ViewVC Help
View File | Revision Log | Show Annotations | Root Listing
root/cvsroot/COMP/CRAB/python/cms_cmssw.py
(Generate patch)

Comparing COMP/CRAB/python/cms_cmssw.py (file contents):
Revision 1.13 by gutsche, Tue Jun 27 02:31:31 2006 UTC vs.
Revision 1.128 by fanzago, Thu Oct 11 16:23:44 2007 UTC

# Line 2 | Line 2 | from JobType import JobType
2   from crab_logger import Logger
3   from crab_exceptions import *
4   from crab_util import *
5 + from BlackWhiteListParser import BlackWhiteListParser
6   import common
6 – import PsetManipulator  
7 –
8 – import DBSInfo_EDM
9 – import DataDiscovery_EDM
10 – import DataLocation_EDM
7   import Scram
8  
9 < import os, string, re
9 > import os, string, glob
10  
11   class Cmssw(JobType):
12 <    def __init__(self, cfg_params):
12 >    def __init__(self, cfg_params, ncjobs):
13          JobType.__init__(self, 'CMSSW')
14          common.logger.debug(3,'CMSSW::__init__')
15  
20 –        self.analisys_common_info = {}
21 –        # Marco.
16          self._params = {}
17          self.cfg_params = cfg_params
18 +
19 +        # init BlackWhiteListParser
20 +        self.blackWhiteListParser = BlackWhiteListParser(cfg_params)
21 +
22 +        try:
23 +            self.MaxTarBallSize = float(self.cfg_params['EDG.maxtarballsize'])
24 +        except KeyError:
25 +            self.MaxTarBallSize = 9.5
26 +
27 +        # number of jobs requested to be created, limit obj splitting
28 +        self.ncjobs = ncjobs
29 +
30          log = common.logger
31          
32          self.scram = Scram.Scram(cfg_params)
27 –        scramArea = ''
33          self.additional_inbox_files = []
34          self.scriptExe = ''
35          self.executable = ''
36 +        self.executable_arch = self.scram.getArch()
37          self.tgz_name = 'default.tgz'
38 +        self.additional_tgz_name = 'additional.tgz'
39 +        self.scriptName = 'CMSSW.sh'
40 +        self.pset = ''      #scrip use case Da  
41 +        self.datasetPath = '' #scrip use case Da
42  
43 +        # set FJR file name
44 +        self.fjrFileName = 'crab_fjr.xml'
45  
46          self.version = self.scram.getSWVersion()
47 +        
48 +        #
49 +        # Try to block creation in case of arch/version mismatch
50 +        #
51 +
52 +        a = string.split(self.version, "_")
53 +
54 +        if int(a[1]) == 1 and (int(a[2]) < 5 and self.executable_arch.find('slc4') == 0):
55 +            msg = "Error: CMS does not support %s with %s architecture"%(self.version, self.executable_arch)
56 +            raise CrabException(msg)
57 +        if int(a[1]) == 1 and (int(a[2]) >= 5 and self.executable_arch.find('slc3') == 0):
58 +            msg = "Error: CMS does not support %s with %s architecture"%(self.version, self.executable_arch)
59 +            raise CrabException(msg)
60 +        
61 +        common.taskDB.setDict('codeVersion',self.version)
62          self.setParam_('application', self.version)
36 –        common.analisys_common_info['sw_version'] = self.version
37 –        ### FEDE
38 –        common.analisys_common_info['copy_input_data'] = 0
39 –        common.analisys_common_info['events_management'] = 1
63  
64          ### collect Data cards
65 +
66 +        ## get DBS mode
67 +        try:
68 +            self.use_dbs_1 = int(self.cfg_params['CMSSW.use_dbs_1'])
69 +        except KeyError:
70 +            self.use_dbs_1 = 0
71 +            
72          try:
73              tmp =  cfg_params['CMSSW.datasetpath']
74              log.debug(6, "CMSSW::CMSSW(): datasetPath = "+tmp)
75              if string.lower(tmp)=='none':
76                  self.datasetPath = None
77 +                self.selectNoInput = 1
78              else:
79                  self.datasetPath = tmp
80 +                self.selectNoInput = 0
81          except KeyError:
82              msg = "Error: datasetpath not defined "  
83              raise CrabException(msg)
# Line 56 | Line 88 | class Cmssw(JobType):
88              self.setParam_('dataset', 'None')
89              self.setParam_('owner', 'None')
90          else:
91 <            datasetpath_split = self.datasetPath.split("/")
92 <            self.setParam_('dataset', datasetpath_split[1])
93 <            self.setParam_('owner', datasetpath_split[-1])
94 <
91 >            try:
92 >                datasetpath_split = self.datasetPath.split("/")
93 >                # standard style
94 >                self.setParam_('datasetFull', self.datasetPath)
95 >                if self.use_dbs_1 == 1 :
96 >                    self.setParam_('dataset', datasetpath_split[1])
97 >                    self.setParam_('owner', datasetpath_split[-1])
98 >                else:
99 >                    self.setParam_('dataset', datasetpath_split[1])
100 >                    self.setParam_('owner', datasetpath_split[2])
101 >            except:
102 >                self.setParam_('dataset', self.datasetPath)
103 >                self.setParam_('owner', self.datasetPath)
104 >                
105          self.setTaskid_()
106          self.setParam_('taskId', self.cfg_params['taskId'])
107  
# Line 82 | Line 124 | class Cmssw(JobType):
124          try:
125              self.pset = cfg_params['CMSSW.pset']
126              log.debug(6, "Cmssw::Cmssw(): PSet file = "+self.pset)
127 <            if (not os.path.exists(self.pset)):
128 <                raise CrabException("User defined PSet file "+self.pset+" does not exist")
127 >            if self.pset.lower() != 'none' :
128 >                if (not os.path.exists(self.pset)):
129 >                    raise CrabException("User defined PSet file "+self.pset+" does not exist")
130 >            else:
131 >                self.pset = None
132          except KeyError:
133              raise CrabException("PSet file missing. Cannot run cmsRun ")
134  
135          # output files
136 +        ## stuff which must be returned always via sandbox
137 +        self.output_file_sandbox = []
138 +
139 +        # add fjr report by default via sandbox
140 +        self.output_file_sandbox.append(self.fjrFileName)
141 +
142 +        # other output files to be returned via sandbox or copied to SE
143          try:
144              self.output_file = []
93 –
145              tmp = cfg_params['CMSSW.output_file']
146              if tmp != '':
147                  tmpOutFiles = string.split(cfg_params['CMSSW.output_file'],',')
# Line 100 | Line 151 | class Cmssw(JobType):
151                      self.output_file.append(tmp)
152                      pass
153              else:
154 <                log.message("No output file defined: only stdout/err will be available")
154 >                log.message("No output file defined: only stdout/err and the CRAB Framework Job Report will be available\n")
155                  pass
156              pass
157          except KeyError:
158 <            log.message("No output file defined: only stdout/err will be available")
158 >            log.message("No output file defined: only stdout/err and the CRAB Framework Job Report will be available\n")
159              pass
160  
161          # script_exe file as additional file in inputSandbox
162          try:
163              self.scriptExe = cfg_params['USER.script_exe']
113 –            self.additional_inbox_files.append(self.scriptExe)
164              if self.scriptExe != '':
165                 if not os.path.isfile(self.scriptExe):
166 <                  msg ="WARNING. file "+self.scriptExe+" not found"
166 >                  msg ="ERROR. file "+self.scriptExe+" not found"
167                    raise CrabException(msg)
168 +               self.additional_inbox_files.append(string.strip(self.scriptExe))
169          except KeyError:
170 <           pass
171 <                  
170 >            self.scriptExe = ''
171 >
172 >        #CarlosDaniele
173 >        if self.datasetPath == None and self.pset == None and self.scriptExe == '' :
174 >           msg ="Error. script_exe  not defined"
175 >           raise CrabException(msg)
176 >
177          ## additional input files
178          try:
179 <            tmpAddFiles = string.split(cfg_params['CMSSW.additional_input_files'],',')
179 >            tmpAddFiles = string.split(cfg_params['USER.additional_input_files'],',')
180              for tmp in tmpAddFiles:
181 <                if not os.path.exists(tmp):
182 <                    raise CrabException("Additional input file not found: "+tmp)
183 <                tmp=string.strip(tmp)
184 <                self.additional_inbox_files.append(tmp)
181 >                tmp = string.strip(tmp)
182 >                dirname = ''
183 >                if not tmp[0]=="/": dirname = "."
184 >                files = []
185 >                if string.find(tmp,"*")>-1:
186 >                    files = glob.glob(os.path.join(dirname, tmp))
187 >                    if len(files)==0:
188 >                        raise CrabException("No additional input file found with this pattern: "+tmp)
189 >                else:
190 >                    files.append(tmp)
191 >                for file in files:
192 >                    if not os.path.exists(file):
193 >                        raise CrabException("Additional input file not found: "+file)
194 >                    pass
195 >                    # fname = string.split(file, '/')[-1]
196 >                    # storedFile = common.work_space.pathForTgz()+'share/'+fname
197 >                    # shutil.copyfile(file, storedFile)
198 >                    self.additional_inbox_files.append(string.strip(file))
199                  pass
200              pass
201 +            common.logger.debug(5,"Additional input files: "+str(self.additional_inbox_files))
202          except KeyError:
203              pass
204  
205          # files per job
206          try:
207 <            self.filesPerJob = int(cfg_params['CMSSW.files_per_jobs']) #Daniele
208 <            self.selectFilesPerJob = 1
207 >            if (cfg_params['CMSSW.files_per_jobs']):
208 >                raise CrabException("files_per_jobs no longer supported.  Quitting.")
209          except KeyError:
210 <            self.filesPerJob = 0
140 <            self.selectFilesPerJob = 0
210 >            pass
211  
212          ## Events per job
213          try:
# Line 147 | Line 217 | class Cmssw(JobType):
217              self.eventsPerJob = -1
218              self.selectEventsPerJob = 0
219      
220 <        # To be implemented
221 <        # ## number of jobs
222 <        # try:
223 <        #     self.numberOfJobs =int( cfg_params['CMSSW.number_of_job'])
224 <        #     self.selectNumberOfJobs = 1
225 <        # except KeyError:
226 <        #     self.selectNumberOfJobs = 0
157 <
158 <        if (self.selectFilesPerJob == self.selectEventsPerJob):
159 <            msg = 'Must define either files_per_jobs or events_per_job'
160 <            raise CrabException(msg)
220 >        ## number of jobs
221 >        try:
222 >            self.theNumberOfJobs =int( cfg_params['CMSSW.number_of_jobs'])
223 >            self.selectNumberOfJobs = 1
224 >        except KeyError:
225 >            self.theNumberOfJobs = 0
226 >            self.selectNumberOfJobs = 0
227  
162 –        if (self.selectEventsPerJob  and not self.datasetPath == None):
163 –            msg = 'Splitting according to events_per_job available only with None as datasetpath'
164 –            raise CrabException(msg)
165 –    
228          try:
229              self.total_number_of_events = int(cfg_params['CMSSW.total_number_of_events'])
230 +            self.selectTotalNumberEvents = 1
231          except KeyError:
232 <            msg = 'Must define total_number_of_events'
233 <            raise CrabException(msg)
234 <        
235 <        CEBlackList = []
232 >            self.total_number_of_events = 0
233 >            self.selectTotalNumberEvents = 0
234 >
235 >        if self.pset != None: #CarlosDaniele
236 >             if ( (self.selectTotalNumberEvents + self.selectEventsPerJob + self.selectNumberOfJobs) != 2 ):
237 >                 msg = 'Must define exactly two of total_number_of_events, events_per_job, or number_of_jobs.'
238 >                 raise CrabException(msg)
239 >        else:
240 >             if (self.selectNumberOfJobs == 0):
241 >                 msg = 'Must specify  number_of_jobs.'
242 >                 raise CrabException(msg)
243 >
244 >        ## source seed for pythia
245          try:
246 <            tmpBad = string.split(cfg_params['EDG.ce_black_list'],',')
175 <            for tmp in tmpBad:
176 <                tmp=string.strip(tmp)
177 <                CEBlackList.append(tmp)
246 >            self.sourceSeed = int(cfg_params['CMSSW.pythia_seed'])
247          except KeyError:
248 <            pass
248 >            self.sourceSeed = None
249 >            common.logger.debug(5,"No seed given")
250  
251 <        self.reCEBlackList=[]
252 <        for bad in CEBlackList:
183 <            self.reCEBlackList.append(re.compile( bad ))
184 <
185 <        common.logger.debug(5,'CEBlackList: '+str(CEBlackList))
186 <
187 <        CEWhiteList = []
188 <        try:
189 <            tmpGood = string.split(cfg_params['EDG.ce_white_list'],',')
190 <            for tmp in tmpGood:
191 <                tmp=string.strip(tmp)
192 <                CEWhiteList.append(tmp)
251 >        try:
252 >            self.sourceSeedVtx = int(cfg_params['CMSSW.vtx_seed'])
253          except KeyError:
254 <            pass
254 >            self.sourceSeedVtx = None
255 >            common.logger.debug(5,"No vertex seed given")
256  
257 <        #print 'CEWhiteList: ',CEWhiteList
258 <        self.reCEWhiteList=[]
259 <        for Good in CEWhiteList:
260 <            self.reCEWhiteList.append(re.compile( Good ))
257 >        try:
258 >            self.sourceSeedG4 = int(cfg_params['CMSSW.g4_seed'])
259 >        except KeyError:
260 >            self.sourceSeedG4 = None
261 >            common.logger.debug(5,"No g4 sim hits seed given")
262  
263 <        common.logger.debug(5,'CEWhiteList: '+str(CEWhiteList))
263 >        try:
264 >            self.sourceSeedMix = int(cfg_params['CMSSW.mix_seed'])
265 >        except KeyError:
266 >            self.sourceSeedMix = None
267 >            common.logger.debug(5,"No mix seed given")
268  
269 <        self.PsetEdit = PsetManipulator.PsetManipulator(self.pset) #Daniele Pset
269 >        try:
270 >            self.firstRun = int(cfg_params['CMSSW.first_run'])
271 >        except KeyError:
272 >            self.firstRun = None
273 >            common.logger.debug(5,"No first run given")
274 >        if self.pset != None: #CarlosDaniele
275 >            ver = string.split(self.version,"_")
276 >            if (int(ver[1])>=1 and int(ver[2])>=5):
277 >                import PsetManipulator150 as pp
278 >            else:
279 >                import PsetManipulator as pp
280 >            PsetEdit = pp.PsetManipulator(self.pset) #Daniele Pset
281  
282          #DBSDLS-start
283          ## Initialize the variables that are extracted from DBS/DLS and needed in other places of the code
284          self.maxEvents=0  # max events available   ( --> check the requested nb. of evts in Creator.py)
285          self.DBSPaths={}  # all dbs paths requested ( --> input to the site local discovery script)
286 +        self.jobDestination=[]  # Site destination(s) for each job (list of lists)
287          ## Perform the data location and discovery (based on DBS/DLS)
288          ## SL: Don't if NONE is specified as input (pythia use case)
289 <        common.analisys_common_info['sites']=None
289 >        blockSites = {}
290          if self.datasetPath:
291 <            self.DataDiscoveryAndLocation(cfg_params)
291 >            blockSites = self.DataDiscoveryAndLocation(cfg_params)
292          #DBSDLS-end          
293  
294          self.tgzNameWithPath = self.getTarBall(self.executable)
217 –
218 –        # modify Pset
219 –        if (self.datasetPath): # standard job
220 –            self.PsetEdit.maxEvent(self.eventsPerJob) #Daniele  
221 –            self.PsetEdit.inputModule("INPUT") #Daniele
222 –
223 –        else:  # pythia like job
224 –            self.PsetEdit.maxEvent(self.eventsPerJob) #Daniele  
225 –            self.PsetEdit.pythiaSeed("INPUT") #Daniele
226 –            try:
227 –                self.sourceSeed = int(cfg_params['CMSSW.pythia_seed'])
228 –            except KeyError:
229 –                self.sourceSeed = 123456
230 –                common.logger.message("No seed given, will use "+str(self.sourceSeed))
231 –        
232 –        self.PsetEdit.psetWriter(self.configFilename())
295      
296          ## Select Splitting
297 <        if self.selectFilesPerJob: self.jobSplittingPerFiles()
298 <        elif self.selectEventsPerJob: self.jobSplittingPerEvents()
297 >        if self.selectNoInput:
298 >            if self.pset == None: #CarlosDaniele
299 >                self.jobSplittingForScript()
300 >            else:
301 >                self.jobSplittingNoInput()
302          else:
303 <            msg = 'Don\'t know how to split...'
239 <            raise CrabException(msg)
303 >            self.jobSplittingByBlocks(blockSites)
304  
305 +        # modify Pset
306 +        if self.pset != None: #CarlosDaniele
307 +            try:
308 +                if (self.datasetPath): # standard job
309 +                    # allow to processa a fraction of events in a file
310 +                    PsetEdit.inputModule("INPUT")
311 +                    PsetEdit.maxEvent("INPUTMAXEVENTS")
312 +                    PsetEdit.skipEvent("INPUTSKIPEVENTS")
313 +                else:  # pythia like job
314 +                    PsetEdit.maxEvent(self.eventsPerJob)
315 +                    if (self.firstRun):
316 +                        PsetEdit.pythiaFirstRun("INPUTFIRSTRUN")  #First Run
317 +                    if (self.sourceSeed) :
318 +                        PsetEdit.pythiaSeed("INPUT")
319 +                        if (self.sourceSeedVtx) :
320 +                            PsetEdit.vtxSeed("INPUTVTX")
321 +                        if (self.sourceSeedG4) :
322 +                            PsetEdit.g4Seed("INPUTG4")
323 +                        if (self.sourceSeedMix) :
324 +                            PsetEdit.mixSeed("INPUTMIX")
325 +                # add FrameworkJobReport to parameter-set
326 +                PsetEdit.addCrabFJR(self.fjrFileName)
327 +                PsetEdit.psetWriter(self.configFilename())
328 +            except:
329 +                msg='Error while manipuliating ParameterSet: exiting...'
330 +                raise CrabException(msg)
331  
332      def DataDiscoveryAndLocation(self, cfg_params):
333  
334 +        import DataDiscovery
335 +        import DataDiscovery_DBS2
336 +        import DataLocation
337          common.logger.debug(10,"CMSSW::DataDiscoveryAndLocation()")
338  
339          datasetPath=self.datasetPath
340  
248 –        ## TODO
249 –        dataTiersList = ""
250 –        dataTiers = dataTiersList.split(',')
251 –
341          ## Contact the DBS
342 +        common.logger.message("Contacting Data Discovery Services ...")
343          try:
344 <            self.pubdata=DataDiscovery_EDM.DataDiscovery_EDM(datasetPath, dataTiers, cfg_params)
344 >
345 >            if self.use_dbs_1 == 1 :
346 >                self.pubdata=DataDiscovery.DataDiscovery(datasetPath, cfg_params)
347 >            else :
348 >                self.pubdata=DataDiscovery_DBS2.DataDiscovery_DBS2(datasetPath, cfg_params)
349              self.pubdata.fetchDBSInfo()
350  
351 <        except DataDiscovery_EDM.NotExistingDatasetError, ex :
351 >        except DataDiscovery.NotExistingDatasetError, ex :
352              msg = 'ERROR ***: failed Data Discovery in DBS : %s'%ex.getErrorMessage()
353              raise CrabException(msg)
354 <
261 <        except DataDiscovery_EDM.NoDataTierinProvenanceError, ex :
354 >        except DataDiscovery.NoDataTierinProvenanceError, ex :
355              msg = 'ERROR ***: failed Data Discovery in DBS : %s'%ex.getErrorMessage()
356              raise CrabException(msg)
357 <        except DataDiscovery_EDM.DataDiscoveryError, ex:
358 <            msg = 'ERROR ***: failed Data Discovery in DBS  %s'%ex.getErrorMessage()
357 >        except DataDiscovery.DataDiscoveryError, ex:
358 >            msg = 'ERROR ***: failed Data Discovery in DBS :  %s'%ex.getErrorMessage()
359 >            raise CrabException(msg)
360 >        except DataDiscovery_DBS2.NotExistingDatasetError_DBS2, ex :
361 >            msg = 'ERROR ***: failed Data Discovery in DBS : %s'%ex.getErrorMessage()
362 >            raise CrabException(msg)
363 >        except DataDiscovery_DBS2.NoDataTierinProvenanceError_DBS2, ex :
364 >            msg = 'ERROR ***: failed Data Discovery in DBS : %s'%ex.getErrorMessage()
365 >            raise CrabException(msg)
366 >        except DataDiscovery_DBS2.DataDiscoveryError_DBS2, ex:
367 >            msg = 'ERROR ***: failed Data Discovery in DBS :  %s'%ex.getErrorMessage()
368              raise CrabException(msg)
369  
370 <        ## get list of all required data in the form of dbs paths  (dbs path = /dataset/datatier/owner)
371 <        ## self.DBSPaths=self.pubdata.getDBSPaths()
372 <        common.logger.message("Required data are :"+self.datasetPath)
271 <
272 <        filesbyblock=self.pubdata.getFiles()
273 <        self.AllInputFiles=filesbyblock.values()
274 <        self.files = self.AllInputFiles        
275 <
276 <        ## TEMP
277 <    #    self.filesTmp = filesbyblock.values()
278 <    #    self.files = []
279 <    #    locPath='rfio:cmsbose2.bo.infn.it:/flatfiles/SE00/cms/fanfani/ProdTest/'
280 <    #    locPath=''
281 <    #    tmp = []
282 <    #    for file in self.filesTmp[0]:
283 <    #        tmp.append(locPath+file)
284 <    #    self.files.append(tmp)
285 <        ## END TEMP
370 >        self.filesbyblock=self.pubdata.getFiles()
371 >        self.eventsbyblock=self.pubdata.getEventsPerBlock()
372 >        self.eventsbyfile=self.pubdata.getEventsPerFile()
373  
374          ## get max number of events
288 –        #common.logger.debug(10,"number of events for primary fileblocks %i"%self.pubdata.getMaxEvents())
375          self.maxEvents=self.pubdata.getMaxEvents() ##  self.maxEvents used in Creator.py
290 –        common.logger.message("\nThe number of available events is %s"%self.maxEvents)
376  
377          ## Contact the DLS and build a list of sites hosting the fileblocks
378          try:
379 <            dataloc=DataLocation_EDM.DataLocation_EDM(filesbyblock.keys(),cfg_params)
379 >            dataloc=DataLocation.DataLocation(self.filesbyblock.keys(),cfg_params)
380              dataloc.fetchDLSInfo()
381 <        except DataLocation_EDM.DataLocationError , ex:
381 >        except DataLocation.DataLocationError , ex:
382              msg = 'ERROR ***: failed Data Location in DLS \n %s '%ex.getErrorMessage()
383              raise CrabException(msg)
384          
300 –        allsites=dataloc.getSites()
301 –        common.logger.debug(5,"sites are %s"%allsites)
302 –        sites=self.checkBlackList(allsites)
303 –        common.logger.debug(5,"sites are (after black list) %s"%sites)
304 –        sites=self.checkWhiteList(sites)
305 –        common.logger.debug(5,"sites are (after white list) %s"%sites)
385  
386 <        if len(sites)==0:
387 <            msg = 'No sites hosting all the needed data! Exiting... '
388 <            raise CrabException(msg)
386 >        sites = dataloc.getSites()
387 >        allSites = []
388 >        listSites = sites.values()
389 >        for listSite in listSites:
390 >            for oneSite in listSite:
391 >                allSites.append(oneSite)
392 >        allSites = self.uniquelist(allSites)
393  
394 <        common.logger.message("List of Sites hosting the data : "+str(sites))
395 <        common.logger.debug(6, "List of Sites: "+str(sites))
396 <        common.analisys_common_info['sites']=sites    ## used in SchedulerEdg.py in createSchScript
397 <        self.setParam_('TargetCE', ','.join(sites))
315 <        return
394 >        # screen output
395 >        common.logger.message("Requested dataset: " + datasetPath + " has " + str(self.maxEvents) + " events in " + str(len(self.filesbyblock.keys())) + " blocks.\n")
396 >
397 >        return sites
398      
399 <    def jobSplittingPerFiles(self):
318 <        """
319 <        Perform job splitting based on number of files to be accessed per job
399 >    def jobSplittingByBlocks(self, blockSites):
400          """
401 <        common.logger.debug(5,'Splitting per input files')
402 <        common.logger.message('Required '+str(self.filesPerJob)+' files per job ')
403 <        common.logger.message('Required '+str(self.total_number_of_events)+' events in total ')
404 <
405 <        ## TODO: SL need to have (from DBS) a detailed list of how many events per each file
406 <        n_tot_files = (len(self.files[0]))
407 <        #print "n_tot_files = ", n_tot_files
408 <        ## SL: this is wrong if the files have different number of events
409 <        #print "self.maxEvents = ", self.maxEvents
410 <        evPerFile = int(self.maxEvents)/n_tot_files
411 <        #print "evPerFile = int(self.maxEvents)/n_tot_files =  ", evPerFile
412 <
413 <        common.logger.debug(5,'Events per File '+str(evPerFile))
414 <
415 <        ## if asked to process all events, do it
416 <        if self.total_number_of_events == -1:
417 <            self.total_number_of_events=self.maxEvents
418 <            self.total_number_of_jobs = int(n_tot_files)*1/int(self.filesPerJob)
419 <            common.logger.message(str(self.total_number_of_jobs)+' jobs will be created for all available events '+str(self.total_number_of_events)+' events')
420 <        
401 >        Perform job splitting. Jobs run over an integer number of files
402 >        and no more than one block.
403 >        ARGUMENT: blockSites: dictionary with blocks as keys and list of host sites as values
404 >        REQUIRES: self.selectTotalNumberEvents, self.selectEventsPerJob, self.selectNumberofJobs,
405 >                  self.total_number_of_events, self.eventsPerJob, self.theNumberOfJobs,
406 >                  self.maxEvents, self.filesbyblock
407 >        SETS: self.jobDestination - Site destination(s) for each job (a list of lists)
408 >              self.total_number_of_jobs - Total # of jobs
409 >              self.list_of_args - File(s) job will run on (a list of lists)
410 >        """
411 >
412 >        # ---- Handle the possible job splitting configurations ---- #
413 >        if (self.selectTotalNumberEvents):
414 >            totalEventsRequested = self.total_number_of_events
415 >        if (self.selectEventsPerJob):
416 >            eventsPerJobRequested = self.eventsPerJob
417 >            if (self.selectNumberOfJobs):
418 >                totalEventsRequested = self.theNumberOfJobs * self.eventsPerJob
419 >
420 >        # If user requested all the events in the dataset
421 >        if (totalEventsRequested == -1):
422 >            eventsRemaining=self.maxEvents
423 >        # If user requested more events than are in the dataset
424 >        elif (totalEventsRequested > self.maxEvents):
425 >            eventsRemaining = self.maxEvents
426 >            common.logger.message("Requested "+str(self.total_number_of_events)+ " events, but only "+str(self.maxEvents)+" events are available.")
427 >        # If user requested less events than are in the dataset
428          else:
429 <            #print "self.total_number_of_events = ", self.total_number_of_events
430 <            #print "evPerFile = ", evPerFile
431 <            self.total_number_of_files = int(self.total_number_of_events/evPerFile)
432 <            #print "self.total_number_of_files = int(self.total_number_of_events/evPerFile) = " , self.total_number_of_files
433 <            ## SL: if ask for less event than what is computed to be available on a
434 <            ##     file, process the first file anyhow.
435 <            if self.total_number_of_files == 0:
436 <                self.total_number_of_files = self.total_number_of_files + 1
350 <                
429 >            eventsRemaining = totalEventsRequested
430 >
431 >        # If user requested more events per job than are in the dataset
432 >        if (self.selectEventsPerJob and eventsPerJobRequested > self.maxEvents):
433 >            eventsPerJobRequested = self.maxEvents
434 >
435 >        # For user info at end
436 >        totalEventCount = 0
437  
438 <            common.logger.debug(5,'N files  '+str(self.total_number_of_files))
438 >        if (self.selectTotalNumberEvents and self.selectNumberOfJobs):
439 >            eventsPerJobRequested = int(eventsRemaining/self.theNumberOfJobs)
440  
441 <            check = 0
441 >        if (self.selectNumberOfJobs):
442 >            common.logger.message("May not create the exact number_of_jobs requested.")
443 >
444 >        if ( self.ncjobs == 'all' ) :
445 >            totalNumberOfJobs = 999999999
446 >        else :
447 >            totalNumberOfJobs = self.ncjobs
448              
356 –            ## Compute the number of jobs
357 –            #self.total_number_of_jobs = int(n_tot_files)*1/int(self.filesPerJob)
358 –            #print "self.total_number_of_files = ", self.total_number_of_files
359 –            #print "self.filesPerJob = ", self.filesPerJob
360 –            self.total_number_of_jobs = int(self.total_number_of_files/self.filesPerJob)
361 –            #print "self.total_number_of_jobs = ", self.total_number_of_jobs
362 –            common.logger.debug(5,'N jobs  '+str(self.total_number_of_jobs))
363 –
364 –            ## is there any remainder?
365 –            check = int(self.total_number_of_files) - (int(self.total_number_of_jobs)*self.filesPerJob)
366 –
367 –            common.logger.debug(5,'Check  '+str(check))
368 –
369 –            if check > 0:
370 –                self.total_number_of_jobs =  self.total_number_of_jobs + 1
371 –                common.logger.message('Warning: last job will be created with '+str(check)+' files')
449  
450 <            #common.logger.message(str(self.total_number_of_jobs)+' jobs will be created for a total of '+str((self.total_number_of_jobs-1)*self.filesPerJob*evPerFile + check*evPerFile)+' events')
451 <            common.logger.message(str(self.total_number_of_jobs)+' jobs will be created for a total of '+str((self.total_number_of_jobs)*self.filesPerJob*evPerFile + check*evPerFile)+' events')
452 <            pass
450 >        blocks = blockSites.keys()
451 >        blockCount = 0
452 >        # Backup variable in case self.maxEvents counted events in a non-included block
453 >        numBlocksInDataset = len(blocks)
454  
455 +        jobCount = 0
456          list_of_lists = []
457 <        for i in xrange(0, int(n_tot_files), self.filesPerJob):
458 <            parString = "\\{"
457 >
458 >        # list tracking which jobs are in which jobs belong to which block
459 >        jobsOfBlock = {}
460 >
461 >        # ---- Iterate over the blocks in the dataset until ---- #
462 >        # ---- we've met the requested total # of events    ---- #
463 >        while ( (eventsRemaining > 0) and (blockCount < numBlocksInDataset) and (jobCount < totalNumberOfJobs)):
464 >            block = blocks[blockCount]
465 >            blockCount += 1
466 >            if block not in jobsOfBlock.keys() :
467 >                jobsOfBlock[block] = []
468              
469 <            params = self.files[0][i: i+self.filesPerJob]
470 <            for i in range(len(params) - 1):
471 <                parString += '\\\"' + params[i] + '\\\"\,'
469 >            if self.eventsbyblock.has_key(block) :
470 >                numEventsInBlock = self.eventsbyblock[block]
471 >                common.logger.debug(5,'Events in Block File '+str(numEventsInBlock))
472              
473 <            parString += '\\\"' + params[len(params) - 1] + '\\\"\\}'
474 <            list_of_lists.append(parString)
475 <            pass
473 >                files = self.filesbyblock[block]
474 >                numFilesInBlock = len(files)
475 >                if (numFilesInBlock <= 0):
476 >                    continue
477 >                fileCount = 0
478 >
479 >                # ---- New block => New job ---- #
480 >                parString = "\\{"
481 >                # counter for number of events in files currently worked on
482 >                filesEventCount = 0
483 >                # flag if next while loop should touch new file
484 >                newFile = 1
485 >                # job event counter
486 >                jobSkipEventCount = 0
487 >            
488 >                # ---- Iterate over the files in the block until we've met the requested ---- #
489 >                # ---- total # of events or we've gone over all the files in this block  ---- #
490 >                while ( (eventsRemaining > 0) and (fileCount < numFilesInBlock) and (jobCount < totalNumberOfJobs) ):
491 >                    file = files[fileCount]
492 >                    if newFile :
493 >                        try:
494 >                            numEventsInFile = self.eventsbyfile[file]
495 >                            common.logger.debug(6, "File "+str(file)+" has "+str(numEventsInFile)+" events")
496 >                            # increase filesEventCount
497 >                            filesEventCount += numEventsInFile
498 >                            # Add file to current job
499 >                            parString += '\\\"' + file + '\\\"\,'
500 >                            newFile = 0
501 >                        except KeyError:
502 >                            common.logger.message("File "+str(file)+" has unknown number of events: skipping")
503 >                        
504 >
505 >                    # if less events in file remain than eventsPerJobRequested
506 >                    if ( filesEventCount - jobSkipEventCount < eventsPerJobRequested ) :
507 >                        # if last file in block
508 >                        if ( fileCount == numFilesInBlock-1 ) :
509 >                            # end job using last file, use remaining events in block
510 >                            # close job and touch new file
511 >                            fullString = parString[:-2]
512 >                            fullString += '\\}'
513 >                            list_of_lists.append([fullString,str(-1),str(jobSkipEventCount)])
514 >                            common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(filesEventCount - jobSkipEventCount)+" events (last file in block).")
515 >                            self.jobDestination.append(blockSites[block])
516 >                            common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
517 >                            # fill jobs of block dictionary
518 >                            jobsOfBlock[block].append(jobCount+1)
519 >                            # reset counter
520 >                            jobCount = jobCount + 1
521 >                            totalEventCount = totalEventCount + filesEventCount - jobSkipEventCount
522 >                            eventsRemaining = eventsRemaining - filesEventCount + jobSkipEventCount
523 >                            jobSkipEventCount = 0
524 >                            # reset file
525 >                            parString = "\\{"
526 >                            filesEventCount = 0
527 >                            newFile = 1
528 >                            fileCount += 1
529 >                        else :
530 >                            # go to next file
531 >                            newFile = 1
532 >                            fileCount += 1
533 >                    # if events in file equal to eventsPerJobRequested
534 >                    elif ( filesEventCount - jobSkipEventCount == eventsPerJobRequested ) :
535 >                        # close job and touch new file
536 >                        fullString = parString[:-2]
537 >                        fullString += '\\}'
538 >                        list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
539 >                        common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.")
540 >                        self.jobDestination.append(blockSites[block])
541 >                        common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
542 >                        jobsOfBlock[block].append(jobCount+1)
543 >                        # reset counter
544 >                        jobCount = jobCount + 1
545 >                        totalEventCount = totalEventCount + eventsPerJobRequested
546 >                        eventsRemaining = eventsRemaining - eventsPerJobRequested
547 >                        jobSkipEventCount = 0
548 >                        # reset file
549 >                        parString = "\\{"
550 >                        filesEventCount = 0
551 >                        newFile = 1
552 >                        fileCount += 1
553 >                        
554 >                    # if more events in file remain than eventsPerJobRequested
555 >                    else :
556 >                        # close job but don't touch new file
557 >                        fullString = parString[:-2]
558 >                        fullString += '\\}'
559 >                        list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
560 >                        common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.")
561 >                        self.jobDestination.append(blockSites[block])
562 >                        common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
563 >                        jobsOfBlock[block].append(jobCount+1)
564 >                        # increase counter
565 >                        jobCount = jobCount + 1
566 >                        totalEventCount = totalEventCount + eventsPerJobRequested
567 >                        eventsRemaining = eventsRemaining - eventsPerJobRequested
568 >                        # calculate skip events for last file
569 >                        # use filesEventCount (contains several files), jobSkipEventCount and eventsPerJobRequest
570 >                        jobSkipEventCount = eventsPerJobRequested - (filesEventCount - jobSkipEventCount - self.eventsbyfile[file])
571 >                        # remove all but the last file
572 >                        filesEventCount = self.eventsbyfile[file]
573 >                        parString = "\\{"
574 >                        parString += '\\\"' + file + '\\\"\,'
575 >                    pass # END if
576 >                pass # END while (iterate over files in the block)
577 >        pass # END while (iterate over blocks in the dataset)
578 >        self.ncjobs = self.total_number_of_jobs = jobCount
579 >        if (eventsRemaining > 0 and jobCount < totalNumberOfJobs ):
580 >            common.logger.message("Could not run on all requested events because some blocks not hosted at allowed sites.")
581 >        common.logger.message(str(jobCount)+" job(s) can run on "+str(totalEventCount)+" events.\n")
582 >        
583 >        # screen output
584 >        screenOutput = "List of jobs and available destination sites:\n\n"
585 >
586 >        # keep trace of block with no sites to print a warning at the end
587 >        noSiteBlock = []
588 >        bloskNoSite = []
589 >
590 >        blockCounter = 0
591 >        for block in blocks:
592 >            if block in jobsOfBlock.keys() :
593 >                blockCounter += 1
594 >                screenOutput += "Block %5i: jobs %20s: sites: %s\n" % (blockCounter,spanRanges(jobsOfBlock[block]),','.join(self.blackWhiteListParser.checkWhiteList(self.blackWhiteListParser.checkBlackList(blockSites[block],block),block)))
595 >                if len(self.blackWhiteListParser.checkWhiteList(self.blackWhiteListParser.checkBlackList(blockSites[block],block),block)) == 0:
596 >                    noSiteBlock.append( spanRanges(jobsOfBlock[block]) )
597 >                    bloskNoSite.append( blockCounter )
598 >        
599 >        common.logger.message(screenOutput)
600 >        if len(noSiteBlock) > 0 and len(bloskNoSite) > 0:
601 >            msg = 'WARNING: No sites are hosting any part of data for block:\n                '
602 >            virgola = ""
603 >            if len(bloskNoSite) > 1:
604 >                virgola = ","
605 >            for block in bloskNoSite:
606 >                msg += ' ' + str(block) + virgola
607 >            msg += '\n               Related jobs:\n                 '
608 >            virgola = ""
609 >            if len(noSiteBlock) > 1:
610 >                virgola = ","
611 >            for range_jobs in noSiteBlock:
612 >                msg += str(range_jobs) + virgola
613 >            msg += '\n               will not be submitted and this block of data can not be analyzed!\n'
614 >            common.logger.message(msg)
615  
616          self.list_of_args = list_of_lists
390 –        #print self.list_of_args
617          return
618  
619 <    def jobSplittingPerEvents(self):
619 >    def jobSplittingNoInput(self):
620          """
621          Perform job splitting based on number of event per job
622          """
623          common.logger.debug(5,'Splitting per events')
624          common.logger.message('Required '+str(self.eventsPerJob)+' events per job ')
625 +        common.logger.message('Required '+str(self.theNumberOfJobs)+' jobs in total ')
626          common.logger.message('Required '+str(self.total_number_of_events)+' events in total ')
627  
628          if (self.total_number_of_events < 0):
629              msg='Cannot split jobs per Events with "-1" as total number of events'
630              raise CrabException(msg)
631  
632 <        self.total_number_of_jobs = int(self.total_number_of_events/self.eventsPerJob)
633 <
634 <        print "@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@"
635 <        print "self.total_number_of_events = ", self.total_number_of_events
636 <        print "self.eventsPerJob = ", self.eventsPerJob
637 <        print "self.total_number_of_jobs = ", self.total_number_of_jobs
638 <        print "@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@"
639 <        
632 >        if (self.selectEventsPerJob):
633 >            if (self.selectTotalNumberEvents):
634 >                self.total_number_of_jobs = int(self.total_number_of_events/self.eventsPerJob)
635 >            elif(self.selectNumberOfJobs) :  
636 >                self.total_number_of_jobs =self.theNumberOfJobs
637 >                self.total_number_of_events =int(self.theNumberOfJobs*self.eventsPerJob)
638 >
639 >        elif (self.selectNumberOfJobs) :
640 >            self.total_number_of_jobs = self.theNumberOfJobs
641 >            self.eventsPerJob = int(self.total_number_of_events/self.total_number_of_jobs)
642 >
643          common.logger.debug(5,'N jobs  '+str(self.total_number_of_jobs))
644  
645          # is there any remainder?
# Line 417 | Line 647 | class Cmssw(JobType):
647  
648          common.logger.debug(5,'Check  '+str(check))
649  
650 +        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')
651          if check > 0:
652 <            common.logger.message('Warning: asked '+self.total_number_of_events+' but will do only '+(int(self.total_number_of_jobs)*self.eventsPerJob))
422 <
423 <        common.logger.message(str(self.total_number_of_jobs)+' jobs will be created for a total of '+str(self.total_number_of_jobs*self.eventsPerJob)+' events')
652 >            common.logger.message('Warning: asked '+str(self.total_number_of_events)+' but can do only '+str(int(self.total_number_of_jobs)*self.eventsPerJob))
653  
654          # argument is seed number.$i
655          self.list_of_args = []
656          for i in range(self.total_number_of_jobs):
657 <            self.list_of_args.append(int(str(self.sourceSeed)+str(i)))
658 <        print self.list_of_args
657 >            ## Since there is no input, any site is good
658 >           # self.jobDestination.append(["Any"])
659 >            self.jobDestination.append([""]) #must be empty to write correctly the xml
660 >            args=[]
661 >            if (self.firstRun):
662 >                    ## pythia first run
663 >                #self.list_of_args.append([(str(self.firstRun)+str(i))])
664 >                args.append(str(self.firstRun)+str(i))
665 >            else:
666 >                ## no first run
667 >                #self.list_of_args.append([str(i)])
668 >                args.append(str(i))
669 >            if (self.sourceSeed):
670 >                args.append(str(self.sourceSeed)+str(i))
671 >                if (self.sourceSeedVtx):
672 >                    ## + vtx random seed
673 >                    args.append(str(self.sourceSeedVtx)+str(i))
674 >                if (self.sourceSeedG4):
675 >                    ## + G4 random seed
676 >                    args.append(str(self.sourceSeedG4)+str(i))
677 >                if (self.sourceSeedMix):    
678 >                    ## + Mix random seed
679 >                    args.append(str(self.sourceSeedMix)+str(i))
680 >                pass
681 >            pass
682 >            self.list_of_args.append(args)
683 >        pass
684 >            
685 >        # print self.list_of_args
686 >
687 >        return
688 >
689 >
690 >    def jobSplittingForScript(self):#CarlosDaniele
691 >        """
692 >        Perform job splitting based on number of job
693 >        """
694 >        common.logger.debug(5,'Splitting per job')
695 >        common.logger.message('Required '+str(self.theNumberOfJobs)+' jobs in total ')
696  
697 +        self.total_number_of_jobs = self.theNumberOfJobs
698 +
699 +        common.logger.debug(5,'N jobs  '+str(self.total_number_of_jobs))
700 +
701 +        common.logger.message(str(self.total_number_of_jobs)+' jobs can be created')
702 +
703 +        # argument is seed number.$i
704 +        self.list_of_args = []
705 +        for i in range(self.total_number_of_jobs):
706 +            ## Since there is no input, any site is good
707 +           # self.jobDestination.append(["Any"])
708 +            self.jobDestination.append([""])
709 +            ## no random seed
710 +            self.list_of_args.append([str(i)])
711          return
712  
713      def split(self, jobParams):
# Line 441 | Line 721 | class Cmssw(JobType):
721              jobParams.append("")
722          
723          for job in range(njobs):
724 <            jobParams[job] = str(arglist[job])
724 >            jobParams[job] = arglist[job]
725 >            # print str(arglist[job])
726 >            # print jobParams[job]
727              common.jobDB.setArguments(job, jobParams[job])
728 +            common.logger.debug(5,"Job "+str(job)+" Destination: "+str(self.jobDestination[job]))
729 +            common.jobDB.setDestination(job, self.jobDestination[job])
730  
731          common.jobDB.save()
732          return
733      
734      def getJobTypeArguments(self, nj, sched):
735 <        return common.jobDB.arguments(nj)
735 >        result = ''
736 >        for i in common.jobDB.arguments(nj):
737 >            result=result+str(i)+" "
738 >        return result
739    
740      def numberOfJobs(self):
741          # Fabio
742          return self.total_number_of_jobs
743  
457 –    def checkBlackList(self, allSites):
458 –        if len(self.reCEBlackList)==0: return allSites
459 –        sites = []
460 –        for site in allSites:
461 –            common.logger.debug(10,'Site '+site)
462 –            good=1
463 –            for re in self.reCEBlackList:
464 –                if re.search(site):
465 –                    common.logger.message('CE in black list, skipping site '+site)
466 –                    good=0
467 –                pass
468 –            if good: sites.append(site)
469 –        if len(sites) == 0:
470 –            common.logger.debug(3,"No sites found after BlackList")
471 –        return sites
472 –
473 –    def checkWhiteList(self, allSites):
474 –
475 –        if len(self.reCEWhiteList)==0: return allSites
476 –        sites = []
477 –        for site in allSites:
478 –            good=0
479 –            for re in self.reCEWhiteList:
480 –                if re.search(site):
481 –                    common.logger.debug(5,'CE in white list, adding site '+site)
482 –                    good=1
483 –                if not good: continue
484 –                sites.append(site)
485 –        if len(sites) == 0:
486 –            common.logger.message("No sites found after WhiteList\n")
487 –        else:
488 –            common.logger.debug(5,"Selected sites via WhiteList are "+str(sites)+"\n")
489 –        return sites
490 –
744      def getTarBall(self, exe):
745          """
746          Return the TarBall with lib and exe
747          """
748          
749          # if it exist, just return it
750 <        self.tgzNameWithPath = common.work_space.shareDir()+self.tgz_name
750 >        #
751 >        # Marco. Let's start to use relative path for Boss XML files
752 >        #
753 >        self.tgzNameWithPath = common.work_space.pathForTgz()+'share/'+self.tgz_name
754          if os.path.exists(self.tgzNameWithPath):
755              return self.tgzNameWithPath
756  
# Line 508 | Line 764 | class Cmssw(JobType):
764          # First of all declare the user Scram area
765          swArea = self.scram.getSWArea_()
766          #print "swArea = ", swArea
767 <        swVersion = self.scram.getSWVersion()
768 <        #print "swVersion = ", swVersion
767 >        # swVersion = self.scram.getSWVersion()
768 >        # print "swVersion = ", swVersion
769          swReleaseTop = self.scram.getReleaseTop_()
770          #print "swReleaseTop = ", swReleaseTop
771          
# Line 517 | Line 773 | class Cmssw(JobType):
773          if swReleaseTop == '' or swArea == swReleaseTop:
774              return
775  
776 <        filesToBeTarred = []
777 <        ## First find the executable
778 <        if (self.executable != ''):
779 <            exeWithPath = self.scram.findFile_(executable)
780 < #           print exeWithPath
781 <            if ( not exeWithPath ):
782 <                raise CrabException('User executable '+executable+' not found')
783 <
784 <            ## then check if it's private or not
785 <            if exeWithPath.find(swReleaseTop) == -1:
786 <                # the exe is private, so we must ship
787 <                common.logger.debug(5,"Exe "+exeWithPath+" to be tarred")
788 <                path = swArea+'/'
789 <                exe = string.replace(exeWithPath, path,'')
790 <                filesToBeTarred.append(exe)
791 <                pass
792 <            else:
793 <                # the exe is from release, we'll find it on WN
794 <                pass
795 <
796 <        ## Now get the libraries: only those in local working area
797 <        libDir = 'lib'
798 <        lib = swArea+'/' +libDir
799 <        common.logger.debug(5,"lib "+lib+" to be tarred")
800 <        if os.path.exists(lib):
801 <            filesToBeTarred.append(libDir)
802 <
803 <        ## Now check if module dir is present
804 <        moduleDir = 'module'
805 <        if os.path.isdir(swArea+'/'+moduleDir):
806 <            filesToBeTarred.append(moduleDir)
807 <
808 <        ## Now check if the Data dir is present
809 <        dataDir = 'src/Data/'
810 <        if os.path.isdir(swArea+'/'+dataDir):
811 <            filesToBeTarred.append(dataDir)
812 <
813 <        ## Create the tar-ball
814 <        if len(filesToBeTarred)>0:
815 <            cwd = os.getcwd()
816 <            os.chdir(swArea)
817 <            tarcmd = 'tar zcvf ' + self.tgzNameWithPath + ' '
818 <            for line in filesToBeTarred:
819 <                tarcmd = tarcmd + line + ' '
820 <            cout = runCommand(tarcmd)
821 <            if not cout:
822 <                raise CrabException('Could not create tar-ball')
823 <            os.chdir(cwd)
824 <        else:
825 <            common.logger.debug(5,"No files to be to be tarred")
776 >        import tarfile
777 >        try: # create tar ball
778 >            tar = tarfile.open(self.tgzNameWithPath, "w:gz")
779 >            ## First find the executable
780 >            if (self.executable != ''):
781 >                exeWithPath = self.scram.findFile_(executable)
782 >                if ( not exeWithPath ):
783 >                    raise CrabException('User executable '+executable+' not found')
784 >    
785 >                ## then check if it's private or not
786 >                if exeWithPath.find(swReleaseTop) == -1:
787 >                    # the exe is private, so we must ship
788 >                    common.logger.debug(5,"Exe "+exeWithPath+" to be tarred")
789 >                    path = swArea+'/'
790 >                    # distinguish case when script is in user project area or given by full path somewhere else
791 >                    if exeWithPath.find(path) >= 0 :
792 >                        exe = string.replace(exeWithPath, path,'')
793 >                        tar.add(path+exe,os.path.basename(executable))
794 >                    else :
795 >                        tar.add(exeWithPath,os.path.basename(executable))
796 >                    pass
797 >                else:
798 >                    # the exe is from release, we'll find it on WN
799 >                    pass
800 >    
801 >            ## Now get the libraries: only those in local working area
802 >            libDir = 'lib'
803 >            lib = swArea+'/' +libDir
804 >            common.logger.debug(5,"lib "+lib+" to be tarred")
805 >            if os.path.exists(lib):
806 >                tar.add(lib,libDir)
807 >    
808 >            ## Now check if module dir is present
809 >            moduleDir = 'module'
810 >            module = swArea + '/' + moduleDir
811 >            if os.path.isdir(module):
812 >                tar.add(module,moduleDir)
813 >
814 >            ## Now check if any data dir(s) is present
815 >            swAreaLen=len(swArea)
816 >            for root, dirs, files in os.walk(swArea):
817 >                if "data" in dirs:
818 >                    common.logger.debug(5,"data "+root+"/data"+" to be tarred")
819 >                    tar.add(root+"/data",root[swAreaLen:]+"/data")
820 >
821 >            ## Add ProdAgent dir to tar
822 >            paDir = 'ProdAgentApi'
823 >            pa = os.environ['CRABDIR'] + '/' + 'ProdAgentApi'
824 >            if os.path.isdir(pa):
825 >                tar.add(pa,paDir)
826 >
827 >            ### FEDE FOR DBS PUBLICATION
828 >            ## Add PRODCOMMON dir to tar
829 >            prodcommonDir = 'ProdCommon'
830 >            prodcommonPath = os.environ['CRABDIR'] + '/' + 'ProdCommon'
831 >            if os.path.isdir(prodcommonPath):
832 >                tar.add(prodcommonPath,prodcommonDir)
833 >            #############################    
834 >        
835 >            common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
836 >            tar.close()
837 >        except :
838 >            raise CrabException('Could not create tar-ball')
839 >
840 >        ## check for tarball size
841 >        tarballinfo = os.stat(self.tgzNameWithPath)
842 >        if ( tarballinfo.st_size > self.MaxTarBallSize*1024*1024 ) :
843 >            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.')
844 >
845 >        ## create tar-ball with ML stuff
846 >        self.MLtgzfile =  common.work_space.pathForTgz()+'share/MLfiles.tgz'
847 >        try:
848 >            tar = tarfile.open(self.MLtgzfile, "w:gz")
849 >            path=os.environ['CRABDIR'] + '/python/'
850 >            for file in ['report.py', 'DashboardAPI.py', 'Logger.py', 'ProcInfo.py', 'apmon.py', 'parseCrabFjr.py']:
851 >                tar.add(path+file,file)
852 >            common.logger.debug(5,"Files added to "+self.MLtgzfile+" : "+str(tar.getnames()))
853 >            tar.close()
854 >        except :
855 >            raise CrabException('Could not create ML files tar-ball')
856          
857          return
858          
859 +    def additionalInputFileTgz(self):
860 +        """
861 +        Put all additional files into a tar ball and return its name
862 +        """
863 +        import tarfile
864 +        tarName=  common.work_space.pathForTgz()+'share/'+self.additional_tgz_name
865 +        tar = tarfile.open(tarName, "w:gz")
866 +        for file in self.additional_inbox_files:
867 +            tar.add(file,string.split(file,'/')[-1])
868 +        common.logger.debug(5,"Files added to "+self.additional_tgz_name+" : "+str(tar.getnames()))
869 +        tar.close()
870 +        return tarName
871 +
872      def wsSetupEnvironment(self, nj):
873          """
874          Returns part of a job script which prepares
# Line 581 | Line 880 | class Cmssw(JobType):
880          ## OLI_Daniele at this level  middleware already known
881  
882          txt += 'if [ $middleware == LCG ]; then \n'
883 +        txt += '    echo "### First set SCRAM ARCH and BUILD_ARCH to ' + self.executable_arch + ' ###"\n'
884 +        txt += '    export SCRAM_ARCH='+self.executable_arch+'\n'
885 +        txt += '    export BUILD_ARCH='+self.executable_arch+'\n'
886          txt += self.wsSetupCMSLCGEnvironment_()
887          txt += 'elif [ $middleware == OSG ]; then\n'
888 <        txt += '    time=`date -u +"%s"`\n'
889 <        txt += '    WORKING_DIR=$OSG_WN_TMP/cms_$time\n'
588 <        txt += '    echo "Creating working directory: $WORKING_DIR"\n'
589 <        txt += '    /bin/mkdir -p $WORKING_DIR\n'
888 >        txt += '    WORKING_DIR=`/bin/mktemp  -d $OSG_WN_TMP/cms_XXXXXXXXXXXX`\n'
889 >        txt += '    echo "Created working directory: $WORKING_DIR"\n'
890          txt += '    if [ ! -d $WORKING_DIR ] ;then\n'
891          txt += '        echo "SET_CMS_ENV 10016 ==> OSG $WORKING_DIR could not be created on WN `hostname`"\n'
892 <        txt += '        echo "JOB_EXIT_STATUS = 10016"\n'
893 <        txt += '        echo "JobExitCode=10016" | tee -a $RUNTIME_AREA/$repo\n'
894 <        txt += '        dumpStatus $RUNTIME_AREA/$repo\n'
892 >        txt += '    echo "JOB_EXIT_STATUS = 10016"\n'
893 >        txt += '    echo "JobExitCode=10016" | tee -a $RUNTIME_AREA/$repo\n'
894 >        txt += '    dumpStatus $RUNTIME_AREA/$repo\n'
895          txt += '        rm -f $RUNTIME_AREA/$repo \n'
896          txt += '        echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
597 –        txt += '        echo "SyncGridJobId=`echo $SyncGridJobId`" | tee -a $RUNTIME_AREA/$repo \n'
897          txt += '        echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
898          txt += '        exit 1\n'
899          txt += '    fi\n'
# Line 602 | Line 901 | class Cmssw(JobType):
901          txt += '    echo "Change to working directory: $WORKING_DIR"\n'
902          txt += '    cd $WORKING_DIR\n'
903          txt += self.wsSetupCMSOSGEnvironment_()
904 +        txt += '    echo "### Set SCRAM ARCH to ' + self.executable_arch + ' ###"\n'
905 +        txt += '    export SCRAM_ARCH='+self.executable_arch+'\n'
906          txt += 'fi\n'
907  
908          # Prepare JobType-specific part
# Line 617 | Line 918 | class Cmssw(JobType):
918          txt += '   dumpStatus $RUNTIME_AREA/$repo\n'
919          txt += '   rm -f $RUNTIME_AREA/$repo \n'
920          txt += '   echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
620 –        txt += '   echo "SyncGridJobId=`echo $SyncGridJobId`" | tee -a $RUNTIME_AREA/$repo \n'
921          txt += '   echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
922          ## OLI_Daniele
923          txt += '    if [ $middleware == OSG ]; then \n'
# Line 625 | 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_CMS_ENV 10018 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after CMSSW CMSSW_0_6_1 not found on `hostname`"\n'
929 <        txt += '            echo "JOB_EXIT_STATUS = 10018"\n'
930 <        txt += '            echo "JobExitCode=10018" | tee -a $RUNTIME_AREA/$repo\n'
931 <        txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
928 >        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'
929 >        txt += '            echo "JOB_EXIT_STATUS = 10018"\n'
930 >        txt += '            echo "JobExitCode=10018" | 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'
634 –        txt += '            echo "SyncGridJobId=`echo $SyncGridJobId`" | tee -a $RUNTIME_AREA/$repo \n'
934          txt += '            echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
935          txt += '        fi\n'
936          txt += '    fi \n'
# Line 639 | Line 938 | class Cmssw(JobType):
938          txt += 'fi \n'
939          txt += 'echo "CMSSW_VERSION =  '+self.version+'"\n'
940          txt += 'cd '+self.version+'\n'
941 +        ########## FEDE FOR DBS2 ######################
942 +        txt += 'SOFTWARE_DIR=`pwd`\n'
943 +        txt += 'echo SOFTWARE_DIR=$SOFTWARE_DIR \n'
944 +        ###############################################
945          ### needed grep for bug in scramv1 ###
946 +        txt += scram+' runtime -sh\n'
947          txt += 'eval `'+scram+' runtime -sh | grep -v SCRAMRT_LSB_JOBNAME`\n'
948 +        txt += 'echo $PATH\n'
949  
950          # Handle the arguments:
951          txt += "\n"
952          txt += "## number of arguments (first argument always jobnumber)\n"
953          txt += "\n"
954 <        txt += "narg=$#\n"
955 <        txt += "if [ $narg -lt 2 ]\n"
954 > #        txt += "narg=$#\n"
955 >        txt += "if [ $nargs -lt 2 ]\n"
956          txt += "then\n"
957 <        txt += "    echo 'SET_EXE_ENV 1 ==> ERROR Too few arguments' +$narg+ \n"
957 >        txt += "    echo 'SET_EXE_ENV 1 ==> ERROR Too few arguments' +$nargs+ \n"
958          txt += '    echo "JOB_EXIT_STATUS = 50113"\n'
959          txt += '    echo "JobExitCode=50113" | tee -a $RUNTIME_AREA/$repo\n'
960          txt += '    dumpStatus $RUNTIME_AREA/$repo\n'
961          txt += '    rm -f $RUNTIME_AREA/$repo \n'
962          txt += '    echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
658 –        txt += '    echo "SyncGridJobId=`echo $SyncGridJobId`" | tee -a $RUNTIME_AREA/$repo \n'
963          txt += '    echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
964          ## OLI_Daniele
965          txt += '    if [ $middleware == OSG ]; then \n'
# Line 663 | Line 967 | class Cmssw(JobType):
967          txt += '        cd $RUNTIME_AREA\n'
968          txt += '        /bin/rm -rf $WORKING_DIR\n'
969          txt += '        if [ -d $WORKING_DIR ] ;then\n'
970 <        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'
971 <        txt += '            echo "JOB_EXIT_STATUS = 50114"\n'
972 <        txt += '            echo "JobExitCode=50114" | tee -a $RUNTIME_AREA/$repo\n'
973 <        txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
970 >        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'
971 >        txt += '            echo "JOB_EXIT_STATUS = 50114"\n'
972 >        txt += '            echo "JobExitCode=50114" | tee -a $RUNTIME_AREA/$repo\n'
973 >        txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
974          txt += '            rm -f $RUNTIME_AREA/$repo \n'
975          txt += '            echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
672 –        txt += '            echo "SyncGridJobId=`echo $SyncGridJobId`" | tee -a $RUNTIME_AREA/$repo \n'
976          txt += '            echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
977          txt += '        fi\n'
978          txt += '    fi \n'
# Line 679 | Line 982 | class Cmssw(JobType):
982  
983          # Prepare job-specific part
984          job = common.job_list[nj]
985 <        pset = os.path.basename(job.configFilename())
986 <        txt += '\n'
987 <        if (self.datasetPath): # standard job
988 <            txt += 'InputFiles=$2\n'
989 <            txt += 'echo "Inputfiles:<$InputFiles>"\n'
990 <            txt += 'sed "s#{\'INPUT\'}#$InputFiles#" $RUNTIME_AREA/'+pset+' > pset.cfg\n'
991 <        else:  # pythia like job
992 <            txt += 'Seed=$2\n'
993 <            txt += 'echo "Seed: <$Seed>"\n'
994 <            txt += 'sed "s#INPUT#$Seed#" $RUNTIME_AREA/'+pset+' > pset.cfg\n'
985 >        ### FEDE FOR DBS OUTPUT PUBLICATION
986 >        if (self.datasetPath):
987 >            txt += '\n'
988 >            txt += 'DatasetPath='+self.datasetPath+'\n'
989 >
990 >            datasetpath_split = self.datasetPath.split("/")
991 >            
992 >            txt += 'PrimaryDataset='+datasetpath_split[1]+'\n'
993 >            txt += 'DataTier='+datasetpath_split[2]+'\n'
994 >            #txt += 'ProcessedDataset='+datasetpath_split[3]+'\n'
995 >            txt += 'ApplicationFamily=cmsRun\n'
996 >
997 >        else:
998 >            txt += 'DatasetPath=MCDataTier\n'
999 >            txt += 'PrimaryDataset=null\n'
1000 >            txt += 'DataTier=null\n'
1001 >            #txt += 'ProcessedDataset=null\n'
1002 >            txt += 'ApplicationFamily=MCDataTier\n'
1003 >        if self.pset != None: #CarlosDaniele
1004 >            pset = os.path.basename(job.configFilename())
1005 >            txt += '\n'
1006 >            txt += 'cp  $RUNTIME_AREA/'+pset+' .\n'
1007 >            if (self.datasetPath): # standard job
1008 >                #txt += 'InputFiles=$2\n'
1009 >                txt += 'InputFiles=${args[1]}\n'
1010 >                txt += 'MaxEvents=${args[2]}\n'
1011 >                txt += 'SkipEvents=${args[3]}\n'
1012 >                txt += 'echo "Inputfiles:<$InputFiles>"\n'
1013 >                txt += 'sed "s#{\'INPUT\'}#$InputFiles#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
1014 >                txt += 'echo "MaxEvents:<$MaxEvents>"\n'
1015 >                txt += 'sed "s#INPUTMAXEVENTS#$MaxEvents#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
1016 >                txt += 'echo "SkipEvents:<$SkipEvents>"\n'
1017 >                txt += 'sed "s#INPUTSKIPEVENTS#$SkipEvents#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
1018 >            else:  # pythia like job
1019 >                seedIndex=1
1020 >                if (self.firstRun):
1021 >                    txt += 'FirstRun=${args['+str(seedIndex)+']}\n'
1022 >                    txt += 'echo "FirstRun: <$FirstRun>"\n'
1023 >                    txt += 'sed "s#\<INPUTFIRSTRUN\>#$FirstRun#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
1024 >                    seedIndex=seedIndex+1
1025 >
1026 >                if (self.sourceSeed):
1027 >                    txt += 'Seed=${args['+str(seedIndex)+']}\n'
1028 >                    txt += 'sed "s#\<INPUT\>#$Seed#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
1029 >                    seedIndex=seedIndex+1
1030 >                    ## the following seeds are not always present
1031 >                    if (self.sourceSeedVtx):
1032 >                        txt += 'VtxSeed=${args['+str(seedIndex)+']}\n'
1033 >                        txt += 'echo "VtxSeed: <$VtxSeed>"\n'
1034 >                        txt += 'sed "s#\<INPUTVTX\>#$VtxSeed#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
1035 >                        seedIndex += 1
1036 >                    if (self.sourceSeedG4):
1037 >                        txt += 'G4Seed=${args['+str(seedIndex)+']}\n'
1038 >                        txt += 'echo "G4Seed: <$G4Seed>"\n'
1039 >                        txt += 'sed "s#\<INPUTG4\>#$G4Seed#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
1040 >                        seedIndex += 1
1041 >                    if (self.sourceSeedMix):
1042 >                        txt += 'mixSeed=${args['+str(seedIndex)+']}\n'
1043 >                        txt += 'echo "MixSeed: <$mixSeed>"\n'
1044 >                        txt += 'sed "s#\<INPUTMIX\>#$mixSeed#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
1045 >                        seedIndex += 1
1046 >                    pass
1047 >                pass
1048 >            txt += 'mv -f '+pset+' pset.cfg\n'
1049  
1050          if len(self.additional_inbox_files) > 0:
1051 <            for file in self.additional_inbox_files:
1052 <                txt += 'if [ -e $RUNTIME_AREA/'+file+' ] ; then\n'
1053 <                txt += '   cp $RUNTIME_AREA/'+file+' .\n'
697 <                txt += '   chmod +x '+file+'\n'
698 <                txt += 'fi\n'
1051 >            txt += 'if [ -e $RUNTIME_AREA/'+self.additional_tgz_name+' ] ; then\n'
1052 >            txt += '  tar xzvf $RUNTIME_AREA/'+self.additional_tgz_name+'\n'
1053 >            txt += 'fi\n'
1054              pass
1055  
1056 <        txt += 'echo "### END JOB SETUP ENVIRONMENT ###"\n\n'
1057 <
1058 <        txt += '\n'
1059 <        txt += 'echo "***** cat pset.cfg *********"\n'
1060 <        txt += 'cat pset.cfg\n'
1061 <        txt += 'echo "****** end pset.cfg ********"\n'
1062 <        txt += '\n'
1063 <        # txt += 'echo "***** cat pset1.cfg *********"\n'
1064 <        # txt += 'cat pset1.cfg\n'
1065 <        # txt += 'echo "****** end pset1.cfg ********"\n'
1056 >        if self.pset != None: #CarlosDaniele
1057 >            txt += 'echo "### END JOB SETUP ENVIRONMENT ###"\n\n'
1058 >        
1059 >            txt += '\n'
1060 >            txt += 'echo "***** cat pset.cfg *********"\n'
1061 >            txt += 'cat pset.cfg\n'
1062 >            txt += 'echo "****** end pset.cfg ********"\n'
1063 >            txt += '\n'
1064 >            ### FEDE FOR DBS OUTPUT PUBLICATION
1065 >            txt += 'PSETHASH=`EdmConfigHash < pset.cfg` \n'
1066 >            txt += 'echo "PSETHASH = $PSETHASH" \n'
1067 >            ##############
1068 >            txt += '\n'
1069 >            # txt += 'echo "***** cat pset1.cfg *********"\n'
1070 >            # txt += 'cat pset1.cfg\n'
1071 >            # txt += 'echo "****** end pset1.cfg ********"\n'
1072          return txt
1073  
1074 <    def wsBuildExe(self, nj):
1074 >    def wsBuildExe(self, nj=0):
1075          """
1076          Put in the script the commands to build an executable
1077          or a library.
# Line 737 | Line 1098 | class Cmssw(JobType):
1098              txt += '           dumpStatus $RUNTIME_AREA/$repo\n'
1099              txt += '           rm -f $RUNTIME_AREA/$repo \n'
1100              txt += '           echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
740 –            txt += '           echo "SyncGridJobId=`echo $SyncGridJobId`" | tee -a $RUNTIME_AREA/$repo \n'
1101              txt += '           echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1102              txt += '       fi\n'
1103              txt += '   fi \n'
# Line 746 | Line 1106 | class Cmssw(JobType):
1106              txt += 'else \n'
1107              txt += '   echo "Successful untar" \n'
1108              txt += 'fi \n'
1109 +            txt += '\n'
1110 +            txt += 'echo "Include ProdAgentApi and PRODCOMMON in PYTHONPATH"\n'
1111 +            txt += 'if [ -z "$PYTHONPATH" ]; then\n'
1112 +            #### FEDE FOR DBS OUTPUT PUBLICATION
1113 +            txt += '   export PYTHONPATH=$SOFTWARE_DIR/ProdAgentApi:$SOFTWARE_DIR/ProdCommon\n'
1114 +            #txt += '   export PYTHONPATH=`pwd`/ProdAgentApi:`pwd`/ProdCommon\n'
1115 +            #txt += '   export PYTHONPATH=ProdAgentApi\n'
1116 +            txt += 'else\n'
1117 +            txt += '   export PYTHONPATH=$SOFTWARE_DIR/ProdAgentApi:$SOFTWARE_DIR/ProdCommon:${PYTHONPATH}\n'
1118 +            #txt += '   export PYTHONPATH=`pwd`/ProdAgentApi:`pwd`/ProdCommon:${PYTHONPATH}\n'
1119 +            #txt += '   export PYTHONPATH=ProdAgentApi:${PYTHONPATH}\n'
1120 +            txt += 'echo "PYTHONPATH=$PYTHONPATH"\n'
1121 +            ###################  
1122 +            txt += 'fi\n'
1123 +            txt += '\n'
1124 +
1125              pass
1126          
1127          return txt
# Line 757 | Line 1133 | class Cmssw(JobType):
1133          """
1134          
1135      def executableName(self):
1136 <        return self.executable
1136 >        if self.scriptExe: #CarlosDaniele
1137 >            return "sh "
1138 >        else:
1139 >            return self.executable
1140  
1141      def executableArgs(self):
1142 <        return " -p pset.cfg"
1142 >        if self.scriptExe:#CarlosDaniele
1143 >            return   self.scriptExe + " $NJob"
1144 >        else:
1145 >            # if >= CMSSW_1_5_X, add -e
1146 >            version_array = self.scram.getSWVersion().split('_')
1147 >            major = 0
1148 >            minor = 0
1149 >            try:
1150 >                major = int(version_array[1])
1151 >                minor = int(version_array[2])
1152 >            except:
1153 >                msg = "Cannot parse CMSSW version string: " + "_".join(version_array) + " for major and minor release number!"  
1154 >                raise CrabException(msg)
1155 >            if major >= 1 and minor >= 5 :
1156 >                return " -e -p pset.cfg"
1157 >            else:
1158 >                return " -p pset.cfg"
1159  
1160      def inputSandbox(self, nj):
1161          """
1162          Returns a list of filenames to be put in JDL input sandbox.
1163          """
1164          inp_box = []
1165 <        # dict added to delete duplicate from input sandbox file list
1166 <        seen = {}
1165 >        # # dict added to delete duplicate from input sandbox file list
1166 >        # seen = {}
1167          ## code
1168          if os.path.isfile(self.tgzNameWithPath):
1169              inp_box.append(self.tgzNameWithPath)
1170 +        if os.path.isfile(self.MLtgzfile):
1171 +            inp_box.append(self.MLtgzfile)
1172          ## config
1173 <        inp_box.append(common.job_list[nj].configFilename())
1173 >        if not self.pset is None:
1174 >            inp_box.append(common.work_space.pathForTgz() + 'job/' + self.configFilename())
1175          ## additional input files
1176 <        #for file in self.additional_inbox_files:
1177 <        #    inp_box.append(common.work_space.cwdDir()+file)
1176 >        tgz = self.additionalInputFileTgz()
1177 >        inp_box.append(tgz)
1178          return inp_box
1179  
1180      def outputSandbox(self, nj):
# Line 785 | Line 1183 | class Cmssw(JobType):
1183          """
1184          out_box = []
1185  
788 –        stdout=common.job_list[nj].stdout()
789 –        stderr=common.job_list[nj].stderr()
790 –
1186          ## User Declared output files
1187 <        for out in self.output_file:
1187 >        for out in (self.output_file+self.output_file_sandbox):
1188              n_out = nj + 1
1189              out_box.append(self.numberFile_(out,str(n_out)))
1190          return out_box
796 –        return []
1191  
1192      def prepareSteeringCards(self):
1193          """
# Line 809 | Line 1203 | class Cmssw(JobType):
1203          txt = '\n'
1204          txt += '# directory content\n'
1205          txt += 'ls \n'
1206 <        file_list = ''
1207 <        for fileWithSuffix in self.output_file:
1206 >
1207 >        txt += 'output_exit_status=0\n'
1208 >        
1209 >        for fileWithSuffix in (self.output_file_sandbox):
1210              output_file_num = self.numberFile_(fileWithSuffix, '$NJob')
815 –            file_list=file_list+output_file_num+' '
1211              txt += '\n'
1212              txt += '# check output file\n'
1213 <            txt += 'ls '+fileWithSuffix+'\n'
1214 <            txt += 'exe_result=$?\n'
1215 <            txt += 'if [ $exe_result -ne 0 ] ; then\n'
1216 <            txt += '   echo "ERROR: No output file to manage"\n'
1217 <            txt += '   echo "JOB_EXIT_STATUS = $exe_result"\n'
1218 <            txt += '   echo "JobExitCode=60302" | tee -a $RUNTIME_AREA/$repo\n'
824 <            txt += '   dumpStatus $RUNTIME_AREA/$repo\n'
825 <            txt += '   rm -f $RUNTIME_AREA/$repo \n'
826 <            txt += '   echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
827 <            txt += '   echo "SyncGridJobId=`echo $SyncGridJobId`" | tee -a $RUNTIME_AREA/$repo \n'
828 <            txt += '   echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
829 <            ### OLI_DANIELE
1213 >            txt += 'if [ -e ./'+fileWithSuffix+' ] ; then\n'
1214 >            txt += '    mv '+fileWithSuffix+' $RUNTIME_AREA\n'
1215 >            txt += '    cp $RUNTIME_AREA/'+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n'
1216 >            txt += 'else\n'
1217 >            txt += '    exit_status=60302\n'
1218 >            txt += '    echo "ERROR: Problem with output file '+fileWithSuffix+'"\n'
1219              if common.scheduler.boss_scheduler_name == 'condor_g':
1220                  txt += '    if [ $middleware == OSG ]; then \n'
1221                  txt += '        echo "prepare dummy output file"\n'
1222                  txt += '        echo "Processing of job output failed" > $RUNTIME_AREA/'+output_file_num+'\n'
1223                  txt += '    fi \n'
1224 +            txt += 'fi\n'
1225 +        
1226 +        for fileWithSuffix in (self.output_file):
1227 +            output_file_num = self.numberFile_(fileWithSuffix, '$NJob')
1228 +            txt += '\n'
1229 +            txt += '# check output file\n'
1230 +            txt += 'if [ -e ./'+fileWithSuffix+' ] ; then\n'
1231 +            txt += '    mv '+fileWithSuffix+' $RUNTIME_AREA\n'
1232 +            txt += '    cp $RUNTIME_AREA/'+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n'
1233              txt += 'else\n'
1234 <            txt += '   cp '+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n'
1234 >            txt += '    exit_status=60302\n'
1235 >            txt += '    echo "ERROR: Problem with output file '+fileWithSuffix+'"\n'
1236 >            txt += '    echo "JOB_EXIT_STATUS = $exit_status"\n'
1237 >            txt += '    output_exit_status=$exit_status\n'
1238 >            if common.scheduler.boss_scheduler_name == 'condor_g':
1239 >                txt += '    if [ $middleware == OSG ]; then \n'
1240 >                txt += '        echo "prepare dummy output file"\n'
1241 >                txt += '        echo "Processing of job output failed" > $RUNTIME_AREA/'+output_file_num+'\n'
1242 >                txt += '    fi \n'
1243              txt += 'fi\n'
1244 <      
1244 >        file_list = []
1245 >        for fileWithSuffix in (self.output_file):
1246 >             file_list.append(self.numberFile_(fileWithSuffix, '$NJob'))
1247 >            
1248 >        txt += 'file_list="'+string.join(file_list,' ')+'"\n'
1249          txt += 'cd $RUNTIME_AREA\n'
840 –        file_list=file_list[:-1]
841 –        txt += 'file_list="'+file_list+'"\n'
842 –        ### OLI_DANIELE
843 –        txt += 'if [ $middleware == OSG ]; then\n'  
844 –        txt += '    cd $RUNTIME_AREA\n'
845 –        txt += '    echo "Remove working directory: $WORKING_DIR"\n'
846 –        txt += '    /bin/rm -rf $WORKING_DIR\n'
847 –        txt += '    if [ -d $WORKING_DIR ] ;then\n'
848 –        txt += '        echo "SET_EXE 60999 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after cleanup of WN"\n'
849 –        txt += '        echo "JOB_EXIT_STATUS = 60999"\n'
850 –        txt += '        echo "JobExitCode=60999" | tee -a $RUNTIME_AREA/$repo\n'
851 –        txt += '        dumpStatus $RUNTIME_AREA/$repo\n'
852 –        txt += '        rm -f $RUNTIME_AREA/$repo \n'
853 –        txt += '        echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
854 –        txt += '        echo "SyncGridJobId=`echo $SyncGridJobId`" | tee -a $RUNTIME_AREA/$repo \n'
855 –        txt += '        echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
856 –        txt += '    fi\n'
857 –        txt += 'fi\n'
858 –        txt += '\n'
1250          return txt
1251  
1252      def numberFile_(self, file, txt):
# Line 866 | Line 1257 | class Cmssw(JobType):
1257          # take away last extension
1258          name = p[0]
1259          for x in p[1:-1]:
1260 <           name=name+"."+x
1260 >            name=name+"."+x
1261          # add "_txt"
1262          if len(p)>1:
1263 <          ext = p[len(p)-1]
1264 <          #result = name + '_' + str(txt) + "." + ext
874 <          result = name + '_' + txt + "." + ext
1263 >            ext = p[len(p)-1]
1264 >            result = name + '_' + txt + "." + ext
1265          else:
1266 <          #result = name + '_' + str(txt)
877 <          result = name + '_' + txt
1266 >            result = name + '_' + txt
1267          
1268          return result
1269  
1270 <    def getRequirements(self):
1270 >    def getRequirements(self, nj=[]):
1271          """
1272          return job requirements to add to jdl files
1273          """
1274          req = ''
1275 <        if common.analisys_common_info['sw_version']:
1275 >        if self.version:
1276              req='Member("VO-cms-' + \
1277 <                 common.analisys_common_info['sw_version'] + \
1277 >                 self.version + \
1278 >                 '", other.GlueHostApplicationSoftwareRunTimeEnvironment)'
1279 >        ## SL add requirement for OS version only if SL4
1280 >        #reSL4 = re.compile( r'slc4' )
1281 >        if self.executable_arch: # and reSL4.search(self.executable_arch):
1282 >            req+=' && Member("VO-cms-' + \
1283 >                 self.executable_arch + \
1284                   '", other.GlueHostApplicationSoftwareRunTimeEnvironment)'
1285 <        if common.analisys_common_info['sites']:
1286 <            if len(common.analisys_common_info['sites'])>0:
1287 <                req = req + ' && ('
893 <                for i in range(len(common.analisys_common_info['sites'])):
894 <                    req = req + 'other.GlueCEInfoHostName == "' \
895 <                         + common.analisys_common_info['sites'][i] + '"'
896 <                    if ( i < (int(len(common.analisys_common_info['sites']) - 1)) ):
897 <                        req = req + ' || '
898 <            req = req + ')'
899 <        #print "req = ", req
1285 >
1286 >        req = req + ' && (other.GlueHostNetworkAdapterOutboundIP)'
1287 >
1288          return req
1289  
1290      def configFilename(self):
# Line 913 | Line 1301 | class Cmssw(JobType):
1301          txt += '   echo "### SETUP CMS OSG  ENVIRONMENT ###"\n'
1302          txt += '   if [ -f $GRID3_APP_DIR/cmssoft/cmsset_default.sh ] ;then\n'
1303          txt += '      # Use $GRID3_APP_DIR/cmssoft/cmsset_default.sh to setup cms software\n'
1304 +        txt += '       export SCRAM_ARCH='+self.executable_arch+'\n'
1305          txt += '       source $GRID3_APP_DIR/cmssoft/cmsset_default.sh '+self.version+'\n'
1306 <        txt += '   elif [ -f $OSG_APP/cmssoft/cmsset_default.sh ] ;then\n'
1307 <        txt += '      # Use $OSG_APP/cmssoft/cmsset_default.sh to setup cms software\n'
1308 <        txt += '       source $OSG_APP/cmssoft/cmsset_default.sh '+self.version+'\n'
1306 >        txt += '   elif [ -f $OSG_APP/cmssoft/cms/cmsset_default.sh ] ;then\n'
1307 >        txt += '      # Use $OSG_APP/cmssoft/cms/cmsset_default.sh to setup cms software\n'
1308 >        txt += '       export SCRAM_ARCH='+self.executable_arch+'\n'
1309 >        txt += '       source $OSG_APP/cmssoft/cms/cmsset_default.sh '+self.version+'\n'
1310          txt += '   else\n'
1311 <        txt += '       echo "SET_CMS_ENV 10020 ==> ERROR $GRID3_APP_DIR/cmssoft/cmsset_default.sh and $OSG_APP/cmssoft/cmsset_default.sh file not found"\n'
1311 >        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'
1312          txt += '       echo "JOB_EXIT_STATUS = 10020"\n'
1313          txt += '       echo "JobExitCode=10020" | tee -a $RUNTIME_AREA/$repo\n'
1314          txt += '       dumpStatus $RUNTIME_AREA/$repo\n'
1315          txt += '       rm -f $RUNTIME_AREA/$repo \n'
1316          txt += '       echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
927 –        txt += '       echo "SyncGridJobId=`echo $SyncGridJobId`" | tee -a $RUNTIME_AREA/$repo \n'
1317          txt += '       echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1318          txt += '       exit 1\n'
1319          txt += '\n'
# Line 932 | Line 1321 | class Cmssw(JobType):
1321          txt += '       cd $RUNTIME_AREA\n'
1322          txt += '       /bin/rm -rf $WORKING_DIR\n'
1323          txt += '       if [ -d $WORKING_DIR ] ;then\n'
1324 <        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/cmsset_default.sh file not found"\n'
1325 <        txt += '            echo "JOB_EXIT_STATUS = 10017"\n'
1326 <        txt += '            echo "JobExitCode=10017" | tee -a $RUNTIME_AREA/$repo\n'
1327 <        txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
1328 <        txt += '            rm -f $RUNTIME_AREA/$repo \n'
1329 <        txt += '            echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1330 <        txt += '            echo "SyncGridJobId=`echo $SyncGridJobId`" | tee -a $RUNTIME_AREA/$repo \n'
942 <        txt += '            echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1324 >        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'
1325 >        txt += '           echo "JOB_EXIT_STATUS = 10017"\n'
1326 >        txt += '           echo "JobExitCode=10017" | tee -a $RUNTIME_AREA/$repo\n'
1327 >        txt += '           dumpStatus $RUNTIME_AREA/$repo\n'
1328 >        txt += '           rm -f $RUNTIME_AREA/$repo \n'
1329 >        txt += '           echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1330 >        txt += '           echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1331          txt += '       fi\n'
1332          txt += '\n'
1333          txt += '       exit 1\n'
# Line 965 | Line 1353 | class Cmssw(JobType):
1353          txt += '       dumpStatus $RUNTIME_AREA/$repo\n'
1354          txt += '       rm -f $RUNTIME_AREA/$repo \n'
1355          txt += '       echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
968 –        txt += '       echo "SyncGridJobId=`echo $SyncGridJobId`" | tee -a $RUNTIME_AREA/$repo \n'
1356          txt += '       echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1357          txt += '       exit 1\n'
1358          txt += '   else\n'
# Line 977 | Line 1364 | class Cmssw(JobType):
1364          txt += '           dumpStatus $RUNTIME_AREA/$repo\n'
1365          txt += '           rm -f $RUNTIME_AREA/$repo \n'
1366          txt += '           echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
980 –        txt += '           echo "SyncGridJobId=`echo $SyncGridJobId`" | tee -a $RUNTIME_AREA/$repo \n'
1367          txt += '           echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1368          txt += '           exit 1\n'
1369          txt += '       fi\n'
# Line 991 | Line 1377 | class Cmssw(JobType):
1377          txt += '           dumpStatus $RUNTIME_AREA/$repo\n'
1378          txt += '           rm -f $RUNTIME_AREA/$repo \n'
1379          txt += '           echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
994 –        txt += '           echo "SyncGridJobId=`echo $SyncGridJobId`" | tee -a $RUNTIME_AREA/$repo \n'
1380          txt += '           echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1381          txt += '           exit 1\n'
1382          txt += '       fi\n'
1383          txt += '   fi\n'
1384          txt += '   \n'
1000 –        txt += '   string=`cat /etc/redhat-release`\n'
1001 –        txt += '   echo $string\n'
1002 –        txt += '   if [[ $string = *alhalla* ]]; then\n'
1003 –        txt += '       echo "SCRAM_ARCH= $SCRAM_ARCH"\n'
1004 –        txt += '   elif [[ $string = *Enterprise* ]] || [[ $string = *cientific* ]]; then\n'
1005 –        txt += '       export SCRAM_ARCH=slc3_ia32_gcc323\n'
1006 –        txt += '       echo "SCRAM_ARCH= $SCRAM_ARCH"\n'
1007 –        txt += '   else\n'
1008 –        txt += '       echo "SET_CMS_ENV 10033 ==> ERROR OS unknown, LCG environment not initialized"\n'
1009 –        txt += '       echo "JOB_EXIT_STATUS = 10033"\n'
1010 –        txt += '       echo "JobExitCode=10033" | tee -a $RUNTIME_AREA/$repo\n'
1011 –        txt += '       dumpStatus $RUNTIME_AREA/$repo\n'
1012 –        txt += '       rm -f $RUNTIME_AREA/$repo \n'
1013 –        txt += '       echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1014 –        txt += '       echo "SyncGridJobId=`echo $SyncGridJobId`" | tee -a $RUNTIME_AREA/$repo \n'
1015 –        txt += '       echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1016 –        txt += '       exit 1\n'
1017 –        txt += '   fi\n'
1385          txt += '   echo "SET_CMS_ENV 0 ==> setup cms environment ok"\n'
1386          txt += '   echo "### END SETUP CMS LCG ENVIRONMENT ###"\n'
1387          return txt
1388  
1389 +    ### FEDE FOR DBS OUTPUT PUBLICATION
1390 +    def modifyReport(self, nj):
1391 +        """
1392 +        insert the part of the script that modifies the FrameworkJob Report
1393 +        """
1394 +
1395 +        txt = ''
1396 +        try:
1397 +            publish_data = int(self.cfg_params['USER.publish_data'])          
1398 +        except KeyError:
1399 +            publish_data = 0
1400 +        if (publish_data == 1):  
1401 +            txt += 'echo "Modify Job Report" \n'
1402 +            #txt += 'chmod a+x $RUNTIME_AREA/'+self.version+'/ProdAgentApi/FwkJobRep/ModifyJobReport.py\n'
1403 +            ################ FEDE FOR DBS2 #############################################
1404 +            txt += 'chmod a+x $SOFTWARE_DIR/ProdAgentApi/FwkJobRep/ModifyJobReport.py\n'
1405 +            #############################################################################
1406 +            #try:
1407 +            #    publish_data = int(self.cfg_params['USER.publish_data'])          
1408 +            #except KeyError:
1409 +            #    publish_data = 0
1410 +
1411 +            txt += 'if [ -z "$SE" ]; then\n'
1412 +            txt += '    SE="" \n'
1413 +            txt += 'fi \n'
1414 +            txt += 'if [ -z "$SE_PATH" ]; then\n'
1415 +            txt += '    SE_PATH="" \n'
1416 +            txt += 'fi \n'
1417 +            txt += 'echo "SE = $SE"\n'
1418 +            txt += 'echo "SE_PATH = $SE_PATH"\n'
1419 +
1420 +        #if (publish_data == 1):  
1421 +            #processedDataset = self.cfg_params['USER.processed_datasetname']
1422 +            processedDataset = self.cfg_params['USER.publish_data_name']
1423 +            txt += 'ProcessedDataset='+processedDataset+'\n'
1424 +            #### LFN=/store/user/<user>/processedDataset_PSETHASH
1425 +            txt += 'if [ "$SE_PATH" == "" ]; then\n'
1426 +            #### FEDE: added slash in LFN ##############
1427 +            txt += '    FOR_LFN=/copy_problems/ \n'
1428 +            txt += 'else \n'
1429 +            txt += '    tmp=`echo $SE_PATH | awk -F \'store\' \'{print$2}\'` \n'
1430 +            #####  FEDE TO BE CHANGED, BECAUSE STORE IS HARDCODED!!!! ########
1431 +            txt += '    FOR_LFN=/store$tmp \n'
1432 +            txt += 'fi \n'
1433 +            txt += 'echo "ProcessedDataset = $ProcessedDataset"\n'
1434 +            txt += 'echo "FOR_LFN = $FOR_LFN" \n'
1435 +            txt += 'echo "CMSSW_VERSION = $CMSSW_VERSION"\n\n'
1436 +            #txt += 'echo "$RUNTIME_AREA/'+self.version+'/ProdAgentApi/FwkJobRep/ModifyJobReport.py crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH"\n'
1437 +            txt += 'echo "$SOFTWARE_DIR/ProdAgentApi/FwkJobRep/ModifyJobReport.py crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH"\n'
1438 +            txt += '$SOFTWARE_DIR/ProdAgentApi/FwkJobRep/ModifyJobReport.py crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH\n'
1439 +            #txt += '$RUNTIME_AREA/'+self.version+'/ProdAgentApi/FwkJobRep/ModifyJobReport.py crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH\n'
1440 +      
1441 +            txt += 'modifyReport_result=$?\n'
1442 +            txt += 'echo modifyReport_result = $modifyReport_result\n'
1443 +            txt += 'if [ $modifyReport_result -ne 0 ]; then\n'
1444 +            txt += '    exit_status=1\n'
1445 +            txt += '    echo "ERROR: Problem with ModifyJobReport"\n'
1446 +            txt += 'else\n'
1447 +            txt += '    mv NewFrameworkJobReport.xml crab_fjr_$NJob.xml\n'
1448 +            txt += 'fi\n'
1449 +        else:
1450 +            txt += 'echo "no data publication required"\n'
1451 +            #txt += 'ProcessedDataset=no_data_to_publish \n'
1452 +            #### FEDE: added slash in LFN ##############
1453 +            #txt += 'FOR_LFN=/local/ \n'
1454 +            #txt += 'echo "ProcessedDataset = $ProcessedDataset"\n'
1455 +            #txt += 'echo "FOR_LFN = $FOR_LFN" \n'
1456 +        return txt
1457 +
1458 +    def cleanEnv(self):
1459 +        ### OLI_DANIELE
1460 +        txt = ''
1461 +        txt += 'if [ $middleware == OSG ]; then\n'  
1462 +        txt += '    cd $RUNTIME_AREA\n'
1463 +        txt += '    echo "Remove working directory: $WORKING_DIR"\n'
1464 +        txt += '    /bin/rm -rf $WORKING_DIR\n'
1465 +        txt += '    if [ -d $WORKING_DIR ] ;then\n'
1466 +        txt += '              echo "SET_EXE 60999 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after cleanup of WN"\n'
1467 +        txt += '              echo "JOB_EXIT_STATUS = 60999"\n'
1468 +        txt += '              echo "JobExitCode=60999" | tee -a $RUNTIME_AREA/$repo\n'
1469 +        txt += '              dumpStatus $RUNTIME_AREA/$repo\n'
1470 +        txt += '        rm -f $RUNTIME_AREA/$repo \n'
1471 +        txt += '        echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1472 +        txt += '        echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1473 +        txt += '    fi\n'
1474 +        txt += 'fi\n'
1475 +        txt += '\n'
1476 +        return txt
1477 +
1478      def setParam_(self, param, value):
1479          self._params[param] = value
1480  
# Line 1030 | Line 1486 | class Cmssw(JobType):
1486          
1487      def getTaskid(self):
1488          return self._taskId
1489 +
1490 +    def uniquelist(self, old):
1491 +        """
1492 +        remove duplicates from a list
1493 +        """
1494 +        nd={}
1495 +        for e in old:
1496 +            nd[e]=0
1497 +        return nd.keys()
1498 +
1499 +
1500 +    def checkOut(self, limit):
1501 +        """
1502 +        check the dimension of the output files
1503 +        """
1504 +        txt = 'echo "*****************************************"\n'
1505 +        txt += 'echo "** Starting output sandbox limit check **"\n'
1506 +        txt += 'echo "*****************************************"\n'
1507 +        allOutFiles = ""
1508 +        listOutFiles = []
1509 +        for fileOut in (self.output_file+self.output_file_sandbox):
1510 +             if fileOut.find('crab_fjr') == -1:
1511 +                 allOutFiles = allOutFiles + " " + self.numberFile_(fileOut, '$NJob')
1512 +                 listOutFiles.append(self.numberFile_(fileOut, '$NJob'))
1513 +        txt += 'echo "OUTPUT files: '+str(allOutFiles)+'";\n'
1514 +        txt += 'ls -gGhrta;\n'
1515 +        txt += 'sum=0;\n'
1516 +        txt += 'for file in '+str(allOutFiles)+' ; do\n'
1517 +        txt += '    if [ -e $file ]; then\n'
1518 +        txt += '        tt=`ls -gGrta $file | awk \'{ print $3 }\'`\n'
1519 +        txt += '        sum=`expr $sum + $tt`\n'
1520 +        txt += '    else\n'
1521 +        txt += '        echo "WARNING: output file $file not found!"\n'
1522 +        txt += '    fi\n'
1523 +        txt += 'done\n'
1524 +        txt += 'echo "Total Output dimension: $sum";\n'
1525 +        txt += 'limit='+str(limit)+';\n'
1526 +        txt += 'echo "OUTPUT FILES LIMIT SET TO: $limit";\n'
1527 +        txt += 'if [ $limit -lt $sum ]; then\n'
1528 +        txt += '    echo "WARNING: output files have to big size - something will be lost;"\n'
1529 +        txt += '    echo "         checking the output file sizes..."\n'
1530 +        """
1531 +        txt += '    dim=0;\n'
1532 +        txt += '    exclude=0;\n'
1533 +        txt += '    for files in '+str(allOutFiles)+' ; do\n'
1534 +        txt += '        sumTemp=0;\n'
1535 +        txt += '        for file2 in '+str(allOutFiles)+' ; do\n'
1536 +        txt += '            if [ $file != $file2 ]; then\n'
1537 +        txt += '                tt=`ls -gGrta $file2 | awk \'{ print $3 }\';`\n'
1538 +        txt += '                sumTemp=`expr $sumTemp + $tt`;\n'
1539 +        txt += '            fi\n'
1540 +        txt += '        done\n'
1541 +        txt += '        if [ $sumTemp -lt $limit ]; then\n'
1542 +        txt += '            if [ $dim -lt $sumTemp ]; then\n'
1543 +        txt += '                dim=$sumTemp;\n'
1544 +        txt += '                exclude=$file;\n'
1545 +        txt += '            fi\n'
1546 +        txt += '        fi\n'
1547 +        txt += '    done\n'
1548 +        txt += '    echo "Dimension calculated: $dim"; echo "File to exclude: $exclude";\n'
1549 +        """
1550 +        txt += '    tot=0;\n'
1551 +        txt += '    for file2 in '+str(allOutFiles)+' ; do\n'
1552 +        txt += '        tt=`ls -gGrta $file2 | awk \'{ print $3 }\';`\n'
1553 +        txt += '        tot=`expr $tot + $tt`;\n'
1554 +        txt += '        if [ $limit -lt $tot ]; then\n'
1555 +        txt += '            tot=`expr $tot - $tt`;\n'
1556 +        txt += '            fileLast=$file;\n'
1557 +        txt += '            break;\n'
1558 +        txt += '        fi\n'
1559 +        txt += '    done\n'
1560 +        txt += '    echo "Dimension calculated: $tot"; echo "First file to exclude: $file";\n'
1561 +        txt += '    flag=0;\n'    
1562 +        txt += '    for filess in '+str(allOutFiles)+' ; do\n'
1563 +        txt += '        if [ $fileLast = $filess ]; then\n'
1564 +        txt += '            flag=1;\n'
1565 +        txt += '        fi\n'
1566 +        txt += '        if [ $flag -eq 1 ]; then\n'
1567 +        txt += '            rm -f $filess;\n'
1568 +        txt += '        fi\n'
1569 +        txt += '    done\n'
1570 +        txt += '    ls -agGhrt;\n'
1571 +        txt += '    echo "WARNING: output files are too big in dimension: can not put in the output_sandbox.";\n'
1572 +        txt += '    echo "JOB_EXIT_STATUS = 70000";\n'
1573 +        txt += '    exit_status=70000;\n'
1574 +        txt += 'else'
1575 +        txt += '    echo "Total Output dimension $sum is fine.";\n'
1576 +        txt += 'fi\n'
1577 +        txt += 'echo "*****************************************"\n'
1578 +        txt += 'echo "*** Ending output sandbox limit check ***"\n'
1579 +        txt += 'echo "*****************************************"\n'
1580 +        return txt

Diff Legend

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