| 1 |
from JobType import JobType
|
| 2 |
from crab_logger import Logger
|
| 3 |
from crab_exceptions import *
|
| 4 |
from crab_util import *
|
| 5 |
import common
|
| 6 |
import DataDiscovery
|
| 7 |
import DataLocation
|
| 8 |
import Scram
|
| 9 |
import TarBall
|
| 10 |
|
| 11 |
import os, string, re
|
| 12 |
|
| 13 |
###Fabio
|
| 14 |
import math
|
| 15 |
|
| 16 |
class Orca(JobType):
|
| 17 |
def __init__(self, cfg_params, ncjobs):
|
| 18 |
JobType.__init__(self, 'ORCA')
|
| 19 |
common.logger.debug(3,'ORCA::__init__')
|
| 20 |
|
| 21 |
self.analisys_common_info = {}
|
| 22 |
# Marco.
|
| 23 |
self._params = {}
|
| 24 |
self.cfg_params = cfg_params
|
| 25 |
|
| 26 |
# number of jobs requested to be created, limit ojb splitting
|
| 27 |
self.ncjobs = ncjobs
|
| 28 |
|
| 29 |
log = common.logger
|
| 30 |
|
| 31 |
self.scram = Scram.Scram(cfg_params)
|
| 32 |
scramArea = ''
|
| 33 |
self.additional_inbox_files = []
|
| 34 |
self.scriptExe = ''
|
| 35 |
|
| 36 |
self.version = self.scram.getSWVersion()
|
| 37 |
self.setParam_('application', self.version)
|
| 38 |
common.analisys_common_info['sw_version'] = self.version
|
| 39 |
common.analisys_common_info['copy_input_data'] = 0
|
| 40 |
|
| 41 |
self.total_number_of_jobs = 0
|
| 42 |
self.job_number_of_events = 0
|
| 43 |
|
| 44 |
### collect Data cards
|
| 45 |
try:
|
| 46 |
self.owner = cfg_params['ORCA.owner']
|
| 47 |
self.setParam_('owner', self.owner)
|
| 48 |
log.debug(6, "Orca::Orca(): owner = "+self.owner)
|
| 49 |
self.dataset = cfg_params['ORCA.dataset']
|
| 50 |
self.setParam_('dataset', self.dataset)
|
| 51 |
log.debug(6, "Orca::Orca(): dataset = "+self.dataset)
|
| 52 |
except KeyError:
|
| 53 |
msg = "Error: owner and/or dataset not defined "
|
| 54 |
raise CrabException(msg)
|
| 55 |
|
| 56 |
self.dataTiers = []
|
| 57 |
try:
|
| 58 |
tmpDataTiers = string.split(cfg_params['ORCA.data_tier'],',')
|
| 59 |
for tmp in tmpDataTiers:
|
| 60 |
tmp=string.strip(tmp)
|
| 61 |
self.dataTiers.append(tmp)
|
| 62 |
pass
|
| 63 |
pass
|
| 64 |
except KeyError:
|
| 65 |
pass
|
| 66 |
log.debug(6, "Orca::Orca(): dataTiers = "+str(self.dataTiers))
|
| 67 |
|
| 68 |
## now the application
|
| 69 |
try:
|
| 70 |
self.executable = cfg_params['ORCA.executable']
|
| 71 |
log.debug(6, "Orca::Orca(): executable = "+self.executable)
|
| 72 |
self.setParam_('exe', self.executable)
|
| 73 |
except KeyError:
|
| 74 |
msg = "Error: executable not defined "
|
| 75 |
raise CrabException(msg)
|
| 76 |
|
| 77 |
try:
|
| 78 |
self.orcarc_file = cfg_params['ORCA.orcarc_file']
|
| 79 |
log.debug(6, "Orca::Orca(): orcarc file = "+self.orcarc_file)
|
| 80 |
if (not os.path.exists(self.orcarc_file)):
|
| 81 |
raise CrabException("User defined .orcarc file "+self.orcarc_file+" does not exist")
|
| 82 |
except KeyError:
|
| 83 |
log.message("Using empty orcarc file")
|
| 84 |
self.orcarc_file = ''
|
| 85 |
|
| 86 |
# output files
|
| 87 |
try:
|
| 88 |
self.output_file = []
|
| 89 |
|
| 90 |
tmp = cfg_params['ORCA.output_file']
|
| 91 |
if tmp != '':
|
| 92 |
tmpOutFiles = string.split(cfg_params['ORCA.output_file'],',')
|
| 93 |
log.debug(7, 'Orca::Orca(): output files '+str(tmpOutFiles))
|
| 94 |
for tmp in tmpOutFiles:
|
| 95 |
tmp=string.strip(tmp)
|
| 96 |
self.output_file.append(tmp)
|
| 97 |
pass
|
| 98 |
|
| 99 |
else:
|
| 100 |
log.message("No output file defined: only stdout/err will be available")
|
| 101 |
pass
|
| 102 |
pass
|
| 103 |
except KeyError:
|
| 104 |
log.message("No output file defined: only stdout/err will be available")
|
| 105 |
pass
|
| 106 |
|
| 107 |
# script_exe file as additional file in inputSandbox
|
| 108 |
try:
|
| 109 |
self.scriptExe = cfg_params['ORCA.script_exe']
|
| 110 |
self.additional_inbox_files.append(self.scriptExe)
|
| 111 |
except KeyError:
|
| 112 |
pass
|
| 113 |
if self.scriptExe != '':
|
| 114 |
if os.path.isfile(self.scriptExe):
|
| 115 |
pass
|
| 116 |
else:
|
| 117 |
log.message("WARNING. file "+self.scriptExe+" not found")
|
| 118 |
sys.exit()
|
| 119 |
|
| 120 |
## additional input files
|
| 121 |
try:
|
| 122 |
tmpAddFiles = string.split(cfg_params['USER.additional_input_files'],',')
|
| 123 |
for tmp in tmpAddFiles:
|
| 124 |
tmp=string.strip(tmp)
|
| 125 |
self.additional_inbox_files.append(tmp)
|
| 126 |
pass
|
| 127 |
pass
|
| 128 |
except KeyError:
|
| 129 |
pass
|
| 130 |
|
| 131 |
try:
|
| 132 |
self.total_number_of_events = int(cfg_params['ORCA.total_number_of_events'])
|
| 133 |
except KeyError:
|
| 134 |
msg = 'Must define total_number_of_events and job_number_of_events'
|
| 135 |
raise CrabException(msg)
|
| 136 |
|
| 137 |
try:
|
| 138 |
self.first_event = int(cfg_params['ORCA.first_event'])
|
| 139 |
except KeyError:
|
| 140 |
self.first_event = 0
|
| 141 |
pass
|
| 142 |
log.debug(6, "Orca::Orca(): total number of events = "+`self.total_number_of_events`)
|
| 143 |
#log.debug(6, "Orca::Orca(): events per job = "+`self.job_number_of_events`)
|
| 144 |
log.debug(6, "Orca::Orca(): first event = "+`self.first_event`)
|
| 145 |
|
| 146 |
#DBSDLS-start
|
| 147 |
## Initialize the variables that are extracted from DBS/DLS and needed in other places of the code
|
| 148 |
self.maxEvents=0 # max events available ( --> check the requested nb. of evts in Creator.py)
|
| 149 |
self.DBSPaths={} # all dbs paths requested ( --> input to the site local discovery script)
|
| 150 |
## Perform the data location and discovery (based on DBS/DLS)
|
| 151 |
self.DataDiscoveryAndLocation(cfg_params)
|
| 152 |
#DBSDLS-end
|
| 153 |
|
| 154 |
self.TarBaller = TarBall.TarBall(self.executable, self.scram)
|
| 155 |
self.tgzNameWithPath = self.TarBaller.prepareTarBall()
|
| 156 |
|
| 157 |
try:
|
| 158 |
self.ML = int(cfg_params['USER.activate_monalisa'])
|
| 159 |
except KeyError:
|
| 160 |
self.ML = 0
|
| 161 |
pass
|
| 162 |
|
| 163 |
self.setTaskid_()
|
| 164 |
self.setParam_('taskId', self.cfg_params['taskId'])
|
| 165 |
|
| 166 |
return
|
| 167 |
|
| 168 |
def split(self, jobParams):
|
| 169 |
"""
|
| 170 |
This method returns the list of orca specific job type items
|
| 171 |
needed to run the jobs
|
| 172 |
"""
|
| 173 |
common.jobDB.load()
|
| 174 |
njobs = self.total_number_of_jobs
|
| 175 |
# create the empty structure
|
| 176 |
for i in range(njobs):
|
| 177 |
jobParams.append("")
|
| 178 |
|
| 179 |
# fill the both the list and the DB (part of the code taken from jobsToDB)
|
| 180 |
firstEvent = self.first_event
|
| 181 |
lastJobsNumberOfEvents = str(self.job_number_of_events)
|
| 182 |
# last jobs is different...
|
| 183 |
for job in range(njobs-1):
|
| 184 |
jobParams[job] = [str(firstEvent), str(lastJobsNumberOfEvents)]
|
| 185 |
common.jobDB.setArguments(job, jobParams[job])
|
| 186 |
common.jobDB.setDestination(job, self.sites)
|
| 187 |
firstEvent += self.job_number_of_events
|
| 188 |
|
| 189 |
# this is the last job
|
| 190 |
lastJobsNumberOfEvents = (self.total_number_of_events + self.first_event) - firstEvent
|
| 191 |
status = common.jobDB.status(njobs - 1)
|
| 192 |
jobParams[njobs - 1] = [str(firstEvent), str(lastJobsNumberOfEvents)]
|
| 193 |
common.jobDB.setArguments(njobs - 1, jobParams[njobs - 1])
|
| 194 |
|
| 195 |
if (lastJobsNumberOfEvents!=self.job_number_of_events):
|
| 196 |
common.logger.message(str(self.total_number_of_jobs-1)+' jobs will be created for '+str(self.job_number_of_events)+' events each plus 1 for '+str(lastJobsNumberOfEvents)+' events for a total of '+str(self.job_number_of_events*(self.total_number_of_jobs-1)+lastJobsNumberOfEvents)+' events')
|
| 197 |
else:
|
| 198 |
common.logger.message(str(self.total_number_of_jobs)+' jobs will be created for '+str(self.job_number_of_events)+' events each for a total of '+str(self.job_number_of_events*(self.total_number_of_jobs-1)+lastJobsNumberOfEvents)+' events')
|
| 199 |
|
| 200 |
common.jobDB.save()
|
| 201 |
return
|
| 202 |
|
| 203 |
def getJobTypeArguments(self, nj, sched):
|
| 204 |
params = common.jobDB.arguments(nj)
|
| 205 |
|
| 206 |
if sched=="EDG" or sched=="CONDOR" or sched=="GRID":
|
| 207 |
parString = "" + str(params[0])+' '+str(params[1])
|
| 208 |
elif sched=="BOSS":
|
| 209 |
parString = "" + str(params[0])+' '+str(params[1])
|
| 210 |
else:
|
| 211 |
return ""
|
| 212 |
return parString
|
| 213 |
|
| 214 |
def numberOfJobs(self):
|
| 215 |
first_event = self.first_event
|
| 216 |
maxAvailableEvents = int(self.maxEvents)
|
| 217 |
common.logger.debug(1,"Available events: "+str(maxAvailableEvents))
|
| 218 |
if first_event>=maxAvailableEvents:
|
| 219 |
raise CrabException('First event is bigger than maximum number of available events!')
|
| 220 |
|
| 221 |
try:
|
| 222 |
n = self.total_number_of_events
|
| 223 |
if n == 'all': n = '-1'
|
| 224 |
if n == '-1':
|
| 225 |
tot_num_events = (maxAvailableEvents - first_event)
|
| 226 |
common.logger.debug(1,"Analysing all available events "+str(tot_num_events))
|
| 227 |
else:
|
| 228 |
if maxAvailableEvents<(int(n)+ first_event): # + self.first_event):
|
| 229 |
raise CrabException('(First event + total events)='+str(int(n)+first_event)+' is bigger than maximum number of available events '+str(maxAvailableEvents)+' !! Use "total_number_of_events=-1" to analyze to whole dataset')
|
| 230 |
tot_num_events = int(n)
|
| 231 |
except KeyError:
|
| 232 |
common.logger.message("total_number_of_events not defined, set it to maximum available")
|
| 233 |
tot_num_events = (maxAvailableEvents - first_event)
|
| 234 |
pass
|
| 235 |
common.logger.message("Total number of events to be analyzed: "+str(self.total_number_of_events))
|
| 236 |
|
| 237 |
|
| 238 |
# read user directives
|
| 239 |
eventPerJob=0
|
| 240 |
try:
|
| 241 |
eventPerJob = self.cfg_params['ORCA.job_number_of_events']
|
| 242 |
except KeyError:
|
| 243 |
pass
|
| 244 |
|
| 245 |
jobsPerTask=0
|
| 246 |
try:
|
| 247 |
jobsPerTask = int(self.cfg_params['ORCA.total_number_of_jobs'])
|
| 248 |
except KeyError:
|
| 249 |
pass
|
| 250 |
|
| 251 |
# If both the above set, complain and use event per jobs
|
| 252 |
if eventPerJob>0 and jobsPerTask>0:
|
| 253 |
msg = 'Warning. '
|
| 254 |
msg += 'job_number_of_events and total_number_of_jobs are both defined '
|
| 255 |
msg += 'Using job_number_of_events.'
|
| 256 |
common.logger.message(msg)
|
| 257 |
jobsPerTask = 0
|
| 258 |
if eventPerJob==0 and jobsPerTask==0:
|
| 259 |
msg = 'Warning. '
|
| 260 |
msg += 'job_number_of_events and total_number_of_jobs are not defined '
|
| 261 |
msg += 'Creating just one job for all events.'
|
| 262 |
common.logger.message(msg)
|
| 263 |
jobsPerTask = 1
|
| 264 |
|
| 265 |
# first case: events per job defined
|
| 266 |
if eventPerJob>0:
|
| 267 |
n=eventPerJob
|
| 268 |
#if n == 'all' or n == '-1' or (int(n)>self.total_number_of_events and self.total_number_of_events>0):
|
| 269 |
if n == 'all' or n == '-1' or (int(n)>tot_num_events and tot_num_events>0):
|
| 270 |
common.logger.message("Asking more events than available: set it to maximum available")
|
| 271 |
job_num_events = tot_num_events
|
| 272 |
tot_num_jobs = 1
|
| 273 |
else:
|
| 274 |
job_num_events = int(n)
|
| 275 |
tot_num_jobs = int((tot_num_events-1)/job_num_events)+1
|
| 276 |
|
| 277 |
elif jobsPerTask>0:
|
| 278 |
common.logger.debug(2,"total number of events: "+str(tot_num_events)+" JobPerTask "+str(jobsPerTask))
|
| 279 |
job_num_events = int(math.floor((tot_num_events)/jobsPerTask))
|
| 280 |
tot_num_jobs = jobsPerTask
|
| 281 |
|
| 282 |
# should not happen...
|
| 283 |
else:
|
| 284 |
raise CrabException('Something wrong with splitting')
|
| 285 |
|
| 286 |
common.logger.debug(2,"total number of events: "+str(tot_num_events)+" events per job: "+str(job_num_events))
|
| 287 |
|
| 288 |
#used by jobsToDB for logs
|
| 289 |
self.job_number_of_events = job_num_events
|
| 290 |
self.total_number_of_jobs = tot_num_jobs
|
| 291 |
return tot_num_jobs
|
| 292 |
|
| 293 |
|
| 294 |
def jobsToDB(self, nJobs):
|
| 295 |
"""
|
| 296 |
Fill the DB with proper entries for ORCA
|
| 297 |
"""
|
| 298 |
|
| 299 |
firstEvent = self.first_event
|
| 300 |
lastJobsNumberOfEvents = self.job_number_of_events
|
| 301 |
|
| 302 |
# last jobs is different...
|
| 303 |
for job in range(nJobs-1):
|
| 304 |
common.jobDB.setFirstEvent(job, firstEvent)
|
| 305 |
common.jobDB.setMaxEvents(job, self.job_number_of_events)
|
| 306 |
firstEvent=firstEvent+self.job_number_of_events
|
| 307 |
|
| 308 |
# this is the last job
|
| 309 |
common.jobDB.setFirstEvent(nJobs-1, firstEvent)
|
| 310 |
lastJobsNumberOfEvents= (self.total_number_of_events+self.first_event)-firstEvent
|
| 311 |
common.jobDB.setMaxEvents(nJobs-1, lastJobsNumberOfEvents)
|
| 312 |
|
| 313 |
if (lastJobsNumberOfEvents!=self.job_number_of_events):
|
| 314 |
common.logger.message(str(self.total_number_of_jobs-1)+' jobs will be created for '+str(self.job_number_of_events)+' events each plus 1 for '+str(lastJobsNumberOfEvents)+' events for a total of '+str(self.job_number_of_events*(self.total_number_of_jobs-1)+lastJobsNumberOfEvents)+' events')
|
| 315 |
else:
|
| 316 |
common.logger.message(str(self.total_number_of_jobs)+' jobs will be created for '+str(self.job_number_of_events)+' events each for a total of '+str(self.job_number_of_events*(self.total_number_of_jobs-1)+lastJobsNumberOfEvents)+' events')
|
| 317 |
|
| 318 |
return
|
| 319 |
|
| 320 |
|
| 321 |
def wsSetupEnvironment(self, nj):
|
| 322 |
"""
|
| 323 |
Returns part of a job script which prepares
|
| 324 |
the execution environment for the job 'nj'.
|
| 325 |
"""
|
| 326 |
|
| 327 |
# Prepare JobType-independent part
|
| 328 |
txt = ''
|
| 329 |
|
| 330 |
## OLI_Daniele at this level middleware already known
|
| 331 |
|
| 332 |
txt += 'if [ $middleware == LCG ]; then \n'
|
| 333 |
txt += self.wsSetupCMSLCGEnvironment_()
|
| 334 |
txt += 'elif [ $middleware == OSG ]; then\n'
|
| 335 |
txt += ' WORKING_DIR=`/bin/mktemp -d $OSG_WN_TMP/cms_XXXXXXXXXXXX`\n'
|
| 336 |
txt += ' echo "Created working directory: $WORKING_DIR"\n'
|
| 337 |
txt += ' if [ ! -d $WORKING_DIR ] ;then\n'
|
| 338 |
txt += ' echo "SET_CMS_ENV 10016 ==> OSG $WORKING_DIR could not be created on WN `hostname`"\n'
|
| 339 |
txt += ' echo "JOB_EXIT_STATUS = 10016"\n'
|
| 340 |
txt += ' echo "JobExitCode=10016" | tee -a $RUNTIME_AREA/$repo\n'
|
| 341 |
txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
|
| 342 |
txt += ' rm -f $RUNTIME_AREA/$repo \n'
|
| 343 |
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
|
| 344 |
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
|
| 345 |
txt += ' exit 1\n'
|
| 346 |
txt += ' fi\n'
|
| 347 |
txt += '\n'
|
| 348 |
txt += ' echo "Change to working directory: $WORKING_DIR"\n'
|
| 349 |
txt += ' cd $WORKING_DIR\n'
|
| 350 |
txt += self.wsSetupCMSOSGEnvironment_()
|
| 351 |
txt += 'fi\n'
|
| 352 |
|
| 353 |
# Prepare JobType-specific part
|
| 354 |
scram = self.scram.commandName()
|
| 355 |
txt += '\n\n'
|
| 356 |
txt += 'echo "### SPECIFIC JOB SETUP ENVIRONMENT ###"\n'
|
| 357 |
txt += scram+' project ORCA '+self.version+'\n'
|
| 358 |
txt += 'status=$?\n'
|
| 359 |
txt += 'if [ $status != 0 ] ; then\n'
|
| 360 |
txt += ' echo "SET_EXE_ENV 10034 ==>ERROR ORCA '+self.version+' not found on `hostname`" \n'
|
| 361 |
txt += ' echo "JOB_EXIT_STATUS = 10034"\n'
|
| 362 |
txt += ' echo "JobExitCode=10034" | tee -a $RUNTIME_AREA/$repo\n'
|
| 363 |
txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
|
| 364 |
txt += ' rm -f $RUNTIME_AREA/$repo \n'
|
| 365 |
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
|
| 366 |
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
|
| 367 |
## OLI_Daniele
|
| 368 |
txt += ' if [ $middleware == OSG ]; then \n'
|
| 369 |
txt += ' echo "Remove working directory: $WORKING_DIR"\n'
|
| 370 |
txt += ' cd $RUNTIME_AREA\n'
|
| 371 |
txt += ' /bin/rm -rf $WORKING_DIR\n'
|
| 372 |
txt += ' if [ -d $WORKING_DIR ] ;then\n'
|
| 373 |
txt += ' echo "SET_CMS_ENV 10018 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after ORCA not found on `hostname`"\n'
|
| 374 |
txt += ' echo "JOB_EXIT_STATUS = 10018"\n'
|
| 375 |
txt += ' echo "JobExitCode=10018" | tee -a $RUNTIME_AREA/$repo\n'
|
| 376 |
txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
|
| 377 |
txt += ' rm -f $RUNTIME_AREA/$repo \n'
|
| 378 |
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
|
| 379 |
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
|
| 380 |
txt += ' fi\n'
|
| 381 |
txt += ' fi \n'
|
| 382 |
txt += ' exit 1\n'
|
| 383 |
txt += 'fi \n'
|
| 384 |
txt += 'echo "ORCA_VERSION = '+self.version+'"\n'
|
| 385 |
txt += 'cd '+self.version+'\n'
|
| 386 |
### needed grep for bug in scramv1 ###
|
| 387 |
|
| 388 |
#txt += 'eval `'+scram+' runtime -sh | grep -v SCRAMRT_LSB_JOBNAME`\n'
|
| 389 |
|
| 390 |
# Handle the arguments:
|
| 391 |
txt += "\n"
|
| 392 |
# txt += "## ARGUMNETS: $1 Job Number\n"
|
| 393 |
# txt += "## ARGUMNETS: $2 First Event for this job\n"
|
| 394 |
# txt += "## ARGUMNETS: $3 Max Event for this job\n"
|
| 395 |
txt += "## ARGUMENTS: ${args[0]} Job Number\n"
|
| 396 |
txt += "## ARGUMENTS: ${args[1]} First Event for this job\n"
|
| 397 |
txt += "## ARGUMENTS: ${args[2]} Max Event for this job\n"
|
| 398 |
txt += "\n"
|
| 399 |
# txt += "narg=$#\n"
|
| 400 |
# txt += "NJob=$1\n"
|
| 401 |
# txt += "FirstEvent=$2\n"
|
| 402 |
# txt += "MaxEvents=$3\n"
|
| 403 |
txt += "NJob=${args[0]}\n"
|
| 404 |
txt += "FirstEvent=${args[1]}\n"
|
| 405 |
txt += "MaxEvents=${args[2]}\n"
|
| 406 |
txt += "if [ $nargs -lt 3 ]\n"
|
| 407 |
txt += "then\n"
|
| 408 |
txt += " echo 'SET_EXE_ENV 50113 ==> ERROR Too few arguments' +$nargs+ \n"
|
| 409 |
txt += ' echo "JOB_EXIT_STATUS = 50113"\n'
|
| 410 |
txt += ' echo "JobExitCode=50113" | tee -a $RUNTIME_AREA/$repo\n'
|
| 411 |
txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
|
| 412 |
txt += ' rm -f $RUNTIME_AREA/$repo \n'
|
| 413 |
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
|
| 414 |
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
|
| 415 |
|
| 416 |
## OLI_Daniele
|
| 417 |
txt += ' if [ $middleware == OSG ]; then \n'
|
| 418 |
txt += ' echo "Remove working directory: $WORKING_DIR"\n'
|
| 419 |
txt += ' cd $RUNTIME_AREA\n'
|
| 420 |
txt += ' /bin/rm -rf $WORKING_DIR\n'
|
| 421 |
txt += ' if [ -d $WORKING_DIR ] ;then\n'
|
| 422 |
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'
|
| 423 |
txt += ' echo "JOB_EXIT_STATUS = 50114"\n'
|
| 424 |
txt += ' echo "JobExitCode=50114" | tee -a $RUNTIME_AREA/$repo\n'
|
| 425 |
txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
|
| 426 |
txt += ' rm -f $RUNTIME_AREA/$repo \n'
|
| 427 |
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
|
| 428 |
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
|
| 429 |
txt += ' fi\n'
|
| 430 |
txt += ' fi \n'
|
| 431 |
txt += " exit 1\n"
|
| 432 |
txt += "fi\n"
|
| 433 |
txt += "\n"
|
| 434 |
|
| 435 |
# Prepare job-specific part
|
| 436 |
job = common.job_list[nj]
|
| 437 |
orcarc = os.path.basename(job.configFilename())
|
| 438 |
txt += '\n'
|
| 439 |
#DBSDLS-start
|
| 440 |
#### site-local catalogue discovery mechanism:
|
| 441 |
## check that the site configuration file exists
|
| 442 |
txt += 'echo "### Site Local Catalogue Discovery ### "\n'
|
| 443 |
txt += 'if [ $middleware == LCG ]; then \n'
|
| 444 |
txt += ' if [ -f $VO_CMS_SW_DIR/cms_site_config ]; then \n'
|
| 445 |
txt += ' dbsdls_cms_site_config=$VO_CMS_SW_DIR/cms_site_config\n'
|
| 446 |
txt += ' else\n'
|
| 447 |
txt += ' echo "Site Local Catalogue Discovery Failed: No site configuration file $VO_CMS_SW_DIR/cms_site_config !" \n'
|
| 448 |
txt += ' echo "JOB_EXIT_STATUS = 10037"\n'
|
| 449 |
txt += ' echo "JobExitCode=10037" | tee -a $RUNTIME_AREA/$repo\n'
|
| 450 |
txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
|
| 451 |
txt += ' rm -f $RUNTIME_AREA/$repo \n'
|
| 452 |
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
|
| 453 |
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
|
| 454 |
txt += ' exit 1\n'
|
| 455 |
txt += ' fi \n'
|
| 456 |
txt += 'elif [ $middleware == OSG ]; then\n'
|
| 457 |
txt += ' if [ -f $GRID3_APP_DIR/cmssoft/cms_site_config ]; then \n'
|
| 458 |
txt += ' dbsdls_cms_site_config=$GRID3_APP_DIR/cmssoft/cms_site_config\n'
|
| 459 |
txt += ' elif [ -f $OSG_APP/cmssoft/cms_site_config ]; then \n'
|
| 460 |
txt += ' dbsdls_cms_site_config=$OSG_APP/cmssoft/cms_site_config\n'
|
| 461 |
txt += ' else\n'
|
| 462 |
txt += ' echo "Site Local Catalogue Discovery Failed: No site configuration file $GRID3_APP_DIR/cmssoft/cms_site_config or $OSG_APP/cmssoft/cms_site_config !" \n'
|
| 463 |
txt += ' echo "JOB_EXIT_STATUS = 10037"\n'
|
| 464 |
txt += ' echo "JobExitCode=10037" | tee -a $RUNTIME_AREA/$repo\n'
|
| 465 |
txt += ' if [ $middleware == OSG ]; then \n'
|
| 466 |
txt += ' echo "Remove working directory: $WORKING_DIR"\n'
|
| 467 |
txt += ' cd $RUNTIME_AREA\n'
|
| 468 |
txt += ' /bin/rm -rf $WORKING_DIR\n'
|
| 469 |
txt += ' if [ -d $WORKING_DIR ] ;then\n'
|
| 470 |
txt += ' echo "SET_EXE_ENV 10019 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after No site configuration file $GRID3_APP_DIR/cmssoft/cms_site_config or $OSG_APP/cmssoft/cms_site_config"\n'
|
| 471 |
txt += ' echo "JOB_EXIT_STATUS = 10019"\n'
|
| 472 |
txt += ' echo "JobExitCode=10019" | tee -a $RUNTIME_AREA/$repo\n'
|
| 473 |
txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
|
| 474 |
txt += ' rm -f $RUNTIME_AREA/$repo \n'
|
| 475 |
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
|
| 476 |
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
|
| 477 |
txt += ' fi\n'
|
| 478 |
txt += ' fi \n'
|
| 479 |
txt += ' exit 1\n'
|
| 480 |
txt += ' fi \n'
|
| 481 |
txt += 'fi\n'
|
| 482 |
txt += 'echo "Site Local Catalog Discovery, selected site configuration: $dbsdls_cms_site_config"\n'
|
| 483 |
## look for a site local script sent as inputsandbox otherwise use the default one under $VO_CMS_SW_DIR for LCG or $GRID3_APP_DIR/cmssoft or $OSG_APP/cmssoft for OSG
|
| 484 |
txt += 'if [ -f $RUNTIME_AREA/cms_site_catalogue.sh ]; then \n'
|
| 485 |
txt += ' sitelocalscript=$RUNTIME_AREA/cms_site_catalogue.sh \n'
|
| 486 |
txt += 'elif [ $middleware == LCG ]; then \n'
|
| 487 |
txt += ' if [ -f $VO_CMS_SW_DIR/cms_site_catalogue.sh ]; then \n'
|
| 488 |
txt += ' sitelocalscript=$VO_CMS_SW_DIR/cms_site_catalogue.sh \n'
|
| 489 |
txt += ' else \n'
|
| 490 |
txt += ' echo "Site Local Catalogue Discovery Failed: No site local script cms_site_catalogue.sh !"\n'
|
| 491 |
txt += ' echo "JOB_EXIT_STATUS = 10038"\n'
|
| 492 |
txt += ' echo "JobExitCode=10038" | tee -a $RUNTIME_AREA/$repo\n'
|
| 493 |
txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
|
| 494 |
txt += ' rm -f $RUNTIME_AREA/$repo \n'
|
| 495 |
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
|
| 496 |
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
|
| 497 |
txt += ' exit 1\n'
|
| 498 |
txt += ' fi \n'
|
| 499 |
txt += 'elif [ $middleware == OSG ]; then\n'
|
| 500 |
txt += ' if [ -f $GRID3_APP_DIR/cmssoft/cms_site_catalogue.sh ]; then \n'
|
| 501 |
txt += ' sitelocalscript=$GRID3_APP_DIR/cmssoft/cms_site_catalogue.sh\n'
|
| 502 |
txt += ' elif [ -f $OSG_APP/cmssoft/cms_site_catalogue.sh ]; then \n'
|
| 503 |
txt += ' sitelocalscript=$OSG_APP/cmssoft/cms_site_catalogue.sh\n'
|
| 504 |
txt += ' else\n'
|
| 505 |
txt += ' echo "Site Local Catalogue Discovery Failed: No site local script cms_site_catalogue.sh !"\n'
|
| 506 |
txt += ' echo "JOB_EXIT_STATUS = 10038"\n'
|
| 507 |
txt += ' echo "JobExitCode=10038" | tee -a $RUNTIME_AREA/$repo\n'
|
| 508 |
txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
|
| 509 |
txt += ' rm -f $RUNTIME_AREA/$repo \n'
|
| 510 |
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
|
| 511 |
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
|
| 512 |
txt += ' if [ $middleware == OSG ]; then \n'
|
| 513 |
txt += ' echo "Remove working directory: $WORKING_DIR"\n'
|
| 514 |
txt += ' cd $RUNTIME_AREA\n'
|
| 515 |
txt += ' /bin/rm -rf $WORKING_DIR\n'
|
| 516 |
txt += ' if [ -d $WORKING_DIR ] ;then\n'
|
| 517 |
txt += ' echo "SET_EXE_ENV 10014 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after No site local script cms_site_catalogue.sh"\n'
|
| 518 |
txt += ' echo "JOB_EXIT_STATUS = 10014"\n'
|
| 519 |
txt += ' echo "JobExitCode=10014" | tee -a $RUNTIME_AREA/$repo\n'
|
| 520 |
txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
|
| 521 |
txt += ' rm -f $RUNTIME_AREA/$repo \n'
|
| 522 |
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
|
| 523 |
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
|
| 524 |
txt += ' fi\n'
|
| 525 |
txt += ' fi \n'
|
| 526 |
txt += ' exit 1\n'
|
| 527 |
txt += ' fi \n'
|
| 528 |
txt += 'fi\n'
|
| 529 |
txt += 'echo "Site Local Catalog Discovery, selected local script: $sitelocalscript"\n'
|
| 530 |
## execute the site local configuration script with the user requied data as input
|
| 531 |
inputdata=string.join(self.DBSPaths,' ')
|
| 532 |
sitecatalog_cmd='$sitelocalscript -c $dbsdls_cms_site_config '+inputdata
|
| 533 |
txt += sitecatalog_cmd+'\n'
|
| 534 |
txt += 'sitestatus=$?\n'
|
| 535 |
txt += 'if [ ! -f inputurl_orcarc ] || [ $sitestatus -ne 0 ]; then\n'
|
| 536 |
txt += ' echo "Site Local Catalogue Discovery Failed: exiting with $sitestatus"\n'
|
| 537 |
txt += ' echo "JOB_EXIT_STATUS = 1"\n'
|
| 538 |
txt += ' echo "JobExitCode=10039" | tee -a $RUNTIME_AREA/$repo\n'
|
| 539 |
txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
|
| 540 |
txt += ' rm -f $RUNTIME_AREA/$repo \n'
|
| 541 |
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
|
| 542 |
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
|
| 543 |
txt += ' if [ $middleware == OSG ]; then \n'
|
| 544 |
txt += ' echo "Remove working directory: $WORKING_DIR"\n'
|
| 545 |
txt += ' cd $RUNTIME_AREA\n'
|
| 546 |
txt += ' /bin/rm -rf $WORKING_DIR\n'
|
| 547 |
txt += ' if [ -d $WORKING_DIR ] ;then\n'
|
| 548 |
txt += ' echo "SET_EXE_ENV 10013 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after Site Local Catalogue Discovery Failed: exiting with $sitestatus"\n'
|
| 549 |
txt += ' echo "JOB_EXIT_STATUS = 10013"\n'
|
| 550 |
txt += ' echo "JobExitCode=10013" | tee -a $RUNTIME_AREA/$repo\n'
|
| 551 |
txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
|
| 552 |
txt += ' rm -f $RUNTIME_AREA/$repo \n'
|
| 553 |
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
|
| 554 |
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
|
| 555 |
txt += ' fi\n'
|
| 556 |
txt += ' fi \n'
|
| 557 |
txt += ' exit 1\n'
|
| 558 |
txt += 'fi \n'
|
| 559 |
## append the orcarc fragment about the Input catalogues to the .orcarc
|
| 560 |
txt += 'cp $RUNTIME_AREA/'+orcarc+' .orcarc\n'
|
| 561 |
txt +=' cat inputurl_orcarc >> .orcarc\n'
|
| 562 |
#DBSDLS-end
|
| 563 |
|
| 564 |
if len(self.additional_inbox_files) > 0:
|
| 565 |
for file in self.additional_inbox_files:
|
| 566 |
file = os.path.basename(file)
|
| 567 |
txt += 'if [ -e $RUNTIME_AREA/'+file+' ] ; then\n'
|
| 568 |
txt += ' cp $RUNTIME_AREA/'+file+' .\n'
|
| 569 |
txt += ' chmod +x '+file+'\n'
|
| 570 |
txt += 'fi\n'
|
| 571 |
pass
|
| 572 |
|
| 573 |
### OLI_DANIELE
|
| 574 |
txt += "echo 'SET_EXE_ENV 0 ==> job setup ok'\n"
|
| 575 |
txt += 'echo "### END JOB SETUP ENVIRONMENT ###"\n\n'
|
| 576 |
|
| 577 |
txt += 'echo "FirstEvent=$FirstEvent" >> .orcarc\n'
|
| 578 |
txt += 'echo "MaxEvents=$MaxEvents" >> .orcarc\n'
|
| 579 |
if self.ML:
|
| 580 |
txt += 'echo "MonalisaJobId=$NJob" >> .orcarc\n'
|
| 581 |
|
| 582 |
txt += '\n'
|
| 583 |
txt += 'echo "***** cat .orcarc *********"\n'
|
| 584 |
txt += 'cat .orcarc\n'
|
| 585 |
txt += 'echo "****** end .orcarc ********"\n'
|
| 586 |
return txt
|
| 587 |
|
| 588 |
def wsBuildExe(self, nj):
|
| 589 |
"""
|
| 590 |
Put in the script the commands to build an executable
|
| 591 |
or a library.
|
| 592 |
"""
|
| 593 |
|
| 594 |
txt = ""
|
| 595 |
|
| 596 |
if os.path.isfile(self.tgzNameWithPath):
|
| 597 |
txt += 'echo "tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+'"\n'
|
| 598 |
txt += 'tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+'\n'
|
| 599 |
txt += 'untar_status=$? \n'
|
| 600 |
txt += 'if [ $untar_status -ne 0 ]; then \n'
|
| 601 |
txt += ' echo "SET_EXE 1 ==> ERROR Untarring .tgz file failed"\n'
|
| 602 |
txt += ' echo "JOB_EXIT_STATUS = $untar_status" \n'
|
| 603 |
txt += ' echo "JobExitCode=$untar_status" | tee -a $repo\n'
|
| 604 |
|
| 605 |
### OLI_DANIELE
|
| 606 |
txt += ' if [ $middleware == OSG ]; then \n'
|
| 607 |
txt += ' echo "Remove working directory: $WORKING_DIR"\n'
|
| 608 |
txt += ' cd $RUNTIME_AREA\n'
|
| 609 |
txt += ' /bin/rm -rf $WORKING_DIR\n'
|
| 610 |
txt += ' if [ -d $WORKING_DIR ] ;then\n'
|
| 611 |
txt += ' echo "SET_EXE 50999 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after Untarring .tgz file failed"\n'
|
| 612 |
txt += ' echo "JOB_EXIT_STATUS = 50999"\n'
|
| 613 |
txt += ' echo "JobExitCode=50999" | tee -a $RUNTIME_AREA/$repo\n'
|
| 614 |
txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
|
| 615 |
txt += ' rm -f $RUNTIME_AREA/$repo \n'
|
| 616 |
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
|
| 617 |
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
|
| 618 |
txt += ' fi\n'
|
| 619 |
txt += ' fi \n'
|
| 620 |
txt += ' \n'
|
| 621 |
txt += ' exit 1 \n'
|
| 622 |
txt += 'else \n'
|
| 623 |
txt += ' echo "Successful untar" \n'
|
| 624 |
txt += 'fi \n'
|
| 625 |
# TODO: what does this code do here ?
|
| 626 |
# SL check that lib/Linux__... is present
|
| 627 |
txt += 'mkdir -p lib/${SCRAM_ARCH} \n'
|
| 628 |
pass
|
| 629 |
txt += 'eval `'+self.scram.commandName()+' runtime -sh |grep -v SCRAMRT_LSB_JOBNAME`'+'\n'
|
| 630 |
|
| 631 |
return txt
|
| 632 |
|
| 633 |
def wsRenameOutput(self, nj):
|
| 634 |
"""
|
| 635 |
Returns part of a job script which renames the produced files.
|
| 636 |
"""
|
| 637 |
|
| 638 |
txt = '\n'
|
| 639 |
txt += '# directory content\n'
|
| 640 |
txt += 'ls \n'
|
| 641 |
file_list = ''
|
| 642 |
for fileWithSuffix in self.output_file:
|
| 643 |
output_file_num = self.numberFile_(fileWithSuffix, '$NJob')
|
| 644 |
file_list=file_list+output_file_num+' '
|
| 645 |
txt += '\n'
|
| 646 |
txt += 'ls '+fileWithSuffix+'\n'
|
| 647 |
#txt += 'exe_result=$?\n'
|
| 648 |
txt += 'ls_result=$?\n'
|
| 649 |
#txt += 'if [ $exe_result -ne 0 ] ; then\n'
|
| 650 |
txt += 'if [ $ls_result -ne 0 ] ; then\n'
|
| 651 |
txt += ' echo "ERROR: Problem with output file"\n'
|
| 652 |
#txt += ' echo "ERROR: No output file to manage"\n'
|
| 653 |
#txt += ' echo "JOB_EXIT_STATUS = $exe_result"\n'
|
| 654 |
#txt += ' echo "JobExitCode=60302" | tee -a $RUNTIME_AREA/$repo\n'
|
| 655 |
#txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
|
| 656 |
# txt += 'exe_result=$?\n'
|
| 657 |
# txt += 'if [ $exe_result -ne 0 ] ; then\n'
|
| 658 |
# txt += ' echo "ERROR: No output file to manage"\n'
|
| 659 |
# txt += ' echo "JOB_EXIT_STATUS = $exe_result"\n'
|
| 660 |
# txt += ' echo "JobExitCode=60302" | tee -a $RUNTIME_AREA/$repo\n'
|
| 661 |
# txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
|
| 662 |
# txt += ' rm -f $RUNTIME_AREA/$repo \n'
|
| 663 |
# txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
|
| 664 |
# txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
|
| 665 |
|
| 666 |
### OLI_DANIELE
|
| 667 |
if common.scheduler.boss_scheduler_name == 'condor_g':
|
| 668 |
txt += ' if [ $middleware == OSG ]; then \n'
|
| 669 |
txt += ' echo "prepare dummy output file"\n'
|
| 670 |
txt += ' echo "Processing of job output failed" > $RUNTIME_AREA/'+output_file_num+'\n'
|
| 671 |
txt += ' fi \n'
|
| 672 |
txt += 'else\n'
|
| 673 |
txt += ' cp '+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n'
|
| 674 |
txt += 'fi\n'
|
| 675 |
|
| 676 |
|
| 677 |
txt += 'cd $RUNTIME_AREA\n'
|
| 678 |
file_list=file_list[:-1]
|
| 679 |
txt += 'file_list="'+file_list+'"\n'
|
| 680 |
### OLI_DANIELE
|
| 681 |
txt += 'if [ $middleware == OSG ]; then\n'
|
| 682 |
txt += ' cd $RUNTIME_AREA\n'
|
| 683 |
txt += ' echo "Remove working directory: $WORKING_DIR"\n'
|
| 684 |
txt += ' /bin/rm -rf $WORKING_DIR\n'
|
| 685 |
txt += ' if [ -d $WORKING_DIR ] ;then\n'
|
| 686 |
txt += ' echo "SET_EXE 60999 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after cleanup of WN"\n'
|
| 687 |
txt += ' echo "JOB_EXIT_STATUS = 60999"\n'
|
| 688 |
txt += ' echo "JobExitCode=60999" | tee -a $RUNTIME_AREA/$repo\n'
|
| 689 |
txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
|
| 690 |
txt += ' rm -f $RUNTIME_AREA/$repo \n'
|
| 691 |
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
|
| 692 |
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
|
| 693 |
txt += ' fi\n'
|
| 694 |
txt += 'fi\n'
|
| 695 |
txt += '\n'
|
| 696 |
|
| 697 |
return txt
|
| 698 |
|
| 699 |
def executableName(self):
|
| 700 |
if self.scriptExe != '':
|
| 701 |
return "./" + os.path.basename(self.scriptExe)
|
| 702 |
else:
|
| 703 |
return self.executable
|
| 704 |
|
| 705 |
#DBSDLS-start
|
| 706 |
def DataDiscoveryAndLocation(self, cfg_params):
|
| 707 |
|
| 708 |
fun = "Orca::DataDiscoveryAndLocation()"
|
| 709 |
|
| 710 |
## Contact the DBS
|
| 711 |
try:
|
| 712 |
self.pubdata=DataDiscovery.DataDiscovery(self.owner,
|
| 713 |
self.dataset,
|
| 714 |
self.dataTiers,
|
| 715 |
cfg_params)
|
| 716 |
self.pubdata.fetchDBSInfo()
|
| 717 |
|
| 718 |
except DataDiscovery.NotExistingDatasetError, ex :
|
| 719 |
msg = 'ERROR ***: failed Data Discovery in DBS : %s'%ex.getErrorMessage()
|
| 720 |
raise CrabException(msg)
|
| 721 |
|
| 722 |
except DataDiscovery.NoDataTierinProvenanceError, ex :
|
| 723 |
msg = 'ERROR ***: failed Data Discovery in DBS : %s'%ex.getErrorMessage()
|
| 724 |
raise CrabException(msg)
|
| 725 |
except DataDiscovery.DataDiscoveryError, ex:
|
| 726 |
msg = 'ERROR ***: failed Data Discovery in DBS %s'%ex.getErrorMessage()
|
| 727 |
raise CrabException(msg)
|
| 728 |
|
| 729 |
## get list of all required data in the form of dbs paths (dbs path = /dataset/datatier/owner)
|
| 730 |
self.DBSPaths=self.pubdata.getDBSPaths()
|
| 731 |
common.logger.message("Required data are : ")
|
| 732 |
for path in self.DBSPaths:
|
| 733 |
common.logger.message(" --> "+path )
|
| 734 |
|
| 735 |
## get max number of events
|
| 736 |
#common.logger.debug(10,"number of events for primary fileblocks %i"%self.pubdata.getMaxEvents())
|
| 737 |
self.maxEvents=self.pubdata.getMaxEvents() ## self.maxEvents used in Creator.py
|
| 738 |
common.logger.message("\nThe number of available events is %s"%self.maxEvents)
|
| 739 |
|
| 740 |
## get fileblocks corresponding to the required data
|
| 741 |
fb=self.pubdata.getFileBlocks()
|
| 742 |
|
| 743 |
|
| 744 |
## Contact the DLS and build a list of sites hosting the fileblocks
|
| 745 |
try:
|
| 746 |
dataloc=DataLocation.DataLocation(self.pubdata.getFileBlocks(),cfg_params)
|
| 747 |
dataloc.fetchDLSInfo()
|
| 748 |
except DataLocation.DataLocationError , ex:
|
| 749 |
msg = 'ERROR ***: failes Data Location in DLS \n %s '%ex.getErrorMessage()
|
| 750 |
raise CrabException(msg)
|
| 751 |
|
| 752 |
|
| 753 |
self.sites=dataloc.getSites()
|
| 754 |
|
| 755 |
if len(self.sites)==0:
|
| 756 |
msg = 'No sites hosting all the needed data! Exiting... '
|
| 757 |
raise CrabException(msg)
|
| 758 |
common.logger.message("List of Sites ("+str(len(self.sites))+") hosting the data : "+str(self.sites))
|
| 759 |
common.logger.debug(6, "List of Sites: "+str(self.sites))
|
| 760 |
|
| 761 |
return
|
| 762 |
|
| 763 |
#DBDDLS-stop
|
| 764 |
|
| 765 |
|
| 766 |
def nJobs(self):
|
| 767 |
# TODO: should not be here !
|
| 768 |
# JobType should have no internal knowledge about submitted jobs
|
| 769 |
# One possibility is to use len(common.job_list).
|
| 770 |
""" return the number of job to be created """
|
| 771 |
return len(common.job_list)
|
| 772 |
|
| 773 |
def prepareSteeringCards(self):
|
| 774 |
"""
|
| 775 |
modify the orcarc card provided by the user,
|
| 776 |
writing a new card into share dir
|
| 777 |
"""
|
| 778 |
infile = ''
|
| 779 |
try:
|
| 780 |
infile = open(self.orcarc_file,'r')
|
| 781 |
except:
|
| 782 |
self.orcarc_file = 'empty.orcarc'
|
| 783 |
cmd='touch '+self.orcarc_file
|
| 784 |
runCommand(cmd)
|
| 785 |
infile = open(self.orcarc_file,'r')
|
| 786 |
|
| 787 |
outfile = open(common.work_space.jobDir()+self.name()+'.orcarc', 'w')
|
| 788 |
|
| 789 |
inline=infile.readlines()
|
| 790 |
### remove from user card these lines ###
|
| 791 |
wordRemove=['InputFileCatalogURL', 'InputCollections', 'FirstEvent', 'MaxEvents', 'TFileAdaptor']
|
| 792 |
for line in inline:
|
| 793 |
word = string.strip(string.split(line,'=')[0])
|
| 794 |
|
| 795 |
if word not in wordRemove:
|
| 796 |
outfile.write(line)
|
| 797 |
else:
|
| 798 |
continue
|
| 799 |
pass
|
| 800 |
|
| 801 |
outfile.write('\n\n##### The following cards have been created by CRAB: DO NOT TOUCH #####\n')
|
| 802 |
outfile.write('TFileAdaptor = true\n')
|
| 803 |
|
| 804 |
outfile.write('MonRecAlisaBuilder=false\n')
|
| 805 |
|
| 806 |
outfile.write('InputCollections=/System/'+self.owner+'/'+self.dataset+'/'+self.dataset+'\n')
|
| 807 |
|
| 808 |
infile.close()
|
| 809 |
outfile.close()
|
| 810 |
return
|
| 811 |
|
| 812 |
def modifySteeringCards(self, nj):
|
| 813 |
"""
|
| 814 |
Creates steering cards file modifying a template file
|
| 815 |
"""
|
| 816 |
return
|
| 817 |
|
| 818 |
def cardsBaseName(self):
|
| 819 |
"""
|
| 820 |
Returns name of user orcarc card-file
|
| 821 |
"""
|
| 822 |
return os.path.split (self.orcarc_file)[1]
|
| 823 |
|
| 824 |
### content of input_sanbdox ...
|
| 825 |
def inputSandbox(self, nj):
|
| 826 |
"""
|
| 827 |
Returns a list of filenames to be put in JDL input sandbox.
|
| 828 |
"""
|
| 829 |
inp_box = []
|
| 830 |
# dict added to delete duplicate from input sandbox file list
|
| 831 |
seen = {}
|
| 832 |
## code
|
| 833 |
if os.path.isfile(self.tgzNameWithPath):
|
| 834 |
inp_box.append(self.tgzNameWithPath)
|
| 835 |
|
| 836 |
##DBSDLS: no orcarc_CE and init_CE.sh produced on UI , thus not inserting them in inputSandbox
|
| 837 |
# ## orcarc
|
| 838 |
# for o in self.allOrcarcs:
|
| 839 |
# for f in o.fileList():
|
| 840 |
# if (f not in seen.keys()):
|
| 841 |
# inp_box.append(common.work_space.jobDir()+f)
|
| 842 |
# seen[f] = 1
|
| 843 |
|
| 844 |
## config
|
| 845 |
inp_box.append(common.job_list[nj].configFilename())
|
| 846 |
## additional input files
|
| 847 |
#inp_box = inp_box + self.additional_inbox_files
|
| 848 |
return inp_box
|
| 849 |
|
| 850 |
### and of output_sandbox
|
| 851 |
def outputSandbox(self, nj):
|
| 852 |
"""
|
| 853 |
Returns a list of filenames to be put in JDL output sandbox.
|
| 854 |
"""
|
| 855 |
out_box = []
|
| 856 |
|
| 857 |
stdout=common.job_list[nj].stdout()
|
| 858 |
stderr=common.job_list[nj].stderr()
|
| 859 |
#out_box.append(stdout)
|
| 860 |
#out_box.append(stderr)
|
| 861 |
|
| 862 |
## User Declared output files
|
| 863 |
for out in self.output_file:
|
| 864 |
n_out = nj + 1
|
| 865 |
#FEDE
|
| 866 |
#out_box.append(self.version+'/'+self.numberFile_(out,str(n_out)))
|
| 867 |
out_box.append(self.numberFile_(out,str(n_out)))
|
| 868 |
return out_box
|
| 869 |
|
| 870 |
def getRequirements(self, nj):
|
| 871 |
"""
|
| 872 |
return job requirements to add to jdl files
|
| 873 |
"""
|
| 874 |
req = ''
|
| 875 |
#if common.analisys_common_info['sites']:
|
| 876 |
if common.analisys_common_info['sw_version']:
|
| 877 |
req='Member("VO-cms-' + \
|
| 878 |
common.analisys_common_info['sw_version'] + \
|
| 879 |
'", other.GlueHostApplicationSoftwareRunTimeEnvironment)'
|
| 880 |
|
| 881 |
sites = common.jobDB.destination(nj)
|
| 882 |
|
| 883 |
if len(sites)>0:
|
| 884 |
req = req + ' && ('
|
| 885 |
for site in sites:
|
| 886 |
req = req + 'other.GlueCEInfoHostName == "' + site + '" || '
|
| 887 |
pass
|
| 888 |
# remove last ||
|
| 889 |
req = req[0:-4]
|
| 890 |
req = req + ')'
|
| 891 |
#print "req = ", req
|
| 892 |
|
| 893 |
return req
|
| 894 |
|
| 895 |
def numberFile_(self, file, txt):
|
| 896 |
"""
|
| 897 |
append _'txt' before last extension of a file
|
| 898 |
"""
|
| 899 |
p = string.split(file,".")
|
| 900 |
# take away last extension
|
| 901 |
name = p[0]
|
| 902 |
for x in p[1:-1]:
|
| 903 |
name=name+"."+x
|
| 904 |
# add "_txt"
|
| 905 |
if len(p)>1:
|
| 906 |
ext = p[len(p)-1]
|
| 907 |
#result = name + '_' + str(txt) + "." + ext
|
| 908 |
result = name + '_' + txt + "." + ext
|
| 909 |
else:
|
| 910 |
#result = name + '_' + str(txt)
|
| 911 |
result = name + '_' + txt
|
| 912 |
|
| 913 |
return result
|
| 914 |
|
| 915 |
|
| 916 |
def stdOut(self):
|
| 917 |
return self.stdOut_
|
| 918 |
|
| 919 |
def stdErr(self):
|
| 920 |
return self.stdErr_
|
| 921 |
|
| 922 |
# marco
|
| 923 |
def setParam_(self, param, value):
|
| 924 |
self._params[param] = value
|
| 925 |
|
| 926 |
def getParams(self):
|
| 927 |
return self._params
|
| 928 |
|
| 929 |
def setTaskid_(self):
|
| 930 |
self._taskId = self.cfg_params['taskId']
|
| 931 |
|
| 932 |
def getTaskid(self):
|
| 933 |
return self._taskId
|
| 934 |
# marco
|
| 935 |
|
| 936 |
def configFilename(self):
|
| 937 |
""" return the config filename """
|
| 938 |
return self.name()+'.orcarc'
|
| 939 |
|
| 940 |
### OLI_DANIELE
|
| 941 |
def wsSetupCMSOSGEnvironment_(self):
|
| 942 |
"""
|
| 943 |
Returns part of a job script which is prepares
|
| 944 |
the execution environment and which is common for all CMS jobs.
|
| 945 |
"""
|
| 946 |
txt = '\n'
|
| 947 |
txt += ' echo "### SETUP CMS OSG ENVIRONMENT ###"\n'
|
| 948 |
txt += ' if [ -f $GRID3_APP_DIR/cmssoft/cmsset_default.sh ] ;then\n'
|
| 949 |
txt += ' # Use $GRID3_APP_DIR/cmssoft/cmsset_default.sh to setup cms software\n'
|
| 950 |
txt += ' source $GRID3_APP_DIR/cmssoft/cmsset_default.sh '+self.version+'\n'
|
| 951 |
txt += ' elif [ -f $OSG_APP/cmssoft/cmsset_default.sh ] ;then\n'
|
| 952 |
txt += ' # Use $OSG_APP/cmssoft/cmsset_default.sh to setup cms software\n'
|
| 953 |
txt += ' source $OSG_APP/cmssoft/cmsset_default.sh '+self.version+'\n'
|
| 954 |
txt += ' else\n'
|
| 955 |
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'
|
| 956 |
txt += ' echo "JOB_EXIT_STATUS = 10020"\n'
|
| 957 |
txt += ' echo "JobExitCode=10020" | tee -a $RUNTIME_AREA/$repo\n'
|
| 958 |
txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
|
| 959 |
txt += ' rm -f $RUNTIME_AREA/$repo \n'
|
| 960 |
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
|
| 961 |
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
|
| 962 |
txt += ' exit 1\n'
|
| 963 |
txt += '\n'
|
| 964 |
txt += ' echo "Remove working directory: $WORKING_DIR"\n'
|
| 965 |
txt += ' cd $RUNTIME_AREA\n'
|
| 966 |
txt += ' /bin/rm -rf $WORKING_DIR\n'
|
| 967 |
txt += ' if [ -d $WORKING_DIR ] ;then\n'
|
| 968 |
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'
|
| 969 |
txt += ' echo "JOB_EXIT_STATUS = 10017"\n'
|
| 970 |
txt += ' echo "JobExitCode=10017" | tee -a $RUNTIME_AREA/$repo\n'
|
| 971 |
txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
|
| 972 |
txt += ' rm -f $RUNTIME_AREA/$repo \n'
|
| 973 |
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
|
| 974 |
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
|
| 975 |
txt += ' fi\n'
|
| 976 |
txt += '\n'
|
| 977 |
txt += ' exit 1\n'
|
| 978 |
txt += ' fi\n'
|
| 979 |
txt += '\n'
|
| 980 |
txt += ' echo "SET_CMS_ENV 0 ==> setup cms environment ok"\n'
|
| 981 |
txt += ' echo " END SETUP CMS OSG ENVIRONMENT "\n'
|
| 982 |
|
| 983 |
return txt
|
| 984 |
|
| 985 |
### OLI_DANIELE
|
| 986 |
def wsSetupCMSLCGEnvironment_(self):
|
| 987 |
"""
|
| 988 |
Returns part of a job script which is prepares
|
| 989 |
the execution environment and which is common for all CMS jobs.
|
| 990 |
"""
|
| 991 |
txt = ' \n'
|
| 992 |
txt += ' echo " ### SETUP CMS LCG ENVIRONMENT ### "\n'
|
| 993 |
txt += ' if [ ! $VO_CMS_SW_DIR ] ;then\n'
|
| 994 |
txt += ' echo "SET_CMS_ENV 10031 ==> ERROR CMS software dir not found on WN `hostname`"\n'
|
| 995 |
txt += ' echo "JOB_EXIT_STATUS = 10031" \n'
|
| 996 |
txt += ' echo "JobExitCode=10031" | tee -a $RUNTIME_AREA/$repo\n'
|
| 997 |
txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
|
| 998 |
txt += ' rm -f $RUNTIME_AREA/$repo \n'
|
| 999 |
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
|
| 1000 |
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
|
| 1001 |
txt += ' exit 1\n'
|
| 1002 |
txt += ' else\n'
|
| 1003 |
txt += ' echo "Sourcing environment... "\n'
|
| 1004 |
txt += ' if [ ! -s $VO_CMS_SW_DIR/cmsset_default.sh ] ;then\n'
|
| 1005 |
txt += ' echo "SET_CMS_ENV 10020 ==> ERROR cmsset_default.sh file not found into dir $VO_CMS_SW_DIR"\n'
|
| 1006 |
txt += ' echo "JOB_EXIT_STATUS = 10020"\n'
|
| 1007 |
txt += ' echo "JobExitCode=10020" | tee -a $RUNTIME_AREA/$repo\n'
|
| 1008 |
txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
|
| 1009 |
txt += ' rm -f $RUNTIME_AREA/$repo \n'
|
| 1010 |
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
|
| 1011 |
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
|
| 1012 |
txt += ' exit 1\n'
|
| 1013 |
txt += ' fi\n'
|
| 1014 |
txt += ' echo "sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
|
| 1015 |
txt += ' source $VO_CMS_SW_DIR/cmsset_default.sh\n'
|
| 1016 |
txt += ' result=$?\n'
|
| 1017 |
txt += ' if [ $result -ne 0 ]; then\n'
|
| 1018 |
txt += ' echo "SET_CMS_ENV 10032 ==> ERROR problem sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
|
| 1019 |
txt += ' echo "JOB_EXIT_STATUS = 10032"\n'
|
| 1020 |
txt += ' echo "JobExitCode=10032" | tee -a $RUNTIME_AREA/$repo\n'
|
| 1021 |
txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
|
| 1022 |
txt += ' rm -f $RUNTIME_AREA/$repo \n'
|
| 1023 |
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
|
| 1024 |
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
|
| 1025 |
txt += ' exit 1\n'
|
| 1026 |
txt += ' fi\n'
|
| 1027 |
txt += ' fi\n'
|
| 1028 |
txt += ' \n'
|
| 1029 |
txt += ' string=`cat /etc/redhat-release`\n'
|
| 1030 |
txt += ' echo $string\n'
|
| 1031 |
txt += ' if [[ $string = *alhalla* ]]; then\n'
|
| 1032 |
txt += ' echo "SCRAM_ARCH= $SCRAM_ARCH"\n'
|
| 1033 |
txt += ' elif [[ $string = *Enterprise* ]] || [[ $string = *cientific* ]]; then\n'
|
| 1034 |
txt += ' export SCRAM_ARCH=slc3_ia32_gcc323\n'
|
| 1035 |
txt += ' echo "SCRAM_ARCH= $SCRAM_ARCH"\n'
|
| 1036 |
txt += ' else\n'
|
| 1037 |
txt += ' echo "SET_CMS_ENV 10033 ==> ERROR OS unknown, LCG environment not initialized"\n'
|
| 1038 |
txt += ' echo "JOB_EXIT_STATUS = 10033"\n'
|
| 1039 |
txt += ' echo "JobExitCode=10033" | tee -a $RUNTIME_AREA/$repo\n'
|
| 1040 |
txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
|
| 1041 |
txt += ' rm -f $RUNTIME_AREA/$repo \n'
|
| 1042 |
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
|
| 1043 |
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
|
| 1044 |
txt += ' exit 1\n'
|
| 1045 |
txt += ' fi\n'
|
| 1046 |
txt += ' echo "SET_CMS_ENV 0 ==> setup cms environment ok"\n'
|
| 1047 |
txt += ' echo "### END SETUP CMS LCG ENVIRONMENT ###"\n'
|
| 1048 |
return txt
|
| 1049 |
|
