148		raise CrabException(msg)
149		self.additional_inbox_files.append(string.strip(self.scriptExe))
150		except KeyError:
151		self.scriptExe = ''
152	+
153		#CarlosDaniele
154		if self.datasetPath == None and self.pset == None and self.scriptExe == '' :
155	<	msg ="WARNING. script_exe  not defined"
155	>	msg ="Error. script_exe  not defined"
156		raise CrabException(msg)
157
158		## additional input files
#	Line 147 | Line 162 | class Cmssw(JobType):
162		tmp = string.strip(tmp)
163		dirname = ''
164		if not tmp[0]=="/": dirname = "."
165	<	files = glob.glob(os.path.join(dirname, tmp))
165	>	files = []
166	>	if string.find(tmp,"*")>-1:
167	>	files = glob.glob(os.path.join(dirname, tmp))
168	>	if len(files)==0:
169	>	raise CrabException("No additional input file found with this pattern: "+tmp)
170	>	else:
171	>	files.append(tmp)
172		for file in files:
173		if not os.path.exists(file):
174		raise CrabException("Additional input file not found: "+file)
175		pass
176	+	# fname = string.split(file, '/')[-1]
177	+	# storedFile = common.work_space.pathForTgz()+'share/'+fname
178	+	# shutil.copyfile(file, storedFile)
179		self.additional_inbox_files.append(string.strip(file))
180		pass
181		pass
#	Line 210 | Line 234 | class Cmssw(JobType):
234		except KeyError:
235		self.sourceSeedVtx = None
236		common.logger.debug(5,"No vertex seed given")
237	+
238	+	try:
239	+	self.sourceSeedG4 = int(cfg_params['CMSSW.g4_seed'])
240	+	except KeyError:
241	+	self.sourceSeedG4 = None
242	+	common.logger.debug(5,"No g4 sim hits seed given")
243	+
244	+	try:
245	+	self.sourceSeedMix = int(cfg_params['CMSSW.mix_seed'])
246	+	except KeyError:
247	+	self.sourceSeedMix = None
248	+	common.logger.debug(5,"No mix seed given")
249	+
250	+	try:
251	+	self.firstRun = int(cfg_params['CMSSW.first_run'])
252	+	except KeyError:
253	+	self.firstRun = None
254	+	common.logger.debug(5,"No first run given")
255		if self.pset != None: #CarlosDaniele
256	<	self.PsetEdit = PsetManipulator.PsetManipulator(self.pset) #Daniele Pset
256	>	ver = string.split(self.version,"_")
257	>	if (int(ver[1])>=1 and int(ver[2])>=5):
258	>	import PsetManipulator150 as pp
259	>	else:
260	>	import PsetManipulator as pp
261	>	PsetEdit = pp.PsetManipulator(self.pset) #Daniele Pset
262
263		#DBSDLS-start
264		## Initialize the variables that are extracted from DBS/DLS and needed in other places of the code
#	Line 233 | Line 280 | class Cmssw(JobType):
280		self.jobSplittingForScript()
281		else:
282		self.jobSplittingNoInput()
283	<	else: self.jobSplittingByBlocks(blockSites)
283	>	else:
284	>	self.jobSplittingByBlocks(blockSites)
285
286		# modify Pset
287		if self.pset != None: #CarlosDaniele
288		try:
289		if (self.datasetPath): # standard job
290		# allow to processa a fraction of events in a file
291	<	self.PsetEdit.inputModule("INPUT")
292	<	self.PsetEdit.maxEvent("INPUTMAXEVENTS")
293	<	self.PsetEdit.skipEvent("INPUTSKIPEVENTS")
291	>	PsetEdit.inputModule("INPUT")
292	>	PsetEdit.maxEvent("INPUTMAXEVENTS")
293	>	PsetEdit.skipEvent("INPUTSKIPEVENTS")
294		else:  # pythia like job
295	<	self.PsetEdit.maxEvent(self.eventsPerJob)
295	>	PsetEdit.maxEvent(self.eventsPerJob)
296	>	if (self.firstRun):
297	>	PsetEdit.pythiaFirstRun("INPUTFIRSTRUN")  #First Run
298		if (self.sourceSeed) :
299	<	self.PsetEdit.pythiaSeed("INPUT")
299	>	PsetEdit.pythiaSeed("INPUT")
300		if (self.sourceSeedVtx) :
301	<	self.PsetEdit.pythiaSeedVtx("INPUTVTX")
301	>	PsetEdit.vtxSeed("INPUTVTX")
302	>	if (self.sourceSeedG4) :
303	>	self.PsetEdit.g4Seed("INPUTG4")
304	>	if (self.sourceSeedMix) :
305	>	self.PsetEdit.mixSeed("INPUTMIX")
306		# add FrameworkJobReport to parameter-set
307	<	self.PsetEdit.addCrabFJR(self.fjrFileName)
308	<	self.PsetEdit.psetWriter(self.configFilename())
307	>	PsetEdit.addCrabFJR(self.fjrFileName)
308	>	PsetEdit.psetWriter(self.configFilename())
309		except:
310		msg='Error while manipuliating ParameterSet: exiting...'
311		raise CrabException(msg)
312
313		def DataDiscoveryAndLocation(self, cfg_params):
314
315	+	import DataDiscovery
316	+	import DataDiscovery_DBS2
317	+	import DataLocation
318		common.logger.debug(10,"CMSSW::DataDiscoveryAndLocation()")
319
320		datasetPath=self.datasetPath
321
265	–	## TODO
266	–	dataTiersList = ""
267	–	dataTiers = dataTiersList.split(',')
268	–
322		## Contact the DBS
323	<	common.logger.message("Contacting DBS...")
323	>	common.logger.message("Contacting Data Discovery Services ...")
324		try:
325	<	self.pubdata=DataDiscovery.DataDiscovery(datasetPath, dataTiers, cfg_params)
325	>
326	>	if self.use_dbs_1 == 1 :
327	>	self.pubdata=DataDiscovery.DataDiscovery(datasetPath, cfg_params)
328	>	else :
329	>	self.pubdata=DataDiscovery_DBS2.DataDiscovery_DBS2(datasetPath, cfg_params)
330		self.pubdata.fetchDBSInfo()
331
332		except DataDiscovery.NotExistingDatasetError, ex :
333		msg = 'ERROR ***: failed Data Discovery in DBS : %s'%ex.getErrorMessage()
334		raise CrabException(msg)
278	–
335		except DataDiscovery.NoDataTierinProvenanceError, ex :
336		msg = 'ERROR ***: failed Data Discovery in DBS : %s'%ex.getErrorMessage()
337		raise CrabException(msg)
338		except DataDiscovery.DataDiscoveryError, ex:
339	<	msg = 'ERROR ***: failed Data Discovery in DBS  %s'%ex.getErrorMessage()
339	>	msg = 'ERROR ***: failed Data Discovery in DBS :  %s'%ex.getErrorMessage()
340	>	raise CrabException(msg)
341	>	except DataDiscovery_DBS2.NotExistingDatasetError_DBS2, ex :
342	>	msg = 'ERROR ***: failed Data Discovery in DBS : %s'%ex.getErrorMessage()
343	>	raise CrabException(msg)
344	>	except DataDiscovery_DBS2.NoDataTierinProvenanceError_DBS2, ex :
345	>	msg = 'ERROR ***: failed Data Discovery in DBS : %s'%ex.getErrorMessage()
346	>	raise CrabException(msg)
347	>	except DataDiscovery_DBS2.DataDiscoveryError_DBS2, ex:
348	>	msg = 'ERROR ***: failed Data Discovery in DBS :  %s'%ex.getErrorMessage()
349		raise CrabException(msg)
285	–
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)
350
351		self.filesbyblock=self.pubdata.getFiles()
352		self.eventsbyblock=self.pubdata.getEventsPerBlock()
353		self.eventsbyfile=self.pubdata.getEventsPerFile()
293	–	# print str(self.filesbyblock)
294	–	# print 'self.eventsbyfile',len(self.eventsbyfile)
295	–	# print str(self.eventsbyfile)
354
355		## get max number of events
356		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)
357
301	–	common.logger.message("Contacting DLS...")
358		## Contact the DLS and build a list of sites hosting the fileblocks
359		try:
360		dataloc=DataLocation.DataLocation(self.filesbyblock.keys(),cfg_params)
#	Line 311 | Line 367 | class Cmssw(JobType):
367		sites = dataloc.getSites()
368		allSites = []
369		listSites = sites.values()
370	<	for list in listSites:
371	<	for oneSite in list:
370	>	for listSite in listSites:
371	>	for oneSite in listSite:
372		allSites.append(oneSite)
373		allSites = self.uniquelist(allSites)
374
375	<	common.logger.message("Sites ("+str(len(allSites))+") hosting part/all of dataset: "+str(allSites))
376	<	common.logger.debug(6, "List of Sites: "+str(allSites))
375	>	# screen output
376	>	common.logger.message("Requested dataset: " + datasetPath + " has " + str(self.maxEvents) + " events in " + str(len(self.filesbyblock.keys())) + " blocks.\n")
377	>
378		return sites
379
380		def jobSplittingByBlocks(self, blockSites):
#	Line 379 | Line 436 | class Cmssw(JobType):
436		jobCount = 0
437		list_of_lists = []
438
439	+	# list tracking which jobs are in which jobs belong to which block
440	+	jobsOfBlock = {}
441	+
442		# ---- Iterate over the blocks in the dataset until ---- #
443		# ---- we've met the requested total # of events    ---- #
444		while ( (eventsRemaining > 0) and (blockCount < numBlocksInDataset) and (jobCount < totalNumberOfJobs)):
445		block = blocks[blockCount]
446		blockCount += 1
447	+	if block not in jobsOfBlock.keys() :
448	+	jobsOfBlock[block] = []
449
450	<
451	<	numEventsInBlock = self.eventsbyblock[block]
452	<	common.logger.debug(5,'Events in Block File '+str(numEventsInBlock))
450	>	if self.eventsbyblock.has_key(block) :
451	>	numEventsInBlock = self.eventsbyblock[block]
452	>	common.logger.debug(5,'Events in Block File '+str(numEventsInBlock))
453
454	<	files = self.filesbyblock[block]
455	<	numFilesInBlock = len(files)
456	<	if (numFilesInBlock <= 0):
457	<	continue
458	<	fileCount = 0
459	<
460	<	# ---- New block => New job ---- #
461	<	parString = "\\{"
462	<	# counter for number of events in files currently worked on
463	<	filesEventCount = 0
464	<	# flag if next while loop should touch new file
465	<	newFile = 1
466	<	# job event counter
467	<	jobSkipEventCount = 0
454	>	files = self.filesbyblock[block]
455	>	numFilesInBlock = len(files)
456	>	if (numFilesInBlock <= 0):
457	>	continue
458	>	fileCount = 0
459	>
460	>	# ---- New block => New job ---- #
461	>	parString = "\\{"
462	>	# counter for number of events in files currently worked on
463	>	filesEventCount = 0
464	>	# flag if next while loop should touch new file
465	>	newFile = 1
466	>	# job event counter
467	>	jobSkipEventCount = 0
468
469	<	# ---- Iterate over the files in the block until we've met the requested ---- #
470	<	# ---- total # of events or we've gone over all the files in this block  ---- #
471	<	while ( (eventsRemaining > 0) and (fileCount < numFilesInBlock) and (jobCount < totalNumberOfJobs) ):
472	<	file = files[fileCount]
473	<	if newFile :
474	<	try:
475	<	numEventsInFile = self.eventsbyfile[file]
476	<	common.logger.debug(6, "File "+str(file)+" has "+str(numEventsInFile)+" events")
477	<	# increase filesEventCount
478	<	filesEventCount += numEventsInFile
479	<	# Add file to current job
480	<	parString += '\\\"' + file + '\\\"\,'
481	<	newFile = 0
482	<	except KeyError:
483	<	common.logger.message("File "+str(file)+" has unknown number of events: skipping")
469	>	# ---- Iterate over the files in the block until we've met the requested ---- #
470	>	# ---- total # of events or we've gone over all the files in this block  ---- #
471	>	while ( (eventsRemaining > 0) and (fileCount < numFilesInBlock) and (jobCount < totalNumberOfJobs) ):
472	>	file = files[fileCount]
473	>	if newFile :
474	>	try:
475	>	numEventsInFile = self.eventsbyfile[file]
476	>	common.logger.debug(6, "File "+str(file)+" has "+str(numEventsInFile)+" events")
477	>	# increase filesEventCount
478	>	filesEventCount += numEventsInFile
479	>	# Add file to current job
480	>	parString += '\\\"' + file + '\\\"\,'
481	>	newFile = 0
482	>	except KeyError:
483	>	common.logger.message("File "+str(file)+" has unknown number of events: skipping")
484
485
486	<	# if less events in file remain than eventsPerJobRequested
487	<	if ( filesEventCount - jobSkipEventCount < eventsPerJobRequested ) :
488	<	# if last file in block
489	<	if ( fileCount == numFilesInBlock-1 ) :
490	<	# end job using last file, use remaining events in block
486	>	# if less events in file remain than eventsPerJobRequested
487	>	if ( filesEventCount - jobSkipEventCount < eventsPerJobRequested ) :
488	>	# if last file in block
489	>	if ( fileCount == numFilesInBlock-1 ) :
490	>	# end job using last file, use remaining events in block
491	>	# close job and touch new file
492	>	fullString = parString[:-2]
493	>	fullString += '\\}'
494	>	list_of_lists.append([fullString,str(-1),str(jobSkipEventCount)])
495	>	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(filesEventCount - jobSkipEventCount)+" events (last file in block).")
496	>	self.jobDestination.append(blockSites[block])
497	>	common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
498	>	# fill jobs of block dictionary
499	>	jobsOfBlock[block].append(jobCount+1)
500	>	# reset counter
501	>	jobCount = jobCount + 1
502	>	totalEventCount = totalEventCount + filesEventCount - jobSkipEventCount
503	>	eventsRemaining = eventsRemaining - filesEventCount + jobSkipEventCount
504	>	jobSkipEventCount = 0
505	>	# reset file
506	>	parString = "\\{"
507	>	filesEventCount = 0
508	>	newFile = 1
509	>	fileCount += 1
510	>	else :
511	>	# go to next file
512	>	newFile = 1
513	>	fileCount += 1
514	>	# if events in file equal to eventsPerJobRequested
515	>	elif ( filesEventCount - jobSkipEventCount == eventsPerJobRequested ) :
516		# close job and touch new file
517		fullString = parString[:-2]
518		fullString += '\\}'
519	<	list_of_lists.append([fullString,str(-1),str(jobSkipEventCount)])
520	<	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(filesEventCount - jobSkipEventCount)+" events (last file in block).")
519	>	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
520	>	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.")
521		self.jobDestination.append(blockSites[block])
522		common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
523	+	jobsOfBlock[block].append(jobCount+1)
524		# reset counter
525		jobCount = jobCount + 1
526	<	totalEventCount = totalEventCount + filesEventCount - jobSkipEventCount
527	<	eventsRemaining = eventsRemaining - filesEventCount + jobSkipEventCount
526	>	totalEventCount = totalEventCount + eventsPerJobRequested
527	>	eventsRemaining = eventsRemaining - eventsPerJobRequested
528		jobSkipEventCount = 0
529		# reset file
530		parString = "\\{"
531		filesEventCount = 0
532		newFile = 1
533		fileCount += 1
534	+
535	+	# if more events in file remain than eventsPerJobRequested
536		else :
537	<	# go to next file
538	<	newFile = 1
539	<	fileCount += 1
540	<	# if events in file equal to eventsPerJobRequested
541	<	elif ( filesEventCount - jobSkipEventCount == eventsPerJobRequested ) :
542	<	# close job and touch new file
543	<	fullString = parString[:-2]
544	<	fullString += '\\}'
545	<	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
546	<	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.")
547	<	self.jobDestination.append(blockSites[block])
548	<	common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
549	<	# reset counter
550	<	jobCount = jobCount + 1
551	<	totalEventCount = totalEventCount + eventsPerJobRequested
552	<	eventsRemaining = eventsRemaining - eventsPerJobRequested
553	<	jobSkipEventCount = 0
554	<	# reset file
555	<	parString = "\\{"
556	<	filesEventCount = 0
557	<	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)
537	>	# close job but don't touch new file
538	>	fullString = parString[:-2]
539	>	fullString += '\\}'
540	>	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
541	>	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.")
542	>	self.jobDestination.append(blockSites[block])
543	>	common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
544	>	jobsOfBlock[block].append(jobCount+1)
545	>	# increase counter
546	>	jobCount = jobCount + 1
547	>	totalEventCount = totalEventCount + eventsPerJobRequested
548	>	eventsRemaining = eventsRemaining - eventsPerJobRequested
549	>	# calculate skip events for last file
550	>	# use filesEventCount (contains several files), jobSkipEventCount and eventsPerJobRequest
551	>	jobSkipEventCount = eventsPerJobRequested - (filesEventCount - jobSkipEventCount - self.eventsbyfile[file])
552	>	# remove all but the last file
553	>	filesEventCount = self.eventsbyfile[file]
554	>	parString = "\\{"
555	>	parString += '\\\"' + file + '\\\"\,'
556	>	pass # END if
557	>	pass # END while (iterate over files in the block)
558		pass # END while (iterate over blocks in the dataset)
559		self.ncjobs = self.total_number_of_jobs = jobCount
560		if (eventsRemaining > 0 and jobCount < totalNumberOfJobs ):
561		common.logger.message("Could not run on all requested events because some blocks not hosted at allowed sites.")
562	<	common.logger.message("\n"+str(jobCount)+" job(s) can run on "+str(totalEventCount)+" events.\n")
562	>	common.logger.message(str(jobCount)+" job(s) can run on "+str(totalEventCount)+" events.\n")
563
564	+	# screen output
565	+	screenOutput = "List of jobs and available destination sites:\n\n"
566	+
567	+	blockCounter = 0
568	+	for block in blocks:
569	+	if block in jobsOfBlock.keys() :
570	+	blockCounter += 1
571	+	screenOutput += "Block %5i: jobs %20s: sites: %s\n" % (blockCounter,spanRanges(jobsOfBlock[block]),','.join(blockSites[block]))
572	+
573	+	common.logger.message(screenOutput)
574	+
575		self.list_of_args = list_of_lists
576		return
577
#	Line 512 | Line 589 | class Cmssw(JobType):
589		raise CrabException(msg)
590
591		if (self.selectEventsPerJob):
592	<	self.total_number_of_jobs = int(self.total_number_of_events/self.eventsPerJob)
592	>	if (self.selectTotalNumberEvents):
593	>	self.total_number_of_jobs = int(self.total_number_of_events/self.eventsPerJob)
594	>	elif(self.selectNumberOfJobs) :
595	>	self.total_number_of_jobs =self.theNumberOfJobs
596	>	self.total_number_of_events =int(self.theNumberOfJobs*self.eventsPerJob)
597	>
598		elif (self.selectNumberOfJobs) :
599		self.total_number_of_jobs = self.theNumberOfJobs
600		self.eventsPerJob = int(self.total_number_of_events/self.total_number_of_jobs)
601	<
601	>
602		common.logger.debug(5,'N jobs  '+str(self.total_number_of_jobs))
603
604		# is there any remainder?
#	Line 534 | Line 616 | class Cmssw(JobType):
616		## Since there is no input, any site is good
617		# self.jobDestination.append(["Any"])
618		self.jobDestination.append([""]) #must be empty to write correctly the xml
619	+	args=[]
620	+	if (self.firstRun):
621	+	## pythia first run
622	+	#self.list_of_args.append([(str(self.firstRun)+str(i))])
623	+	args.append(str(self.firstRun)+str(i))
624	+	else:
625	+	## no first run
626	+	#self.list_of_args.append([str(i)])
627	+	args.append(str(i))
628		if (self.sourceSeed):
629	+	args.append(str(self.sourceSeed)+str(i))
630		if (self.sourceSeedVtx):
631	<	## pythia + vtx random seed
632	<	self.list_of_args.append([
633	<	str(self.sourceSeed)+str(i),
634	<	str(self.sourceSeedVtx)+str(i)
635	<	])
636	<	else:
637	<	## only pythia random seed
638	<	self.list_of_args.append([(str(self.sourceSeed)+str(i))])
639	<	else:
640	<	## no random seed
641	<	self.list_of_args.append([str(i)])
642	<	#print self.list_of_args
631	>	## + vtx random seed
632	>	args.append(str(self.sourceSeedVtx)+str(i))
633	>	if (self.sourceSeedG4):
634	>	## + G4 random seed
635	>	args.append(str(self.sourceSeedG4)+str(i))
636	>	if (self.sourceSeedMix):
637	>	## + Mix random seed
638	>	args.append(str(self.sourceSeedMix)+str(i))
639	>	pass
640	>	pass
641	>	self.list_of_args.append(args)
642	>	pass
643	>
644	>	# print self.list_of_args
645
646		return
647
#	Line 612 | Line 706 | class Cmssw(JobType):
706		"""
707
708		# if it exist, just return it
709	<	self.tgzNameWithPath = common.work_space.shareDir()+self.tgz_name
709	>	#
710	>	# Marco. Let's start to use relative path for Boss XML files
711	>	#
712	>	self.tgzNameWithPath = common.work_space.pathForTgz()+'share/'+self.tgz_name
713		if os.path.exists(self.tgzNameWithPath):
714		return self.tgzNameWithPath
715
#	Line 626 | Line 723 | class Cmssw(JobType):
723		# First of all declare the user Scram area
724		swArea = self.scram.getSWArea_()
725		#print "swArea = ", swArea
726	<	swVersion = self.scram.getSWVersion()
727	<	#print "swVersion = ", swVersion
726	>	# swVersion = self.scram.getSWVersion()
727	>	# print "swVersion = ", swVersion
728		swReleaseTop = self.scram.getReleaseTop_()
729		#print "swReleaseTop = ", swReleaseTop
730
#	Line 635 | Line 732 | class Cmssw(JobType):
732		if swReleaseTop == '' or swArea == swReleaseTop:
733		return
734
735	<	filesToBeTarred = []
736	<	## First find the executable
737	<	if (self.executable != ''):
738	<	exeWithPath = self.scram.findFile_(executable)
739	<	#           print exeWithPath
740	<	if ( not exeWithPath ):
741	<	raise CrabException('User executable '+executable+' not found')
742	<
743	<	## then check if it's private or not
744	<	if exeWithPath.find(swReleaseTop) == -1:
745	<	# the exe is private, so we must ship
746	<	common.logger.debug(5,"Exe "+exeWithPath+" to be tarred")
747	<	path = swArea+'/'
748	<	exe = string.replace(exeWithPath, path,'')
749	<	filesToBeTarred.append(exe)
750	<	pass
751	<	else:
752	<	# the exe is from release, we'll find it on WN
753	<	pass
754	<
755	<	## Now get the libraries: only those in local working area
756	<	libDir = 'lib'
757	<	lib = swArea+'/' +libDir
758	<	common.logger.debug(5,"lib "+lib+" to be tarred")
759	<	if os.path.exists(lib):
760	<	filesToBeTarred.append(libDir)
761	<
762	<	## Now check if module dir is present
763	<	moduleDir = 'module'
764	<	if os.path.isdir(swArea+'/'+moduleDir):
765	<	filesToBeTarred.append(moduleDir)
766	<
767	<	## Now check if the Data dir is present
768	<	dataDir = 'src/Data/'
769	<	if os.path.isdir(swArea+'/'+dataDir):
770	<	filesToBeTarred.append(dataDir)
771	<
772	<	## copy ProdAgent dir to swArea
773	<	cmd = '\cp -rf ' + os.environ['CRABDIR'] + '/ProdAgentApi ' + swArea
774	<	cmd_out = runCommand(cmd)
775	<	if cmd_out != '':
776	<	common.logger.message('ProdAgentApi directory could not be copied to local CMSSW project directory.')
777	<	common.logger.message('No FrameworkJobreport parsing is possible on the WorkerNode.')
778	<
779	<	## Now check if the Data dir is present
780	<	paDir = 'ProdAgentApi'
781	<	if os.path.isdir(swArea+'/'+paDir):
782	<	filesToBeTarred.append(paDir)
783	<
784	<	## Create the tar-ball
785	<	if len(filesToBeTarred)>0:
786	<	cwd = os.getcwd()
787	<	os.chdir(swArea)
788	<	tarcmd = 'tar zcvf ' + self.tgzNameWithPath + ' '
789	<	for line in filesToBeTarred:
790	<	tarcmd = tarcmd + line + ' '
791	<	cout = runCommand(tarcmd)
792	<	if not cout:
793	<	raise CrabException('Could not create tar-ball')
794	<	os.chdir(cwd)
795	<	else:
796	<	common.logger.debug(5,"No files to be to be tarred")
735	>	import tarfile
736	>	try: # create tar ball
737	>	tar = tarfile.open(self.tgzNameWithPath, "w:gz")
738	>	## First find the executable
739	>	if (self.executable != ''):
740	>	exeWithPath = self.scram.findFile_(executable)
741	>	if ( not exeWithPath ):
742	>	raise CrabException('User executable '+executable+' not found')
743	>
744	>	## then check if it's private or not
745	>	if exeWithPath.find(swReleaseTop) == -1:
746	>	# the exe is private, so we must ship
747	>	common.logger.debug(5,"Exe "+exeWithPath+" to be tarred")
748	>	path = swArea+'/'
749	>	# distinguish case when script is in user project area or given by full path somewhere else
750	>	if exeWithPath.find(path) >= 0 :
751	>	exe = string.replace(exeWithPath, path,'')
752	>	tar.add(path+exe,os.path.basename(executable))
753	>	else :
754	>	tar.add(exeWithPath,os.path.basename(executable))
755	>	pass
756	>	else:
757	>	# the exe is from release, we'll find it on WN
758	>	pass
759	>
760	>	## Now get the libraries: only those in local working area
761	>	libDir = 'lib'
762	>	lib = swArea+'/' +libDir
763	>	common.logger.debug(5,"lib "+lib+" to be tarred")
764	>	if os.path.exists(lib):
765	>	tar.add(lib,libDir)
766	>
767	>	## Now check if module dir is present
768	>	moduleDir = 'module'
769	>	module = swArea + '/' + moduleDir
770	>	if os.path.isdir(module):
771	>	tar.add(module,moduleDir)
772	>
773	>	## Now check if any data dir(s) is present
774	>	swAreaLen=len(swArea)
775	>	for root, dirs, files in os.walk(swArea):
776	>	if "data" in dirs:
777	>	common.logger.debug(5,"data "+root+"/data"+" to be tarred")
778	>	tar.add(root+"/data",root[swAreaLen:]+"/data")
779	>
780	>	## Add ProdAgent dir to tar
781	>	paDir = 'ProdAgentApi'
782	>	pa = os.environ['CRABDIR'] + '/' + 'ProdAgentApi'
783	>	if os.path.isdir(pa):
784	>	tar.add(pa,paDir)
785	>
786	>	### FEDE FOR DBS PUBLICATION
787	>	## Add PRODCOMMON dir to tar
788	>	prodcommonDir = 'ProdCommon'
789	>	prodcommonPath = os.environ['CRABDIR'] + '/' + 'ProdCommon'
790	>	if os.path.isdir(prodcommonPath):
791	>	tar.add(prodcommonPath,prodcommonDir)
792	>	#############################
793	>
794	>	common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
795	>	tar.close()
796	>	except :
797	>	raise CrabException('Could not create tar-ball')
798	>
799	>	## check for tarball size
800	>	tarballinfo = os.stat(self.tgzNameWithPath)
801	>	if ( tarballinfo.st_size > self.MaxTarBallSize*1024*1024 ) :
802	>	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.')
803	>
804	>	## create tar-ball with ML stuff
805	>	self.MLtgzfile =  common.work_space.pathForTgz()+'share/MLfiles.tgz'
806	>	try:
807	>	tar = tarfile.open(self.MLtgzfile, "w:gz")
808	>	path=os.environ['CRABDIR'] + '/python/'
809	>	for file in ['report.py', 'DashboardAPI.py', 'Logger.py', 'ProcInfo.py', 'apmon.py', 'parseCrabFjr.py']:
810	>	tar.add(path+file,file)
811	>	common.logger.debug(5,"Files added to "+self.MLtgzfile+" : "+str(tar.getnames()))
812	>	tar.close()
813	>	except :
814	>	raise CrabException('Could not create ML files tar-ball')
815
816		return
817
818	+	def additionalInputFileTgz(self):
819	+	"""
820	+	Put all additional files into a tar ball and return its name
821	+	"""
822	+	import tarfile
823	+	tarName=  common.work_space.pathForTgz()+'share/'+self.additional_tgz_name
824	+	tar = tarfile.open(tarName, "w:gz")
825	+	for file in self.additional_inbox_files:
826	+	tar.add(file,string.split(file,'/')[-1])
827	+	common.logger.debug(5,"Files added to "+self.additional_tgz_name+" : "+str(tar.getnames()))
828	+	tar.close()
829	+	return tarName
830	+
831		def wsSetupEnvironment(self, nj):
832		"""
833		Returns part of a job script which prepares
#	Line 710 | Line 838 | class Cmssw(JobType):
838
839		## OLI_Daniele at this level  middleware already known
840
841	+	txt += 'echo "### Firtst set SCRAM ARCH and BUILD_ARCH ###"\n'
842	+	txt += 'echo "Setting SCRAM_ARCH='+self.executable_arch+'"\n'
843	+	txt += 'export SCRAM_ARCH='+self.executable_arch+'\n'
844	+	txt += 'export BUILD_ARCH='+self.executable_arch+'\n'
845		txt += 'if [ $middleware == LCG ]; then \n'
846		txt += self.wsSetupCMSLCGEnvironment_()
847		txt += 'elif [ $middleware == OSG ]; then\n'
#	Line 717 | Line 849 | class Cmssw(JobType):
849		txt += '    echo "Created working directory: $WORKING_DIR"\n'
850		txt += '    if [ ! -d $WORKING_DIR ] ;then\n'
851		txt += '        echo "SET_CMS_ENV 10016 ==> OSG $WORKING_DIR could not be created on WN `hostname`"\n'
852	<	txt += '        echo "JOB_EXIT_STATUS = 10016"\n'
853	<	txt += '        echo "JobExitCode=10016" | tee -a $RUNTIME_AREA/$repo\n'
854	<	txt += '        dumpStatus $RUNTIME_AREA/$repo\n'
852	>	txt += '    echo "JOB_EXIT_STATUS = 10016"\n'
853	>	txt += '    echo "JobExitCode=10016" | tee -a $RUNTIME_AREA/$repo\n'
854	>	txt += '    dumpStatus $RUNTIME_AREA/$repo\n'
855		txt += '        rm -f $RUNTIME_AREA/$repo \n'
856		txt += '        echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
857		txt += '        echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
#	Line 751 | Line 883 | class Cmssw(JobType):
883		txt += '        cd $RUNTIME_AREA\n'
884		txt += '        /bin/rm -rf $WORKING_DIR\n'
885		txt += '        if [ -d $WORKING_DIR ] ;then\n'
886	<	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'
887	<	txt += '            echo "JOB_EXIT_STATUS = 10018"\n'
888	<	txt += '            echo "JobExitCode=10018" | tee -a $RUNTIME_AREA/$repo\n'
889	<	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
886	>	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'
887	>	txt += '            echo "JOB_EXIT_STATUS = 10018"\n'
888	>	txt += '            echo "JobExitCode=10018" | tee -a $RUNTIME_AREA/$repo\n'
889	>	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
890		txt += '            rm -f $RUNTIME_AREA/$repo \n'
891		txt += '            echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
892		txt += '            echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
#	Line 764 | Line 896 | class Cmssw(JobType):
896		txt += 'fi \n'
897		txt += 'echo "CMSSW_VERSION =  '+self.version+'"\n'
898		txt += 'cd '+self.version+'\n'
899	+	########## FEDE FOR DBS2 ######################
900	+	txt += 'SOFTWARE_DIR=`pwd`\n'
901	+	txt += 'echo SOFTWARE_DIR=$SOFTWARE_DIR \n'
902	+	###############################################
903		### needed grep for bug in scramv1 ###
904	+	txt += scram+' runtime -sh\n'
905		txt += 'eval `'+scram+' runtime -sh | grep -v SCRAMRT_LSB_JOBNAME`\n'
906	+	txt += 'echo $PATH\n'
907
908		# Handle the arguments:
909		txt += "\n"
#	Line 787 | Line 925 | class Cmssw(JobType):
925		txt += '        cd $RUNTIME_AREA\n'
926		txt += '        /bin/rm -rf $WORKING_DIR\n'
927		txt += '        if [ -d $WORKING_DIR ] ;then\n'
928	<	txt += '            echo "SET_EXE_ENV 50114 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after Too few arguments for CRAB job wrapper"\n'
929	<	txt += '            echo "JOB_EXIT_STATUS = 50114"\n'
930	<	txt += '            echo "JobExitCode=50114" | tee -a $RUNTIME_AREA/$repo\n'
931	<	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
928	>	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'
929	>	txt += '            echo "JOB_EXIT_STATUS = 50114"\n'
930	>	txt += '            echo "JobExitCode=50114" | tee -a $RUNTIME_AREA/$repo\n'
931	>	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
932		txt += '            rm -f $RUNTIME_AREA/$repo \n'
933		txt += '            echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
934		txt += '            echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
#	Line 802 | Line 940 | class Cmssw(JobType):
940
941		# Prepare job-specific part
942		job = common.job_list[nj]
943	+	### FEDE FOR DBS OUTPUT PUBLICATION
944	+	if (self.datasetPath):
945	+	txt += '\n'
946	+	txt += 'DatasetPath='+self.datasetPath+'\n'
947	+
948	+	datasetpath_split = self.datasetPath.split("/")
949	+
950	+	txt += 'PrimaryDataset='+datasetpath_split[1]+'\n'
951	+	txt += 'DataTier='+datasetpath_split[2]+'\n'
952	+	#txt += 'ProcessedDataset='+datasetpath_split[3]+'\n'
953	+	txt += 'ApplicationFamily=cmsRun\n'
954	+
955	+	else:
956	+	txt += 'DatasetPath=MCDataTier\n'
957	+	txt += 'PrimaryDataset=null\n'
958	+	txt += 'DataTier=null\n'
959	+	#txt += 'ProcessedDataset=null\n'
960	+	txt += 'ApplicationFamily=MCDataTier\n'
961		if self.pset != None: #CarlosDaniele
962		pset = os.path.basename(job.configFilename())
963		txt += '\n'
964	+	txt += 'cp  $RUNTIME_AREA/'+pset+' .\n'
965		if (self.datasetPath): # standard job
966		#txt += 'InputFiles=$2\n'
967		txt += 'InputFiles=${args[1]}\n'
968		txt += 'MaxEvents=${args[2]}\n'
969		txt += 'SkipEvents=${args[3]}\n'
970		txt += 'echo "Inputfiles:<$InputFiles>"\n'
971	<	txt += 'sed "s#{\'INPUT\'}#$InputFiles#" $RUNTIME_AREA/'+pset+' > pset_tmp_1.cfg\n'
971	>	txt += 'sed "s#{\'INPUT\'}#$InputFiles#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
972		txt += 'echo "MaxEvents:<$MaxEvents>"\n'
973	<	txt += 'sed "s#INPUTMAXEVENTS#$MaxEvents#" pset_tmp_1.cfg > pset_tmp_2.cfg\n'
973	>	txt += 'sed "s#INPUTMAXEVENTS#$MaxEvents#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
974		txt += 'echo "SkipEvents:<$SkipEvents>"\n'
975	<	txt += 'sed "s#INPUTSKIPEVENTS#$SkipEvents#" pset_tmp_2.cfg > pset.cfg\n'
975	>	txt += 'sed "s#INPUTSKIPEVENTS#$SkipEvents#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
976		else:  # pythia like job
977	+	seedIndex=1
978	+	if (self.firstRun):
979	+	txt += 'FirstRun=${args['+str(seedIndex)+']}\n'
980	+	txt += 'echo "FirstRun: <$FirstRun>"\n'
981	+	txt += 'sed "s#\<INPUTFIRSTRUN\>#$FirstRun#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
982	+	seedIndex=seedIndex+1
983	+
984		if (self.sourceSeed):
985	<	#                    txt += 'Seed=$2\n'
986	<	txt += 'Seed=${args[1]}\n'
987	<	txt += 'echo "Seed: <$Seed>"\n'
988	<	txt += 'sed "s#\<INPUT\>#$Seed#" $RUNTIME_AREA/'+pset+' > tmp.cfg\n'
985	>	txt += 'Seed=${args['+str(seedIndex)+']}\n'
986	>	txt += 'sed "s#\<INPUT\>#$Seed#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
987	>	seedIndex=seedIndex+1
988	>	## the following seeds are not always present
989		if (self.sourceSeedVtx):
990	<	#                        txt += 'VtxSeed=$3\n'
827	<	txt += 'VtxSeed=${args[2]}\n'
990	>	txt += 'VtxSeed=${args['+str(seedIndex)+']}\n'
991		txt += 'echo "VtxSeed: <$VtxSeed>"\n'
992	<	txt += 'sed "s#INPUTVTX#$VtxSeed#" tmp.cfg > pset.cfg\n'
993	<	else:
994	<	txt += 'mv tmp.cfg pset.cfg\n'
995	<	else:
996	<	txt += '# Copy untouched pset\n'
997	<	txt += 'cp $RUNTIME_AREA/'+pset+' pset.cfg\n'
998	<
992	>	txt += 'sed "s#\<INPUTVTX\>#$VtxSeed#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
993	>	seedIndex += 1
994	>	if (self.sourceSeedG4):
995	>	txt += 'G4Seed=${args['+str(seedIndex)+']}\n'
996	>	txt += 'echo "G4Seed: <$G4Seed>"\n'
997	>	txt += 'sed "s#\<INPUTG4\>#$G4Seed#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
998	>	seedIndex += 1
999	>	if (self.sourceSeedMix):
1000	>	txt += 'mixSeed=${args['+str(seedIndex)+']}\n'
1001	>	txt += 'echo "MixSeed: <$mixSeed>"\n'
1002	>	txt += 'sed "s#\<INPUTMIX\>#$mixSeed#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
1003	>	seedIndex += 1
1004	>	pass
1005	>	pass
1006	>	txt += 'mv -f '+pset+' pset.cfg\n'
1007
1008		if len(self.additional_inbox_files) > 0:
1009	<	for file in self.additional_inbox_files:
1010	<	relFile = file.split("/")[-1]
1011	<	txt += 'if [ -e $RUNTIME_AREA/'+relFile+' ] ; then\n'
841	<	txt += '   cp $RUNTIME_AREA/'+relFile+' .\n'
842	<	txt += '   chmod +x '+relFile+'\n'
843	<	txt += 'fi\n'
1009	>	txt += 'if [ -e $RUNTIME_AREA/'+self.additional_tgz_name+' ] ; then\n'
1010	>	txt += '  tar xzvf $RUNTIME_AREA/'+self.additional_tgz_name+'\n'
1011	>	txt += 'fi\n'
1012		pass
1013
1014		if self.pset != None: #CarlosDaniele
#	Line 851 | Line 1019 | class Cmssw(JobType):
1019		txt += 'cat pset.cfg\n'
1020		txt += 'echo "****** end pset.cfg ********"\n'
1021		txt += '\n'
1022	+	### FEDE FOR DBS OUTPUT PUBLICATION
1023	+	txt += 'PSETHASH=`EdmConfigHash < pset.cfg` \n'
1024	+	txt += 'echo "PSETHASH = $PSETHASH" \n'
1025	+	##############
1026	+	txt += '\n'
1027		# txt += 'echo "***** cat pset1.cfg *********"\n'
1028		# txt += 'cat pset1.cfg\n'
1029		# txt += 'echo "****** end pset1.cfg ********"\n'
1030		return txt
1031
1032	<	def wsBuildExe(self, nj):
1032	>	def wsBuildExe(self, nj=0):
1033		"""
1034		Put in the script the commands to build an executable
1035		or a library.
#	Line 892 | Line 1065 | class Cmssw(JobType):
1065		txt += '   echo "Successful untar" \n'
1066		txt += 'fi \n'
1067		txt += '\n'
1068	<	txt += 'echo "Include ProdAgentApi in PYTHONPATH"\n'
1068	>	txt += 'echo "Include ProdAgentApi and PRODCOMMON in PYTHONPATH"\n'
1069		txt += 'if [ -z "$PYTHONPATH" ]; then\n'
1070	<	txt += '   export PYTHONPATH=ProdAgentApi\n'
1070	>	#### FEDE FOR DBS OUTPUT PUBLICATION
1071	>	txt += '   export PYTHONPATH=$SOFTWARE_DIR/ProdAgentApi:$SOFTWARE_DIR/ProdCommon\n'
1072	>	#txt += '   export PYTHONPATH=`pwd`/ProdAgentApi:`pwd`/ProdCommon\n'
1073	>	#txt += '   export PYTHONPATH=ProdAgentApi\n'
1074		txt += 'else\n'
1075	<	txt += '   export PYTHONPATH=ProdAgentApi:${PYTHONPATH}\n'
1075	>	txt += '   export PYTHONPATH=$SOFTWARE_DIR/ProdAgentApi:$SOFTWARE_DIR/ProdCommon:${PYTHONPATH}\n'
1076	>	#txt += '   export PYTHONPATH=`pwd`/ProdAgentApi:`pwd`/ProdCommon:${PYTHONPATH}\n'
1077	>	#txt += '   export PYTHONPATH=ProdAgentApi:${PYTHONPATH}\n'
1078	>	txt += 'echo "PYTHONPATH=$PYTHONPATH"\n'
1079	>	###################
1080		txt += 'fi\n'
1081		txt += '\n'
1082
#	Line 911 | Line 1091 | class Cmssw(JobType):
1091		"""
1092
1093		def executableName(self):
1094	<	if self.pset == None: #CarlosDaniele
1094	>	if self.scriptExe: #CarlosDaniele
1095		return "sh "
1096		else:
1097		return self.executable
1098
1099		def executableArgs(self):
1100	<	if self.pset == None:#CarlosDaniele
1100	>	if self.scriptExe:#CarlosDaniele
1101		return   self.scriptExe + " $NJob"
1102	<	else:
1103	<	return " -p pset.cfg"
1102	>	else:
1103	>	# if >= CMSSW_1_5_X, add -e
1104	>	version_array = self.scram.getSWVersion().split('_')
1105	>	major = 0
1106	>	minor = 0
1107	>	try:
1108	>	major = int(version_array[1])
1109	>	minor = int(version_array[2])
1110	>	except:
1111	>	msg = "Cannot parse CMSSW version string: " + "_".join(version_array) + " for major and minor release number!"
1112	>	raise CrabException(msg)
1113	>	if major >= 1 and minor >= 5 :
1114	>	return " -e -p pset.cfg"
1115	>	else:
1116	>	return " -p pset.cfg"
1117
1118		def inputSandbox(self, nj):
1119		"""
#	Line 932 | Line 1125 | class Cmssw(JobType):
1125		## code
1126		if os.path.isfile(self.tgzNameWithPath):
1127		inp_box.append(self.tgzNameWithPath)
1128	+	if os.path.isfile(self.MLtgzfile):
1129	+	inp_box.append(self.MLtgzfile)
1130		## config
1131	<	if not self.pset is None: #CarlosDaniele
1132	<	inp_box.append(common.job_list[nj].configFilename())
1131	>	if not self.pset is None:
1132	>	inp_box.append(common.work_space.pathForTgz() + 'job/' + self.configFilename())
1133		## additional input files
1134	<	#for file in self.additional_inbox_files:
1135	<	#    inp_box.append(common.work_space.cwdDir()+file)
1134	>	tgz = self.additionalInputFileTgz()
1135	>	inp_box.append(tgz)
1136		return inp_box
1137
1138		def outputSandbox(self, nj):
#	Line 971 | Line 1166 | class Cmssw(JobType):
1166		output_file_num = self.numberFile_(fileWithSuffix, '$NJob')
1167		txt += '\n'
1168		txt += '# check output file\n'
1169	<	txt += 'ls '+fileWithSuffix+'\n'
1170	<	txt += 'ls_result=$?\n'
1171	<	txt += 'if [ $ls_result -ne 0 ] ; then\n'
1172	<	txt += '   echo "ERROR: Problem with output file"\n'
1169	>	# txt += 'ls '+fileWithSuffix+'\n'
1170	>	# txt += 'ls_result=$?\n'
1171	>	txt += 'if [ -e ./'+fileWithSuffix+' ] ; then\n'
1172	>	txt += '   mv '+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n'
1173	>	txt += 'else\n'
1174	>	txt += '   exit_status=60302\n'
1175	>	txt += '   echo "ERROR: Problem with output file '+fileWithSuffix+'"\n'
1176		if common.scheduler.boss_scheduler_name == 'condor_g':
1177		txt += '    if [ $middleware == OSG ]; then \n'
1178		txt += '        echo "prepare dummy output file"\n'
1179		txt += '        echo "Processing of job output failed" > $RUNTIME_AREA/'+output_file_num+'\n'
1180		txt += '    fi \n'
983	–	txt += 'else\n'
984	–	txt += '   cp '+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n'
1181		txt += 'fi\n'
1182	+	file_list = []
1183	+	for fileWithSuffix in (self.output_file):
1184	+	file_list.append(self.numberFile_(fileWithSuffix, '$NJob'))
1185	+	txt += 'file_list="'+string.join(file_list,' ')+'"\n'
1186
1187		txt += 'cd $RUNTIME_AREA\n'
1188	<	txt += 'cd $RUNTIME_AREA\n'
1188	>	#### FEDE this is the cleanEnv function
1189		### OLI_DANIELE
1190	<	txt += 'if [ $middleware == OSG ]; then\n'
1191	<	txt += '    cd $RUNTIME_AREA\n'
1192	<	txt += '    echo "Remove working directory: $WORKING_DIR"\n'
1193	<	txt += '    /bin/rm -rf $WORKING_DIR\n'
1194	<	txt += '    if [ -d $WORKING_DIR ] ;then\n'
1195	<	txt += '        echo "SET_EXE 60999 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after cleanup of WN"\n'
1196	<	txt += '        echo "JOB_EXIT_STATUS = 60999"\n'
1197	<	txt += '        echo "JobExitCode=60999" | tee -a $RUNTIME_AREA/$repo\n'
1198	<	txt += '        dumpStatus $RUNTIME_AREA/$repo\n'
1199	<	txt += '        rm -f $RUNTIME_AREA/$repo \n'
1200	<	txt += '        echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1201	<	txt += '        echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1202	<	txt += '    fi\n'
1203	<	txt += 'fi\n'
1204	<	txt += '\n'
1190	>	#txt += 'if [ $middleware == OSG ]; then\n'
1191	>	#txt += '    cd $RUNTIME_AREA\n'
1192	>	#txt += '    echo "Remove working directory: $WORKING_DIR"\n'
1193	>	#txt += '    /bin/rm -rf $WORKING_DIR\n'
1194	>	#txt += '    if [ -d $WORKING_DIR ] ;then\n'
1195	>	#txt += '        echo "SET_EXE 60999 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after cleanup of WN"\n'
1196	>	#txt += '        echo "JOB_EXIT_STATUS = 60999"\n'
1197	>	#txt += '        echo "JobExitCode=60999" | tee -a $RUNTIME_AREA/$repo\n'
1198	>	#txt += '        dumpStatus $RUNTIME_AREA/$repo\n'
1199	>	#txt += '        rm -f $RUNTIME_AREA/$repo \n'
1200	>	#txt += '        echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1201	>	#txt += '        echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1202	>	#txt += '    fi\n'
1203	>	#txt += 'fi\n'
1204	>	#txt += '\n'
1205
1006	–	file_list = ''
1007	–	## Add to filelist only files to be possibly copied to SE
1008	–	for fileWithSuffix in self.output_file:
1009	–	output_file_num = self.numberFile_(fileWithSuffix, '$NJob')
1010	–	file_list=file_list+output_file_num+' '
1011	–	file_list=file_list[:-1]
1012	–	txt += 'file_list="'+file_list+'"\n'
1206
1207		return txt
1208
#	Line 1021 | Line 1214 | class Cmssw(JobType):
1214		# take away last extension
1215		name = p[0]
1216		for x in p[1:-1]:
1217	<	name=name+"."+x
1217	>	name=name+"."+x
1218		# add "_txt"
1219		if len(p)>1:
1220	<	ext = p[len(p)-1]
1221	<	result = name + '_' + txt + "." + ext
1220	>	ext = p[len(p)-1]
1221	>	result = name + '_' + txt + "." + ext
1222		else:
1223	<	result = name + '_' + txt
1223	>	result = name + '_' + txt
1224
1225		return result
1226
1227	<	def getRequirements(self):
1227	>	def getRequirements(self, nj=[]):
1228		"""
1229		return job requirements to add to jdl files
1230		"""
#	Line 1040 | Line 1233 | class Cmssw(JobType):
1233		req='Member("VO-cms-' + \
1234		self.version + \
1235		'", other.GlueHostApplicationSoftwareRunTimeEnvironment)'
1236	+	## SL add requirement for OS version only if SL4
1237	+	reSL4 = re.compile( r'slc4' )
1238	+	if self.executable_arch and reSL4.search(self.executable_arch):
1239	+	req+=' && Member("VO-cms-' + \
1240	+	self.executable_arch + \
1241	+	'", other.GlueHostApplicationSoftwareRunTimeEnvironment)'
1242
1243		req = req + ' && (other.GlueHostNetworkAdapterOutboundIP)'
1244
#	Line 1059 | Line 1258 | class Cmssw(JobType):
1258		txt += '   echo "### SETUP CMS OSG  ENVIRONMENT ###"\n'
1259		txt += '   if [ -f $GRID3_APP_DIR/cmssoft/cmsset_default.sh ] ;then\n'
1260		txt += '      # Use $GRID3_APP_DIR/cmssoft/cmsset_default.sh to setup cms software\n'
1261	+	txt += '       export SCRAM_ARCH='+self.executable_arch+'\n'
1262		txt += '       source $GRID3_APP_DIR/cmssoft/cmsset_default.sh '+self.version+'\n'
1263		txt += '   elif [ -f $OSG_APP/cmssoft/cms/cmsset_default.sh ] ;then\n'
1264		txt += '      # Use $OSG_APP/cmssoft/cms/cmsset_default.sh to setup cms software\n'
1265	+	txt += '       export SCRAM_ARCH='+self.executable_arch+'\n'
1266		txt += '       source $OSG_APP/cmssoft/cms/cmsset_default.sh '+self.version+'\n'
1267		txt += '   else\n'
1268		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'
#	Line 1077 | Line 1278 | class Cmssw(JobType):
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'
1287	<	txt += '            echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\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'
1287	>	txt += '           echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1288		txt += '       fi\n'
1289		txt += '\n'
1290		txt += '       exit 1\n'
#	Line 1138 | Line 1339 | class Cmssw(JobType):
1339		txt += '       fi\n'
1340		txt += '   fi\n'
1341		txt += '   \n'
1141	–	txt += '   string=`cat /etc/redhat-release`\n'
1142	–	txt += '   echo $string\n'
1143	–	txt += '   if [[ $string = *alhalla* ]]; then\n'
1144	–	txt += '       echo "SCRAM_ARCH= $SCRAM_ARCH"\n'
1145	–	txt += '   elif [[ $string = *Enterprise* ]] || [[ $string = *cientific* ]]; then\n'
1146	–	txt += '       export SCRAM_ARCH=slc3_ia32_gcc323\n'
1147	–	txt += '       echo "SCRAM_ARCH= $SCRAM_ARCH"\n'
1148	–	txt += '   else\n'
1149	–	txt += '       echo "SET_CMS_ENV 10033 ==> ERROR OS unknown, LCG environment not initialized"\n'
1150	–	txt += '       echo "JOB_EXIT_STATUS = 10033"\n'
1151	–	txt += '       echo "JobExitCode=10033" | tee -a $RUNTIME_AREA/$repo\n'
1152	–	txt += '       dumpStatus $RUNTIME_AREA/$repo\n'
1153	–	txt += '       rm -f $RUNTIME_AREA/$repo \n'
1154	–	txt += '       echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1155	–	txt += '       echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1156	–	txt += '       exit 1\n'
1157	–	txt += '   fi\n'
1342		txt += '   echo "SET_CMS_ENV 0 ==> setup cms environment ok"\n'
1343		txt += '   echo "### END SETUP CMS LCG ENVIRONMENT ###"\n'
1344		return txt
1345
1346	+	### FEDE FOR DBS OUTPUT PUBLICATION
1347	+	def modifyReport(self, nj):
1348	+	"""
1349	+	insert the part of the script that modifies the FrameworkJob Report
1350	+	"""
1351	+
1352	+	txt = ''
1353	+	txt += 'echo "Modify Job Report" \n'
1354	+	#txt += 'chmod a+x $RUNTIME_AREA/'+self.version+'/ProdAgentApi/FwkJobRep/ModifyJobReport.py\n'
1355	+	################ FEDE FOR DBS2 #############################################
1356	+	txt += 'chmod a+x $SOFTWARE_DIR/ProdAgentApi/FwkJobRep/ModifyJobReport.py\n'
1357	+	#############################################################################
1358	+	try:
1359	+	publish_data = int(self.cfg_params['USER.publish_data'])
1360	+	except KeyError:
1361	+	publish_data = 0
1362	+
1363	+	txt += 'if [ -z "$SE" ]; then\n'
1364	+	txt += '    SE="" \n'
1365	+	txt += 'fi \n'
1366	+	txt += 'if [ -z "$SE_PATH" ]; then\n'
1367	+	txt += '    SE_PATH="" \n'
1368	+	txt += 'fi \n'
1369	+	txt += 'echo "SE = $SE"\n'
1370	+	txt += 'echo "SE_PATH = $SE_PATH"\n'
1371	+
1372	+	if (publish_data == 1):
1373	+	#processedDataset = self.cfg_params['USER.processed_datasetname']
1374	+	processedDataset = self.cfg_params['USER.publish_data_name']
1375	+	txt += 'ProcessedDataset='+processedDataset+'\n'
1376	+	#### LFN=/store/user/<user>/processedDataset_PSETHASH
1377	+	txt += 'if [ "$SE_PATH" == "" ]; then\n'
1378	+	#### FEDE: added slash in LFN ##############
1379	+	txt += '    FOR_LFN=/copy_problems/ \n'
1380	+	txt += 'else \n'
1381	+	txt += '    tmp=`echo $SE_PATH | awk -F \'store\' \'{print$2}\'` \n'
1382	+	#####  FEDE TO BE CHANGED, BECAUSE STORE IS HARDCODED!!!! ########
1383	+	txt += '    FOR_LFN=/store$tmp \n'
1384	+	txt += 'fi \n'
1385	+	txt += 'echo "ProcessedDataset = $ProcessedDataset"\n'
1386	+	txt += 'echo "FOR_LFN = $FOR_LFN" \n'
1387	+	txt += 'echo "CMSSW_VERSION = $CMSSW_VERSION"\n\n'
1388	+	#txt += 'echo "$RUNTIME_AREA/'+self.version+'/ProdAgentApi/FwkJobRep/ModifyJobReport.py crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH"\n'
1389	+	txt += 'echo "$SOFTWARE_DIR/ProdAgentApi/FwkJobRep/ModifyJobReport.py crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH"\n'
1390	+	txt += '$SOFTWARE_DIR/ProdAgentApi/FwkJobRep/ModifyJobReport.py crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH\n'
1391	+	#txt += '$RUNTIME_AREA/'+self.version+'/ProdAgentApi/FwkJobRep/ModifyJobReport.py crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH\n'
1392	+
1393	+	txt += 'modifyReport_result=$?\n'
1394	+	txt += 'echo modifyReport_result = $modifyReport_result\n'
1395	+	txt += 'if [ $modifyReport_result -ne 0 ]; then\n'
1396	+	txt += '    exit_status=1\n'
1397	+	txt += '    echo "ERROR: Problem with ModifyJobReport"\n'
1398	+	txt += 'else\n'
1399	+	txt += '    mv NewFrameworkJobReport.xml crab_fjr_$NJob.xml\n'
1400	+	txt += 'fi\n'
1401	+	else:
1402	+	txt += 'ProcessedDataset=no_data_to_publish \n'
1403	+	#### FEDE: added slash in LFN ##############
1404	+	txt += 'FOR_LFN=/local/ \n'
1405	+	txt += 'echo "ProcessedDataset = $ProcessedDataset"\n'
1406	+	txt += 'echo "FOR_LFN = $FOR_LFN" \n'
1407	+	return txt
1408	+
1409	+	def cleanEnv(self):
1410	+	### OLI_DANIELE
1411	+	txt = ''
1412	+	txt += 'if [ $middleware == OSG ]; then\n'
1413	+	txt += '    cd $RUNTIME_AREA\n'
1414	+	txt += '    echo "Remove working directory: $WORKING_DIR"\n'
1415	+	txt += '    /bin/rm -rf $WORKING_DIR\n'
1416	+	txt += '    if [ -d $WORKING_DIR ] ;then\n'
1417	+	txt += '              echo "SET_EXE 60999 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after cleanup of WN"\n'
1418	+	txt += '              echo "JOB_EXIT_STATUS = 60999"\n'
1419	+	txt += '              echo "JobExitCode=60999" | tee -a $RUNTIME_AREA/$repo\n'
1420	+	txt += '              dumpStatus $RUNTIME_AREA/$repo\n'
1421	+	txt += '        rm -f $RUNTIME_AREA/$repo \n'
1422	+	txt += '        echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1423	+	txt += '        echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1424	+	txt += '    fi\n'
1425	+	txt += 'fi\n'
1426	+	txt += '\n'
1427	+	return txt
1428	+
1429		def setParam_(self, param, value):
1430		self._params[param] = value
1431
#	Line 1171 | Line 1438 | class Cmssw(JobType):
1438		def getTaskid(self):
1439		return self._taskId
1440
1174	–	#######################################################################
1441		def uniquelist(self, old):
1442		"""
1443		remove duplicates from a list

Diff Legend

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