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 1.152 by fanzago, Wed Jan 9 16:05:10 2008 UTC vs.
Revision 1.365 by spiga, Tue Nov 9 21:10:07 2010 UTC

#	Line 1 | Line 1
1	+
2	+	__revision__ = "$Id$"
3	+	__version__ = "$Revision$"
4	+
5		from JobType import JobType
2	–	from crab_logger import Logger
6		from crab_exceptions import *
7		from crab_util import *
5	–	from BlackWhiteListParser import BlackWhiteListParser
8		import common
9	+	import re
10		import Scram
11	+	from Splitter import JobSplitter
12	+	from Downloader import Downloader
13	+	try:
14	+	import json
15	+	except:
16	+	import simplejson as json
17
18	+	from IMProv.IMProvNode import IMProvNode
19	+	from IMProv.IMProvLoader import loadIMProvFile
20		import os, string, glob
21	+	from xml.dom import pulldom
22
23		class Cmssw(JobType):
24	<	def __init__(self, cfg_params, ncjobs):
24	>	def __init__(self, cfg_params, ncjobs,skip_blocks, isNew):
25		JobType.__init__(self, 'CMSSW')
26	<	common.logger.debug(3,'CMSSW::__init__')
27	<
28	<	self.argsList = []
29	<
26	>	common.logger.debug('CMSSW::__init__')
27	>	self.skip_blocks = skip_blocks
28	>	self.argsList = 2
29	>	self.NumEvents=0
30		self._params = {}
31		self.cfg_params = cfg_params
32	<	# init BlackWhiteListParser
33	<	self.blackWhiteListParser = BlackWhiteListParser(cfg_params)
32	>	### FEDE FOR MULTI ###
33	>	self.var_filter=''
34
35	<	try:
36	<	self.MaxTarBallSize = float(self.cfg_params['EDG.maxtarballsize'])
37	<	except KeyError:
38	<	self.MaxTarBallSize = 9.5
35	>	### Temporary patch to automatically skip the ISB size check:
36	>	self.server = self.cfg_params.get('CRAB.server_name',None) or \
37	>	self.cfg_params.get('CRAB.use_server',0)
38	>	self.local  = common.scheduler.name().upper() in ['LSF','CAF','CONDOR','SGE','PBS']
39	>	size = 9.5
40	>	if self.server :
41	>	size = 1000
42	>	elif self.local:
43	>	size = 9999999
44	>	self.MaxTarBallSize = float(self.cfg_params.get('GRID.maxtarballsize',size))
45
46		# number of jobs requested to be created, limit obj splitting
47		self.ncjobs = ncjobs
48
31	–	log = common.logger
32	–
49		self.scram = Scram.Scram(cfg_params)
50		self.additional_inbox_files = []
51		self.scriptExe = ''
52		self.executable = ''
53		self.executable_arch = self.scram.getArch()
54		self.tgz_name = 'default.tgz'
39	–	self.additional_tgz_name = 'additional.tgz'
55		self.scriptName = 'CMSSW.sh'
56	<	self.pset = ''      #scrip use case Da
57	<	self.datasetPath = '' #scrip use case Da
56	>	self.pset = ''
57	>	self.datasetPath = ''
58
59	+	self.tgzNameWithPath = common.work_space.pathForTgz()+self.tgz_name
60		# set FJR file name
61		self.fjrFileName = 'crab_fjr.xml'
62
63		self.version = self.scram.getSWVersion()
64	+	common.logger.log(10-1,"CMSSW version is: "+str(self.version))
65	+	version_array = self.version.split('_')
66	+	self.CMSSW_major = 0
67	+	self.CMSSW_minor = 0
68	+	self.CMSSW_patch = 0
69	+	try:
70	+	self.CMSSW_major = int(version_array[1])
71	+	self.CMSSW_minor = int(version_array[2])
72	+	self.CMSSW_patch = int(version_array[3])
73	+	except:
74	+	msg = "Cannot parse CMSSW version string: " + self.version + " for major and minor release number!"
75	+	raise CrabException(msg)
76
77	<	#
78	<	# Try to block creation in case of arch/version mismatch
51	<	#
52	<
53	<	a = string.split(self.version, "_")
54	<
55	<	if int(a[1]) == 1 and (int(a[2]) < 5 and self.executable_arch.find('slc4') == 0):
56	<	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)
57	<	common.logger.message(msg)
58	<	if int(a[1]) == 1 and (int(a[2]) >= 5 and self.executable_arch.find('slc3') == 0):
59	<	msg = "Error: CMS does not support %s with %s architecture"%(self.version, self.executable_arch)
77	>	if self.CMSSW_major < 2 or (self.CMSSW_major == 2 and self.CMSSW_minor < 1):
78	>	msg = "CRAB supports CMSSW >= 2_1_x only. Use an older CRAB version."
79		raise CrabException(msg)
80	+	"""
81	+	As CMSSW versions are dropped we can drop more code:
82	+	2.x dropped: drop check for lumi range setting
83	+	"""
84	+	self.checkCMSSWVersion()
85	+	### collect Data cards
86
87	<	common.taskDB.setDict('codeVersion',self.version)
88	<	self.setParam_('application', self.version)
87	>	### Temporary: added to remove input file control in the case of PU
88	>	self.dataset_pu = cfg_params.get('CMSSW.dataset_pu', None)
89
90	<	### collect Data cards
90	>	tmp =  cfg_params['CMSSW.datasetpath']
91	>	common.logger.log(10-1, "CMSSW::CMSSW(): datasetPath = "+tmp)
92
93	<	try:
68	<	tmp =  cfg_params['CMSSW.datasetpath']
69	<	log.debug(6, "CMSSW::CMSSW(): datasetPath = "+tmp)
70	<	if string.lower(tmp)=='none':
71	<	self.datasetPath = None
72	<	self.selectNoInput = 1
73	<	else:
74	<	self.datasetPath = tmp
75	<	self.selectNoInput = 0
76	<	except KeyError:
93	>	if tmp =='':
94		msg = "Error: datasetpath not defined "
95		raise CrabException(msg)
96	<
97	<	# ML monitoring
98	<	# split dataset path style: /PreProdR3Minbias/SIM/GEN-SIM
99	<	if not self.datasetPath:
83	<	self.setParam_('dataset', 'None')
84	<	self.setParam_('owner', 'None')
96	>	elif string.lower(tmp)=='none':
97	>	self.datasetPath = None
98	>	self.selectNoInput = 1
99	>	self.primaryDataset = 'null'
100		else:
101	<	try:
102	<	datasetpath_split = self.datasetPath.split("/")
103	<	# standard style
104	<	self.setParam_('datasetFull', self.datasetPath)
105	<	self.setParam_('dataset', datasetpath_split[1])
106	<	self.setParam_('owner', datasetpath_split[2])
107	<	except:
108	<	self.setParam_('dataset', self.datasetPath)
109	<	self.setParam_('owner', self.datasetPath)
95	<
96	<	self.setParam_('taskId', common.taskDB.dict('taskId'))
101	>	self.datasetPath = tmp
102	>	self.selectNoInput = 0
103	>	ll = len(self.datasetPath.split("/"))
104	>	if (ll < 4):
105	>	msg = 'Your datasetpath has a invalid format ' + self.datasetPath + '\n'
106	>	msg += 'Expected a path in format /PRIMARY/PROCESSED/TIER1-TIER2 or /PRIMARY/PROCESSED/TIER/METHOD for ADS'
107	>	raise CrabException(msg)
108	>	self.primaryDataset = self.datasetPath.split("/")[1]
109	>	self.dataTier = self.datasetPath.split("/")[2]
110
111	<	self.dataTiers = []
111	>	# Analysis dataset is primary/processed/tier/definition
112	>	self.ads = False
113	>	if self.datasetPath:
114	>	self.ads = len(self.datasetPath.split("/")) > 4
115	>	self.lumiMask = self.cfg_params.get('CMSSW.lumi_mask',None)
116	>	self.lumiParams = self.cfg_params.get('CMSSW.total_number_of_lumis',None) or \
117	>	self.cfg_params.get('CMSSW.lumis_per_job',None)
118	>
119	>	# FUTURE: Can remove this check
120	>	if self.ads and self.CMSSW_major < 3:
121	>	common.logger.info('Warning: Analysis dataset support is incomplete in CMSSW 2_x.')
122	>	common.logger.info('  Only file level, not lumi level, granularity is supported.')
123	>
124	>	self.debugWrap=''
125	>	self.debug_wrapper = int(cfg_params.get('USER.debug_wrapper',0))
126	>	if self.debug_wrapper == 1: self.debugWrap='--debug'
127
128		## now the application
129	<	try:
130	<	self.executable = cfg_params['CMSSW.executable']
131	<	self.setParam_('exe', self.executable)
132	<	log.debug(6, "CMSSW::CMSSW(): executable = "+self.executable)
105	<	msg = "Default executable cmsRun overridden. Switch to " + self.executable
106	<	log.debug(3,msg)
107	<	except KeyError:
108	<	self.executable = 'cmsRun'
109	<	self.setParam_('exe', self.executable)
110	<	msg = "User executable not defined. Use cmsRun"
111	<	log.debug(3,msg)
112	<	pass
129	>	self.managedGenerators = ['madgraph', 'comphep', 'lhe']
130	>	self.generator = cfg_params.get('CMSSW.generator','pythia').lower()
131	>	self.executable = cfg_params.get('CMSSW.executable','cmsRun')
132	>	common.logger.log(10-1, "CMSSW::CMSSW(): executable = "+self.executable)
133
134	<	try:
115	<	self.pset = cfg_params['CMSSW.pset']
116	<	log.debug(6, "Cmssw::Cmssw(): PSet file = "+self.pset)
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:
121	<	self.pset = None
122	<	except KeyError:
134	>	if not cfg_params.has_key('CMSSW.pset'):
135		raise CrabException("PSet file missing. Cannot run cmsRun ")
136	+	self.pset = cfg_params['CMSSW.pset']
137	+	common.logger.log(10-1, "Cmssw::Cmssw(): PSet file = "+self.pset)
138	+	if self.pset.lower() != 'none' :
139	+	if (not os.path.exists(self.pset)):
140	+	raise CrabException("User defined PSet file "+self.pset+" does not exist")
141	+	else:
142	+	self.pset = None
143
144		# output files
145		## stuff which must be returned always via sandbox
#	Line 130 | Line 149 | class Cmssw(JobType):
149		self.output_file_sandbox.append(self.fjrFileName)
150
151		# other output files to be returned via sandbox or copied to SE
152	<	try:
153	<	self.output_file = []
154	<	tmp = cfg_params['CMSSW.output_file']
155	<	if tmp != '':
156	<	tmpOutFiles = string.split(cfg_params['CMSSW.output_file'],',')
157	<	log.debug(7, 'cmssw::cmssw(): output files '+str(tmpOutFiles))
158	<	for tmp in tmpOutFiles:
159	<	tmp=string.strip(tmp)
160	<	self.output_file.append(tmp)
161	<	pass
162	<	else:
163	<	log.message("No output file defined: only stdout/err and the CRAB Framework Job Report will be available\n")
164	<	pass
146	<	pass
147	<	except KeyError:
148	<	log.message("No output file defined: only stdout/err and the CRAB Framework Job Report will be available\n")
149	<	pass
152	>	outfileflag = False
153	>	self.output_file = []
154	>	tmp = cfg_params.get('CMSSW.output_file',None)
155	>	if tmp :
156	>	self.output_file = [x.strip() for x in tmp.split(',')]
157	>	outfileflag = True #output found
158	>
159	>	self.scriptExe = cfg_params.get('USER.script_exe',None)
160	>	if self.scriptExe :
161	>	if not os.path.isfile(self.scriptExe):
162	>	msg ="ERROR. file "+self.scriptExe+" not found"
163	>	raise CrabException(msg)
164	>	self.additional_inbox_files.append(string.strip(self.scriptExe))
165
166	<	# script_exe file as additional file in inputSandbox
167	<	try:
153	<	self.scriptExe = cfg_params['USER.script_exe']
154	<	if self.scriptExe != '':
155	<	if not os.path.isfile(self.scriptExe):
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 = ''
166	>	self.AdditionalArgs = cfg_params.get('USER.script_arguments',None)
167	>	if self.AdditionalArgs : self.AdditionalArgs = string.replace(self.AdditionalArgs,',',' ')
168
162	–	#CarlosDaniele
169		if self.datasetPath == None and self.pset == None and self.scriptExe == '' :
170	<	msg ="Error. script_exe  not defined"
171	<	raise CrabException(msg)
170	>	msg ="Error. script_exe  not defined"
171	>	raise CrabException(msg)
172	>
173	>	# use parent files...
174	>	self.useParent = int(self.cfg_params.get('CMSSW.use_parent',0))
175
176		## additional input files
177	<	try:
177	>	if cfg_params.has_key('USER.additional_input_files'):
178		tmpAddFiles = string.split(cfg_params['USER.additional_input_files'],',')
179		for tmp in tmpAddFiles:
180		tmp = string.strip(tmp)
#	Line 182 | Line 191 | class Cmssw(JobType):
191		if not os.path.exists(file):
192		raise CrabException("Additional input file not found: "+file)
193		pass
185	–	# fname = string.split(file, '/')[-1]
186	–	# storedFile = common.work_space.pathForTgz()+'share/'+fname
187	–	# shutil.copyfile(file, storedFile)
194		self.additional_inbox_files.append(string.strip(file))
195		pass
196		pass
197	<	common.logger.debug(5,"Additional input files: "+str(self.additional_inbox_files))
198	<	except KeyError:
193	<	pass
194	<
195	<	# files per job
196	<	try:
197	<	if (cfg_params['CMSSW.files_per_jobs']):
198	<	raise CrabException("files_per_jobs no longer supported.  Quitting.")
199	<	except KeyError:
200	<	pass
201	<
202	<	## Events per job
203	<	try:
204	<	self.eventsPerJob =int( cfg_params['CMSSW.events_per_job'])
205	<	self.selectEventsPerJob = 1
206	<	except KeyError:
207	<	self.eventsPerJob = -1
208	<	self.selectEventsPerJob = 0
209	<
210	<	## number of jobs
211	<	try:
212	<	self.theNumberOfJobs =int( cfg_params['CMSSW.number_of_jobs'])
213	<	self.selectNumberOfJobs = 1
214	<	except KeyError:
215	<	self.theNumberOfJobs = 0
216	<	self.selectNumberOfJobs = 0
217	<
218	<	try:
219	<	self.total_number_of_events = int(cfg_params['CMSSW.total_number_of_events'])
220	<	self.selectTotalNumberEvents = 1
221	<	except KeyError:
222	<	self.total_number_of_events = 0
223	<	self.selectTotalNumberEvents = 0
224	<
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)
229	<	else:
230	<	if (self.selectNumberOfJobs == 0):
231	<	msg = 'Must specify  number_of_jobs.'
232	<	raise CrabException(msg)
233	<
234	<	## source seed for pythia
235	<	try:
236	<	self.sourceSeed = int(cfg_params['CMSSW.pythia_seed'])
237	<	except KeyError:
238	<	self.sourceSeed = None
239	<	common.logger.debug(5,"No seed given")
240	<
241	<	try:
242	<	self.sourceSeedVtx = int(cfg_params['CMSSW.vtx_seed'])
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	<	import PsetManipulator as pp
266	<	PsetEdit = pp.PsetManipulator(self.pset) #Daniele Pset
197	>	common.logger.debug("Additional input files: "+str(self.additional_inbox_files))
198	>	pass
199
268	–	# Copy/return
200
201	<	try:
202	<	self.copy_data = int(cfg_params['USER.copy_data'])
203	<	except KeyError:
204	<	self.copy_data = 0
205	<	try:
206	<	self.return_data = int(cfg_params['USER.return_data'])
207	<	except KeyError:
208	<	self.return_data = 0
201	>	## New method of dealing with seeds
202	>	self.incrementSeeds = []
203	>	self.preserveSeeds = []
204	>	if cfg_params.has_key('CMSSW.preserve_seeds'):
205	>	tmpList = cfg_params['CMSSW.preserve_seeds'].split(',')
206	>	for tmp in tmpList:
207	>	tmp.strip()
208	>	self.preserveSeeds.append(tmp)
209	>	if cfg_params.has_key('CMSSW.increment_seeds'):
210	>	tmpList = cfg_params['CMSSW.increment_seeds'].split(',')
211	>	for tmp in tmpList:
212	>	tmp.strip()
213	>	self.incrementSeeds.append(tmp)
214	>
215	>	# Copy/return/publish
216	>	self.copy_data = int(cfg_params.get('USER.copy_data',0))
217	>	self.return_data = int(cfg_params.get('USER.return_data',0))
218	>	self.publish_data = int(cfg_params.get('USER.publish_data',0))
219	>	if (self.publish_data == 1):
220	>	if not cfg_params.has_key('USER.publish_data_name'):
221	>	raise CrabException('Cannot publish output data, because you did not specify USER.publish_data_name parameter in the crab.cfg file')
222	>	else:
223	>	self.processedDataset = cfg_params['USER.publish_data_name']
224
225	+	self.conf = {}
226	+	self.conf['pubdata'] = None
227	+	# number of jobs requested to be created, limit obj splitting DD
228		#DBSDLS-start
229		## Initialize the variables that are extracted from DBS/DLS and needed in other places of the code
230		self.maxEvents=0  # max events available   ( --> check the requested nb. of evts in Creator.py)
#	Line 284 | Line 233 | class Cmssw(JobType):
233		## Perform the data location and discovery (based on DBS/DLS)
234		## SL: Don't if NONE is specified as input (pythia use case)
235		blockSites = {}
236	<	if self.datasetPath:
237	<	blockSites = self.DataDiscoveryAndLocation(cfg_params)
238	<	#DBSDLS-end
236	>	#wmbs
237	>	self.automation = int(self.cfg_params.get('WMBS.automation',0))
238	>	if self.automation == 0:
239	>	if self.datasetPath:
240	>	blockSites = self.DataDiscoveryAndLocation(cfg_params)
241	>	#DBSDLS-end
242	>	self.conf['blockSites']=blockSites
243	>
244	>	## Select Splitting
245	>	splitByRun = int(cfg_params.get('CMSSW.split_by_run',0))
246	>
247	>	if self.selectNoInput:
248	>	if self.pset == None:
249	>	self.algo = 'ForScript'
250	>	else:
251	>	self.algo = 'NoInput'
252	>	self.conf['managedGenerators']=self.managedGenerators
253	>	self.conf['generator']=self.generator
254	>	elif self.ads or self.lumiMask or self.lumiParams:
255	>	self.algo = 'LumiBased'
256	>	if splitByRun:
257	>	msg = "Cannot combine split by run with lumi_mask, ADS, " \
258	>	"or lumis_per_job. Use split by lumi mode instead."
259	>	raise CrabException(msg)
260	>
261	>	elif splitByRun ==1:
262	>	self.algo = 'RunBased'
263	>	else:
264	>	self.algo = 'EventBased'
265	>	common.logger.debug("Job splitting method: %s" % self.algo)
266	>
267	>	splitter = JobSplitter(self.cfg_params,self.conf)
268	>	self.dict = splitter.Algos()[self.algo]()
269
270	<	self.tgzNameWithPath = self.getTarBall(self.executable)
270	>	self.argsFile= '%s/arguments.xml'%common.work_space.shareDir()
271	>	self.rootArgsFilename= 'arguments'
272	>	# modify Pset only the first time
273	>	if isNew:
274	>	if self.pset != None: self.ModifyPset()
275	>
276	>	## Prepare inputSandbox TarBall (only the first time)
277	>	self.tarNameWithPath = self.getTarBall(self.executable)
278	>
279	>
280	>	def ModifyPset(self):
281	>	import PsetManipulator as pp
282	>
283	>	# If pycfg_params set, fake out the config script
284	>	# to make it think it was called with those args
285	>	pycfg_params = self.cfg_params.get('CMSSW.pycfg_params',None)
286	>	if pycfg_params:
287	>	trueArgv = sys.argv
288	>	sys.argv = [self.pset]
289	>	sys.argv.extend(pycfg_params.split(' '))
290	>	PsetEdit = pp.PsetManipulator(self.pset)
291	>	if pycfg_params: # Restore original sys.argv
292	>	sys.argv = trueArgv
293	>
294	>	try:
295	>	# Add FrameworkJobReport to parameter-set, set max events.
296	>	# Reset later for data jobs by writeCFG which does all modifications
297	>	PsetEdit.maxEvent(1)
298	>	PsetEdit.skipEvent(0)
299	>	PsetEdit.psetWriter(self.configFilename())
300	>	## If present, add TFileService to output files
301	>	if not int(self.cfg_params.get('CMSSW.skip_tfileservice_output',0)):
302	>	tfsOutput = PsetEdit.getTFileService()
303	>	if tfsOutput:
304	>	if tfsOutput in self.output_file:
305	>	common.logger.debug("Output from TFileService "+tfsOutput+" already in output files")
306	>	else:
307	>	outfileflag = True #output found
308	>	self.output_file.append(tfsOutput)
309	>	common.logger.info("Adding "+tfsOutput+" (from TFileService) to list of output files")
310	>	pass
311	>	pass
312
313	<	## Select Splitting
314	<	if self.selectNoInput:
315	<	if self.pset == None: #CarlosDaniele
316	<	self.jobSplittingForScript()
313	>	# If requested, add PoolOutputModule to output files
314	>	### FEDE FOR MULTI ###
315	>	#edmOutput = PsetEdit.getPoolOutputModule()
316	>	edmOutputDict = PsetEdit.getPoolOutputModule()
317	>	common.logger.debug("(test) edmOutputDict = "+str(edmOutputDict))
318	>	filter_dict = {}
319	>	for key in edmOutputDict.keys():
320	>	filter_dict[key]=edmOutputDict[key]['dataset']
321	>	common.logger.debug("(test) filter_dict for multi =  "+str(filter_dict))
322	>
323	>	#### in CMSSW.sh: export var_filter
324	>
325	>	self.var_filter = json.dumps(filter_dict)
326	>	common.logger.debug("(test) var_filter for multi =  "+self.var_filter)
327	>
328	>	edmOutput = edmOutputDict.keys()
329	>	if int(self.cfg_params.get('CMSSW.get_edm_output',0)):
330	>	if edmOutput:
331	>	for outputFile in edmOutput:
332	>	if outputFile in self.output_file:
333	>	common.logger.debug("Output from PoolOutputModule "+outputFile+" already in output files")
334	>	else:
335	>	self.output_file.append(outputFile)
336	>	common.logger.info("Adding "+outputFile+" (from PoolOutputModule) to list of output files")
337	>	# not requested, check anyhow to avoid accidental T2 overload
338		else:
339	<	self.jobSplittingNoInput()
340	<	else:
341	<	self.jobSplittingByBlocks(blockSites)
339	>	if edmOutput:
340	>	missedFiles = []
341	>	for outputFile in edmOutput:
342	>	if outputFile not in self.output_file:
343	>	missedFiles.append(outputFile)
344	>	if missedFiles:
345	>	msg  = "ERROR: PoolOutputModule(s) are present in your ParameteSet %s \n"%self.pset
346	>	msg += "    but the file(s) produced ( %s ) are not in the list of output files\n" % ', '.join(missedFiles)
347	>	msg += "WARNING: please remove them. If you want to keep them, add the file(s) to output_files or use CMSSW.get_edm_output = 1\n"
348	>	if int(self.cfg_params.get('CMSSW.ignore_edm_output',0)):
349	>	msg += "    CMSSW.ignore_edm_output==1 : Hope you know what you are doing...\n"
350	>	common.logger.info(msg)
351	>	else :
352	>	raise CrabException(msg)
353
354	<	# modify Pset
355	<	if self.pset != None: #CarlosDaniele
356	<	try:
357	<	if (self.datasetPath): # standard job
358	<	# allow to processa a fraction of events in a file
359	<	PsetEdit.inputModule("INPUTFILE")
360	<	PsetEdit.maxEvent(0)
361	<	PsetEdit.skipEvent(0)
362	<	else:  # pythia like job
363	<	PsetEdit.maxEvent(self.eventsPerJob)
364	<	if (self.firstRun):
365	<	PsetEdit.pythiaFirstRun(0)  #First Run
366	<	if (self.sourceSeed) :
315	<	PsetEdit.pythiaSeed(0)
316	<	if (self.sourceSeedVtx) :
317	<	PsetEdit.vtxSeed(0)
318	<	if (self.sourceSeedG4) :
319	<	PsetEdit.g4Seed(0)
320	<	if (self.sourceSeedMix) :
321	<	PsetEdit.mixSeed(0)
322	<	# add FrameworkJobReport to parameter-set
323	<	PsetEdit.addCrabFJR(self.fjrFileName)
324	<	PsetEdit.psetWriter(self.configFilename())
325	<	except:
326	<	msg='Error while manipuliating ParameterSet: exiting...'
354	>	if (PsetEdit.getBadFilesSetting()):
355	>	msg = "WARNING: You have set skipBadFiles to True. This will continue processing on some errors and you may not be notified."
356	>	common.logger.info(msg)
357	>
358	>	except CrabException, msg:
359	>	common.logger.info(str(msg))
360	>	msg='Error while manipulating ParameterSet (see previous message, if any): exiting...'
361	>	raise CrabException(msg)
362	>
363	>	valid = re.compile('^[\w\.\-]+$')
364	>	for fileName in self.output_file:
365	>	if not valid.match(fileName):
366	>	msg = "The file %s may only contain alphanumeric characters and -, _, ." % fileName
367		raise CrabException(msg)
368
369	+
370		def DataDiscoveryAndLocation(self, cfg_params):
371
372		import DataDiscovery
373		import DataLocation
374	<	common.logger.debug(10,"CMSSW::DataDiscoveryAndLocation()")
374	>	common.logger.log(10-1,"CMSSW::DataDiscoveryAndLocation()")
375
376		datasetPath=self.datasetPath
377
378		## Contact the DBS
379	<	common.logger.message("Contacting Data Discovery Services ...")
379	>	common.logger.info("Contacting Data Discovery Services ...")
380		try:
381	<
341	<	self.pubdata=DataDiscovery.DataDiscovery(datasetPath, cfg_params)
381	>	self.pubdata=DataDiscovery.DataDiscovery(datasetPath, cfg_params,self.skip_blocks)
382		self.pubdata.fetchDBSInfo()
383
384		except DataDiscovery.NotExistingDatasetError, ex :
#	Line 352 | Line 392 | class Cmssw(JobType):
392		raise CrabException(msg)
393
394		self.filesbyblock=self.pubdata.getFiles()
395	<	self.eventsbyblock=self.pubdata.getEventsPerBlock()
356	<	self.eventsbyfile=self.pubdata.getEventsPerFile()
395	>	self.conf['pubdata']=self.pubdata
396
397		## get max number of events
398	<	self.maxEvents=self.pubdata.getMaxEvents() ##  self.maxEvents used in Creator.py
398	>	self.maxEvents=self.pubdata.getMaxEvents()
399
400		## Contact the DLS and build a list of sites hosting the fileblocks
401		try:
402		dataloc=DataLocation.DataLocation(self.filesbyblock.keys(),cfg_params)
403		dataloc.fetchDLSInfo()
404	+
405		except DataLocation.DataLocationError , ex:
406		msg = 'ERROR ***: failed Data Location in DLS \n %s '%ex.getErrorMessage()
407		raise CrabException(msg)
408
409
410	<	sites = dataloc.getSites()
410	>	unsorted_sites = dataloc.getSites()
411	>	sites = self.filesbyblock.fromkeys(self.filesbyblock,'')
412	>	for lfn in self.filesbyblock.keys():
413	>	if unsorted_sites.has_key(lfn):
414	>	sites[lfn]=unsorted_sites[lfn]
415	>	else:
416	>	sites[lfn]=[]
417	>
418	>	if len(sites)==0:
419	>	msg = 'ERROR ***: no location for any of the blocks of this dataset: \n\t %s \n'%datasetPath
420	>	msg += "\tMaybe the dataset is located only at T1's (or at T0), where analysis jobs are not allowed\n"
421	>	msg += "\tPlease check DataDiscovery page https://cmsweb.cern.ch/dbs_discovery/\n"
422	>	raise CrabException(msg)
423	>
424		allSites = []
425		listSites = sites.values()
426		for listSite in listSites:
427		for oneSite in listSite:
428		allSites.append(oneSite)
429	<	allSites = self.uniquelist(allSites)
377	<
378	<	# screen output
379	<	common.logger.message("Requested dataset: " + datasetPath + " has " + str(self.maxEvents) + " events in " + str(len(self.filesbyblock.keys())) + " blocks.\n")
380	<
381	<	return sites
429	>	[allSites.append(it) for it in allSites if not allSites.count(it)]
430
383	–	def setArgsList(self, argsList):
384	–	self.argsList = argsList
431
432	<	def jobSplittingByBlocks(self, blockSites):
433	<	"""
434	<	Perform job splitting. Jobs run over an integer number of files
435	<	and no more than one block.
390	<	ARGUMENT: blockSites: dictionary with blocks as keys and list of host sites as values
391	<	REQUIRES: self.selectTotalNumberEvents, self.selectEventsPerJob, self.selectNumberofJobs,
392	<	self.total_number_of_events, self.eventsPerJob, self.theNumberOfJobs,
393	<	self.maxEvents, self.filesbyblock
394	<	SETS: self.jobDestination - Site destination(s) for each job (a list of lists)
395	<	self.total_number_of_jobs - Total # of jobs
396	<	self.list_of_args - File(s) job will run on (a list of lists)
397	<	"""
398	<
399	<	# ---- Handle the possible job splitting configurations ---- #
400	<	if (self.selectTotalNumberEvents):
401	<	totalEventsRequested = self.total_number_of_events
402	<	if (self.selectEventsPerJob):
403	<	eventsPerJobRequested = self.eventsPerJob
404	<	if (self.selectNumberOfJobs):
405	<	totalEventsRequested = self.theNumberOfJobs * self.eventsPerJob
406	<
407	<	# If user requested all the events in the dataset
408	<	if (totalEventsRequested == -1):
409	<	eventsRemaining=self.maxEvents
410	<	# If user requested more events than are in the dataset
411	<	elif (totalEventsRequested > self.maxEvents):
412	<	eventsRemaining = self.maxEvents
413	<	common.logger.message("Requested "+str(self.total_number_of_events)+ " events, but only "+str(self.maxEvents)+" events are available.")
414	<	# If user requested less events than are in the dataset
432	>	# screen output
433	>	if self.ads or self.lumiMask:
434	>	common.logger.info("Requested (A)DS %s has %s block(s)." %
435	>	(datasetPath, len(self.filesbyblock.keys())))
436		else:
437	<	eventsRemaining = totalEventsRequested
438	<
439	<	# If user requested more events per job than are in the dataset
419	<	if (self.selectEventsPerJob and eventsPerJobRequested > self.maxEvents):
420	<	eventsPerJobRequested = self.maxEvents
421	<
422	<	# For user info at end
423	<	totalEventCount = 0
424	<
425	<	if (self.selectTotalNumberEvents and self.selectNumberOfJobs):
426	<	eventsPerJobRequested = int(eventsRemaining/self.theNumberOfJobs)
427	<
428	<	if (self.selectNumberOfJobs):
429	<	common.logger.message("May not create the exact number_of_jobs requested.")
430	<
431	<	if ( self.ncjobs == 'all' ) :
432	<	totalNumberOfJobs = 999999999
433	<	else :
434	<	totalNumberOfJobs = self.ncjobs
437	>	common.logger.info("Requested dataset: " + datasetPath + \
438	>	" has " + str(self.maxEvents) + " events in " + \
439	>	str(len(self.filesbyblock.keys())) + " blocks.\n")
440
441	+	return sites
442
437	–	blocks = blockSites.keys()
438	–	blockCount = 0
439	–	# Backup variable in case self.maxEvents counted events in a non-included block
440	–	numBlocksInDataset = len(blocks)
441	–
442	–	jobCount = 0
443	–	list_of_lists = []
444	–
445	–	# list tracking which jobs are in which jobs belong to which block
446	–	jobsOfBlock = {}
447	–
448	–	# ---- Iterate over the blocks in the dataset until ---- #
449	–	# ---- we've met the requested total # of events    ---- #
450	–	while ( (eventsRemaining > 0) and (blockCount < numBlocksInDataset) and (jobCount < totalNumberOfJobs)):
451	–	block = blocks[blockCount]
452	–	blockCount += 1
453	–	if block not in jobsOfBlock.keys() :
454	–	jobsOfBlock[block] = []
455	–
456	–	if self.eventsbyblock.has_key(block) :
457	–	numEventsInBlock = self.eventsbyblock[block]
458	–	common.logger.debug(5,'Events in Block File '+str(numEventsInBlock))
459	–
460	–	files = self.filesbyblock[block]
461	–	numFilesInBlock = len(files)
462	–	if (numFilesInBlock <= 0):
463	–	continue
464	–	fileCount = 0
465	–
466	–	# ---- New block => New job ---- #
467	–	parString = ""
468	–	# counter for number of events in files currently worked on
469	–	filesEventCount = 0
470	–	# flag if next while loop should touch new file
471	–	newFile = 1
472	–	# job event counter
473	–	jobSkipEventCount = 0
474	–
475	–	# ---- Iterate over the files in the block until we've met the requested ---- #
476	–	# ---- total # of events or we've gone over all the files in this block  ---- #
477	–	while ( (eventsRemaining > 0) and (fileCount < numFilesInBlock) and (jobCount < totalNumberOfJobs) ):
478	–	file = files[fileCount]
479	–	if newFile :
480	–	try:
481	–	numEventsInFile = self.eventsbyfile[file]
482	–	common.logger.debug(6, "File "+str(file)+" has "+str(numEventsInFile)+" events")
483	–	# increase filesEventCount
484	–	filesEventCount += numEventsInFile
485	–	# Add file to current job
486	–	parString += '\\\"' + file + '\\\"\,'
487	–	newFile = 0
488	–	except KeyError:
489	–	common.logger.message("File "+str(file)+" has unknown number of events: skipping")
490	–
491	–
492	–	# if less events in file remain than eventsPerJobRequested
493	–	if ( filesEventCount - jobSkipEventCount < eventsPerJobRequested ) :
494	–	# if last file in block
495	–	if ( fileCount == numFilesInBlock-1 ) :
496	–	# end job using last file, use remaining events in block
497	–	# close job and touch new file
498	–	fullString = parString[:-2]
499	–	list_of_lists.append([fullString,str(-1),str(jobSkipEventCount)])
500	–	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(filesEventCount - jobSkipEventCount)+" events (last file in block).")
501	–	self.jobDestination.append(blockSites[block])
502	–	common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
503	–	# fill jobs of block dictionary
504	–	jobsOfBlock[block].append(jobCount+1)
505	–	# reset counter
506	–	jobCount = jobCount + 1
507	–	totalEventCount = totalEventCount + filesEventCount - jobSkipEventCount
508	–	eventsRemaining = eventsRemaining - filesEventCount + jobSkipEventCount
509	–	jobSkipEventCount = 0
510	–	# reset file
511	–	parString = ""
512	–	filesEventCount = 0
513	–	newFile = 1
514	–	fileCount += 1
515	–	else :
516	–	# go to next file
517	–	newFile = 1
518	–	fileCount += 1
519	–	# if events in file equal to eventsPerJobRequested
520	–	elif ( filesEventCount - jobSkipEventCount == eventsPerJobRequested ) :
521	–	# close job and touch new file
522	–	fullString = parString[:-2]
523	–	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
524	–	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.")
525	–	self.jobDestination.append(blockSites[block])
526	–	common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
527	–	jobsOfBlock[block].append(jobCount+1)
528	–	# reset counter
529	–	jobCount = jobCount + 1
530	–	totalEventCount = totalEventCount + eventsPerJobRequested
531	–	eventsRemaining = eventsRemaining - eventsPerJobRequested
532	–	jobSkipEventCount = 0
533	–	# reset file
534	–	parString = ""
535	–	filesEventCount = 0
536	–	newFile = 1
537	–	fileCount += 1
443
444	<	# if more events in file remain than eventsPerJobRequested
540	<	else :
541	<	# close job but don't touch new file
542	<	fullString = parString[:-2]
543	<	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
544	<	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.")
545	<	self.jobDestination.append(blockSites[block])
546	<	common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
547	<	jobsOfBlock[block].append(jobCount+1)
548	<	# increase counter
549	<	jobCount = jobCount + 1
550	<	totalEventCount = totalEventCount + eventsPerJobRequested
551	<	eventsRemaining = eventsRemaining - eventsPerJobRequested
552	<	# calculate skip events for last file
553	<	# use filesEventCount (contains several files), jobSkipEventCount and eventsPerJobRequest
554	<	jobSkipEventCount = eventsPerJobRequested - (filesEventCount - jobSkipEventCount - self.eventsbyfile[file])
555	<	# remove all but the last file
556	<	filesEventCount = self.eventsbyfile[file]
557	<	parString = ""
558	<	parString += '\\\"' + file + '\\\"\,'
559	<	pass # END if
560	<	pass # END while (iterate over files in the block)
561	<	pass # END while (iterate over blocks in the dataset)
562	<	self.ncjobs = self.total_number_of_jobs = jobCount
563	<	if (eventsRemaining > 0 and jobCount < totalNumberOfJobs ):
564	<	common.logger.message("Could not run on all requested events because some blocks not hosted at allowed sites.")
565	<	common.logger.message(str(jobCount)+" job(s) can run on "+str(totalEventCount)+" events.\n")
444	>	def split(self, jobParams,firstJobID):
445
446	<	# screen output
447	<	screenOutput = "List of jobs and available destination sites:\n\n"
446	>	jobParams = self.dict['args']
447	>	njobs = self.dict['njobs']
448	>	self.jobDestination = self.dict['jobDestination']
449	>
450	>	if njobs == 0:
451	>	raise CrabException("Asked to split zero jobs: aborting")
452	>	if not self.server and not self.local and njobs > 500:
453	>	raise CrabException("The CRAB client will not submit more than 500 jobs. You must use the server mode.")
454
455	<	# keep trace of block with no sites to print a warning at the end
456	<	noSiteBlock = []
457	<	bloskNoSite = []
573	<
574	<	blockCounter = 0
575	<	for block in blocks:
576	<	if block in jobsOfBlock.keys() :
577	<	blockCounter += 1
578	<	screenOutput += "Block %5i: jobs %20s: sites: %s\n" % (blockCounter,spanRanges(jobsOfBlock[block]),','.join(self.blackWhiteListParser.checkWhiteList(self.blackWhiteListParser.checkBlackList(blockSites[block],block),block)))
579	<	if len(self.blackWhiteListParser.checkWhiteList(self.blackWhiteListParser.checkBlackList(blockSites[block],block),block)) == 0:
580	<	noSiteBlock.append( spanRanges(jobsOfBlock[block]) )
581	<	bloskNoSite.append( blockCounter )
582	<
583	<	common.logger.message(screenOutput)
584	<	if len(noSiteBlock) > 0 and len(bloskNoSite) > 0:
585	<	msg = 'WARNING: No sites are hosting any part of data for block:\n                '
586	<	virgola = ""
587	<	if len(bloskNoSite) > 1:
588	<	virgola = ","
589	<	for block in bloskNoSite:
590	<	msg += ' ' + str(block) + virgola
591	<	msg += '\n               Related jobs:\n                 '
592	<	virgola = ""
593	<	if len(noSiteBlock) > 1:
594	<	virgola = ","
595	<	for range_jobs in noSiteBlock:
596	<	msg += str(range_jobs) + virgola
597	<	msg += '\n               will not be submitted and this block of data can not be analyzed!\n'
598	<	common.logger.message(msg)
455	>	# create the empty structure
456	>	for i in range(njobs):
457	>	jobParams.append("")
458
459	<	self.list_of_args = list_of_lists
459	>	listID=[]
460	>	listField=[]
461	>	listDictions=[]
462	>	exist= os.path.exists(self.argsFile)
463	>	for id in range(njobs):
464	>	job = id + int(firstJobID)
465	>	listID.append(job+1)
466	>	job_ToSave ={}
467	>	concString = ' '
468	>	argu=''
469	>	str_argu = str(job+1)
470	>	if len(jobParams[id]):
471	>	argu = {'JobID': job+1}
472	>	for i in range(len(jobParams[id])):
473	>	argu[self.dict['params'][i]]=jobParams[id][i]
474	>	if len(jobParams[id])==1: self.NumEvents = jobParams[id][i]
475	>	# just for debug
476	>	str_argu += concString.join(jobParams[id])
477	>	if argu != '': listDictions.append(argu)
478	>	job_ToSave['arguments']= '%d %d'%( (job+1), 0)
479	>	job_ToSave['dlsDestination']= self.jobDestination[id]
480	>	listField.append(job_ToSave)
481	>	from ProdCommon.SiteDB.CmsSiteMapper import CmsSEMap
482	>	cms_se = CmsSEMap()
483	>	msg="Job  %s  Arguments:  %s\n"%(str(job+1),str_argu)
484	>	msg+="\t  Destination: %s "%(str(self.jobDestination[id]))
485	>	SEDestination = [cms_se[dest] for dest in self.jobDestination[id]]
486	>	msg+="\t  CMSDestination: %s "%(str(SEDestination))
487	>	common.logger.log(10-1,msg)
488	>	# write xml
489	>	if len(listDictions):
490	>	if exist==False: self.CreateXML()
491	>	self.addEntry(listDictions)
492	>	common._db.updateJob_(listID,listField)
493		return
494
495	<	def jobSplittingNoInput(self):
495	>	def CreateXML(self):
496		"""
605	–	Perform job splitting based on number of event per job
497		"""
498	<	common.logger.debug(5,'Splitting per events')
499	<
609	<	if (self.selectEventsPerJob):
610	<	common.logger.message('Required '+str(self.eventsPerJob)+' events per job ')
611	<	if (self.selectNumberOfJobs):
612	<	common.logger.message('Required '+str(self.theNumberOfJobs)+' jobs in total ')
613	<	if (self.selectTotalNumberEvents):
614	<	common.logger.message('Required '+str(self.total_number_of_events)+' events in total ')
615	<
616	<	if (self.total_number_of_events < 0):
617	<	msg='Cannot split jobs per Events with "-1" as total number of events'
618	<	raise CrabException(msg)
619	<
620	<	if (self.selectEventsPerJob):
621	<	if (self.selectTotalNumberEvents):
622	<	self.total_number_of_jobs = int(self.total_number_of_events/self.eventsPerJob)
623	<	elif(self.selectNumberOfJobs) :
624	<	self.total_number_of_jobs =self.theNumberOfJobs
625	<	self.total_number_of_events =int(self.theNumberOfJobs*self.eventsPerJob)
626	<
627	<	elif (self.selectNumberOfJobs) :
628	<	self.total_number_of_jobs = self.theNumberOfJobs
629	<	self.eventsPerJob = int(self.total_number_of_events/self.total_number_of_jobs)
630	<
631	<	common.logger.debug(5,'N jobs  '+str(self.total_number_of_jobs))
632	<
633	<	# is there any remainder?
634	<	check = int(self.total_number_of_events) - (int(self.total_number_of_jobs)*self.eventsPerJob)
635	<
636	<	common.logger.debug(5,'Check  '+str(check))
637	<
638	<	common.logger.message(str(self.total_number_of_jobs)+' jobs can be created, each for '+str(self.eventsPerJob)+' for a total of '+str(self.total_number_of_jobs*self.eventsPerJob)+' events')
639	<	if check > 0:
640	<	common.logger.message('Warning: asked '+str(self.total_number_of_events)+' but can do only '+str(int(self.total_number_of_jobs)*self.eventsPerJob))
641	<
642	<	# argument is seed number.$i
643	<	self.list_of_args = []
644	<	for i in range(self.total_number_of_jobs):
645	<	## Since there is no input, any site is good
646	<	self.jobDestination.append([""]) #must be empty to write correctly the xml
647	<	args=[]
648	<	if (self.firstRun):
649	<	## pythia first run
650	<	args.append(str(self.firstRun)+str(i))
651	<	if (self.sourceSeed):
652	<	args.append(str(self.sourceSeed)+str(i))
653	<	if (self.sourceSeedVtx):
654	<	## + vtx random seed
655	<	args.append(str(self.sourceSeedVtx)+str(i))
656	<	if (self.sourceSeedG4):
657	<	## + G4 random seed
658	<	args.append(str(self.sourceSeedG4)+str(i))
659	<	if (self.sourceSeedMix):
660	<	## + Mix random seed
661	<	args.append(str(self.sourceSeedMix)+str(i))
662	<	pass
663	<	pass
664	<	self.list_of_args.append(args)
665	<	pass
666	<
498	>	result = IMProvNode( self.rootArgsFilename )
499	>	outfile = file( self.argsFile, 'w').write(str(result))
500		return
501
502	<
670	<	def jobSplittingForScript(self):#CarlosDaniele
671	<	"""
672	<	Perform job splitting based on number of job
502	>	def addEntry(self, listDictions):
503		"""
504	<	common.logger.debug(5,'Splitting per job')
675	<	common.logger.message('Required '+str(self.theNumberOfJobs)+' jobs in total ')
504	>	_addEntry_
505
506	<	self.total_number_of_jobs = self.theNumberOfJobs
507	<
508	<	common.logger.debug(5,'N jobs  '+str(self.total_number_of_jobs))
509	<
510	<	common.logger.message(str(self.total_number_of_jobs)+' jobs can be created')
511	<
512	<	# argument is seed number.$i
513	<	self.list_of_args = []
514	<	for i in range(self.total_number_of_jobs):
686	<	## Since there is no input, any site is good
687	<	# self.jobDestination.append(["Any"])
688	<	self.jobDestination.append([""])
689	<	## no random seed
690	<	self.list_of_args.append([str(i)])
691	<	return
692	<
693	<	def split(self, jobParams):
694	<
695	<	common.jobDB.load()
696	<	#### Fabio
697	<	njobs = self.total_number_of_jobs
698	<	arglist = self.list_of_args
699	<	# create the empty structure
700	<	for i in range(njobs):
701	<	jobParams.append("")
702	<
703	<	for job in range(njobs):
704	<	jobParams[job] = arglist[job]
705	<	# print str(arglist[job])
706	<	# print jobParams[job]
707	<	common.jobDB.setArguments(job, jobParams[job])
708	<	common.logger.debug(5,"Job "+str(job)+" Destination: "+str(self.jobDestination[job]))
709	<	common.jobDB.setDestination(job, self.jobDestination[job])
710	<
711	<	common.jobDB.save()
506	>	add an entry to the xml file
507	>	"""
508	>	## load xml
509	>	improvDoc = loadIMProvFile(self.argsFile)
510	>	entrname= 'Job'
511	>	for dictions in listDictions:
512	>	report = IMProvNode(entrname , None, **dictions)
513	>	improvDoc.addNode(report)
514	>	outfile = file( self.argsFile, 'w').write(str(improvDoc))
515		return
516
714	–	def getJobTypeArguments(self, nj, sched):
715	–	result = ''
716	–	for i in common.jobDB.arguments(nj):
717	–	result=result+str(i)+" "
718	–	return result
719	–
517		def numberOfJobs(self):
518	<	# Fabio
519	<	return self.total_number_of_jobs
518	>	#wmbs
519	>	if self.automation==0:
520	>	return self.dict['njobs']
521	>	else:
522	>	return None
523
524		def getTarBall(self, exe):
525		"""
526		Return the TarBall with lib and exe
527		"""
528	<
729	<	# if it exist, just return it
730	<	#
731	<	# Marco. Let's start to use relative path for Boss XML files
732	<	#
733	<	self.tgzNameWithPath = common.work_space.pathForTgz()+'share/'+self.tgz_name
528	>	self.tgzNameWithPath = common.work_space.pathForTgz()+self.tgz_name
529		if os.path.exists(self.tgzNameWithPath):
530		return self.tgzNameWithPath
531
#	Line 743 | Line 538 | class Cmssw(JobType):
538
539		# First of all declare the user Scram area
540		swArea = self.scram.getSWArea_()
746	–	#print "swArea = ", swArea
747	–	# swVersion = self.scram.getSWVersion()
748	–	# print "swVersion = ", swVersion
541		swReleaseTop = self.scram.getReleaseTop_()
750	–	#print "swReleaseTop = ", swReleaseTop
542
543		## check if working area is release top
544		if swReleaseTop == '' or swArea == swReleaseTop:
545	+	common.logger.debug("swArea = "+swArea+" swReleaseTop ="+swReleaseTop)
546		return
547
548		import tarfile
#	Line 765 | Line 557 | class Cmssw(JobType):
557		## then check if it's private or not
558		if exeWithPath.find(swReleaseTop) == -1:
559		# the exe is private, so we must ship
560	<	common.logger.debug(5,"Exe "+exeWithPath+" to be tarred")
560	>	common.logger.debug("Exe "+exeWithPath+" to be tarred")
561		path = swArea+'/'
562		# distinguish case when script is in user project area or given by full path somewhere else
563		if exeWithPath.find(path) >= 0 :
#	Line 779 | Line 571 | class Cmssw(JobType):
571		pass
572
573		## Now get the libraries: only those in local working area
574	+	tar.dereference=True
575		libDir = 'lib'
576		lib = swArea+'/' +libDir
577	<	common.logger.debug(5,"lib "+lib+" to be tarred")
577	>	common.logger.debug("lib "+lib+" to be tarred")
578		if os.path.exists(lib):
579		tar.add(lib,libDir)
580
#	Line 790 | Line 583 | class Cmssw(JobType):
583		module = swArea + '/' + moduleDir
584		if os.path.isdir(module):
585		tar.add(module,moduleDir)
586	+	tar.dereference=False
587
588		## Now check if any data dir(s) is present
589	<	swAreaLen=len(swArea)
590	<	for root, dirs, files in os.walk(swArea):
591	<	if "data" in dirs:
592	<	common.logger.debug(5,"data "+root+"/data"+" to be tarred")
593	<	tar.add(root+"/data",root[swAreaLen:]+"/data")
594	<
595	<	### Removed ProdAgent Api dependencies ###
596	<	### Add ProdAgent dir to tar
597	<	#paDir = 'ProdAgentApi'
598	<	#pa = os.environ['CRABDIR'] + '/' + 'ProdAgentApi'
599	<	#if os.path.isdir(pa):
600	<	#    tar.add(pa,paDir)
589	>	self.dataExist = False
590	>	todo_list = [(i, i) for i in  os.listdir(swArea+"/src")]
591	>	while len(todo_list):
592	>	entry, name = todo_list.pop()
593	>	if name.startswith('crab_0_') or  name.startswith('.') or name == 'CVS':
594	>	continue
595	>	if os.path.isdir(swArea+"/src/"+entry):
596	>	entryPath = entry + '/'
597	>	todo_list += [(entryPath + i, i) for i in  os.listdir(swArea+"/src/"+entry)]
598	>	if name == 'data':
599	>	self.dataExist=True
600	>	common.logger.debug("data "+entry+" to be tarred")
601	>	tar.add(swArea+"/src/"+entry,"src/"+entry)
602	>	pass
603	>	pass
604
605	<	## Add ProdCommon dir to tar
606	<	prodcommonDir = 'ProdCommon'
607	<	prodcommonPath = os.environ['CRABDIR'] + '/' + 'ProdCommon'
608	<	if os.path.isdir(prodcommonPath):
609	<	tar.add(prodcommonPath,prodcommonDir)
605	>	### CMSSW ParameterSet
606	>	if not self.pset is None:
607	>	cfg_file = common.work_space.jobDir()+self.configFilename()
608	>	pickleFile = common.work_space.jobDir()+self.configFilename() + '.pkl'
609	>	tar.add(cfg_file,self.configFilename())
610	>	tar.add(pickleFile,self.configFilename() + '.pkl')
611
612	<	common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
613	<	tar.close()
614	<	except :
615	<	raise CrabException('Could not create tar-ball')
612	>	try:
613	>	crab_cfg_file = common.work_space.shareDir()+'/crab.cfg'
614	>	tar.add(crab_cfg_file,'crab.cfg')
615	>	except:
616	>	pass
617
618	<	## check for tarball size
619	<	tarballinfo = os.stat(self.tgzNameWithPath)
620	<	if ( tarballinfo.st_size > self.MaxTarBallSize*1024*1024 ) :
621	<	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.')
618	>	## Add ProdCommon dir to tar
619	>	prodcommonDir = './'
620	>	prodcommonPath = os.environ['CRABDIR'] + '/' + 'external/'
621	>	neededStuff = ['ProdCommon/__init__.py','ProdCommon/FwkJobRep', 'ProdCommon/CMSConfigTools', \
622	>	'ProdCommon/Core', 'ProdCommon/MCPayloads', 'IMProv', 'ProdCommon/Storage', \
623	>	'WMCore/__init__.py','WMCore/Algorithms']
624	>	for file in neededStuff:
625	>	tar.add(prodcommonPath+file,prodcommonDir+file)
626
627	<	## create tar-ball with ML stuff
628	<	self.MLtgzfile =  common.work_space.pathForTgz()+'share/MLfiles.tgz'
826	<	try:
827	<	tar = tarfile.open(self.MLtgzfile, "w:gz")
627	>	##### ML stuff
628	>	ML_file_list=['report.py', 'DashboardAPI.py', 'Logger.py', 'ProcInfo.py', 'apmon.py']
629		path=os.environ['CRABDIR'] + '/python/'
630	<	for file in ['report.py', 'DashboardAPI.py', 'Logger.py', 'ProcInfo.py', 'apmon.py', 'parseCrabFjr.py']:
630	>	for file in ML_file_list:
631		tar.add(path+file,file)
632	<	common.logger.debug(5,"Files added to "+self.MLtgzfile+" : "+str(tar.getnames()))
632	>
633	>	##### Utils
634	>	Utils_file_list=['parseCrabFjr.py','writeCfg.py', 'fillCrabFjr.py','cmscp.py']
635	>	for file in Utils_file_list:
636	>	tar.add(path+file,file)
637	>
638	>	##### AdditionalFiles
639	>	tar.dereference=True
640	>	for file in self.additional_inbox_files:
641	>	tar.add(file,string.split(file,'/')[-1])
642	>	tar.dereference=False
643	>	common.logger.log(10-1,"Files in "+self.tgzNameWithPath+" : "+str(tar.getnames()))
644	>
645		tar.close()
646	<	except :
647	<	raise CrabException('Could not create ML files tar-ball')
646	>	except IOError, exc:
647	>	msg = 'Could not create tar-ball %s \n'%self.tgzNameWithPath
648	>	msg += str(exc)
649	>	raise CrabException(msg)
650	>	except tarfile.TarError, exc:
651	>	msg = 'Could not create tar-ball %s \n'%self.tgzNameWithPath
652	>	msg += str(exc)
653	>	raise CrabException(msg)
654
655	<	return
655	>	tarballinfo = os.stat(self.tgzNameWithPath)
656	>	if ( tarballinfo.st_size > self.MaxTarBallSize*1024*1024 ) :
657	>	if not self.server:
658	>	msg  = 'Input sandbox size of ' + str(float(tarballinfo.st_size)/1024.0/1024.0) + ' MB is larger than the allowed ' + \
659	>	str(self.MaxTarBallSize) +'MB input sandbox limit \n'
660	>	msg += '      and not supported by the direct GRID submission system.\n'
661	>	msg += '      Please use the CRAB server mode by setting server_name=<NAME> in section [CRAB] of your crab.cfg.\n'
662	>	msg += '      For further infos please see https://twiki.cern.ch/twiki/bin/view/CMS/SWGuideCrabServerForUsers#Server_available_for_users'
663	>	else:
664	>	msg  = 'Input sandbox size of ' + str(float(tarballinfo.st_size)/1024.0/1024.0) + ' MB is larger than the allowed ' +  \
665	>	str(self.MaxTarBallSize) +'MB input sandbox limit in the server.'
666	>	raise CrabException(msg)
667
668	<	def additionalInputFileTgz(self):
839	<	"""
840	<	Put all additional files into a tar ball and return its name
841	<	"""
842	<	import tarfile
843	<	tarName=  common.work_space.pathForTgz()+'share/'+self.additional_tgz_name
844	<	tar = tarfile.open(tarName, "w:gz")
845	<	for file in self.additional_inbox_files:
846	<	tar.add(file,string.split(file,'/')[-1])
847	<	common.logger.debug(5,"Files added to "+self.additional_tgz_name+" : "+str(tar.getnames()))
848	<	tar.close()
849	<	return tarName
668	>	## create tar-ball with ML stuff
669
670	<	def wsSetupEnvironment(self, nj):
670	>	def wsSetupEnvironment(self, nj=0):
671		"""
672		Returns part of a job script which prepares
673		the execution environment for the job 'nj'.
674		"""
675	+	psetName = 'pset.py'
676	+
677		# Prepare JobType-independent part
678	<	txt = ''
678	>	txt = '\n#Written by cms_cmssw::wsSetupEnvironment\n'
679		txt += 'echo ">>> setup environment"\n'
680	<	txt += 'if [ $middleware == LCG ]; then \n'
680	>	txt += 'echo "set SCRAM ARCH to ' + self.executable_arch + '"\n'
681	>	txt += 'export SCRAM_ARCH=' + self.executable_arch + '\n'
682	>	txt += 'echo "SCRAM_ARCH = $SCRAM_ARCH"\n'
683	>	txt += 'if [ $middleware == LCG ] || [ $middleware == CAF ] || [ $middleware == LSF ]; then \n'
684		txt += self.wsSetupCMSLCGEnvironment_()
685		txt += 'elif [ $middleware == OSG ]; then\n'
686		txt += '    WORKING_DIR=`/bin/mktemp  -d $OSG_WN_TMP/cms_XXXXXXXXXXXX`\n'
687		txt += '    if [ ! $? == 0 ] ;then\n'
688	<	txt += '        echo "SET_CMS_ENV 10016 ==> OSG $WORKING_DIR could not be created on WN `hostname`"\n'
689	<	txt += '        echo "JOB_EXIT_STATUS = 10016"\n'
690	<	txt += '        echo "JobExitCode=10016" | tee -a $RUNTIME_AREA/$repo\n'
867	<	txt += '        dumpStatus $RUNTIME_AREA/$repo\n'
868	<	txt += '        exit 1\n'
688	>	txt += '        echo "ERROR ==> OSG $WORKING_DIR could not be created on WN `hostname`"\n'
689	>	txt += '        job_exit_code=10016\n'
690	>	txt += '        func_exit\n'
691		txt += '    fi\n'
692		txt += '    echo ">>> Created working directory: $WORKING_DIR"\n'
693		txt += '\n'
#	Line 873 | Line 695 | class Cmssw(JobType):
695		txt += '    cd $WORKING_DIR\n'
696		txt += '    echo ">>> current directory (WORKING_DIR): $WORKING_DIR"\n'
697		txt += self.wsSetupCMSOSGEnvironment_()
698	<	#txt += '    echo "### Set SCRAM ARCH to ' + self.executable_arch + ' ###"\n'
699	<	#txt += '    export SCRAM_ARCH='+self.executable_arch+'\n'
698	>	#Setup SGE Environment
699	>	txt += 'elif [ $middleware == SGE ]; then\n'
700	>	txt += self.wsSetupCMSLCGEnvironment_()
701	>
702	>	txt += 'elif [ $middleware == ARC ]; then\n'
703	>	txt += self.wsSetupCMSLCGEnvironment_()
704	>
705	>	#Setup PBS Environment
706	>	txt += 'elif [ $middleware == PBS ]; then\n'
707	>	txt += self.wsSetupCMSLCGEnvironment_()
708	>
709		txt += 'fi\n'
710
711		# Prepare JobType-specific part
#	Line 885 | Line 716 | class Cmssw(JobType):
716		txt += scram+' project CMSSW '+self.version+'\n'
717		txt += 'status=$?\n'
718		txt += 'if [ $status != 0 ] ; then\n'
719	<	txt += '    echo "SET_EXE_ENV 10034 ==>ERROR CMSSW '+self.version+' not found on `hostname`" \n'
720	<	txt += '    echo "JOB_EXIT_STATUS = 10034"\n'
721	<	txt += '    echo "JobExitCode=10034" | tee -a $RUNTIME_AREA/$repo\n'
891	<	txt += '    dumpStatus $RUNTIME_AREA/$repo\n'
892	<	txt += '    if [ $middleware == OSG ]; then \n'
893	<	txt += '        cd $RUNTIME_AREA\n'
894	<	txt += '        echo ">>> current directory (RUNTIME_AREA): $RUNTIME_AREA"\n'
895	<	txt += '        echo ">>> Remove working directory: $WORKING_DIR"\n'
896	<	txt += '        /bin/rm -rf $WORKING_DIR\n'
897	<	txt += '        if [ -d $WORKING_DIR ] ;then\n'
898	<	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'
899	<	txt += '            echo "JOB_EXIT_STATUS = 10018"\n'
900	<	txt += '            echo "JobExitCode=10018" | tee -a $RUNTIME_AREA/$repo\n'
901	<	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
902	<	txt += '        fi\n'
903	<	txt += '    fi \n'
904	<	txt += '    exit 1 \n'
719	>	txt += '    echo "ERROR ==> CMSSW '+self.version+' not found on `hostname`" \n'
720	>	txt += '    job_exit_code=10034\n'
721	>	txt += '    func_exit\n'
722		txt += 'fi \n'
723		txt += 'cd '+self.version+'\n'
724	<	########## FEDE FOR DBS2 ######################
908	<	txt += 'SOFTWARE_DIR=`pwd`\n'
724	>	txt += 'SOFTWARE_DIR=`pwd`; export SOFTWARE_DIR\n'
725		txt += 'echo ">>> current directory (SOFTWARE_DIR): $SOFTWARE_DIR" \n'
910	–	###############################################
911	–	### needed grep for bug in scramv1 ###
726		txt += 'eval `'+scram+' runtime -sh | grep -v SCRAMRT_LSB_JOBNAME`\n'
727	+	txt += 'if [ $? != 0 ] ; then\n'
728	+	txt += '    echo "ERROR ==> Problem with the command: "\n'
729	+	txt += '    echo "eval \`'+scram+' runtime -sh | grep -v SCRAMRT_LSB_JOBNAME \` at `hostname`"\n'
730	+	txt += '    job_exit_code=10034\n'
731	+	txt += '    func_exit\n'
732	+	txt += 'fi \n'
733		# Handle the arguments:
734		txt += "\n"
735	<	txt += "## number of arguments (first argument always jobnumber)\n"
735	>	txt += "## number of arguments (first argument always jobnumber, the second is the resubmission number)\n"
736		txt += "\n"
737	<	txt += "if [ $nargs -lt "+str(len(self.argsList[nj].split()))+" ]\n"
737	>	txt += "if [ $nargs -lt "+str(self.argsList)+" ]\n"
738		txt += "then\n"
739	<	txt += "    echo 'SET_EXE_ENV 1 ==> ERROR Too few arguments' +$nargs+ \n"
740	<	txt += '    echo "JOB_EXIT_STATUS = 50113"\n'
741	<	txt += '    echo "JobExitCode=50113" | tee -a $RUNTIME_AREA/$repo\n'
922	<	txt += '    dumpStatus $RUNTIME_AREA/$repo\n'
923	<	txt += '    if [ $middleware == OSG ]; then \n'
924	<	txt += '        cd $RUNTIME_AREA\n'
925	<	txt += '        echo ">>> current directory (RUNTIME_AREA): $RUNTIME_AREA"\n'
926	<	txt += '        echo ">>> Remove working directory: $WORKING_DIR"\n'
927	<	txt += '        /bin/rm -rf $WORKING_DIR\n'
928	<	txt += '        if [ -d $WORKING_DIR ] ;then\n'
929	<	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'
930	<	txt += '            echo "JOB_EXIT_STATUS = 50114"\n'
931	<	txt += '            echo "JobExitCode=50114" | tee -a $RUNTIME_AREA/$repo\n'
932	<	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
933	<	txt += '        fi\n'
934	<	txt += '    fi \n'
935	<	txt += "    exit 1\n"
739	>	txt += "    echo 'ERROR ==> Too few arguments' +$nargs+ \n"
740	>	txt += '    job_exit_code=50113\n'
741	>	txt += "    func_exit\n"
742		txt += "fi\n"
743		txt += "\n"
744
745		# Prepare job-specific part
746		job = common.job_list[nj]
941	–	### FEDE FOR DBS OUTPUT PUBLICATION
747		if (self.datasetPath):
748		txt += '\n'
749		txt += 'DatasetPath='+self.datasetPath+'\n'
750
751	<	datasetpath_split = self.datasetPath.split("/")
752	<
948	<	txt += 'PrimaryDataset='+datasetpath_split[1]+'\n'
949	<	txt += 'DataTier='+datasetpath_split[2]+'\n'
751	>	txt += 'PrimaryDataset='+self.primaryDataset +'\n'
752	>	txt += 'DataTier='+self.dataTier+'\n'
753		txt += 'ApplicationFamily=cmsRun\n'
754
755		else:
#	Line 954 | Line 757 | class Cmssw(JobType):
757		txt += 'PrimaryDataset=null\n'
758		txt += 'DataTier=null\n'
759		txt += 'ApplicationFamily=MCDataTier\n'
760	<	if self.pset != None: #CarlosDaniele
760	>	if self.pset != None:
761		pset = os.path.basename(job.configFilename())
762	+	pkl  = os.path.basename(job.configFilename()) + '.pkl'
763		txt += '\n'
764		txt += 'cp  $RUNTIME_AREA/'+pset+' .\n'
765	<	if (self.datasetPath): # standard job
962	<	txt += 'InputFiles=${args[1]}\n'
963	<	txt += 'MaxEvents=${args[2]}\n'
964	<	txt += 'SkipEvents=${args[3]}\n'
965	<	txt += 'echo "Inputfiles:<$InputFiles>"\n'
966	<	txt += 'sed "s#\'INPUTFILE\'#$InputFiles#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
967	<	txt += 'echo "MaxEvents:<$MaxEvents>"\n'
968	<	txt += 'sed "s#int32 input = 0#int32 input = $MaxEvents#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
969	<	txt += 'echo "SkipEvents:<$SkipEvents>"\n'
970	<	txt += 'sed "s#uint32 skipEvents = 0#uint32 skipEvents = $SkipEvents#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
971	<	else:  # pythia like job
972	<	seedIndex=1
973	<	if (self.firstRun):
974	<	txt += 'FirstRun=${args['+str(seedIndex)+']}\n'
975	<	txt += 'echo "FirstRun: <$FirstRun>"\n'
976	<	txt += 'sed "s#uint32 firstRun = 0#uint32 firstRun = $FirstRun#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
977	<	seedIndex=seedIndex+1
978	<
979	<	if (self.sourceSeed):
980	<	txt += 'Seed=${args['+str(seedIndex)+']}\n'
981	<	txt += 'sed "s#uint32 sourceSeed = 0#uint32 sourceSeed = $Seed#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
982	<	seedIndex=seedIndex+1
983	<	## the following seeds are not always present
984	<	if (self.sourceSeedVtx):
985	<	txt += 'VtxSeed=${args['+str(seedIndex)+']}\n'
986	<	txt += 'echo "VtxSeed: <$VtxSeed>"\n'
987	<	txt += 'sed "s#uint32 VtxSmeared = 0#uint32 VtxSmeared = $VtxSeed#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
988	<	seedIndex += 1
989	<	if (self.sourceSeedG4):
990	<	txt += 'G4Seed=${args['+str(seedIndex)+']}\n'
991	<	txt += 'echo "G4Seed: <$G4Seed>"\n'
992	<	txt += 'sed "s#uint32 g4SimHits = 0#uint32 g4SimHits = $G4Seed#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
993	<	seedIndex += 1
994	<	if (self.sourceSeedMix):
995	<	txt += 'mixSeed=${args['+str(seedIndex)+']}\n'
996	<	txt += 'echo "MixSeed: <$mixSeed>"\n'
997	<	txt += 'sed "s#uint32 mix = 0#uint32 mix = $mixSeed#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
998	<	seedIndex += 1
999	<	pass
1000	<	pass
1001	<	txt += 'mv -f '+pset+' pset.cfg\n'
765	>	txt += 'cp  $RUNTIME_AREA/'+pkl+' .\n'
766
767	<	if len(self.additional_inbox_files) > 0:
768	<	txt += 'if [ -e $RUNTIME_AREA/'+self.additional_tgz_name+' ] ; then\n'
769	<	txt += '  tar xzvf $RUNTIME_AREA/'+self.additional_tgz_name+'\n'
770	<	txt += 'fi\n'
771	<	pass
772	<
773	<	if self.pset != None: #CarlosDaniele
774	<	txt += '\n'
775	<	txt += 'echo "***** cat pset.cfg *********"\n'
776	<	txt += 'cat pset.cfg\n'
777	<	txt += 'echo "****** end pset.cfg ********"\n'
1014	<	txt += '\n'
1015	<	### FEDE FOR DBS OUTPUT PUBLICATION
1016	<	txt += 'PSETHASH=`EdmConfigHash < pset.cfg` \n'
1017	<	txt += 'echo "PSETHASH = $PSETHASH" \n'
1018	<	##############
767	>	txt += 'PreserveSeeds='  + ','.join(self.preserveSeeds)  + '; export PreserveSeeds\n'
768	>	txt += 'IncrementSeeds=' + ','.join(self.incrementSeeds) + '; export IncrementSeeds\n'
769	>	txt += 'echo "PreserveSeeds: <$PreserveSeeds>"\n'
770	>	txt += 'echo "IncrementSeeds:<$IncrementSeeds>"\n'
771	>
772	>	txt += 'mv -f ' + pset + ' ' + psetName + '\n'
773	>	if self.var_filter:
774	>	#print "self.var_filter = ",self.var_filter
775	>	txt += "export var_filter="+"'"+self.var_filter+"'\n"
776	>	txt += 'echo $var_filter'
777	>	else:
778		txt += '\n'
779	+	if self.AdditionalArgs: txt += 'export AdditionalArgs=\"%s\"\n'%(self.AdditionalArgs)
780	+	if int(self.NumEvents) != 0: txt += 'export MaxEvents=%s\n'%str(self.NumEvents)
781		return txt
782
783	<	def wsBuildExe(self, nj=0):
783	>	def wsUntarSoftware(self, nj=0):
784		"""
785		Put in the script the commands to build an executable
786		or a library.
787		"""
788
789	<	txt = ""
789	>	txt = '\n#Written by cms_cmssw::wsUntarSoftware\n'
790
791		if os.path.isfile(self.tgzNameWithPath):
792	<	txt += 'echo ">>> tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+' :" \n'
793	<	txt += 'tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+'\n'
792	>	txt += 'echo ">>> tar xzf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+' :" \n'
793	>	if  self.debug_wrapper==1 :
794	>	txt += 'tar zxvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+'\n'
795	>	txt += 'ls -Al \n'
796	>	else:
797	>	txt += 'tar zxf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+'\n'
798		txt += 'untar_status=$? \n'
799		txt += 'if [ $untar_status -ne 0 ]; then \n'
800	<	txt += '   echo "SET_EXE 1 ==> ERROR Untarring .tgz file failed"\n'
801	<	txt += '   echo "JOB_EXIT_STATUS = $untar_status" \n'
802	<	txt += '   echo "JobExitCode=$untar_status" | tee -a $RUNTIME_AREA/$repo\n'
1038	<	txt += '   if [ $middleware == OSG ]; then \n'
1039	<	txt += '       cd $RUNTIME_AREA\n'
1040	<	txt += '        echo ">>> current directory (RUNTIME_AREA): $RUNTIME_AREA"\n'
1041	<	txt += '        echo ">>> Remove working directory: $WORKING_DIR"\n'
1042	<	txt += '       /bin/rm -rf $WORKING_DIR\n'
1043	<	txt += '       if [ -d $WORKING_DIR ] ;then\n'
1044	<	txt += '           echo "SET_EXE 50999 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after Untarring .tgz file failed"\n'
1045	<	txt += '           echo "JOB_EXIT_STATUS = 50999"\n'
1046	<	txt += '           echo "JobExitCode=50999" | tee -a $RUNTIME_AREA/$repo\n'
1047	<	txt += '           dumpStatus $RUNTIME_AREA/$repo\n'
1048	<	txt += '       fi\n'
1049	<	txt += '   fi \n'
1050	<	txt += '   \n'
1051	<	txt += '   exit 1 \n'
800	>	txt += '   echo "ERROR ==> Untarring .tgz file failed"\n'
801	>	txt += '   job_exit_code=$untar_status\n'
802	>	txt += '   func_exit\n'
803		txt += 'else \n'
804		txt += '   echo "Successful untar" \n'
805		txt += 'fi \n'
806		txt += '\n'
807	<	#### Removed ProdAgent API dependencies
1057	<	txt += 'echo ">>> Include ProdCommon in PYTHONPATH:"\n'
807	>	txt += 'echo ">>> Include $RUNTIME_AREA in PYTHONPATH:"\n'
808		txt += 'if [ -z "$PYTHONPATH" ]; then\n'
809	<	#### FEDE FOR DBS OUTPUT PUBLICATION
1060	<	txt += '   export PYTHONPATH=$SOFTWARE_DIR/ProdCommon\n'
809	>	txt += '   export PYTHONPATH=$RUNTIME_AREA/\n'
810		txt += 'else\n'
811	<	txt += '   export PYTHONPATH=$SOFTWARE_DIR/ProdCommon:${PYTHONPATH}\n'
811	>	txt += '   export PYTHONPATH=$RUNTIME_AREA/:${PYTHONPATH}\n'
812		txt += 'echo "PYTHONPATH=$PYTHONPATH"\n'
1064	–	###################
813		txt += 'fi\n'
814		txt += '\n'
815
#	Line 1069 | Line 817 | class Cmssw(JobType):
817
818		return txt
819
820	<	def modifySteeringCards(self, nj):
820	>	def wsBuildExe(self, nj=0):
821		"""
822	<	modify the card provided by the user,
823	<	writing a new card into share dir
822	>	Put in the script the commands to build an executable
823	>	or a library.
824		"""
825
826	+	txt = '\n#Written by cms_cmssw::wsBuildExe\n'
827	+	txt += 'echo ">>> moving CMSSW software directories in `pwd`" \n'
828	+
829	+	txt += 'rm -r lib/ module/ \n'
830	+	txt += 'mv $RUNTIME_AREA/lib/ . \n'
831	+	txt += 'mv $RUNTIME_AREA/module/ . \n'
832	+	if self.dataExist == True:
833	+	txt += 'rm -r src/ \n'
834	+	txt += 'mv $RUNTIME_AREA/src/ . \n'
835	+	if len(self.additional_inbox_files)>0:
836	+	for file in self.additional_inbox_files:
837	+	txt += 'mv $RUNTIME_AREA/'+os.path.basename(file)+' . \n'
838	+
839	+	txt += 'echo ">>> Include $RUNTIME_AREA in PYTHONPATH:"\n'
840	+	txt += 'if [ -z "$PYTHONPATH" ]; then\n'
841	+	txt += '   export PYTHONPATH=$RUNTIME_AREA/\n'
842	+	txt += 'else\n'
843	+	txt += '   export PYTHONPATH=$RUNTIME_AREA/:${PYTHONPATH}\n'
844	+	txt += 'echo "PYTHONPATH=$PYTHONPATH"\n'
845	+	txt += 'fi\n'
846	+	txt += '\n'
847	+
848	+	if self.pset != None:
849	+	psetName = 'pset.py'
850	+
851	+	txt += '\n'
852	+	if self.debug_wrapper == 1:
853	+	txt += 'echo "***** cat ' + psetName + ' *********"\n'
854	+	txt += 'cat ' + psetName + '\n'
855	+	txt += 'echo "****** end ' + psetName + ' ********"\n'
856	+	txt += '\n'
857	+	txt += 'echo "***********************" \n'
858	+	txt += 'which edmConfigHash \n'
859	+	txt += 'echo "***********************" \n'
860	+	txt += 'edmConfigHash ' + psetName + ' \n'
861	+	txt += 'PSETHASH=`edmConfigHash ' + psetName + '` \n'
862	+	txt += 'echo "PSETHASH = $PSETHASH" \n'
863	+	#### FEDE temporary fix for noEdm files #####
864	+	txt += 'if [ -z "$PSETHASH" ]; then \n'
865	+	txt += '   export PSETHASH=null\n'
866	+	txt += 'fi \n'
867	+	#############################################
868	+	txt += '\n'
869	+	return txt
870	+
871	+
872		def executableName(self):
873	<	if self.scriptExe: #CarlosDaniele
873	>	if self.scriptExe:
874		return "sh "
875		else:
876		return self.executable
877
878		def executableArgs(self):
879	<	if self.scriptExe:#CarlosDaniele
880	<	return   self.scriptExe + " $NJob"
879	>	if self.scriptExe:
880	>	return self.scriptExe + " $NJob $AdditionalArgs"
881		else:
882	<	# if >= CMSSW_1_5_X, add -j crab_fjr.xml
1089	<	version_array = self.scram.getSWVersion().split('_')
1090	<	major = 0
1091	<	minor = 0
1092	<	try:
1093	<	major = int(version_array[1])
1094	<	minor = int(version_array[2])
1095	<	except:
1096	<	msg = "Cannot parse CMSSW version string: " + "_".join(version_array) + " for major and minor release number!"
1097	<	raise CrabException(msg)
1098	<	if major >= 1 and minor >= 5 :
1099	<	return " -j " + self.fjrFileName + " -p pset.cfg"
1100	<	else:
1101	<	return " -p pset.cfg"
882	>	return " -j $RUNTIME_AREA/crab_fjr_$NJob.xml -p pset.py"
883
884		def inputSandbox(self, nj):
885		"""
886		Returns a list of filenames to be put in JDL input sandbox.
887		"""
888		inp_box = []
1108	–	# # dict added to delete duplicate from input sandbox file list
1109	–	# seen = {}
1110	–	## code
889		if os.path.isfile(self.tgzNameWithPath):
890		inp_box.append(self.tgzNameWithPath)
891	<	if os.path.isfile(self.MLtgzfile):
892	<	inp_box.append(self.MLtgzfile)
893	<	## config
1116	<	if not self.pset is None:
1117	<	inp_box.append(common.work_space.pathForTgz() + 'job/' + self.configFilename())
1118	<	## additional input files
1119	<	tgz = self.additionalInputFileTgz()
1120	<	inp_box.append(tgz)
891	>	if os.path.isfile(self.argsFile):
892	>	inp_box.append(self.argsFile)
893	>	inp_box.append(common.work_space.jobDir() + self.scriptName)
894		return inp_box
895
896		def outputSandbox(self, nj):
#	Line 1129 | Line 902 | class Cmssw(JobType):
902		## User Declared output files
903		for out in (self.output_file+self.output_file_sandbox):
904		n_out = nj + 1
905	<	out_box.append(self.numberFile_(out,str(n_out)))
905	>	out_box.append(numberFile(out,str(n_out)))
906		return out_box
907
1135	–	def prepareSteeringCards(self):
1136	–	"""
1137	–	Make initial modifications of the user's steering card file.
1138	–	"""
1139	–	return
908
909		def wsRenameOutput(self, nj):
910		"""
911		Returns part of a job script which renames the produced files.
912		"""
913
914	<	txt = '\n'
914	>	txt = '\n#Written by cms_cmssw::wsRenameOutput\n'
915		txt += 'echo ">>> current directory (SOFTWARE_DIR): $SOFTWARE_DIR" \n'
916		txt += 'echo ">>> current directory content:"\n'
917	<	txt += 'ls \n'
917	>	if self.debug_wrapper==1:
918	>	txt += 'ls -Al\n'
919		txt += '\n'
920
1152	–	txt += 'output_exit_status=0\n'
1153	–
1154	–	for fileWithSuffix in (self.output_file_sandbox):
1155	–	output_file_num = self.numberFile_(fileWithSuffix, '$NJob')
1156	–	txt += '\n'
1157	–	txt += '# check output file\n'
1158	–	txt += 'if [ -e ./'+fileWithSuffix+' ] ; then\n'
1159	–	txt += '    mv '+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n'
1160	–	txt += '    ln -s $RUNTIME_AREA/'+output_file_num+' $RUNTIME_AREA/'+fileWithSuffix+'\n'
1161	–	txt += 'else\n'
1162	–	txt += '    exit_status=60302\n'
1163	–	txt += '    echo "ERROR: Output file '+fileWithSuffix+' not found"\n'
1164	–	if common.scheduler.boss_scheduler_name == 'condor_g':
1165	–	txt += '    if [ $middleware == OSG ]; then \n'
1166	–	txt += '        echo "prepare dummy output file"\n'
1167	–	txt += '        echo "Processing of job output failed" > $RUNTIME_AREA/'+output_file_num+'\n'
1168	–	txt += '    fi \n'
1169	–	txt += 'fi\n'
1170	–
921		for fileWithSuffix in (self.output_file):
922	<	output_file_num = self.numberFile_(fileWithSuffix, '$NJob')
922	>	output_file_num = numberFile(fileWithSuffix, '$OutUniqueID')
923		txt += '\n'
924		txt += '# check output file\n'
925		txt += 'if [ -e ./'+fileWithSuffix+' ] ; then\n'
#	Line 1180 | Line 930 | class Cmssw(JobType):
930		txt += '    mv '+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n'
931		txt += '    ln -s $RUNTIME_AREA/'+output_file_num+' $RUNTIME_AREA/'+fileWithSuffix+'\n'
932		txt += 'else\n'
933	<	txt += '    exit_status=60302\n'
934	<	txt += '    echo "ERROR: Output file '+fileWithSuffix+' not found"\n'
935	<	txt += '    echo "JOB_EXIT_STATUS = $exit_status"\n'
1186	<	txt += '    output_exit_status=$exit_status\n'
1187	<	if common.scheduler.boss_scheduler_name == 'condor_g':
933	>	txt += '    job_exit_code=60302\n'
934	>	txt += '    echo "WARNING: Output file '+fileWithSuffix+' not found"\n'
935	>	if common.scheduler.name().upper() == 'CONDOR_G':
936		txt += '    if [ $middleware == OSG ]; then \n'
937		txt += '        echo "prepare dummy output file"\n'
938		txt += '        echo "Processing of job output failed" > $RUNTIME_AREA/'+output_file_num+'\n'
#	Line 1192 | Line 940 | class Cmssw(JobType):
940		txt += 'fi\n'
941		file_list = []
942		for fileWithSuffix in (self.output_file):
943	<	file_list.append(self.numberFile_(fileWithSuffix, '$NJob'))
943	>	file_list.append(numberFile('$SOFTWARE_DIR/'+fileWithSuffix, '$OutUniqueID'))
944
945	<	txt += 'file_list="'+string.join(file_list,' ')+'"\n'
945	>	txt += 'file_list="'+string.join(file_list,',')+'"\n'
946		txt += '\n'
947		txt += 'echo ">>> current directory (SOFTWARE_DIR): $SOFTWARE_DIR" \n'
948		txt += 'echo ">>> current directory content:"\n'
949	<	txt += 'ls \n'
949	>	if self.debug_wrapper==1:
950	>	txt += 'ls -Al\n'
951		txt += '\n'
952		txt += 'cd $RUNTIME_AREA\n'
953		txt += 'echo ">>> current directory (RUNTIME_AREA):  $RUNTIME_AREA"\n'
954		return txt
955
1207	–	def numberFile_(self, file, txt):
1208	–	"""
1209	–	append _'txt' before last extension of a file
1210	–	"""
1211	–	p = string.split(file,".")
1212	–	# take away last extension
1213	–	name = p[0]
1214	–	for x in p[1:-1]:
1215	–	name=name+"."+x
1216	–	# add "_txt"
1217	–	if len(p)>1:
1218	–	ext = p[len(p)-1]
1219	–	result = name + '_' + txt + "." + ext
1220	–	else:
1221	–	result = name + '_' + txt
1222	–
1223	–	return result
1224	–
956		def getRequirements(self, nj=[]):
957		"""
958		return job requirements to add to jdl files
#	Line 1231 | Line 962 | class Cmssw(JobType):
962		req='Member("VO-cms-' + \
963		self.version + \
964		'", other.GlueHostApplicationSoftwareRunTimeEnvironment)'
965	<	## SL add requirement for OS version only if SL4
1235	<	#reSL4 = re.compile( r'slc4' )
1236	<	if self.executable_arch: # and reSL4.search(self.executable_arch):
965	>	if self.executable_arch:
966		req+=' && Member("VO-cms-' + \
967		self.executable_arch + \
968		'", other.GlueHostApplicationSoftwareRunTimeEnvironment)'
969
970		req = req + ' && (other.GlueHostNetworkAdapterOutboundIP)'
971	+	if ( common.scheduler.name() in ["glite"] ):
972	+	## 25-Jun-2009 SL: patch to use Cream enabled WMS
973	+	if ( self.cfg_params.get('GRID.use_cream',None) ):
974	+	req += ' && (other.GlueCEStateStatus == "Production" || other.GlueCEStateStatus == "Special")'
975	+	else:
976	+	req += ' && other.GlueCEStateStatus == "Production" '
977
978		return req
979
980		def configFilename(self):
981		""" return the config filename """
982	<	return self.name()+'.cfg'
982	>	return self.name()+'.py'
983
984		def wsSetupCMSOSGEnvironment_(self):
985		"""
986		Returns part of a job script which is prepares
987		the execution environment and which is common for all CMS jobs.
988		"""
989	<	txt = '    echo ">>> setup CMS OSG environment:"\n'
989	>	txt = '\n#Written by cms_cmssw::wsSetupCMSOSGEnvironment_\n'
990	>	txt += '    echo ">>> setup CMS OSG environment:"\n'
991		txt += '    echo "set SCRAM ARCH to ' + self.executable_arch + '"\n'
992		txt += '    export SCRAM_ARCH='+self.executable_arch+'\n'
993		txt += '    echo "SCRAM_ARCH = $SCRAM_ARCH"\n'
#	Line 1259 | Line 995 | class Cmssw(JobType):
995		txt += '      # Use $OSG_APP/cmssoft/cms/cmsset_default.sh to setup cms software\n'
996		txt += '        source $OSG_APP/cmssoft/cms/cmsset_default.sh '+self.version+'\n'
997		txt += '    else\n'
998	<	txt += '        echo "SET_CMS_ENV 10020 ==> ERROR $OSG_APP/cmssoft/cms/cmsset_default.sh file not found"\n'
999	<	txt += '        echo "JOB_EXIT_STATUS = 10020"\n'
1000	<	txt += '        echo "JobExitCode=10020" | tee -a $RUNTIME_AREA/$repo\n'
1265	<	txt += '        dumpStatus $RUNTIME_AREA/$repo\n'
1266	<	txt += '\n'
1267	<	txt += '        cd $RUNTIME_AREA\n'
1268	<	txt += '        echo ">>> current directory (RUNTIME_AREA): $RUNTIME_AREA"\n'
1269	<	txt += '        echo ">>> Remove working directory: $WORKING_DIR"\n'
1270	<	txt += '        /bin/rm -rf $WORKING_DIR\n'
1271	<	txt += '        if [ -d $WORKING_DIR ] ;then\n'
1272	<	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'
1273	<	txt += '            echo "JOB_EXIT_STATUS = 10017"\n'
1274	<	txt += '            echo "JobExitCode=10017" | tee -a $RUNTIME_AREA/$repo\n'
1275	<	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
1276	<	txt += '        fi\n'
1277	<	txt += '\n'
1278	<	txt += '        exit 1\n'
998	>	txt += '        echo "ERROR ==> $OSG_APP/cmssoft/cms/cmsset_default.sh file not found"\n'
999	>	txt += '        job_exit_code=10020\n'
1000	>	txt += '        func_exit\n'
1001		txt += '    fi\n'
1002		txt += '\n'
1003	<	txt += '    echo "SET_CMS_ENV 0 ==> setup cms environment ok"\n'
1003	>	txt += '    echo "==> setup cms environment ok"\n'
1004		txt += '    echo "SCRAM_ARCH = $SCRAM_ARCH"\n'
1005
1006		return txt
1007
1286	–	### OLI_DANIELE
1008		def wsSetupCMSLCGEnvironment_(self):
1009		"""
1010		Returns part of a job script which is prepares
1011		the execution environment and which is common for all CMS jobs.
1012		"""
1013	<	txt = '    echo ">>> setup CMS LCG environment:"\n'
1013	>	txt = '\n#Written by cms_cmssw::wsSetupCMSLCGEnvironment_\n'
1014	>	txt += '    echo ">>> setup CMS LCG environment:"\n'
1015		txt += '    echo "set SCRAM ARCH and BUILD_ARCH to ' + self.executable_arch + ' ###"\n'
1016		txt += '    export SCRAM_ARCH='+self.executable_arch+'\n'
1017		txt += '    export BUILD_ARCH='+self.executable_arch+'\n'
1018		txt += '    if [ ! $VO_CMS_SW_DIR ] ;then\n'
1019	<	txt += '        echo "SET_CMS_ENV 10031 ==> ERROR CMS software dir not found on WN `hostname`"\n'
1020	<	txt += '        echo "JOB_EXIT_STATUS = 10031" \n'
1021	<	txt += '        echo "JobExitCode=10031" | tee -a $RUNTIME_AREA/$repo\n'
1300	<	txt += '        dumpStatus $RUNTIME_AREA/$repo\n'
1301	<	txt += '        exit 1\n'
1019	>	txt += '        echo "ERROR ==> CMS software dir not found on WN `hostname`"\n'
1020	>	txt += '        job_exit_code=10031\n'
1021	>	txt += '        func_exit\n'
1022		txt += '    else\n'
1023		txt += '        echo "Sourcing environment... "\n'
1024		txt += '        if [ ! -s $VO_CMS_SW_DIR/cmsset_default.sh ] ;then\n'
1025	<	txt += '            echo "SET_CMS_ENV 10020 ==> ERROR cmsset_default.sh file not found into dir $VO_CMS_SW_DIR"\n'
1026	<	txt += '            echo "JOB_EXIT_STATUS = 10020"\n'
1027	<	txt += '            echo "JobExitCode=10020" | tee -a $RUNTIME_AREA/$repo\n'
1308	<	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
1309	<	txt += '            exit 1\n'
1025	>	txt += '            echo "ERROR ==> cmsset_default.sh file not found into dir $VO_CMS_SW_DIR"\n'
1026	>	txt += '            job_exit_code=10020\n'
1027	>	txt += '            func_exit\n'
1028		txt += '        fi\n'
1029		txt += '        echo "sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
1030		txt += '        source $VO_CMS_SW_DIR/cmsset_default.sh\n'
1031		txt += '        result=$?\n'
1032		txt += '        if [ $result -ne 0 ]; then\n'
1033	<	txt += '            echo "SET_CMS_ENV 10032 ==> ERROR problem sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
1034	<	txt += '            echo "JOB_EXIT_STATUS = 10032"\n'
1035	<	txt += '            echo "JobExitCode=10032" | tee -a $RUNTIME_AREA/$repo\n'
1318	<	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
1319	<	txt += '            exit 1\n'
1033	>	txt += '            echo "ERROR ==> problem sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
1034	>	txt += '            job_exit_code=10032\n'
1035	>	txt += '            func_exit\n'
1036		txt += '        fi\n'
1037		txt += '    fi\n'
1038		txt += '    \n'
1039	<	txt += '    echo "SET_CMS_ENV 0 ==> setup cms environment ok"\n'
1039	>	txt += '    echo "==> setup cms environment ok"\n'
1040		return txt
1041
1042	<	### FEDE FOR DBS OUTPUT PUBLICATION
1327	<	def modifyReport(self, nj):
1042	>	def wsModifyReport(self, nj):
1043		"""
1044		insert the part of the script that modifies the FrameworkJob Report
1045		"""
1046
1047		txt = ''
1048	<	try:
1049	<	publish_data = int(self.cfg_params['USER.publish_data'])
1050	<	except KeyError:
1336	<	publish_data = 0
1337	<	if (publish_data == 1):
1048	>	if (self.copy_data == 1):
1049	>	txt = '\n#Written by cms_cmssw::wsModifyReport\n'
1050	>
1051		txt += 'echo ">>> Modify Job Report:" \n'
1052	<	################ FEDE FOR DBS2 #############################################
1053	<	#txt += 'chmod a+x $SOFTWARE_DIR/ProdAgentApi/FwkJobRep/ModifyJobReport.py\n'
1341	<	txt += 'chmod a+x $SOFTWARE_DIR/ProdCommon/ProdCommon/FwkJobRep/ModifyJobReport.py\n'
1342	<	#############################################################################
1052	>	txt += 'chmod a+x $RUNTIME_AREA/ProdCommon/FwkJobRep/ModifyJobReport.py\n'
1053	>	txt += 'echo "CMSSW_VERSION = $CMSSW_VERSION"\n\n'
1054
1055	<	txt += 'if [ -z "$SE" ]; then\n'
1345	<	txt += '    SE="" \n'
1346	<	txt += 'fi \n'
1347	<	txt += 'if [ -z "$SE_PATH" ]; then\n'
1348	<	txt += '    SE_PATH="" \n'
1349	<	txt += 'fi \n'
1350	<	txt += 'echo "SE = $SE"\n'
1351	<	txt += 'echo "SE_PATH = $SE_PATH"\n'
1055	>	args = 'fjr $RUNTIME_AREA/crab_fjr_$NJob.xml json $RUNTIME_AREA/resultCopyFile n_job $OutUniqueID PrimaryDataset $PrimaryDataset  ApplicationFamily $ApplicationFamily ApplicationName $executable cmssw_version $CMSSW_VERSION psethash $PSETHASH'
1056
1057	<	processedDataset = self.cfg_params['USER.publish_data_name']
1058	<	txt += 'ProcessedDataset='+processedDataset+'\n'
1059	<	#### LFN=/store/user/<user>/processedDataset_PSETHASH
1060	<	txt += 'if [ "$SE_PATH" == "" ]; then\n'
1357	<	#### FEDE: added slash in LFN ##############
1358	<	txt += '    FOR_LFN=/copy_problems/ \n'
1359	<	txt += 'else \n'
1360	<	txt += '    tmp=`echo $SE_PATH | awk -F \'store\' \'{print$2}\'` \n'
1361	<	#####  FEDE TO BE CHANGED, BECAUSE STORE IS HARDCODED!!!! ########
1362	<	txt += '    FOR_LFN=/store$tmp \n'
1363	<	txt += 'fi \n'
1364	<	txt += 'echo "ProcessedDataset = $ProcessedDataset"\n'
1365	<	txt += 'echo "FOR_LFN = $FOR_LFN" \n'
1366	<	txt += 'echo "CMSSW_VERSION = $CMSSW_VERSION"\n\n'
1367	<	#txt += 'echo "$SOFTWARE_DIR/ProdAgentApi/FwkJobRep/ModifyJobReport.py crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH"\n'
1368	<	#txt += '$SOFTWARE_DIR/ProdAgentApi/FwkJobRep/ModifyJobReport.py crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH\n'
1369	<	txt += 'echo "$SOFTWARE_DIR/ProdCommon/ProdCommon/FwkJobRep/ModifyJobReport.py crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH"\n'
1370	<	txt += '$SOFTWARE_DIR/ProdCommon/ProdCommon/FwkJobRep/ModifyJobReport.py crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH\n'
1057	>	if (self.publish_data == 1):
1058	>	txt += 'ProcessedDataset='+self.processedDataset+'\n'
1059	>	txt += 'echo "ProcessedDataset = $ProcessedDataset"\n'
1060	>	args += ' UserProcessedDataset $USER-$ProcessedDataset-$PSETHASH'
1061
1062	+	txt += 'echo "$RUNTIME_AREA/ProdCommon/FwkJobRep/ModifyJobReport.py '+str(args)+'"\n'
1063	+	txt += '$RUNTIME_AREA/ProdCommon/FwkJobRep/ModifyJobReport.py '+str(args)+'\n'
1064		txt += 'modifyReport_result=$?\n'
1373	–	txt += 'echo modifyReport_result = $modifyReport_result\n'
1065		txt += 'if [ $modifyReport_result -ne 0 ]; then\n'
1066	<	txt += '    exit_status=1\n'
1067	<	txt += '    echo "ERROR: Problem with ModifyJobReport"\n'
1066	>	txt += '    modifyReport_result=70500\n'
1067	>	txt += '    job_exit_code=$modifyReport_result\n'
1068	>	txt += '    echo "ModifyReportResult=$modifyReport_result" | tee -a $RUNTIME_AREA/$repo\n'
1069	>	txt += '    echo "WARNING: Problem with ModifyJobReport"\n'
1070		txt += 'else\n'
1071	<	txt += '    mv NewFrameworkJobReport.xml crab_fjr_$NJob.xml\n'
1071	>	txt += '    mv NewFrameworkJobReport.xml $RUNTIME_AREA/crab_fjr_$NJob.xml\n'
1072		txt += 'fi\n'
1380	–	else:
1381	–	txt += 'echo "no data publication required"\n'
1073		return txt
1074
1075	<	def cleanEnv(self):
1076	<	txt = ''
1077	<	txt += 'if [ $middleware == OSG ]; then\n'
1078	<	txt += '    cd $RUNTIME_AREA\n'
1079	<	txt += '    echo ">>> current directory (RUNTIME_AREA): $RUNTIME_AREA"\n'
1080	<	txt += '    echo ">>> Remove working directory: $WORKING_DIR"\n'
1081	<	txt += '    /bin/rm -rf $WORKING_DIR\n'
1082	<	txt += '    if [ -d $WORKING_DIR ] ;then\n'
1083	<	txt += '        echo "SET_EXE 60999 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after cleanup of WN"\n'
1084	<	txt += '        echo "JOB_EXIT_STATUS = 60999"\n'
1085	<	txt += '        echo "JobExitCode=60999" | tee -a $RUNTIME_AREA/$repo\n'
1086	<	txt += '        dumpStatus $RUNTIME_AREA/$repo\n'
1075	>	def wsParseFJR(self):
1076	>	"""
1077	>	Parse the FrameworkJobReport to obtain useful infos
1078	>	"""
1079	>	txt = '\n#Written by cms_cmssw::wsParseFJR\n'
1080	>	txt += 'echo ">>> Parse FrameworkJobReport crab_fjr.xml"\n'
1081	>	txt += 'if [ -s $RUNTIME_AREA/crab_fjr_$NJob.xml ]; then\n'
1082	>	txt += '    if [ -s $RUNTIME_AREA/parseCrabFjr.py ]; then\n'
1083	>	txt += '        cmd_out=`python $RUNTIME_AREA/parseCrabFjr.py --input $RUNTIME_AREA/crab_fjr_$NJob.xml --dashboard $MonitorID,$MonitorJobID '+self.debugWrap+'`\n'
1084	>	if self.debug_wrapper==1 :
1085	>	txt += '        echo "Result of parsing the FrameworkJobReport crab_fjr.xml: $cmd_out"\n'
1086	>	txt += '        executable_exit_status=`python $RUNTIME_AREA/parseCrabFjr.py --input $RUNTIME_AREA/crab_fjr_$NJob.xml --exitcode`\n'
1087	>	txt += '        if [ $executable_exit_status -eq 50115 ];then\n'
1088	>	txt += '            echo ">>> crab_fjr.xml contents: "\n'
1089	>	txt += '            cat $RUNTIME_AREA/crab_fjr_$NJob.xml\n'
1090	>	txt += '            echo "Wrong FrameworkJobReport --> does not contain useful info. ExitStatus: $executable_exit_status"\n'
1091	>	txt += '        elif [ $executable_exit_status -eq -999 ];then\n'
1092	>	txt += '            echo "ExitStatus from FrameworkJobReport not available. not available. Using exit code of executable from command line."\n'
1093	>	txt += '        else\n'
1094	>	txt += '            echo "Extracted ExitStatus from FrameworkJobReport parsing output: $executable_exit_status"\n'
1095	>	txt += '        fi\n'
1096	>	txt += '    else\n'
1097	>	txt += '        echo "CRAB python script to parse CRAB FrameworkJobReport crab_fjr.xml is not available, using exit code of executable from command line."\n'
1098	>	txt += '    fi\n'
1099	>	#### Patch to check input data reading for CMSSW16x Hopefully we-ll remove it asap
1100	>	txt += '    if [ $executable_exit_status -eq 0 ];then\n'
1101	>	txt += '        echo ">>> Executable succeded  $executable_exit_status"\n'
1102		txt += '    fi\n'
1103	+	txt += 'else\n'
1104	+	txt += '    echo "CRAB FrameworkJobReport crab_fjr.xml is not available, using exit code of executable from command line."\n'
1105		txt += 'fi\n'
1106		txt += '\n'
1107	+	txt += 'if [ $executable_exit_status -ne 0 ];then\n'
1108	+	txt += '    echo ">>> Executable failed  $executable_exit_status"\n'
1109	+	txt += '    echo "ExeExitCode=$executable_exit_status" | tee -a $RUNTIME_AREA/$repo\n'
1110	+	txt += '    echo "EXECUTABLE_EXIT_STATUS = $executable_exit_status"\n'
1111	+	txt += '    job_exit_code=$executable_exit_status\n'
1112	+	txt += '    func_exit\n'
1113	+	txt += 'fi\n\n'
1114	+	txt += 'echo "ExeExitCode=$executable_exit_status" | tee -a $RUNTIME_AREA/$repo\n'
1115	+	txt += 'echo "EXECUTABLE_EXIT_STATUS = $executable_exit_status"\n'
1116	+	txt += 'job_exit_code=$executable_exit_status\n'
1117	+
1118		return txt
1119
1120		def setParam_(self, param, value):
#	Line 1404 | Line 1123 | class Cmssw(JobType):
1123		def getParams(self):
1124		return self._params
1125
1126	<	def uniquelist(self, old):
1408	<	"""
1409	<	remove duplicates from a list
1410	<	"""
1411	<	nd={}
1412	<	for e in old:
1413	<	nd[e]=0
1414	<	return nd.keys()
1415	<
1416	<
1417	<	def checkOut(self, limit):
1126	>	def outList(self,list=False):
1127		"""
1128		check the dimension of the output files
1129		"""
1130	<	txt = 'echo ">>> Starting output sandbox limit check :"\n'
1131	<	allOutFiles = ""
1130	>	txt = ''
1131	>	txt += 'echo ">>> list of expected files on output sandbox"\n'
1132		listOutFiles = []
1133	<	txt += 'stdoutFile=`ls *stdout` \n'
1134	<	txt += 'stderrFile=`ls *stderr` \n'
1133	>	stdout = 'CMSSW_$NJob.stdout'
1134	>	stderr = 'CMSSW_$NJob.stderr'
1135	>	if len(self.output_file) <= 0:
1136	>	msg ="WARNING: no output files name have been defined!!\n"
1137	>	msg+="\tno output files will be reported back/staged\n"
1138	>	common.logger.info(msg)
1139	>
1140		if (self.return_data == 1):
1141	<	for fileOut in (self.output_file+self.output_file_sandbox):
1142	<	allOutFiles = allOutFiles + " " + self.numberFile_(fileOut, '$NJob') + " $stdoutFile $stderrFile"
1143	<	else:
1144	<	for fileOut in (self.output_file_sandbox):
1145	<	txt += 'echo " '+fileOut+'";\n'
1146	<	allOutFiles = allOutFiles + " " + self.numberFile_(fileOut, '$NJob') + " $stdoutFile $stderrFile"
1147	<	txt += 'echo "OUTPUT files: '+str(allOutFiles)+'";\n'
1148	<	txt += 'ls -gGhrta;\n'
1149	<	txt += 'sum=0;\n'
1150	<	txt += 'for file in '+str(allOutFiles)+' ; do\n'
1151	<	txt += '    if [ -e $file ]; then\n'
1152	<	txt += '        tt=`ls -gGrta $file | awk \'{ print $3 }\'`\n'
1153	<	txt += '        sum=`expr $sum + $tt`\n'
1440	<	txt += '    else\n'
1441	<	txt += '        echo "WARNING: output file $file not found!"\n'
1442	<	txt += '    fi\n'
1443	<	txt += 'done\n'
1444	<	txt += 'echo "Total Output dimension: $sum";\n'
1445	<	txt += 'limit='+str(limit)+';\n'
1446	<	txt += 'echo "OUTPUT FILES LIMIT SET TO: $limit";\n'
1447	<	txt += 'if [ $limit -lt $sum ]; then\n'
1448	<	txt += '    echo "WARNING: output files have to big size - something will be lost;"\n'
1449	<	txt += '    echo "         checking the output file sizes..."\n'
1450	<	txt += '    tot=0;\n'
1451	<	txt += '    for filefile in '+str(allOutFiles)+' ; do\n'
1452	<	txt += '        dimFile=`ls -gGrta $filefile | awk \'{ print $3 }\';`\n'
1453	<	txt += '        tot=`expr $tot + $tt`;\n'
1454	<	txt += '        if [ $limit -lt $dimFile ]; then\n'
1455	<	txt += '            echo "deleting file: $filefile";\n'
1456	<	txt += '            rm -f $filefile\n'
1457	<	txt += '        elif [ $limit -lt $tot ]; then\n'
1458	<	txt += '            echo "deleting file: $filefile";\n'
1459	<	txt += '            rm -f $filefile\n'
1460	<	txt += '        else\n'
1461	<	txt += '            echo "saving file: $filefile"\n'
1462	<	txt += '        fi\n'
1463	<	txt += '    done\n'
1141	>	for file in (self.output_file):
1142	>	listOutFiles.append(numberFile(file, '$OutUniqueID'))
1143	>	for file in (self.output_file_sandbox):
1144	>	listOutFiles.append(numberFile(file, '$NJob'))
1145	>	listOutFiles.append(stdout)
1146	>	listOutFiles.append(stderr)
1147	>
1148	>	txt += 'echo "output files: '+string.join(listOutFiles,' ')+'"\n'
1149	>	txt += 'filesToCheck="'+string.join(listOutFiles,' ')+'"\n'
1150	>	txt += 'export filesToCheck\n'
1151	>	taskinfo={}
1152	>	taskinfo['outfileBasename'] = self.output_file
1153	>	common._db.updateTask_(taskinfo)
1154
1155	<	txt += '    ls -agGhrt;\n'
1466	<	txt += '    echo "WARNING: output files are too big in dimension: can not put in the output_sandbox.";\n'
1467	<	txt += '    echo "JOB_EXIT_STATUS = 70000";\n'
1468	<	txt += '    exit_status=70000;\n'
1469	<	txt += 'else'
1470	<	txt += '    echo "Total Output dimension $sum is fine.";\n'
1471	<	txt += 'fi\n'
1472	<	txt += 'echo "Ending output sandbox limit check"\n'
1155	>	if list : return self.output_file
1156		return txt
1157	+
1158	+	def checkCMSSWVersion(self, url = "https://cmstags.cern.ch/cgi-bin/CmsTC/", fileName = "ReleasesXML"):
1159	+	"""
1160	+	compare current CMSSW release and arch with allowed releases
1161	+	"""
1162	+
1163	+	downloader = Downloader(url)
1164	+	goodRelease = False
1165	+
1166	+	try:
1167	+	result = downloader.config(fileName)
1168	+	except:
1169	+	common.logger.info("ERROR: Problem reading file of allowed CMSSW releases.")
1170	+
1171	+	try:
1172	+	events = pulldom.parseString(result)
1173	+
1174	+	arch     = None
1175	+	release  = None
1176	+	relType  = None
1177	+	relState = None
1178	+	for (event, node) in events:
1179	+	if event == pulldom.START_ELEMENT:
1180	+	if node.tagName == 'architecture':
1181	+	arch = node.attributes.getNamedItem('name').nodeValue
1182	+	if node.tagName == 'project':
1183	+	relType = node.attributes.getNamedItem('type').nodeValue
1184	+	relState = node.attributes.getNamedItem('state').nodeValue
1185	+	if relType == 'Production' and relState == 'Announced':
1186	+	release = node.attributes.getNamedItem('label').nodeValue
1187	+	if self.executable_arch == arch and self.version == release:
1188	+	goodRelease = True
1189	+	return goodRelease
1190	+
1191	+	if not goodRelease:
1192	+	msg = "WARNING: %s on %s is not a supported release. " % \
1193	+	(self.version, self.executable_arch)
1194	+	msg += "Submission may fail."
1195	+	common.logger.info(msg)
1196	+	except:
1197	+	common.logger.info("Problems parsing file of allowed CMSSW releases.")
1198	+
1199	+	return goodRelease
1200	+

Diff Legend

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