| 2 |
|
from crab_logger import Logger |
| 3 |
|
from crab_exceptions import * |
| 4 |
|
from crab_util import * |
| 5 |
+ |
import math |
| 6 |
|
import common |
| 7 |
+ |
import PsetManipulator |
| 8 |
|
|
| 9 |
< |
import DataDiscovery |
| 10 |
< |
import DataLocation |
| 9 |
> |
import DBSInfo_EDM |
| 10 |
> |
import DataDiscovery_EDM |
| 11 |
> |
import DataLocation_EDM |
| 12 |
|
import Scram |
| 13 |
|
|
| 14 |
|
import os, string, re |
| 19 |
|
common.logger.debug(3,'CMSSW::__init__') |
| 20 |
|
|
| 21 |
|
self.analisys_common_info = {} |
| 22 |
< |
|
| 22 |
> |
# Marco. |
| 23 |
> |
self._params = {} |
| 24 |
> |
self.cfg_params = cfg_params |
| 25 |
|
log = common.logger |
| 26 |
|
|
| 27 |
|
self.scram = Scram.Scram(cfg_params) |
| 31 |
|
self.executable = '' |
| 32 |
|
self.tgz_name = 'default.tgz' |
| 33 |
|
|
| 34 |
+ |
|
| 35 |
|
self.version = self.scram.getSWVersion() |
| 36 |
+ |
self.setParam_('application', self.version) |
| 37 |
|
common.analisys_common_info['sw_version'] = self.version |
| 38 |
+ |
### FEDE |
| 39 |
+ |
common.analisys_common_info['copy_input_data'] = 0 |
| 40 |
+ |
common.analisys_common_info['events_management'] = 1 |
| 41 |
|
|
| 42 |
|
### collect Data cards |
| 43 |
|
try: |
| 44 |
< |
self.owner = cfg_params['CMSSW.owner'] |
| 45 |
< |
log.debug(6, "CMSSW::CMSSW(): owner = "+self.owner) |
| 46 |
< |
self.dataset = cfg_params['CMSSW.dataset'] |
| 47 |
< |
log.debug(6, "CMSSW::CMSSW(): dataset = "+self.dataset) |
| 44 |
> |
tmp = cfg_params['CMSSW.datasetpath'] |
| 45 |
> |
log.debug(6, "CMSSW::CMSSW(): datasetPath = "+tmp) |
| 46 |
> |
if string.lower(tmp)=='none': |
| 47 |
> |
self.datasetPath = None |
| 48 |
> |
self.selectNoInput = 1 |
| 49 |
> |
else: |
| 50 |
> |
self.datasetPath = tmp |
| 51 |
> |
self.selectNoInput = 0 |
| 52 |
|
except KeyError: |
| 53 |
< |
msg = "Error: owner and/or dataset not defined " |
| 53 |
> |
msg = "Error: datasetpath not defined " |
| 54 |
|
raise CrabException(msg) |
| 55 |
|
|
| 56 |
+ |
# ML monitoring |
| 57 |
+ |
# split dataset path style: /PreProdR3Minbias/SIM/GEN-SIM |
| 58 |
+ |
if not self.datasetPath: |
| 59 |
+ |
self.setParam_('dataset', 'None') |
| 60 |
+ |
self.setParam_('owner', 'None') |
| 61 |
+ |
else: |
| 62 |
+ |
datasetpath_split = self.datasetPath.split("/") |
| 63 |
+ |
self.setParam_('dataset', datasetpath_split[1]) |
| 64 |
+ |
self.setParam_('owner', datasetpath_split[-1]) |
| 65 |
+ |
|
| 66 |
+ |
self.setTaskid_() |
| 67 |
+ |
self.setParam_('taskId', self.cfg_params['taskId']) |
| 68 |
+ |
|
| 69 |
|
self.dataTiers = [] |
| 43 |
– |
try: |
| 44 |
– |
tmpDataTiers = string.split(cfg_params['CMSSW.data_tier'],',') |
| 45 |
– |
for tmp in tmpDataTiers: |
| 46 |
– |
tmp=string.strip(tmp) |
| 47 |
– |
self.dataTiers.append(tmp) |
| 48 |
– |
pass |
| 49 |
– |
pass |
| 50 |
– |
except KeyError: |
| 51 |
– |
pass |
| 52 |
– |
log.debug(6, "Cmssw::Cmssw(): dataTiers = "+str(self.dataTiers)) |
| 70 |
|
|
| 71 |
|
## now the application |
| 72 |
|
try: |
| 73 |
|
self.executable = cfg_params['CMSSW.executable'] |
| 74 |
+ |
self.setParam_('exe', self.executable) |
| 75 |
|
log.debug(6, "CMSSW::CMSSW(): executable = "+self.executable) |
| 76 |
|
msg = "Default executable cmsRun overridden. Switch to " + self.executable |
| 77 |
|
log.debug(3,msg) |
| 78 |
|
except KeyError: |
| 79 |
|
self.executable = 'cmsRun' |
| 80 |
+ |
self.setParam_('exe', self.executable) |
| 81 |
|
msg = "User executable not defined. Use cmsRun" |
| 82 |
|
log.debug(3,msg) |
| 83 |
|
pass |
| 102 |
|
tmp=string.strip(tmp) |
| 103 |
|
self.output_file.append(tmp) |
| 104 |
|
pass |
| 86 |
– |
|
| 105 |
|
else: |
| 106 |
|
log.message("No output file defined: only stdout/err will be available") |
| 107 |
|
pass |
| 112 |
|
|
| 113 |
|
# script_exe file as additional file in inputSandbox |
| 114 |
|
try: |
| 115 |
< |
self.scriptExe = cfg_params['CMSSW.script_exe'] |
| 116 |
< |
self.additional_inbox_files.append(self.scriptExe) |
| 115 |
> |
self.scriptExe = cfg_params['USER.script_exe'] |
| 116 |
> |
self.additional_inbox_files.append(self.scriptExe) |
| 117 |
> |
if self.scriptExe != '': |
| 118 |
> |
if not os.path.isfile(self.scriptExe): |
| 119 |
> |
msg ="WARNING. file "+self.scriptExe+" not found" |
| 120 |
> |
raise CrabException(msg) |
| 121 |
|
except KeyError: |
| 122 |
|
pass |
| 101 |
– |
if self.scriptExe != '': |
| 102 |
– |
if os.path.isfile(self.scriptExe): |
| 103 |
– |
pass |
| 104 |
– |
else: |
| 105 |
– |
log.message("WARNING. file "+self.scriptExe+" not found") |
| 106 |
– |
sys.exit() |
| 123 |
|
|
| 124 |
|
## additional input files |
| 125 |
|
try: |
| 126 |
|
tmpAddFiles = string.split(cfg_params['CMSSW.additional_input_files'],',') |
| 127 |
|
for tmp in tmpAddFiles: |
| 128 |
+ |
if not os.path.exists(tmp): |
| 129 |
+ |
raise CrabException("Additional input file not found: "+tmp) |
| 130 |
|
tmp=string.strip(tmp) |
| 131 |
|
self.additional_inbox_files.append(tmp) |
| 132 |
|
pass |
| 134 |
|
except KeyError: |
| 135 |
|
pass |
| 136 |
|
|
| 137 |
+ |
# files per job |
| 138 |
+ |
try: |
| 139 |
+ |
self.filesPerJob = int(cfg_params['CMSSW.files_per_jobs']) #Daniele |
| 140 |
+ |
self.selectFilesPerJob = 1 |
| 141 |
+ |
except KeyError: |
| 142 |
+ |
self.filesPerJob = 0 |
| 143 |
+ |
self.selectFilesPerJob = 0 |
| 144 |
+ |
|
| 145 |
+ |
## Events per job |
| 146 |
+ |
try: |
| 147 |
+ |
self.eventsPerJob =int( cfg_params['CMSSW.events_per_job']) |
| 148 |
+ |
self.selectEventsPerJob = 1 |
| 149 |
+ |
except KeyError: |
| 150 |
+ |
self.eventsPerJob = -1 |
| 151 |
+ |
self.selectEventsPerJob = 0 |
| 152 |
+ |
|
| 153 |
+ |
## number of jobs |
| 154 |
+ |
try: |
| 155 |
+ |
self.theNumberOfJobs =int( cfg_params['CMSSW.number_of_jobs']) |
| 156 |
+ |
self.selectNumberOfJobs = 1 |
| 157 |
+ |
except KeyError: |
| 158 |
+ |
self.theNumberOfJobs = 0 |
| 159 |
+ |
self.selectNumberOfJobs = 0 |
| 160 |
+ |
|
| 161 |
+ |
## source seed for pythia |
| 162 |
+ |
try: |
| 163 |
+ |
self.sourceSeed = int(cfg_params['CMSSW.pythia_seed']) |
| 164 |
+ |
except KeyError: |
| 165 |
+ |
self.sourceSeed = None |
| 166 |
+ |
common.logger.debug(5,"No seed given") |
| 167 |
+ |
|
| 168 |
+ |
if not (self.selectFilesPerJob + self.selectEventsPerJob + self.selectNumberOfJobs == 1 ): |
| 169 |
+ |
msg = 'Must define either files_per_jobs or events_per_job or number_of_jobs' |
| 170 |
+ |
raise CrabException(msg) |
| 171 |
+ |
|
| 172 |
|
try: |
| 173 |
|
self.total_number_of_events = int(cfg_params['CMSSW.total_number_of_events']) |
| 174 |
|
except KeyError: |
| 175 |
< |
msg = 'Must define total_number_of_events and job_number_of_events' |
| 175 |
> |
msg = 'Must define total_number_of_events' |
| 176 |
|
raise CrabException(msg) |
| 124 |
– |
|
| 125 |
– |
#Marco: FirstEvent is nolonger used inside PSet |
| 126 |
– |
# try: |
| 127 |
– |
# self.first = int(cfg_params['CMSSW.first_event']) |
| 128 |
– |
# except KeyError: |
| 129 |
– |
# self.first = 0 |
| 130 |
– |
# pass |
| 131 |
– |
# log.debug(6, "Orca::Orca(): total number of events = "+`self.total_number_of_events`) |
| 132 |
– |
#log.debug(6, "Orca::Orca(): events per job = "+`self.job_number_of_events`) |
| 133 |
– |
# log.debug(6, "Orca::Orca(): first event = "+`self.first`) |
| 177 |
|
|
| 178 |
|
CEBlackList = [] |
| 179 |
|
try: |
| 193 |
|
CEWhiteList = [] |
| 194 |
|
try: |
| 195 |
|
tmpGood = string.split(cfg_params['EDG.ce_white_list'],',') |
| 153 |
– |
#tmpGood = ['cern'] |
| 196 |
|
for tmp in tmpGood: |
| 197 |
|
tmp=string.strip(tmp) |
| 156 |
– |
#if (tmp == 'cnaf'): tmp = 'webserver' ########## warning: temp. patch |
| 198 |
|
CEWhiteList.append(tmp) |
| 199 |
|
except KeyError: |
| 200 |
|
pass |
| 206 |
|
|
| 207 |
|
common.logger.debug(5,'CEWhiteList: '+str(CEWhiteList)) |
| 208 |
|
|
| 209 |
+ |
self.PsetEdit = PsetManipulator.PsetManipulator(self.pset) #Daniele Pset |
| 210 |
+ |
|
| 211 |
|
#DBSDLS-start |
| 212 |
|
## Initialize the variables that are extracted from DBS/DLS and needed in other places of the code |
| 213 |
|
self.maxEvents=0 # max events available ( --> check the requested nb. of evts in Creator.py) |
| 214 |
|
self.DBSPaths={} # all dbs paths requested ( --> input to the site local discovery script) |
| 215 |
|
## Perform the data location and discovery (based on DBS/DLS) |
| 216 |
< |
self.DataDiscoveryAndLocation(cfg_params) |
| 216 |
> |
## SL: Don't if NONE is specified as input (pythia use case) |
| 217 |
> |
common.analisys_common_info['sites']=None |
| 218 |
> |
if self.datasetPath: |
| 219 |
> |
self.DataDiscoveryAndLocation(cfg_params) |
| 220 |
|
#DBSDLS-end |
| 221 |
|
|
| 222 |
|
self.tgzNameWithPath = self.getTarBall(self.executable) |
| 223 |
+ |
|
| 224 |
+ |
## Select Splitting |
| 225 |
+ |
if self.selectNoInput: self.jobSplittingNoInput() |
| 226 |
+ |
elif self.selectFilesPerJob or self.selectEventsPerJob or self.selectNumberOfJobs: self.jobSplittingPerFiles() |
| 227 |
+ |
else: |
| 228 |
+ |
msg = 'Don\'t know how to split...' |
| 229 |
+ |
raise CrabException(msg) |
| 230 |
+ |
|
| 231 |
+ |
# modify Pset |
| 232 |
+ |
try: |
| 233 |
+ |
if (self.datasetPath): # standard job |
| 234 |
+ |
#self.PsetEdit.maxEvent(self.eventsPerJob) |
| 235 |
+ |
# always process all events in a file |
| 236 |
+ |
self.PsetEdit.maxEvent("-1") |
| 237 |
+ |
self.PsetEdit.inputModule("INPUT") |
| 238 |
+ |
|
| 239 |
+ |
else: # pythia like job |
| 240 |
+ |
self.PsetEdit.maxEvent(self.eventsPerJob) |
| 241 |
+ |
if (self.sourceSeed) : |
| 242 |
+ |
self.PsetEdit.pythiaSeed("INPUT","INPUTVTX") |
| 243 |
+ |
self.PsetEdit.psetWriter(self.configFilename()) |
| 244 |
+ |
except: |
| 245 |
+ |
msg='Error while manipuliating ParameterSet: exiting...' |
| 246 |
+ |
raise CrabException(msg) |
| 247 |
|
|
| 248 |
|
def DataDiscoveryAndLocation(self, cfg_params): |
| 249 |
|
|
| 250 |
< |
fun = "CMSSW::DataDiscoveryAndLocation()" |
| 250 |
> |
common.logger.debug(10,"CMSSW::DataDiscoveryAndLocation()") |
| 251 |
> |
|
| 252 |
> |
datasetPath=self.datasetPath |
| 253 |
> |
|
| 254 |
> |
## TODO |
| 255 |
> |
dataTiersList = "" |
| 256 |
> |
dataTiers = dataTiersList.split(',') |
| 257 |
|
|
| 258 |
|
## Contact the DBS |
| 259 |
|
try: |
| 260 |
< |
self.pubdata=DataDiscovery.DataDiscovery(self.owner, |
| 185 |
< |
self.dataset, |
| 186 |
< |
self.dataTiers, |
| 187 |
< |
cfg_params) |
| 260 |
> |
self.pubdata=DataDiscovery_EDM.DataDiscovery_EDM(datasetPath, dataTiers, cfg_params) |
| 261 |
|
self.pubdata.fetchDBSInfo() |
| 262 |
|
|
| 263 |
< |
except DataDiscovery.NotExistingDatasetError, ex : |
| 263 |
> |
except DataDiscovery_EDM.NotExistingDatasetError, ex : |
| 264 |
|
msg = 'ERROR ***: failed Data Discovery in DBS : %s'%ex.getErrorMessage() |
| 265 |
|
raise CrabException(msg) |
| 266 |
|
|
| 267 |
< |
except DataDiscovery.NoDataTierinProvenanceError, ex : |
| 267 |
> |
except DataDiscovery_EDM.NoDataTierinProvenanceError, ex : |
| 268 |
|
msg = 'ERROR ***: failed Data Discovery in DBS : %s'%ex.getErrorMessage() |
| 269 |
|
raise CrabException(msg) |
| 270 |
< |
except DataDiscovery.DataDiscoveryError, ex: |
| 270 |
> |
except DataDiscovery_EDM.DataDiscoveryError, ex: |
| 271 |
|
msg = 'ERROR ***: failed Data Discovery in DBS %s'%ex.getErrorMessage() |
| 272 |
|
raise CrabException(msg) |
| 273 |
|
|
| 274 |
|
## get list of all required data in the form of dbs paths (dbs path = /dataset/datatier/owner) |
| 275 |
< |
self.DBSPaths=self.pubdata.getDBSPaths() |
| 276 |
< |
common.logger.message("Required data are : ") |
| 277 |
< |
for path in self.DBSPaths: |
| 278 |
< |
common.logger.message(" --> "+path ) |
| 275 |
> |
## self.DBSPaths=self.pubdata.getDBSPaths() |
| 276 |
> |
common.logger.message("Required data are :"+self.datasetPath) |
| 277 |
> |
|
| 278 |
> |
filesbyblock=self.pubdata.getFiles() |
| 279 |
> |
# print filesbyblock |
| 280 |
> |
self.AllInputFiles=filesbyblock.values() |
| 281 |
> |
self.files = self.AllInputFiles |
| 282 |
|
|
| 283 |
|
## get max number of events |
| 284 |
< |
common.logger.debug(10,"number of events for primary fileblocks %i"%self.pubdata.getMaxEvents()) |
| 284 |
> |
#common.logger.debug(10,"number of events for primary fileblocks %i"%self.pubdata.getMaxEvents()) |
| 285 |
|
self.maxEvents=self.pubdata.getMaxEvents() ## self.maxEvents used in Creator.py |
| 286 |
|
common.logger.message("\nThe number of available events is %s"%self.maxEvents) |
| 287 |
|
|
| 212 |
– |
## get fileblocks corresponding to the required data |
| 213 |
– |
fb=self.pubdata.getFileBlocks() |
| 214 |
– |
common.logger.debug(5,"fileblocks are %s"%fb) |
| 215 |
– |
|
| 288 |
|
## Contact the DLS and build a list of sites hosting the fileblocks |
| 289 |
|
try: |
| 290 |
< |
dataloc=DataLocation.DataLocation(self.pubdata.getFileBlocks(),cfg_params) |
| 290 |
> |
dataloc=DataLocation_EDM.DataLocation_EDM(filesbyblock.keys(),cfg_params) |
| 291 |
|
dataloc.fetchDLSInfo() |
| 292 |
< |
except DataLocation.DataLocationError , ex: |
| 292 |
> |
except DataLocation_EDM.DataLocationError , ex: |
| 293 |
|
msg = 'ERROR ***: failed Data Location in DLS \n %s '%ex.getErrorMessage() |
| 294 |
|
raise CrabException(msg) |
| 295 |
|
|
| 303 |
|
if len(sites)==0: |
| 304 |
|
msg = 'No sites hosting all the needed data! Exiting... ' |
| 305 |
|
raise CrabException(msg) |
| 306 |
< |
common.logger.message("List of Sites hosting the data : "+str(sites)) |
| 306 |
> |
|
| 307 |
> |
common.logger.message("List of Sites ("+str(len(sites))+") hosting the data : "+str(sites)) |
| 308 |
|
common.logger.debug(6, "List of Sites: "+str(sites)) |
| 309 |
|
common.analisys_common_info['sites']=sites ## used in SchedulerEdg.py in createSchScript |
| 310 |
+ |
self.setParam_('TargetCE', ','.join(sites)) |
| 311 |
+ |
return |
| 312 |
+ |
|
| 313 |
+ |
def jobSplittingPerFiles(self): |
| 314 |
+ |
""" |
| 315 |
+ |
Perform job splitting based on number of files to be accessed per job |
| 316 |
+ |
""" |
| 317 |
+ |
common.logger.debug(5,'Splitting per input files') |
| 318 |
+ |
common.logger.message('Required '+str(self.total_number_of_events)+' events in total ') |
| 319 |
+ |
common.logger.message('Available '+str(self.maxEvents)+' events in total ') |
| 320 |
+ |
common.logger.message('Required '+str(self.filesPerJob)+' files per job ') |
| 321 |
+ |
common.logger.message('Required '+str(self.theNumberOfJobs)+' jobs in total ') |
| 322 |
+ |
common.logger.message('Required '+str(self.eventsPerJob)+' events per job') |
| 323 |
+ |
|
| 324 |
+ |
## if asked to process all events, do it |
| 325 |
+ |
if self.total_number_of_events == -1: |
| 326 |
+ |
self.total_number_of_events=self.maxEvents |
| 327 |
+ |
else: |
| 328 |
+ |
if self.total_number_of_events>self.maxEvents: |
| 329 |
+ |
common.logger.message("Asked "+str(self.total_number_of_events)+" but only "+str(self.maxEvents)+" available.") |
| 330 |
+ |
self.total_number_of_events=self.maxEvents |
| 331 |
+ |
pass |
| 332 |
+ |
|
| 333 |
+ |
## TODO: SL need to have (from DBS) a detailed list of how many events per each file |
| 334 |
+ |
n_tot_files = (len(self.files[0])) |
| 335 |
+ |
## SL: this is wrong if the files have different number of events |
| 336 |
+ |
evPerFile = int(self.maxEvents)/n_tot_files |
| 337 |
+ |
|
| 338 |
+ |
common.logger.debug(5,'Events per File '+str(evPerFile)) |
| 339 |
+ |
|
| 340 |
+ |
## compute job splitting parameters: filesPerJob, eventsPerJob and theNumberOfJobs |
| 341 |
+ |
if self.selectFilesPerJob: |
| 342 |
+ |
## user define files per event. |
| 343 |
+ |
filesPerJob = self.filesPerJob |
| 344 |
+ |
eventsPerJob = filesPerJob*evPerFile |
| 345 |
+ |
theNumberOfJobs = int(self.total_number_of_events*1./eventsPerJob) |
| 346 |
+ |
check = int(self.total_number_of_events) - (theNumberOfJobs*eventsPerJob) |
| 347 |
+ |
if check > 0: |
| 348 |
+ |
theNumberOfJobs +=1 |
| 349 |
+ |
filesLastJob = int(check*1./evPerFile+0.5) |
| 350 |
+ |
common.logger.message('Warning: last job will be created with '+str(check)+' files') |
| 351 |
+ |
else: |
| 352 |
+ |
filesLastJob = filesPerJob |
| 353 |
+ |
|
| 354 |
+ |
elif self.selectNumberOfJobs: |
| 355 |
+ |
## User select the number of jobs: last might be bigger to match request of events |
| 356 |
+ |
theNumberOfJobs = self.theNumberOfJobs |
| 357 |
+ |
|
| 358 |
+ |
eventsPerJob = self.total_number_of_events/theNumberOfJobs |
| 359 |
+ |
filesPerJob = int(eventsPerJob/evPerFile) |
| 360 |
+ |
if (filesPerJob==0) : filesPerJob=1 |
| 361 |
+ |
check = int(self.total_number_of_events) - (int(theNumberOfJobs)*filesPerJob*evPerFile) |
| 362 |
+ |
if not check == 0: |
| 363 |
+ |
if check<0: |
| 364 |
+ |
missingFiles = int(check/evPerFile) |
| 365 |
+ |
additionalJobs = int(missingFiles/filesPerJob) |
| 366 |
+ |
#print missingFiles, additionalJobs |
| 367 |
+ |
theNumberOfJobs+=additionalJobs |
| 368 |
+ |
common.logger.message('Warning: will create only '+str(theNumberOfJobs)+' jobs') |
| 369 |
+ |
check = int(self.total_number_of_events) - (int(theNumberOfJobs)*filesPerJob*evPerFile) |
| 370 |
+ |
|
| 371 |
+ |
if check >0 : |
| 372 |
+ |
filesLastJob = filesPerJob+int(check*1./evPerFile+0.5) |
| 373 |
+ |
common.logger.message('Warning: last job will be created with '+str(filesLastJob*evPerFile)+' events') |
| 374 |
+ |
else: |
| 375 |
+ |
filesLastJob = filesPerJob |
| 376 |
+ |
else: |
| 377 |
+ |
filesLastJob = filesPerJob |
| 378 |
+ |
elif self.selectEventsPerJob: |
| 379 |
+ |
# SL case if asked events per job |
| 380 |
+ |
## estimate the number of files per job to match the user requirement |
| 381 |
+ |
filesPerJob = int(float(self.eventsPerJob)/float(evPerFile)) |
| 382 |
+ |
if filesPerJob==0: filesPerJob=1 |
| 383 |
+ |
common.logger.debug(5,"filesPerJob "+str(filesPerJob)) |
| 384 |
+ |
if (filesPerJob==0): filesPerJob=1 |
| 385 |
+ |
eventsPerJob=filesPerJob*evPerFile |
| 386 |
+ |
theNumberOfJobs = int(self.total_number_of_events)/int(eventsPerJob) |
| 387 |
+ |
check = int(self.total_number_of_events) - (int(theNumberOfJobs)*eventsPerJob) |
| 388 |
+ |
if not check == 0: |
| 389 |
+ |
missingFiles = int(check/evPerFile) |
| 390 |
+ |
additionalJobs = int(missingFiles/filesPerJob) |
| 391 |
+ |
if ( additionalJobs>0) : theNumberOfJobs+=additionalJobs |
| 392 |
+ |
check = int(self.total_number_of_events) - (int(theNumberOfJobs)*eventsPerJob) |
| 393 |
+ |
if not check == 0: |
| 394 |
+ |
if (check <0 ): |
| 395 |
+ |
filesLastJob = filesPerJob+int(check*1./evPerFile-0.5) |
| 396 |
+ |
else: |
| 397 |
+ |
theNumberOfJobs+=1 |
| 398 |
+ |
filesLastJob = int(check*1./evPerFile+0.5) |
| 399 |
+ |
|
| 400 |
+ |
common.logger.message('Warning: last job will be created with '+str(filesLastJob*evPerFile)+' events') |
| 401 |
+ |
else: |
| 402 |
+ |
filesLastJob = filesPerJob |
| 403 |
+ |
else: |
| 404 |
+ |
filesLastJob = filesPerJob |
| 405 |
+ |
|
| 406 |
+ |
self.total_number_of_jobs = theNumberOfJobs |
| 407 |
+ |
|
| 408 |
+ |
totalEventsToBeUsed=theNumberOfJobs*filesPerJob*evPerFile |
| 409 |
+ |
if not check == 0: |
| 410 |
+ |
# print (theNumberOfJobs-1)*filesPerJob*evPerFile,filesLastJob*evPerFile |
| 411 |
+ |
totalEventsToBeUsed=(theNumberOfJobs-1)*filesPerJob*evPerFile+filesLastJob*evPerFile |
| 412 |
+ |
|
| 413 |
+ |
common.logger.message(str(self.total_number_of_jobs)+' jobs will be created, each for '+str(filesPerJob*evPerFile)+' events, for a total of '+str(totalEventsToBeUsed)+' events') |
| 414 |
+ |
|
| 415 |
+ |
totalFilesToBeUsed=filesPerJob*(theNumberOfJobs-1)+filesLastJob |
| 416 |
+ |
|
| 417 |
+ |
## set job arguments (files) |
| 418 |
+ |
list_of_lists = [] |
| 419 |
+ |
lastFile=0 |
| 420 |
+ |
for i in range(0, int(totalFilesToBeUsed), filesPerJob)[:-1]: |
| 421 |
+ |
parString = "\\{" |
| 422 |
+ |
|
| 423 |
+ |
lastFile=i+filesPerJob |
| 424 |
+ |
params = self.files[0][i: lastFile] |
| 425 |
+ |
for i in range(len(params) - 1): |
| 426 |
+ |
parString += '\\\"' + params[i] + '\\\"\,' |
| 427 |
+ |
|
| 428 |
+ |
parString += '\\\"' + params[len(params) - 1] + '\\\"\\}' |
| 429 |
+ |
list_of_lists.append([parString]) |
| 430 |
+ |
pass |
| 431 |
+ |
|
| 432 |
+ |
## last job |
| 433 |
+ |
parString = "\\{" |
| 434 |
+ |
|
| 435 |
+ |
params = self.files[0][lastFile: lastFile+filesLastJob] |
| 436 |
+ |
for i in range(len(params) - 1): |
| 437 |
+ |
parString += '\\\"' + params[i] + '\\\"\,' |
| 438 |
+ |
|
| 439 |
+ |
parString += '\\\"' + params[len(params) - 1] + '\\\"\\}' |
| 440 |
+ |
list_of_lists.append([parString]) |
| 441 |
+ |
pass |
| 442 |
+ |
|
| 443 |
+ |
self.list_of_args = list_of_lists |
| 444 |
+ |
# print self.list_of_args[0] |
| 445 |
+ |
return |
| 446 |
+ |
|
| 447 |
+ |
def jobSplittingNoInput(self): |
| 448 |
+ |
""" |
| 449 |
+ |
Perform job splitting based on number of event per job |
| 450 |
+ |
""" |
| 451 |
+ |
common.logger.debug(5,'Splitting per events') |
| 452 |
+ |
common.logger.message('Required '+str(self.eventsPerJob)+' events per job ') |
| 453 |
+ |
common.logger.message('Required '+str(self.theNumberOfJobs)+' jobs in total ') |
| 454 |
+ |
common.logger.message('Required '+str(self.total_number_of_events)+' events in total ') |
| 455 |
+ |
|
| 456 |
+ |
if (self.total_number_of_events < 0): |
| 457 |
+ |
msg='Cannot split jobs per Events with "-1" as total number of events' |
| 458 |
+ |
raise CrabException(msg) |
| 459 |
+ |
|
| 460 |
+ |
if (self.selectEventsPerJob): |
| 461 |
+ |
self.total_number_of_jobs = int(self.total_number_of_events/self.eventsPerJob) |
| 462 |
+ |
elif (self.selectNumberOfJobs) : |
| 463 |
+ |
self.total_number_of_jobs = self.theNumberOfJobs |
| 464 |
+ |
self.eventsPerJob = int(self.total_number_of_events/self.total_number_of_jobs) |
| 465 |
+ |
|
| 466 |
+ |
common.logger.debug(5,'N jobs '+str(self.total_number_of_jobs)) |
| 467 |
+ |
|
| 468 |
+ |
# is there any remainder? |
| 469 |
+ |
check = int(self.total_number_of_events) - (int(self.total_number_of_jobs)*self.eventsPerJob) |
| 470 |
+ |
|
| 471 |
+ |
common.logger.debug(5,'Check '+str(check)) |
| 472 |
+ |
|
| 473 |
+ |
common.logger.message(str(self.total_number_of_jobs)+' jobs will be created, each for '+str(self.eventsPerJob)+' for a total of '+str(self.total_number_of_jobs*self.eventsPerJob)+' events') |
| 474 |
+ |
if check > 0: |
| 475 |
+ |
common.logger.message('Warning: asked '+str(self.total_number_of_events)+' but will do only '+str(int(self.total_number_of_jobs)*self.eventsPerJob)) |
| 476 |
+ |
|
| 477 |
+ |
|
| 478 |
+ |
# argument is seed number.$i |
| 479 |
+ |
self.list_of_args = [] |
| 480 |
+ |
for i in range(self.total_number_of_jobs): |
| 481 |
+ |
if (self.sourceSeed): |
| 482 |
+ |
self.list_of_args.append([(str(self.sourceSeed)+str(i))]) |
| 483 |
+ |
else: |
| 484 |
+ |
self.list_of_args.append([str(i)]) |
| 485 |
+ |
#print self.list_of_args |
| 486 |
+ |
|
| 487 |
|
return |
| 488 |
+ |
|
| 489 |
+ |
def split(self, jobParams): |
| 490 |
+ |
|
| 491 |
+ |
common.jobDB.load() |
| 492 |
+ |
#### Fabio |
| 493 |
+ |
njobs = self.total_number_of_jobs |
| 494 |
+ |
arglist = self.list_of_args |
| 495 |
+ |
# create the empty structure |
| 496 |
+ |
for i in range(njobs): |
| 497 |
+ |
jobParams.append("") |
| 498 |
|
|
| 499 |
+ |
for job in range(njobs): |
| 500 |
+ |
jobParams[job] = arglist[job] |
| 501 |
+ |
# print str(arglist[job]) |
| 502 |
+ |
# print jobParams[job] |
| 503 |
+ |
common.jobDB.setArguments(job, jobParams[job]) |
| 504 |
+ |
|
| 505 |
+ |
common.jobDB.save() |
| 506 |
+ |
return |
| 507 |
+ |
|
| 508 |
+ |
def getJobTypeArguments(self, nj, sched): |
| 509 |
+ |
result = '' |
| 510 |
+ |
for i in common.jobDB.arguments(nj): |
| 511 |
+ |
result=result+str(i)+" " |
| 512 |
+ |
return result |
| 513 |
+ |
|
| 514 |
+ |
def numberOfJobs(self): |
| 515 |
+ |
# Fabio |
| 516 |
+ |
return self.total_number_of_jobs |
| 517 |
+ |
|
| 518 |
|
def checkBlackList(self, allSites): |
| 519 |
|
if len(self.reCEBlackList)==0: return allSites |
| 520 |
|
sites = [] |
| 531 |
|
common.logger.debug(3,"No sites found after BlackList") |
| 532 |
|
return sites |
| 533 |
|
|
| 534 |
< |
def checkWhiteList(self, allsites): |
| 534 |
> |
def checkWhiteList(self, allSites): |
| 535 |
|
|
| 536 |
< |
if len(self.reCEWhiteList)==0: return pubDBUrls |
| 536 |
> |
if len(self.reCEWhiteList)==0: return allSites |
| 537 |
|
sites = [] |
| 538 |
< |
for site in allsites: |
| 260 |
< |
#print 'connecting to the URL ',url |
| 538 |
> |
for site in allSites: |
| 539 |
|
good=0 |
| 540 |
|
for re in self.reCEWhiteList: |
| 541 |
|
if re.search(site): |
| 605 |
|
if os.path.exists(lib): |
| 606 |
|
filesToBeTarred.append(libDir) |
| 607 |
|
|
| 608 |
+ |
## Now check if module dir is present |
| 609 |
+ |
moduleDir = 'module' |
| 610 |
+ |
if os.path.isdir(swArea+'/'+moduleDir): |
| 611 |
+ |
filesToBeTarred.append(moduleDir) |
| 612 |
+ |
|
| 613 |
|
## Now check if the Data dir is present |
| 614 |
|
dataDir = 'src/Data/' |
| 615 |
|
if os.path.isdir(swArea+'/'+dataDir): |
| 637 |
|
the execution environment for the job 'nj'. |
| 638 |
|
""" |
| 639 |
|
# Prepare JobType-independent part |
| 640 |
< |
txt = self.wsSetupCMSEnvironment_() |
| 640 |
> |
txt = '' |
| 641 |
> |
|
| 642 |
> |
## OLI_Daniele at this level middleware already known |
| 643 |
> |
|
| 644 |
> |
txt += 'if [ $middleware == LCG ]; then \n' |
| 645 |
> |
txt += self.wsSetupCMSLCGEnvironment_() |
| 646 |
> |
txt += 'elif [ $middleware == OSG ]; then\n' |
| 647 |
> |
txt += ' time=`date -u +"%s"`\n' |
| 648 |
> |
txt += ' WORKING_DIR=$OSG_WN_TMP/cms_$time\n' |
| 649 |
> |
txt += ' echo "Creating working directory: $WORKING_DIR"\n' |
| 650 |
> |
txt += ' /bin/mkdir -p $WORKING_DIR\n' |
| 651 |
> |
txt += ' if [ ! -d $WORKING_DIR ] ;then\n' |
| 652 |
> |
txt += ' echo "SET_CMS_ENV 10016 ==> OSG $WORKING_DIR could not be created on WN `hostname`"\n' |
| 653 |
> |
txt += ' echo "JOB_EXIT_STATUS = 10016"\n' |
| 654 |
> |
txt += ' echo "JobExitCode=10016" | tee -a $RUNTIME_AREA/$repo\n' |
| 655 |
> |
txt += ' dumpStatus $RUNTIME_AREA/$repo\n' |
| 656 |
> |
txt += ' rm -f $RUNTIME_AREA/$repo \n' |
| 657 |
> |
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n' |
| 658 |
> |
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n' |
| 659 |
> |
txt += ' exit 1\n' |
| 660 |
> |
txt += ' fi\n' |
| 661 |
> |
txt += '\n' |
| 662 |
> |
txt += ' echo "Change to working directory: $WORKING_DIR"\n' |
| 663 |
> |
txt += ' cd $WORKING_DIR\n' |
| 664 |
> |
txt += self.wsSetupCMSOSGEnvironment_() |
| 665 |
> |
txt += 'fi\n' |
| 666 |
|
|
| 667 |
|
# Prepare JobType-specific part |
| 668 |
|
scram = self.scram.commandName() |
| 671 |
|
txt += scram+' project CMSSW '+self.version+'\n' |
| 672 |
|
txt += 'status=$?\n' |
| 673 |
|
txt += 'if [ $status != 0 ] ; then\n' |
| 674 |
< |
txt += ' echo "SET_EXE_ENV 1 ==>ERROR CMSSW '+self.version+' not found on `hostname`" \n' |
| 675 |
< |
txt += ' echo "JOB_EXIT_STATUS = 5"\n' |
| 676 |
< |
txt += ' echo "SanityCheckCode = 5" | tee -a $RUNTIME_AREA/$repo\n' |
| 674 |
> |
txt += ' echo "SET_EXE_ENV 10034 ==>ERROR CMSSW '+self.version+' not found on `hostname`" \n' |
| 675 |
> |
txt += ' echo "JOB_EXIT_STATUS = 10034"\n' |
| 676 |
> |
txt += ' echo "JobExitCode=10034" | tee -a $RUNTIME_AREA/$repo\n' |
| 677 |
|
txt += ' dumpStatus $RUNTIME_AREA/$repo\n' |
| 678 |
< |
txt += ' exit 5 \n' |
| 678 |
> |
txt += ' rm -f $RUNTIME_AREA/$repo \n' |
| 679 |
> |
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n' |
| 680 |
> |
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n' |
| 681 |
> |
## OLI_Daniele |
| 682 |
> |
txt += ' if [ $middleware == OSG ]; then \n' |
| 683 |
> |
txt += ' echo "Remove working directory: $WORKING_DIR"\n' |
| 684 |
> |
txt += ' cd $RUNTIME_AREA\n' |
| 685 |
> |
txt += ' /bin/rm -rf $WORKING_DIR\n' |
| 686 |
> |
txt += ' if [ -d $WORKING_DIR ] ;then\n' |
| 687 |
> |
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' |
| 688 |
> |
txt += ' echo "JOB_EXIT_STATUS = 10018"\n' |
| 689 |
> |
txt += ' echo "JobExitCode=10018" | tee -a $RUNTIME_AREA/$repo\n' |
| 690 |
> |
txt += ' dumpStatus $RUNTIME_AREA/$repo\n' |
| 691 |
> |
txt += ' rm -f $RUNTIME_AREA/$repo \n' |
| 692 |
> |
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n' |
| 693 |
> |
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n' |
| 694 |
> |
txt += ' fi\n' |
| 695 |
> |
txt += ' fi \n' |
| 696 |
> |
txt += ' exit 1 \n' |
| 697 |
|
txt += 'fi \n' |
| 698 |
|
txt += 'echo "CMSSW_VERSION = '+self.version+'"\n' |
| 699 |
|
txt += 'cd '+self.version+'\n' |
| 702 |
|
|
| 703 |
|
# Handle the arguments: |
| 704 |
|
txt += "\n" |
| 705 |
< |
txt += "## ARGUMNETS: $1 Job Number\n" |
| 380 |
< |
# txt += "## ARGUMNETS: $2 First Event for this job\n" |
| 381 |
< |
# txt += "## ARGUMNETS: $3 Max Event for this job\n" |
| 705 |
> |
txt += "## number of arguments (first argument always jobnumber)\n" |
| 706 |
|
txt += "\n" |
| 707 |
|
txt += "narg=$#\n" |
| 708 |
< |
txt += "if [ $narg -lt 1 ]\n" |
| 708 |
> |
txt += "if [ $narg -lt 2 ]\n" |
| 709 |
|
txt += "then\n" |
| 710 |
|
txt += " echo 'SET_EXE_ENV 1 ==> ERROR Too few arguments' +$narg+ \n" |
| 711 |
< |
txt += ' echo "JOB_EXIT_STATUS = 1"\n' |
| 712 |
< |
txt += ' echo "SanityCheckCode = 1" | tee -a $RUNTIME_AREA/$repo\n' |
| 711 |
> |
txt += ' echo "JOB_EXIT_STATUS = 50113"\n' |
| 712 |
> |
txt += ' echo "JobExitCode=50113" | tee -a $RUNTIME_AREA/$repo\n' |
| 713 |
|
txt += ' dumpStatus $RUNTIME_AREA/$repo\n' |
| 714 |
+ |
txt += ' rm -f $RUNTIME_AREA/$repo \n' |
| 715 |
+ |
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n' |
| 716 |
+ |
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n' |
| 717 |
+ |
## OLI_Daniele |
| 718 |
+ |
txt += ' if [ $middleware == OSG ]; then \n' |
| 719 |
+ |
txt += ' echo "Remove working directory: $WORKING_DIR"\n' |
| 720 |
+ |
txt += ' cd $RUNTIME_AREA\n' |
| 721 |
+ |
txt += ' /bin/rm -rf $WORKING_DIR\n' |
| 722 |
+ |
txt += ' if [ -d $WORKING_DIR ] ;then\n' |
| 723 |
+ |
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' |
| 724 |
+ |
txt += ' echo "JOB_EXIT_STATUS = 50114"\n' |
| 725 |
+ |
txt += ' echo "JobExitCode=50114" | tee -a $RUNTIME_AREA/$repo\n' |
| 726 |
+ |
txt += ' dumpStatus $RUNTIME_AREA/$repo\n' |
| 727 |
+ |
txt += ' rm -f $RUNTIME_AREA/$repo \n' |
| 728 |
+ |
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n' |
| 729 |
+ |
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n' |
| 730 |
+ |
txt += ' fi\n' |
| 731 |
+ |
txt += ' fi \n' |
| 732 |
|
txt += " exit 1\n" |
| 733 |
|
txt += "fi\n" |
| 734 |
|
txt += "\n" |
| 393 |
– |
txt += "NJob=$1\n" |
| 394 |
– |
# txt += "FirstEvent=$2\n" |
| 395 |
– |
# txt += "MaxEvents=$3\n" |
| 735 |
|
|
| 736 |
|
# Prepare job-specific part |
| 737 |
|
job = common.job_list[nj] |
| 738 |
|
pset = os.path.basename(job.configFilename()) |
| 739 |
|
txt += '\n' |
| 740 |
< |
txt += 'cp $RUNTIME_AREA/'+pset+' pset.cfg\n' |
| 741 |
< |
# txt += 'if [ -e $RUNTIME_AREA/orcarc_$CE ] ; then\n' |
| 742 |
< |
# txt += ' cat $RUNTIME_AREA/orcarc_$CE .orcarc >> .orcarc_tmp\n' |
| 743 |
< |
# txt += ' mv .orcarc_tmp .orcarc\n' |
| 744 |
< |
# txt += 'fi\n' |
| 745 |
< |
# txt += 'if [ -e $RUNTIME_AREA/init_$CE.sh ] ; then\n' |
| 746 |
< |
# txt += ' cp $RUNTIME_AREA/init_$CE.sh init.sh\n' |
| 747 |
< |
# txt += 'fi\n' |
| 740 |
> |
if (self.datasetPath): # standard job |
| 741 |
> |
txt += 'InputFiles=$2\n' |
| 742 |
> |
txt += 'echo "Inputfiles:<$InputFiles>"\n' |
| 743 |
> |
txt += 'sed "s#{\'INPUT\'}#$InputFiles#" $RUNTIME_AREA/'+pset+' > pset.cfg\n' |
| 744 |
> |
else: # pythia like job |
| 745 |
> |
if (self.sourceSeed): |
| 746 |
> |
txt += 'Seed=$2\n' |
| 747 |
> |
txt += 'echo "Seed: <$Seed>"\n' |
| 748 |
> |
txt += 'sed "s#INPUT#$Seed#" $RUNTIME_AREA/'+pset+' > pset.cfg\n' |
| 749 |
> |
else: |
| 750 |
> |
txt += '# Copy untouched pset\n' |
| 751 |
> |
txt += 'cp $RUNTIME_AREA/'+pset+' pset.cfg\n' |
| 752 |
> |
|
| 753 |
|
|
| 754 |
|
if len(self.additional_inbox_files) > 0: |
| 755 |
|
for file in self.additional_inbox_files: |
| 759 |
|
txt += 'fi\n' |
| 760 |
|
pass |
| 761 |
|
|
| 418 |
– |
# txt += '\n' |
| 419 |
– |
# txt += 'chmod +x ./init.sh\n' |
| 420 |
– |
# txt += './init.sh\n' |
| 421 |
– |
# txt += 'exitStatus=$?\n' |
| 422 |
– |
# txt += 'if [ $exitStatus != 0 ] ; then\n' |
| 423 |
– |
# txt += ' echo "SET_EXE_ENV 1 ==> ERROR StageIn init script failed"\n' |
| 424 |
– |
# txt += ' echo "JOB_EXIT_STATUS = $exitStatus" \n' |
| 425 |
– |
# txt += ' echo "SanityCheckCode = $exitStatus" | tee -a $RUNTIME_AREA/$repo\n' |
| 426 |
– |
# txt += ' dumpStatus $RUNTIME_AREA/$repo\n' |
| 427 |
– |
# txt += ' exit $exitStatus\n' |
| 428 |
– |
# txt += 'fi\n' |
| 429 |
– |
# txt += "echo 'SET_EXE_ENV 0 ==> job setup ok'\n" |
| 762 |
|
txt += 'echo "### END JOB SETUP ENVIRONMENT ###"\n\n' |
| 763 |
|
|
| 432 |
– |
# txt += 'echo "FirstEvent=$FirstEvent" >> .orcarc\n' |
| 433 |
– |
# txt += 'echo "MaxEvents=$MaxEvents" >> .orcarc\n' |
| 434 |
– |
# if self.ML: |
| 435 |
– |
# txt += 'echo "MonalisaJobId=$NJob" >> .orcarc\n' |
| 436 |
– |
|
| 764 |
|
txt += '\n' |
| 765 |
|
txt += 'echo "***** cat pset.cfg *********"\n' |
| 766 |
|
txt += 'cat pset.cfg\n' |
| 767 |
|
txt += 'echo "****** end pset.cfg ********"\n' |
| 768 |
+ |
txt += '\n' |
| 769 |
+ |
# txt += 'echo "***** cat pset1.cfg *********"\n' |
| 770 |
+ |
# txt += 'cat pset1.cfg\n' |
| 771 |
+ |
# txt += 'echo "****** end pset1.cfg ********"\n' |
| 772 |
+ |
return txt |
| 773 |
+ |
|
| 774 |
+ |
def wsBuildExe(self, nj): |
| 775 |
+ |
""" |
| 776 |
+ |
Put in the script the commands to build an executable |
| 777 |
+ |
or a library. |
| 778 |
+ |
""" |
| 779 |
+ |
|
| 780 |
+ |
txt = "" |
| 781 |
+ |
|
| 782 |
+ |
if os.path.isfile(self.tgzNameWithPath): |
| 783 |
+ |
txt += 'echo "tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+'"\n' |
| 784 |
+ |
txt += 'tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+'\n' |
| 785 |
+ |
txt += 'untar_status=$? \n' |
| 786 |
+ |
txt += 'if [ $untar_status -ne 0 ]; then \n' |
| 787 |
+ |
txt += ' echo "SET_EXE 1 ==> ERROR Untarring .tgz file failed"\n' |
| 788 |
+ |
txt += ' echo "JOB_EXIT_STATUS = $untar_status" \n' |
| 789 |
+ |
txt += ' echo "JobExitCode=$untar_status" | tee -a $RUNTIME_AREA/$repo\n' |
| 790 |
+ |
txt += ' if [ $middleware == OSG ]; then \n' |
| 791 |
+ |
txt += ' echo "Remove working directory: $WORKING_DIR"\n' |
| 792 |
+ |
txt += ' cd $RUNTIME_AREA\n' |
| 793 |
+ |
txt += ' /bin/rm -rf $WORKING_DIR\n' |
| 794 |
+ |
txt += ' if [ -d $WORKING_DIR ] ;then\n' |
| 795 |
+ |
txt += ' echo "SET_EXE 50999 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after Untarring .tgz file failed"\n' |
| 796 |
+ |
txt += ' echo "JOB_EXIT_STATUS = 50999"\n' |
| 797 |
+ |
txt += ' echo "JobExitCode=50999" | tee -a $RUNTIME_AREA/$repo\n' |
| 798 |
+ |
txt += ' dumpStatus $RUNTIME_AREA/$repo\n' |
| 799 |
+ |
txt += ' rm -f $RUNTIME_AREA/$repo \n' |
| 800 |
+ |
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n' |
| 801 |
+ |
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n' |
| 802 |
+ |
txt += ' fi\n' |
| 803 |
+ |
txt += ' fi \n' |
| 804 |
+ |
txt += ' \n' |
| 805 |
+ |
txt += ' exit 1 \n' |
| 806 |
+ |
txt += 'else \n' |
| 807 |
+ |
txt += ' echo "Successful untar" \n' |
| 808 |
+ |
txt += 'fi \n' |
| 809 |
+ |
pass |
| 810 |
+ |
|
| 811 |
|
return txt |
| 812 |
|
|
| 813 |
|
def modifySteeringCards(self, nj): |
| 820 |
|
return self.executable |
| 821 |
|
|
| 822 |
|
def executableArgs(self): |
| 823 |
< |
return "-p pset.cfg" |
| 823 |
> |
return " -p pset.cfg" |
| 824 |
|
|
| 825 |
|
def inputSandbox(self, nj): |
| 826 |
|
""" |
| 835 |
|
## config |
| 836 |
|
inp_box.append(common.job_list[nj].configFilename()) |
| 837 |
|
## additional input files |
| 838 |
< |
for file in self.additional_inbox_files: |
| 839 |
< |
inp_box.append(common.work_space.cwdDir()+file) |
| 470 |
< |
#print "sono inputSandbox, inp_box = ", inp_box |
| 838 |
> |
#for file in self.additional_inbox_files: |
| 839 |
> |
# inp_box.append(common.work_space.cwdDir()+file) |
| 840 |
|
return inp_box |
| 841 |
|
|
| 842 |
|
def outputSandbox(self, nj): |
| 859 |
|
""" |
| 860 |
|
Make initial modifications of the user's steering card file. |
| 861 |
|
""" |
| 493 |
– |
infile = open(self.pset,'r') |
| 494 |
– |
|
| 495 |
– |
outfile = open(common.work_space.jobDir()+self.name()+'.cfg', 'w') |
| 496 |
– |
|
| 497 |
– |
outfile.write('\n\n##### The following cards have been created by CRAB: DO NOT TOUCH #####\n') |
| 498 |
– |
|
| 499 |
– |
outfile.write('InputCollections=/System/'+self.owner+'/'+self.dataset+'/'+self.dataset+'\n') |
| 500 |
– |
|
| 501 |
– |
infile.close() |
| 502 |
– |
outfile.close() |
| 862 |
|
return |
| 863 |
|
|
| 864 |
|
def wsRenameOutput(self, nj): |
| 867 |
|
""" |
| 868 |
|
|
| 869 |
|
txt = '\n' |
| 870 |
+ |
txt += '# directory content\n' |
| 871 |
+ |
txt += 'ls \n' |
| 872 |
|
file_list = '' |
| 873 |
|
for fileWithSuffix in self.output_file: |
| 874 |
|
output_file_num = self.numberFile_(fileWithSuffix, '$NJob') |
| 875 |
< |
file_list=file_list+output_file_num+',' |
| 515 |
< |
txt += '\n' |
| 516 |
< |
txt += 'ls \n' |
| 875 |
> |
file_list=file_list+output_file_num+' ' |
| 876 |
|
txt += '\n' |
| 877 |
+ |
txt += '# check output file\n' |
| 878 |
|
txt += 'ls '+fileWithSuffix+'\n' |
| 879 |
< |
txt += 'exe_result=$?\n' |
| 880 |
< |
txt += 'if [ $exe_result -ne 0 ] ; then\n' |
| 881 |
< |
txt += ' echo "ERROR: No output file to manage"\n' |
| 882 |
< |
txt += ' echo "JOB_EXIT_STATUS = $exe_result"\n' |
| 883 |
< |
txt += ' echo "SanityCheckCode = $exe_result" | tee -a $RUNTIME_AREA/$repo\n' |
| 884 |
< |
txt += ' dumpStatus $RUNTIME_AREA/$repo\n' |
| 885 |
< |
txt += ' exit $exe_result \n' |
| 879 |
> |
txt += 'ls_result=$?\n' |
| 880 |
> |
#txt += 'exe_result=$?\n' |
| 881 |
> |
txt += 'if [ $ls_result -ne 0 ] ; then\n' |
| 882 |
> |
txt += ' echo "ERROR: Problem with output file"\n' |
| 883 |
> |
#txt += ' echo "JOB_EXIT_STATUS = $exe_result"\n' |
| 884 |
> |
#txt += ' echo "JobExitCode=60302" | tee -a $RUNTIME_AREA/$repo\n' |
| 885 |
> |
#txt += ' dumpStatus $RUNTIME_AREA/$repo\n' |
| 886 |
> |
### OLI_DANIELE |
| 887 |
> |
if common.scheduler.boss_scheduler_name == 'condor_g': |
| 888 |
> |
txt += ' if [ $middleware == OSG ]; then \n' |
| 889 |
> |
txt += ' echo "prepare dummy output file"\n' |
| 890 |
> |
txt += ' echo "Processing of job output failed" > $RUNTIME_AREA/'+output_file_num+'\n' |
| 891 |
> |
txt += ' fi \n' |
| 892 |
|
txt += 'else\n' |
| 893 |
|
txt += ' cp '+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n' |
| 894 |
|
txt += 'fi\n' |
| 529 |
– |
txt += 'cd $RUNTIME_AREA\n' |
| 530 |
– |
|
| 531 |
– |
pass |
| 895 |
|
|
| 896 |
+ |
txt += 'cd $RUNTIME_AREA\n' |
| 897 |
|
file_list=file_list[:-1] |
| 898 |
< |
txt += 'file_list='+file_list+'\n' |
| 898 |
> |
txt += 'file_list="'+file_list+'"\n' |
| 899 |
> |
txt += 'cd $RUNTIME_AREA\n' |
| 900 |
> |
### OLI_DANIELE |
| 901 |
> |
txt += 'if [ $middleware == OSG ]; then\n' |
| 902 |
> |
txt += ' cd $RUNTIME_AREA\n' |
| 903 |
> |
txt += ' echo "Remove working directory: $WORKING_DIR"\n' |
| 904 |
> |
txt += ' /bin/rm -rf $WORKING_DIR\n' |
| 905 |
> |
txt += ' if [ -d $WORKING_DIR ] ;then\n' |
| 906 |
> |
txt += ' echo "SET_EXE 60999 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after cleanup of WN"\n' |
| 907 |
> |
txt += ' echo "JOB_EXIT_STATUS = 60999"\n' |
| 908 |
> |
txt += ' echo "JobExitCode=60999" | tee -a $RUNTIME_AREA/$repo\n' |
| 909 |
> |
txt += ' dumpStatus $RUNTIME_AREA/$repo\n' |
| 910 |
> |
txt += ' rm -f $RUNTIME_AREA/$repo \n' |
| 911 |
> |
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n' |
| 912 |
> |
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n' |
| 913 |
> |
txt += ' fi\n' |
| 914 |
> |
txt += 'fi\n' |
| 915 |
> |
txt += '\n' |
| 916 |
|
return txt |
| 917 |
|
|
| 918 |
|
def numberFile_(self, file, txt): |
| 940 |
|
return job requirements to add to jdl files |
| 941 |
|
""" |
| 942 |
|
req = '' |
| 943 |
+ |
if common.analisys_common_info['sw_version']: |
| 944 |
+ |
req='Member("VO-cms-' + \ |
| 945 |
+ |
common.analisys_common_info['sw_version'] + \ |
| 946 |
+ |
'", other.GlueHostApplicationSoftwareRunTimeEnvironment)' |
| 947 |
|
if common.analisys_common_info['sites']: |
| 563 |
– |
if common.analisys_common_info['sw_version']: |
| 564 |
– |
req='Member("VO-cms-' + \ |
| 565 |
– |
common.analisys_common_info['sw_version'] + \ |
| 566 |
– |
'", other.GlueHostApplicationSoftwareRunTimeEnvironment)' |
| 948 |
|
if len(common.analisys_common_info['sites'])>0: |
| 949 |
|
req = req + ' && (' |
| 950 |
|
for i in range(len(common.analisys_common_info['sites'])): |
| 955 |
|
req = req + ')' |
| 956 |
|
#print "req = ", req |
| 957 |
|
return req |
| 958 |
+ |
|
| 959 |
+ |
def configFilename(self): |
| 960 |
+ |
""" return the config filename """ |
| 961 |
+ |
return self.name()+'.cfg' |
| 962 |
+ |
|
| 963 |
+ |
### OLI_DANIELE |
| 964 |
+ |
def wsSetupCMSOSGEnvironment_(self): |
| 965 |
+ |
""" |
| 966 |
+ |
Returns part of a job script which is prepares |
| 967 |
+ |
the execution environment and which is common for all CMS jobs. |
| 968 |
+ |
""" |
| 969 |
+ |
txt = '\n' |
| 970 |
+ |
txt += ' echo "### SETUP CMS OSG ENVIRONMENT ###"\n' |
| 971 |
+ |
txt += ' if [ -f $GRID3_APP_DIR/cmssoft/cmsset_default.sh ] ;then\n' |
| 972 |
+ |
txt += ' # Use $GRID3_APP_DIR/cmssoft/cmsset_default.sh to setup cms software\n' |
| 973 |
+ |
txt += ' source $GRID3_APP_DIR/cmssoft/cmsset_default.sh '+self.version+'\n' |
| 974 |
+ |
txt += ' elif [ -f $OSG_APP/cmssoft/cmsset_default.sh ] ;then\n' |
| 975 |
+ |
txt += ' # Use $OSG_APP/cmssoft/cmsset_default.sh to setup cms software\n' |
| 976 |
+ |
txt += ' source $OSG_APP/cmssoft/cmsset_default.sh '+self.version+'\n' |
| 977 |
+ |
txt += ' else\n' |
| 978 |
+ |
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' |
| 979 |
+ |
txt += ' echo "JOB_EXIT_STATUS = 10020"\n' |
| 980 |
+ |
txt += ' echo "JobExitCode=10020" | tee -a $RUNTIME_AREA/$repo\n' |
| 981 |
+ |
txt += ' dumpStatus $RUNTIME_AREA/$repo\n' |
| 982 |
+ |
txt += ' rm -f $RUNTIME_AREA/$repo \n' |
| 983 |
+ |
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n' |
| 984 |
+ |
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n' |
| 985 |
+ |
txt += ' exit 1\n' |
| 986 |
+ |
txt += '\n' |
| 987 |
+ |
txt += ' echo "Remove working directory: $WORKING_DIR"\n' |
| 988 |
+ |
txt += ' cd $RUNTIME_AREA\n' |
| 989 |
+ |
txt += ' /bin/rm -rf $WORKING_DIR\n' |
| 990 |
+ |
txt += ' if [ -d $WORKING_DIR ] ;then\n' |
| 991 |
+ |
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' |
| 992 |
+ |
txt += ' echo "JOB_EXIT_STATUS = 10017"\n' |
| 993 |
+ |
txt += ' echo "JobExitCode=10017" | tee -a $RUNTIME_AREA/$repo\n' |
| 994 |
+ |
txt += ' dumpStatus $RUNTIME_AREA/$repo\n' |
| 995 |
+ |
txt += ' rm -f $RUNTIME_AREA/$repo \n' |
| 996 |
+ |
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n' |
| 997 |
+ |
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n' |
| 998 |
+ |
txt += ' fi\n' |
| 999 |
+ |
txt += '\n' |
| 1000 |
+ |
txt += ' exit 1\n' |
| 1001 |
+ |
txt += ' fi\n' |
| 1002 |
+ |
txt += '\n' |
| 1003 |
+ |
txt += ' echo "SET_CMS_ENV 0 ==> setup cms environment ok"\n' |
| 1004 |
+ |
txt += ' echo " END SETUP CMS OSG ENVIRONMENT "\n' |
| 1005 |
+ |
|
| 1006 |
+ |
return txt |
| 1007 |
+ |
|
| 1008 |
+ |
### OLI_DANIELE |
| 1009 |
+ |
def wsSetupCMSLCGEnvironment_(self): |
| 1010 |
+ |
""" |
| 1011 |
+ |
Returns part of a job script which is prepares |
| 1012 |
+ |
the execution environment and which is common for all CMS jobs. |
| 1013 |
+ |
""" |
| 1014 |
+ |
txt = ' \n' |
| 1015 |
+ |
txt += ' echo " ### SETUP CMS LCG ENVIRONMENT ### "\n' |
| 1016 |
+ |
txt += ' if [ ! $VO_CMS_SW_DIR ] ;then\n' |
| 1017 |
+ |
txt += ' echo "SET_CMS_ENV 10031 ==> ERROR CMS software dir not found on WN `hostname`"\n' |
| 1018 |
+ |
txt += ' echo "JOB_EXIT_STATUS = 10031" \n' |
| 1019 |
+ |
txt += ' echo "JobExitCode=10031" | tee -a $RUNTIME_AREA/$repo\n' |
| 1020 |
+ |
txt += ' dumpStatus $RUNTIME_AREA/$repo\n' |
| 1021 |
+ |
txt += ' rm -f $RUNTIME_AREA/$repo \n' |
| 1022 |
+ |
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n' |
| 1023 |
+ |
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n' |
| 1024 |
+ |
txt += ' exit 1\n' |
| 1025 |
+ |
txt += ' else\n' |
| 1026 |
+ |
txt += ' echo "Sourcing environment... "\n' |
| 1027 |
+ |
txt += ' if [ ! -s $VO_CMS_SW_DIR/cmsset_default.sh ] ;then\n' |
| 1028 |
+ |
txt += ' echo "SET_CMS_ENV 10020 ==> ERROR cmsset_default.sh file not found into dir $VO_CMS_SW_DIR"\n' |
| 1029 |
+ |
txt += ' echo "JOB_EXIT_STATUS = 10020"\n' |
| 1030 |
+ |
txt += ' echo "JobExitCode=10020" | tee -a $RUNTIME_AREA/$repo\n' |
| 1031 |
+ |
txt += ' dumpStatus $RUNTIME_AREA/$repo\n' |
| 1032 |
+ |
txt += ' rm -f $RUNTIME_AREA/$repo \n' |
| 1033 |
+ |
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n' |
| 1034 |
+ |
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n' |
| 1035 |
+ |
txt += ' exit 1\n' |
| 1036 |
+ |
txt += ' fi\n' |
| 1037 |
+ |
txt += ' echo "sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n' |
| 1038 |
+ |
txt += ' source $VO_CMS_SW_DIR/cmsset_default.sh\n' |
| 1039 |
+ |
txt += ' result=$?\n' |
| 1040 |
+ |
txt += ' if [ $result -ne 0 ]; then\n' |
| 1041 |
+ |
txt += ' echo "SET_CMS_ENV 10032 ==> ERROR problem sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n' |
| 1042 |
+ |
txt += ' echo "JOB_EXIT_STATUS = 10032"\n' |
| 1043 |
+ |
txt += ' echo "JobExitCode=10032" | tee -a $RUNTIME_AREA/$repo\n' |
| 1044 |
+ |
txt += ' dumpStatus $RUNTIME_AREA/$repo\n' |
| 1045 |
+ |
txt += ' rm -f $RUNTIME_AREA/$repo \n' |
| 1046 |
+ |
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n' |
| 1047 |
+ |
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n' |
| 1048 |
+ |
txt += ' exit 1\n' |
| 1049 |
+ |
txt += ' fi\n' |
| 1050 |
+ |
txt += ' fi\n' |
| 1051 |
+ |
txt += ' \n' |
| 1052 |
+ |
txt += ' string=`cat /etc/redhat-release`\n' |
| 1053 |
+ |
txt += ' echo $string\n' |
| 1054 |
+ |
txt += ' if [[ $string = *alhalla* ]]; then\n' |
| 1055 |
+ |
txt += ' echo "SCRAM_ARCH= $SCRAM_ARCH"\n' |
| 1056 |
+ |
txt += ' elif [[ $string = *Enterprise* ]] || [[ $string = *cientific* ]]; then\n' |
| 1057 |
+ |
txt += ' export SCRAM_ARCH=slc3_ia32_gcc323\n' |
| 1058 |
+ |
txt += ' echo "SCRAM_ARCH= $SCRAM_ARCH"\n' |
| 1059 |
+ |
txt += ' else\n' |
| 1060 |
+ |
txt += ' echo "SET_CMS_ENV 10033 ==> ERROR OS unknown, LCG environment not initialized"\n' |
| 1061 |
+ |
txt += ' echo "JOB_EXIT_STATUS = 10033"\n' |
| 1062 |
+ |
txt += ' echo "JobExitCode=10033" | tee -a $RUNTIME_AREA/$repo\n' |
| 1063 |
+ |
txt += ' dumpStatus $RUNTIME_AREA/$repo\n' |
| 1064 |
+ |
txt += ' rm -f $RUNTIME_AREA/$repo \n' |
| 1065 |
+ |
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n' |
| 1066 |
+ |
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n' |
| 1067 |
+ |
txt += ' exit 1\n' |
| 1068 |
+ |
txt += ' fi\n' |
| 1069 |
+ |
txt += ' echo "SET_CMS_ENV 0 ==> setup cms environment ok"\n' |
| 1070 |
+ |
txt += ' echo "### END SETUP CMS LCG ENVIRONMENT ###"\n' |
| 1071 |
+ |
return txt |
| 1072 |
+ |
|
| 1073 |
+ |
def setParam_(self, param, value): |
| 1074 |
+ |
self._params[param] = value |
| 1075 |
+ |
|
| 1076 |
+ |
def getParams(self): |
| 1077 |
+ |
return self._params |
| 1078 |
+ |
|
| 1079 |
+ |
def setTaskid_(self): |
| 1080 |
+ |
self._taskId = self.cfg_params['taskId'] |
| 1081 |
+ |
|
| 1082 |
+ |
def getTaskid(self): |
| 1083 |
+ |
return self._taskId |
