| 1 |
slacapra |
1.1 |
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 Scram
|
| 7 |
|
|
|
| 8 |
slacapra |
1.70 |
import os, string, re, shutil, glob
|
| 9 |
slacapra |
1.1 |
|
| 10 |
|
|
class Cmssw(JobType):
|
| 11 |
gutsche |
1.38 |
def __init__(self, cfg_params, ncjobs):
|
| 12 |
slacapra |
1.1 |
JobType.__init__(self, 'CMSSW')
|
| 13 |
|
|
common.logger.debug(3,'CMSSW::__init__')
|
| 14 |
|
|
|
| 15 |
gutsche |
1.3 |
self._params = {}
|
| 16 |
|
|
self.cfg_params = cfg_params
|
| 17 |
gutsche |
1.38 |
|
| 18 |
gutsche |
1.72 |
try:
|
| 19 |
|
|
self.MaxTarBallSize = float(self.cfg_params['EDG.maxtarballsize'])
|
| 20 |
|
|
except KeyError:
|
| 21 |
slacapra |
1.86 |
self.MaxTarBallSize = 9.5
|
| 22 |
gutsche |
1.72 |
|
| 23 |
gutsche |
1.44 |
# number of jobs requested to be created, limit obj splitting
|
| 24 |
gutsche |
1.38 |
self.ncjobs = ncjobs
|
| 25 |
|
|
|
| 26 |
slacapra |
1.1 |
log = common.logger
|
| 27 |
|
|
|
| 28 |
|
|
self.scram = Scram.Scram(cfg_params)
|
| 29 |
|
|
self.additional_inbox_files = []
|
| 30 |
|
|
self.scriptExe = ''
|
| 31 |
|
|
self.executable = ''
|
| 32 |
slacapra |
1.71 |
self.executable_arch = self.scram.getArch()
|
| 33 |
slacapra |
1.1 |
self.tgz_name = 'default.tgz'
|
| 34 |
slacapra |
1.97 |
self.additional_tgz_name = 'additional.tgz'
|
| 35 |
corvo |
1.56 |
self.scriptName = 'CMSSW.sh'
|
| 36 |
spiga |
1.42 |
self.pset = '' #scrip use case Da
|
| 37 |
|
|
self.datasetPath = '' #scrip use case Da
|
| 38 |
gutsche |
1.3 |
|
| 39 |
gutsche |
1.50 |
# set FJR file name
|
| 40 |
|
|
self.fjrFileName = 'crab_fjr.xml'
|
| 41 |
|
|
|
| 42 |
slacapra |
1.1 |
self.version = self.scram.getSWVersion()
|
| 43 |
slacapra |
1.55 |
common.taskDB.setDict('codeVersion',self.version)
|
| 44 |
gutsche |
1.5 |
self.setParam_('application', self.version)
|
| 45 |
slacapra |
1.47 |
|
| 46 |
slacapra |
1.1 |
### collect Data cards
|
| 47 |
gutsche |
1.66 |
|
| 48 |
|
|
## get DBS mode
|
| 49 |
|
|
try:
|
| 50 |
slacapra |
1.86 |
self.use_dbs_1 = int(self.cfg_params['CMSSW.use_dbs_1'])
|
| 51 |
gutsche |
1.66 |
except KeyError:
|
| 52 |
slacapra |
1.86 |
self.use_dbs_1 = 0
|
| 53 |
gutsche |
1.66 |
|
| 54 |
slacapra |
1.1 |
try:
|
| 55 |
slacapra |
1.9 |
tmp = cfg_params['CMSSW.datasetpath']
|
| 56 |
|
|
log.debug(6, "CMSSW::CMSSW(): datasetPath = "+tmp)
|
| 57 |
|
|
if string.lower(tmp)=='none':
|
| 58 |
|
|
self.datasetPath = None
|
| 59 |
slacapra |
1.21 |
self.selectNoInput = 1
|
| 60 |
slacapra |
1.9 |
else:
|
| 61 |
|
|
self.datasetPath = tmp
|
| 62 |
slacapra |
1.21 |
self.selectNoInput = 0
|
| 63 |
slacapra |
1.1 |
except KeyError:
|
| 64 |
gutsche |
1.3 |
msg = "Error: datasetpath not defined "
|
| 65 |
slacapra |
1.1 |
raise CrabException(msg)
|
| 66 |
gutsche |
1.5 |
|
| 67 |
|
|
# ML monitoring
|
| 68 |
|
|
# split dataset path style: /PreProdR3Minbias/SIM/GEN-SIM
|
| 69 |
slacapra |
1.9 |
if not self.datasetPath:
|
| 70 |
|
|
self.setParam_('dataset', 'None')
|
| 71 |
|
|
self.setParam_('owner', 'None')
|
| 72 |
|
|
else:
|
| 73 |
gutsche |
1.92 |
try:
|
| 74 |
|
|
datasetpath_split = self.datasetPath.split("/")
|
| 75 |
|
|
# standard style
|
| 76 |
|
|
if self.use_dbs_1 == 1 :
|
| 77 |
|
|
self.setParam_('dataset', datasetpath_split[1])
|
| 78 |
|
|
self.setParam_('owner', datasetpath_split[-1])
|
| 79 |
|
|
else:
|
| 80 |
|
|
self.setParam_('dataset', datasetpath_split[1])
|
| 81 |
|
|
self.setParam_('owner', datasetpath_split[2])
|
| 82 |
|
|
except:
|
| 83 |
|
|
self.setParam_('dataset', self.datasetPath)
|
| 84 |
|
|
self.setParam_('owner', self.datasetPath)
|
| 85 |
|
|
|
| 86 |
gutsche |
1.8 |
self.setTaskid_()
|
| 87 |
|
|
self.setParam_('taskId', self.cfg_params['taskId'])
|
| 88 |
gutsche |
1.5 |
|
| 89 |
slacapra |
1.1 |
self.dataTiers = []
|
| 90 |
|
|
|
| 91 |
|
|
## now the application
|
| 92 |
|
|
try:
|
| 93 |
|
|
self.executable = cfg_params['CMSSW.executable']
|
| 94 |
gutsche |
1.5 |
self.setParam_('exe', self.executable)
|
| 95 |
slacapra |
1.1 |
log.debug(6, "CMSSW::CMSSW(): executable = "+self.executable)
|
| 96 |
|
|
msg = "Default executable cmsRun overridden. Switch to " + self.executable
|
| 97 |
|
|
log.debug(3,msg)
|
| 98 |
|
|
except KeyError:
|
| 99 |
|
|
self.executable = 'cmsRun'
|
| 100 |
gutsche |
1.5 |
self.setParam_('exe', self.executable)
|
| 101 |
slacapra |
1.1 |
msg = "User executable not defined. Use cmsRun"
|
| 102 |
|
|
log.debug(3,msg)
|
| 103 |
|
|
pass
|
| 104 |
|
|
|
| 105 |
|
|
try:
|
| 106 |
|
|
self.pset = cfg_params['CMSSW.pset']
|
| 107 |
|
|
log.debug(6, "Cmssw::Cmssw(): PSet file = "+self.pset)
|
| 108 |
spiga |
1.42 |
if self.pset.lower() != 'none' :
|
| 109 |
|
|
if (not os.path.exists(self.pset)):
|
| 110 |
|
|
raise CrabException("User defined PSet file "+self.pset+" does not exist")
|
| 111 |
|
|
else:
|
| 112 |
|
|
self.pset = None
|
| 113 |
slacapra |
1.1 |
except KeyError:
|
| 114 |
|
|
raise CrabException("PSet file missing. Cannot run cmsRun ")
|
| 115 |
|
|
|
| 116 |
|
|
# output files
|
| 117 |
slacapra |
1.53 |
## stuff which must be returned always via sandbox
|
| 118 |
|
|
self.output_file_sandbox = []
|
| 119 |
|
|
|
| 120 |
|
|
# add fjr report by default via sandbox
|
| 121 |
|
|
self.output_file_sandbox.append(self.fjrFileName)
|
| 122 |
|
|
|
| 123 |
|
|
# other output files to be returned via sandbox or copied to SE
|
| 124 |
slacapra |
1.1 |
try:
|
| 125 |
|
|
self.output_file = []
|
| 126 |
|
|
tmp = cfg_params['CMSSW.output_file']
|
| 127 |
|
|
if tmp != '':
|
| 128 |
|
|
tmpOutFiles = string.split(cfg_params['CMSSW.output_file'],',')
|
| 129 |
|
|
log.debug(7, 'cmssw::cmssw(): output files '+str(tmpOutFiles))
|
| 130 |
|
|
for tmp in tmpOutFiles:
|
| 131 |
|
|
tmp=string.strip(tmp)
|
| 132 |
|
|
self.output_file.append(tmp)
|
| 133 |
|
|
pass
|
| 134 |
|
|
else:
|
| 135 |
gutsche |
1.92 |
log.message("No output file defined: only stdout/err and the CRAB Framework Job Report will be available\n")
|
| 136 |
slacapra |
1.1 |
pass
|
| 137 |
|
|
pass
|
| 138 |
|
|
except KeyError:
|
| 139 |
gutsche |
1.92 |
log.message("No output file defined: only stdout/err and the CRAB Framework Job Report will be available\n")
|
| 140 |
slacapra |
1.1 |
pass
|
| 141 |
|
|
|
| 142 |
|
|
# script_exe file as additional file in inputSandbox
|
| 143 |
|
|
try:
|
| 144 |
slacapra |
1.10 |
self.scriptExe = cfg_params['USER.script_exe']
|
| 145 |
|
|
if self.scriptExe != '':
|
| 146 |
|
|
if not os.path.isfile(self.scriptExe):
|
| 147 |
slacapra |
1.64 |
msg ="ERROR. file "+self.scriptExe+" not found"
|
| 148 |
slacapra |
1.10 |
raise CrabException(msg)
|
| 149 |
spiga |
1.42 |
self.additional_inbox_files.append(string.strip(self.scriptExe))
|
| 150 |
slacapra |
1.1 |
except KeyError:
|
| 151 |
spiga |
1.42 |
self.scriptExe = ''
|
| 152 |
slacapra |
1.70 |
|
| 153 |
spiga |
1.42 |
#CarlosDaniele
|
| 154 |
|
|
if self.datasetPath == None and self.pset == None and self.scriptExe == '' :
|
| 155 |
slacapra |
1.70 |
msg ="Error. script_exe not defined"
|
| 156 |
spiga |
1.42 |
raise CrabException(msg)
|
| 157 |
|
|
|
| 158 |
slacapra |
1.1 |
## additional input files
|
| 159 |
|
|
try:
|
| 160 |
slacapra |
1.29 |
tmpAddFiles = string.split(cfg_params['USER.additional_input_files'],',')
|
| 161 |
slacapra |
1.70 |
for tmp in tmpAddFiles:
|
| 162 |
|
|
tmp = string.strip(tmp)
|
| 163 |
|
|
dirname = ''
|
| 164 |
|
|
if not tmp[0]=="/": dirname = "."
|
| 165 |
corvo |
1.85 |
files = []
|
| 166 |
|
|
if string.find(tmp,"*")>-1:
|
| 167 |
|
|
files = glob.glob(os.path.join(dirname, tmp))
|
| 168 |
|
|
if len(files)==0:
|
| 169 |
|
|
raise CrabException("No additional input file found with this pattern: "+tmp)
|
| 170 |
|
|
else:
|
| 171 |
|
|
files.append(tmp)
|
| 172 |
slacapra |
1.70 |
for file in files:
|
| 173 |
|
|
if not os.path.exists(file):
|
| 174 |
|
|
raise CrabException("Additional input file not found: "+file)
|
| 175 |
slacapra |
1.45 |
pass
|
| 176 |
corvo |
1.85 |
fname = string.split(file, '/')[-1]
|
| 177 |
|
|
storedFile = common.work_space.pathForTgz()+'share/'+fname
|
| 178 |
slacapra |
1.70 |
shutil.copyfile(file, storedFile)
|
| 179 |
|
|
self.additional_inbox_files.append(string.strip(storedFile))
|
| 180 |
slacapra |
1.1 |
pass
|
| 181 |
|
|
pass
|
| 182 |
slacapra |
1.70 |
common.logger.debug(5,"Additional input files: "+str(self.additional_inbox_files))
|
| 183 |
slacapra |
1.1 |
except KeyError:
|
| 184 |
|
|
pass
|
| 185 |
|
|
|
| 186 |
slacapra |
1.9 |
# files per job
|
| 187 |
slacapra |
1.1 |
try:
|
| 188 |
gutsche |
1.35 |
if (cfg_params['CMSSW.files_per_jobs']):
|
| 189 |
|
|
raise CrabException("files_per_jobs no longer supported. Quitting.")
|
| 190 |
gutsche |
1.3 |
except KeyError:
|
| 191 |
gutsche |
1.35 |
pass
|
| 192 |
gutsche |
1.3 |
|
| 193 |
slacapra |
1.9 |
## Events per job
|
| 194 |
gutsche |
1.3 |
try:
|
| 195 |
slacapra |
1.10 |
self.eventsPerJob =int( cfg_params['CMSSW.events_per_job'])
|
| 196 |
slacapra |
1.9 |
self.selectEventsPerJob = 1
|
| 197 |
gutsche |
1.3 |
except KeyError:
|
| 198 |
slacapra |
1.9 |
self.eventsPerJob = -1
|
| 199 |
|
|
self.selectEventsPerJob = 0
|
| 200 |
|
|
|
| 201 |
slacapra |
1.22 |
## number of jobs
|
| 202 |
|
|
try:
|
| 203 |
|
|
self.theNumberOfJobs =int( cfg_params['CMSSW.number_of_jobs'])
|
| 204 |
|
|
self.selectNumberOfJobs = 1
|
| 205 |
|
|
except KeyError:
|
| 206 |
|
|
self.theNumberOfJobs = 0
|
| 207 |
|
|
self.selectNumberOfJobs = 0
|
| 208 |
slacapra |
1.10 |
|
| 209 |
gutsche |
1.35 |
try:
|
| 210 |
|
|
self.total_number_of_events = int(cfg_params['CMSSW.total_number_of_events'])
|
| 211 |
|
|
self.selectTotalNumberEvents = 1
|
| 212 |
|
|
except KeyError:
|
| 213 |
|
|
self.total_number_of_events = 0
|
| 214 |
|
|
self.selectTotalNumberEvents = 0
|
| 215 |
|
|
|
| 216 |
spiga |
1.42 |
if self.pset != None: #CarlosDaniele
|
| 217 |
|
|
if ( (self.selectTotalNumberEvents + self.selectEventsPerJob + self.selectNumberOfJobs) != 2 ):
|
| 218 |
|
|
msg = 'Must define exactly two of total_number_of_events, events_per_job, or number_of_jobs.'
|
| 219 |
|
|
raise CrabException(msg)
|
| 220 |
|
|
else:
|
| 221 |
|
|
if (self.selectNumberOfJobs == 0):
|
| 222 |
|
|
msg = 'Must specify number_of_jobs.'
|
| 223 |
|
|
raise CrabException(msg)
|
| 224 |
gutsche |
1.35 |
|
| 225 |
slacapra |
1.22 |
## source seed for pythia
|
| 226 |
|
|
try:
|
| 227 |
|
|
self.sourceSeed = int(cfg_params['CMSSW.pythia_seed'])
|
| 228 |
|
|
except KeyError:
|
| 229 |
slacapra |
1.23 |
self.sourceSeed = None
|
| 230 |
|
|
common.logger.debug(5,"No seed given")
|
| 231 |
slacapra |
1.22 |
|
| 232 |
slacapra |
1.28 |
try:
|
| 233 |
|
|
self.sourceSeedVtx = int(cfg_params['CMSSW.vtx_seed'])
|
| 234 |
|
|
except KeyError:
|
| 235 |
|
|
self.sourceSeedVtx = None
|
| 236 |
|
|
common.logger.debug(5,"No vertex seed given")
|
| 237 |
slacapra |
1.90 |
|
| 238 |
|
|
try:
|
| 239 |
|
|
self.sourceSeedG4 = int(cfg_params['CMSSW.g4_seed'])
|
| 240 |
|
|
except KeyError:
|
| 241 |
|
|
self.sourceSeedG4 = None
|
| 242 |
|
|
common.logger.debug(5,"No g4 sim hits seed given")
|
| 243 |
|
|
|
| 244 |
|
|
try:
|
| 245 |
|
|
self.sourceSeedMix = int(cfg_params['CMSSW.mix_seed'])
|
| 246 |
|
|
except KeyError:
|
| 247 |
|
|
self.sourceSeedMix = None
|
| 248 |
|
|
common.logger.debug(5,"No mix seed given")
|
| 249 |
|
|
|
| 250 |
spiga |
1.57 |
try:
|
| 251 |
|
|
self.firstRun = int(cfg_params['CMSSW.first_run'])
|
| 252 |
|
|
except KeyError:
|
| 253 |
|
|
self.firstRun = None
|
| 254 |
|
|
common.logger.debug(5,"No first run given")
|
| 255 |
spiga |
1.42 |
if self.pset != None: #CarlosDaniele
|
| 256 |
slacapra |
1.97 |
ver = string.split(self.version,"_")
|
| 257 |
|
|
if (int(ver[1])>=1 and int(ver[2])>=5):
|
| 258 |
|
|
import PsetManipulator150 as pp
|
| 259 |
|
|
else:
|
| 260 |
|
|
import PsetManipulator as pp
|
| 261 |
|
|
PsetEdit = pp.PsetManipulator(self.pset) #Daniele Pset
|
| 262 |
gutsche |
1.3 |
|
| 263 |
slacapra |
1.1 |
#DBSDLS-start
|
| 264 |
|
|
## Initialize the variables that are extracted from DBS/DLS and needed in other places of the code
|
| 265 |
|
|
self.maxEvents=0 # max events available ( --> check the requested nb. of evts in Creator.py)
|
| 266 |
|
|
self.DBSPaths={} # all dbs paths requested ( --> input to the site local discovery script)
|
| 267 |
gutsche |
1.35 |
self.jobDestination=[] # Site destination(s) for each job (list of lists)
|
| 268 |
slacapra |
1.1 |
## Perform the data location and discovery (based on DBS/DLS)
|
| 269 |
slacapra |
1.9 |
## SL: Don't if NONE is specified as input (pythia use case)
|
| 270 |
gutsche |
1.35 |
blockSites = {}
|
| 271 |
slacapra |
1.9 |
if self.datasetPath:
|
| 272 |
gutsche |
1.35 |
blockSites = self.DataDiscoveryAndLocation(cfg_params)
|
| 273 |
slacapra |
1.1 |
#DBSDLS-end
|
| 274 |
|
|
|
| 275 |
|
|
self.tgzNameWithPath = self.getTarBall(self.executable)
|
| 276 |
slacapra |
1.10 |
|
| 277 |
slacapra |
1.9 |
## Select Splitting
|
| 278 |
spiga |
1.42 |
if self.selectNoInput:
|
| 279 |
|
|
if self.pset == None: #CarlosDaniele
|
| 280 |
|
|
self.jobSplittingForScript()
|
| 281 |
|
|
else:
|
| 282 |
|
|
self.jobSplittingNoInput()
|
| 283 |
gutsche |
1.92 |
else:
|
| 284 |
corvo |
1.56 |
self.jobSplittingByBlocks(blockSites)
|
| 285 |
gutsche |
1.5 |
|
| 286 |
slacapra |
1.22 |
# modify Pset
|
| 287 |
spiga |
1.42 |
if self.pset != None: #CarlosDaniele
|
| 288 |
slacapra |
1.86 |
try:
|
| 289 |
|
|
if (self.datasetPath): # standard job
|
| 290 |
|
|
# allow to processa a fraction of events in a file
|
| 291 |
slacapra |
1.90 |
PsetEdit.inputModule("INPUT")
|
| 292 |
|
|
PsetEdit.maxEvent("INPUTMAXEVENTS")
|
| 293 |
|
|
PsetEdit.skipEvent("INPUTSKIPEVENTS")
|
| 294 |
slacapra |
1.86 |
else: # pythia like job
|
| 295 |
slacapra |
1.90 |
PsetEdit.maxEvent(self.eventsPerJob)
|
| 296 |
slacapra |
1.86 |
if (self.firstRun):
|
| 297 |
slacapra |
1.90 |
PsetEdit.pythiaFirstRun("INPUTFIRSTRUN") #First Run
|
| 298 |
slacapra |
1.86 |
if (self.sourceSeed) :
|
| 299 |
slacapra |
1.90 |
PsetEdit.pythiaSeed("INPUT")
|
| 300 |
slacapra |
1.86 |
if (self.sourceSeedVtx) :
|
| 301 |
slacapra |
1.90 |
PsetEdit.vtxSeed("INPUTVTX")
|
| 302 |
|
|
if (self.sourceSeedG4) :
|
| 303 |
|
|
self.PsetEdit.g4Seed("INPUTG4")
|
| 304 |
|
|
if (self.sourceSeedMix) :
|
| 305 |
|
|
self.PsetEdit.mixSeed("INPUTMIX")
|
| 306 |
slacapra |
1.86 |
# add FrameworkJobReport to parameter-set
|
| 307 |
slacapra |
1.90 |
PsetEdit.addCrabFJR(self.fjrFileName)
|
| 308 |
|
|
PsetEdit.psetWriter(self.configFilename())
|
| 309 |
slacapra |
1.86 |
except:
|
| 310 |
|
|
msg='Error while manipuliating ParameterSet: exiting...'
|
| 311 |
|
|
raise CrabException(msg)
|
| 312 |
gutsche |
1.3 |
|
| 313 |
slacapra |
1.1 |
def DataDiscoveryAndLocation(self, cfg_params):
|
| 314 |
|
|
|
| 315 |
slacapra |
1.86 |
import DataDiscovery
|
| 316 |
|
|
import DataDiscovery_DBS2
|
| 317 |
|
|
import DataLocation
|
| 318 |
gutsche |
1.3 |
common.logger.debug(10,"CMSSW::DataDiscoveryAndLocation()")
|
| 319 |
|
|
|
| 320 |
|
|
datasetPath=self.datasetPath
|
| 321 |
|
|
|
| 322 |
slacapra |
1.1 |
## Contact the DBS
|
| 323 |
gutsche |
1.92 |
common.logger.message("Contacting Data Discovery Services ...")
|
| 324 |
slacapra |
1.1 |
try:
|
| 325 |
gutsche |
1.66 |
|
| 326 |
slacapra |
1.86 |
if self.use_dbs_1 == 1 :
|
| 327 |
|
|
self.pubdata=DataDiscovery.DataDiscovery(datasetPath, cfg_params)
|
| 328 |
|
|
else :
|
| 329 |
corvo |
1.85 |
self.pubdata=DataDiscovery_DBS2.DataDiscovery_DBS2(datasetPath, cfg_params)
|
| 330 |
slacapra |
1.1 |
self.pubdata.fetchDBSInfo()
|
| 331 |
|
|
|
| 332 |
slacapra |
1.41 |
except DataDiscovery.NotExistingDatasetError, ex :
|
| 333 |
slacapra |
1.1 |
msg = 'ERROR ***: failed Data Discovery in DBS : %s'%ex.getErrorMessage()
|
| 334 |
|
|
raise CrabException(msg)
|
| 335 |
slacapra |
1.41 |
except DataDiscovery.NoDataTierinProvenanceError, ex :
|
| 336 |
slacapra |
1.1 |
msg = 'ERROR ***: failed Data Discovery in DBS : %s'%ex.getErrorMessage()
|
| 337 |
|
|
raise CrabException(msg)
|
| 338 |
slacapra |
1.41 |
except DataDiscovery.DataDiscoveryError, ex:
|
| 339 |
gutsche |
1.66 |
msg = 'ERROR ***: failed Data Discovery in DBS : %s'%ex.getErrorMessage()
|
| 340 |
slacapra |
1.1 |
raise CrabException(msg)
|
| 341 |
gutsche |
1.67 |
except DataDiscovery_DBS2.NotExistingDatasetError_DBS2, ex :
|
| 342 |
|
|
msg = 'ERROR ***: failed Data Discovery in DBS : %s'%ex.getErrorMessage()
|
| 343 |
|
|
raise CrabException(msg)
|
| 344 |
|
|
except DataDiscovery_DBS2.NoDataTierinProvenanceError_DBS2, ex :
|
| 345 |
|
|
msg = 'ERROR ***: failed Data Discovery in DBS : %s'%ex.getErrorMessage()
|
| 346 |
|
|
raise CrabException(msg)
|
| 347 |
|
|
except DataDiscovery_DBS2.DataDiscoveryError_DBS2, ex:
|
| 348 |
|
|
msg = 'ERROR ***: failed Data Discovery in DBS : %s'%ex.getErrorMessage()
|
| 349 |
|
|
raise CrabException(msg)
|
| 350 |
slacapra |
1.1 |
|
| 351 |
gutsche |
1.35 |
self.filesbyblock=self.pubdata.getFiles()
|
| 352 |
mkirn |
1.37 |
self.eventsbyblock=self.pubdata.getEventsPerBlock()
|
| 353 |
|
|
self.eventsbyfile=self.pubdata.getEventsPerFile()
|
| 354 |
gutsche |
1.3 |
|
| 355 |
slacapra |
1.1 |
## get max number of events
|
| 356 |
|
|
self.maxEvents=self.pubdata.getMaxEvents() ## self.maxEvents used in Creator.py
|
| 357 |
|
|
|
| 358 |
|
|
## Contact the DLS and build a list of sites hosting the fileblocks
|
| 359 |
|
|
try:
|
| 360 |
slacapra |
1.41 |
dataloc=DataLocation.DataLocation(self.filesbyblock.keys(),cfg_params)
|
| 361 |
gutsche |
1.6 |
dataloc.fetchDLSInfo()
|
| 362 |
slacapra |
1.41 |
except DataLocation.DataLocationError , ex:
|
| 363 |
slacapra |
1.1 |
msg = 'ERROR ***: failed Data Location in DLS \n %s '%ex.getErrorMessage()
|
| 364 |
|
|
raise CrabException(msg)
|
| 365 |
|
|
|
| 366 |
|
|
|
| 367 |
gutsche |
1.35 |
sites = dataloc.getSites()
|
| 368 |
|
|
allSites = []
|
| 369 |
|
|
listSites = sites.values()
|
| 370 |
slacapra |
1.63 |
for listSite in listSites:
|
| 371 |
|
|
for oneSite in listSite:
|
| 372 |
gutsche |
1.35 |
allSites.append(oneSite)
|
| 373 |
|
|
allSites = self.uniquelist(allSites)
|
| 374 |
gutsche |
1.3 |
|
| 375 |
gutsche |
1.92 |
# screen output
|
| 376 |
|
|
common.logger.message("Requested dataset: " + datasetPath + " has " + str(self.maxEvents) + " events in " + str(len(self.filesbyblock.keys())) + " blocks.\n")
|
| 377 |
|
|
|
| 378 |
gutsche |
1.35 |
return sites
|
| 379 |
gutsche |
1.3 |
|
| 380 |
gutsche |
1.35 |
def jobSplittingByBlocks(self, blockSites):
|
| 381 |
slacapra |
1.9 |
"""
|
| 382 |
gutsche |
1.35 |
Perform job splitting. Jobs run over an integer number of files
|
| 383 |
|
|
and no more than one block.
|
| 384 |
|
|
ARGUMENT: blockSites: dictionary with blocks as keys and list of host sites as values
|
| 385 |
|
|
REQUIRES: self.selectTotalNumberEvents, self.selectEventsPerJob, self.selectNumberofJobs,
|
| 386 |
|
|
self.total_number_of_events, self.eventsPerJob, self.theNumberOfJobs,
|
| 387 |
|
|
self.maxEvents, self.filesbyblock
|
| 388 |
|
|
SETS: self.jobDestination - Site destination(s) for each job (a list of lists)
|
| 389 |
|
|
self.total_number_of_jobs - Total # of jobs
|
| 390 |
|
|
self.list_of_args - File(s) job will run on (a list of lists)
|
| 391 |
|
|
"""
|
| 392 |
|
|
|
| 393 |
|
|
# ---- Handle the possible job splitting configurations ---- #
|
| 394 |
|
|
if (self.selectTotalNumberEvents):
|
| 395 |
|
|
totalEventsRequested = self.total_number_of_events
|
| 396 |
|
|
if (self.selectEventsPerJob):
|
| 397 |
|
|
eventsPerJobRequested = self.eventsPerJob
|
| 398 |
|
|
if (self.selectNumberOfJobs):
|
| 399 |
|
|
totalEventsRequested = self.theNumberOfJobs * self.eventsPerJob
|
| 400 |
|
|
|
| 401 |
|
|
# If user requested all the events in the dataset
|
| 402 |
|
|
if (totalEventsRequested == -1):
|
| 403 |
|
|
eventsRemaining=self.maxEvents
|
| 404 |
|
|
# If user requested more events than are in the dataset
|
| 405 |
|
|
elif (totalEventsRequested > self.maxEvents):
|
| 406 |
|
|
eventsRemaining = self.maxEvents
|
| 407 |
|
|
common.logger.message("Requested "+str(self.total_number_of_events)+ " events, but only "+str(self.maxEvents)+" events are available.")
|
| 408 |
|
|
# If user requested less events than are in the dataset
|
| 409 |
|
|
else:
|
| 410 |
|
|
eventsRemaining = totalEventsRequested
|
| 411 |
slacapra |
1.22 |
|
| 412 |
slacapra |
1.41 |
# If user requested more events per job than are in the dataset
|
| 413 |
|
|
if (self.selectEventsPerJob and eventsPerJobRequested > self.maxEvents):
|
| 414 |
|
|
eventsPerJobRequested = self.maxEvents
|
| 415 |
|
|
|
| 416 |
gutsche |
1.35 |
# For user info at end
|
| 417 |
|
|
totalEventCount = 0
|
| 418 |
gutsche |
1.3 |
|
| 419 |
gutsche |
1.35 |
if (self.selectTotalNumberEvents and self.selectNumberOfJobs):
|
| 420 |
|
|
eventsPerJobRequested = int(eventsRemaining/self.theNumberOfJobs)
|
| 421 |
slacapra |
1.22 |
|
| 422 |
gutsche |
1.35 |
if (self.selectNumberOfJobs):
|
| 423 |
|
|
common.logger.message("May not create the exact number_of_jobs requested.")
|
| 424 |
slacapra |
1.23 |
|
| 425 |
gutsche |
1.38 |
if ( self.ncjobs == 'all' ) :
|
| 426 |
|
|
totalNumberOfJobs = 999999999
|
| 427 |
|
|
else :
|
| 428 |
|
|
totalNumberOfJobs = self.ncjobs
|
| 429 |
|
|
|
| 430 |
|
|
|
| 431 |
gutsche |
1.35 |
blocks = blockSites.keys()
|
| 432 |
|
|
blockCount = 0
|
| 433 |
|
|
# Backup variable in case self.maxEvents counted events in a non-included block
|
| 434 |
|
|
numBlocksInDataset = len(blocks)
|
| 435 |
gutsche |
1.3 |
|
| 436 |
gutsche |
1.35 |
jobCount = 0
|
| 437 |
|
|
list_of_lists = []
|
| 438 |
gutsche |
1.3 |
|
| 439 |
gutsche |
1.92 |
# list tracking which jobs are in which jobs belong to which block
|
| 440 |
|
|
jobsOfBlock = {}
|
| 441 |
|
|
|
| 442 |
gutsche |
1.35 |
# ---- Iterate over the blocks in the dataset until ---- #
|
| 443 |
|
|
# ---- we've met the requested total # of events ---- #
|
| 444 |
gutsche |
1.38 |
while ( (eventsRemaining > 0) and (blockCount < numBlocksInDataset) and (jobCount < totalNumberOfJobs)):
|
| 445 |
gutsche |
1.35 |
block = blocks[blockCount]
|
| 446 |
gutsche |
1.44 |
blockCount += 1
|
| 447 |
|
|
|
| 448 |
gutsche |
1.68 |
if self.eventsbyblock.has_key(block) :
|
| 449 |
|
|
numEventsInBlock = self.eventsbyblock[block]
|
| 450 |
|
|
common.logger.debug(5,'Events in Block File '+str(numEventsInBlock))
|
| 451 |
slacapra |
1.9 |
|
| 452 |
gutsche |
1.68 |
files = self.filesbyblock[block]
|
| 453 |
|
|
numFilesInBlock = len(files)
|
| 454 |
|
|
if (numFilesInBlock <= 0):
|
| 455 |
|
|
continue
|
| 456 |
|
|
fileCount = 0
|
| 457 |
|
|
|
| 458 |
|
|
# ---- New block => New job ---- #
|
| 459 |
|
|
parString = "\\{"
|
| 460 |
|
|
# counter for number of events in files currently worked on
|
| 461 |
|
|
filesEventCount = 0
|
| 462 |
|
|
# flag if next while loop should touch new file
|
| 463 |
|
|
newFile = 1
|
| 464 |
|
|
# job event counter
|
| 465 |
|
|
jobSkipEventCount = 0
|
| 466 |
slacapra |
1.9 |
|
| 467 |
gutsche |
1.68 |
# ---- Iterate over the files in the block until we've met the requested ---- #
|
| 468 |
|
|
# ---- total # of events or we've gone over all the files in this block ---- #
|
| 469 |
|
|
while ( (eventsRemaining > 0) and (fileCount < numFilesInBlock) and (jobCount < totalNumberOfJobs) ):
|
| 470 |
|
|
file = files[fileCount]
|
| 471 |
|
|
if newFile :
|
| 472 |
|
|
try:
|
| 473 |
|
|
numEventsInFile = self.eventsbyfile[file]
|
| 474 |
|
|
common.logger.debug(6, "File "+str(file)+" has "+str(numEventsInFile)+" events")
|
| 475 |
|
|
# increase filesEventCount
|
| 476 |
|
|
filesEventCount += numEventsInFile
|
| 477 |
|
|
# Add file to current job
|
| 478 |
|
|
parString += '\\\"' + file + '\\\"\,'
|
| 479 |
|
|
newFile = 0
|
| 480 |
|
|
except KeyError:
|
| 481 |
|
|
common.logger.message("File "+str(file)+" has unknown number of events: skipping")
|
| 482 |
slacapra |
1.41 |
|
| 483 |
gutsche |
1.38 |
|
| 484 |
gutsche |
1.68 |
# if less events in file remain than eventsPerJobRequested
|
| 485 |
|
|
if ( filesEventCount - jobSkipEventCount < eventsPerJobRequested ) :
|
| 486 |
|
|
# if last file in block
|
| 487 |
|
|
if ( fileCount == numFilesInBlock-1 ) :
|
| 488 |
|
|
# end job using last file, use remaining events in block
|
| 489 |
|
|
# close job and touch new file
|
| 490 |
|
|
fullString = parString[:-2]
|
| 491 |
|
|
fullString += '\\}'
|
| 492 |
|
|
list_of_lists.append([fullString,str(-1),str(jobSkipEventCount)])
|
| 493 |
|
|
common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(filesEventCount - jobSkipEventCount)+" events (last file in block).")
|
| 494 |
|
|
self.jobDestination.append(blockSites[block])
|
| 495 |
|
|
common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
|
| 496 |
gutsche |
1.92 |
# fill jobs of block dictionary
|
| 497 |
|
|
if block in jobsOfBlock.keys() :
|
| 498 |
|
|
jobsOfBlock[block].append(jobCount+1)
|
| 499 |
|
|
else:
|
| 500 |
|
|
jobsOfBlock[block] = [jobCount+1]
|
| 501 |
gutsche |
1.68 |
# reset counter
|
| 502 |
|
|
jobCount = jobCount + 1
|
| 503 |
|
|
totalEventCount = totalEventCount + filesEventCount - jobSkipEventCount
|
| 504 |
|
|
eventsRemaining = eventsRemaining - filesEventCount + jobSkipEventCount
|
| 505 |
|
|
jobSkipEventCount = 0
|
| 506 |
|
|
# reset file
|
| 507 |
|
|
parString = "\\{"
|
| 508 |
|
|
filesEventCount = 0
|
| 509 |
|
|
newFile = 1
|
| 510 |
|
|
fileCount += 1
|
| 511 |
|
|
else :
|
| 512 |
|
|
# go to next file
|
| 513 |
|
|
newFile = 1
|
| 514 |
|
|
fileCount += 1
|
| 515 |
|
|
# if events in file equal to eventsPerJobRequested
|
| 516 |
|
|
elif ( filesEventCount - jobSkipEventCount == eventsPerJobRequested ) :
|
| 517 |
gutsche |
1.38 |
# close job and touch new file
|
| 518 |
|
|
fullString = parString[:-2]
|
| 519 |
|
|
fullString += '\\}'
|
| 520 |
gutsche |
1.68 |
list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
|
| 521 |
|
|
common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.")
|
| 522 |
gutsche |
1.38 |
self.jobDestination.append(blockSites[block])
|
| 523 |
|
|
common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
|
| 524 |
gutsche |
1.92 |
if block in jobsOfBlock.keys() :
|
| 525 |
|
|
jobsOfBlock[block].append(jobCount+1)
|
| 526 |
|
|
else:
|
| 527 |
|
|
jobsOfBlock[block] = [jobCount+1]
|
| 528 |
gutsche |
1.38 |
# reset counter
|
| 529 |
|
|
jobCount = jobCount + 1
|
| 530 |
gutsche |
1.68 |
totalEventCount = totalEventCount + eventsPerJobRequested
|
| 531 |
|
|
eventsRemaining = eventsRemaining - eventsPerJobRequested
|
| 532 |
gutsche |
1.38 |
jobSkipEventCount = 0
|
| 533 |
|
|
# reset file
|
| 534 |
|
|
parString = "\\{"
|
| 535 |
|
|
filesEventCount = 0
|
| 536 |
|
|
newFile = 1
|
| 537 |
|
|
fileCount += 1
|
| 538 |
gutsche |
1.68 |
|
| 539 |
|
|
# if more events in file remain than eventsPerJobRequested
|
| 540 |
gutsche |
1.38 |
else :
|
| 541 |
gutsche |
1.68 |
# close job but don't touch new file
|
| 542 |
|
|
fullString = parString[:-2]
|
| 543 |
|
|
fullString += '\\}'
|
| 544 |
|
|
list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
|
| 545 |
|
|
common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.")
|
| 546 |
|
|
self.jobDestination.append(blockSites[block])
|
| 547 |
|
|
common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
|
| 548 |
gutsche |
1.92 |
if block in jobsOfBlock.keys() :
|
| 549 |
|
|
jobsOfBlock[block].append(jobCount+1)
|
| 550 |
|
|
else:
|
| 551 |
|
|
jobsOfBlock[block] = [jobCount+1]
|
| 552 |
gutsche |
1.68 |
# increase counter
|
| 553 |
|
|
jobCount = jobCount + 1
|
| 554 |
|
|
totalEventCount = totalEventCount + eventsPerJobRequested
|
| 555 |
|
|
eventsRemaining = eventsRemaining - eventsPerJobRequested
|
| 556 |
|
|
# calculate skip events for last file
|
| 557 |
|
|
# use filesEventCount (contains several files), jobSkipEventCount and eventsPerJobRequest
|
| 558 |
|
|
jobSkipEventCount = eventsPerJobRequested - (filesEventCount - jobSkipEventCount - self.eventsbyfile[file])
|
| 559 |
|
|
# remove all but the last file
|
| 560 |
|
|
filesEventCount = self.eventsbyfile[file]
|
| 561 |
|
|
parString = "\\{"
|
| 562 |
|
|
parString += '\\\"' + file + '\\\"\,'
|
| 563 |
|
|
pass # END if
|
| 564 |
|
|
pass # END while (iterate over files in the block)
|
| 565 |
gutsche |
1.35 |
pass # END while (iterate over blocks in the dataset)
|
| 566 |
slacapra |
1.41 |
self.ncjobs = self.total_number_of_jobs = jobCount
|
| 567 |
gutsche |
1.38 |
if (eventsRemaining > 0 and jobCount < totalNumberOfJobs ):
|
| 568 |
gutsche |
1.35 |
common.logger.message("Could not run on all requested events because some blocks not hosted at allowed sites.")
|
| 569 |
gutsche |
1.92 |
common.logger.message(str(jobCount)+" job(s) can run on "+str(totalEventCount)+" events.\n")
|
| 570 |
slacapra |
1.22 |
|
| 571 |
gutsche |
1.92 |
# screen output
|
| 572 |
|
|
screenOutput = "List of jobs and available destination sites:\n\n"
|
| 573 |
|
|
|
| 574 |
|
|
blockCounter = 0
|
| 575 |
|
|
for block in jobsOfBlock.keys():
|
| 576 |
|
|
blockCounter += 1
|
| 577 |
|
|
screenOutput += "Block %5i: jobs %20s: sites: %s\n" % (blockCounter,spanRanges(jobsOfBlock[block]),','.join(blockSites[block]))
|
| 578 |
|
|
|
| 579 |
|
|
common.logger.message(screenOutput)
|
| 580 |
|
|
|
| 581 |
slacapra |
1.9 |
self.list_of_args = list_of_lists
|
| 582 |
|
|
return
|
| 583 |
|
|
|
| 584 |
slacapra |
1.21 |
def jobSplittingNoInput(self):
|
| 585 |
slacapra |
1.9 |
"""
|
| 586 |
|
|
Perform job splitting based on number of event per job
|
| 587 |
|
|
"""
|
| 588 |
|
|
common.logger.debug(5,'Splitting per events')
|
| 589 |
|
|
common.logger.message('Required '+str(self.eventsPerJob)+' events per job ')
|
| 590 |
slacapra |
1.22 |
common.logger.message('Required '+str(self.theNumberOfJobs)+' jobs in total ')
|
| 591 |
slacapra |
1.9 |
common.logger.message('Required '+str(self.total_number_of_events)+' events in total ')
|
| 592 |
|
|
|
| 593 |
slacapra |
1.10 |
if (self.total_number_of_events < 0):
|
| 594 |
|
|
msg='Cannot split jobs per Events with "-1" as total number of events'
|
| 595 |
|
|
raise CrabException(msg)
|
| 596 |
|
|
|
| 597 |
slacapra |
1.22 |
if (self.selectEventsPerJob):
|
| 598 |
spiga |
1.65 |
if (self.selectTotalNumberEvents):
|
| 599 |
|
|
self.total_number_of_jobs = int(self.total_number_of_events/self.eventsPerJob)
|
| 600 |
|
|
elif(self.selectNumberOfJobs) :
|
| 601 |
|
|
self.total_number_of_jobs =self.theNumberOfJobs
|
| 602 |
|
|
self.total_number_of_events =int(self.theNumberOfJobs*self.eventsPerJob)
|
| 603 |
|
|
|
| 604 |
slacapra |
1.22 |
elif (self.selectNumberOfJobs) :
|
| 605 |
|
|
self.total_number_of_jobs = self.theNumberOfJobs
|
| 606 |
|
|
self.eventsPerJob = int(self.total_number_of_events/self.total_number_of_jobs)
|
| 607 |
spiga |
1.65 |
|
| 608 |
slacapra |
1.9 |
common.logger.debug(5,'N jobs '+str(self.total_number_of_jobs))
|
| 609 |
|
|
|
| 610 |
|
|
# is there any remainder?
|
| 611 |
|
|
check = int(self.total_number_of_events) - (int(self.total_number_of_jobs)*self.eventsPerJob)
|
| 612 |
|
|
|
| 613 |
|
|
common.logger.debug(5,'Check '+str(check))
|
| 614 |
|
|
|
| 615 |
gutsche |
1.35 |
common.logger.message(str(self.total_number_of_jobs)+' jobs can be created, each for '+str(self.eventsPerJob)+' for a total of '+str(self.total_number_of_jobs*self.eventsPerJob)+' events')
|
| 616 |
slacapra |
1.9 |
if check > 0:
|
| 617 |
gutsche |
1.35 |
common.logger.message('Warning: asked '+str(self.total_number_of_events)+' but can do only '+str(int(self.total_number_of_jobs)*self.eventsPerJob))
|
| 618 |
slacapra |
1.9 |
|
| 619 |
slacapra |
1.10 |
# argument is seed number.$i
|
| 620 |
slacapra |
1.9 |
self.list_of_args = []
|
| 621 |
|
|
for i in range(self.total_number_of_jobs):
|
| 622 |
gutsche |
1.35 |
## Since there is no input, any site is good
|
| 623 |
slacapra |
1.86 |
# self.jobDestination.append(["Any"])
|
| 624 |
spiga |
1.42 |
self.jobDestination.append([""]) #must be empty to write correctly the xml
|
| 625 |
slacapra |
1.90 |
args=[]
|
| 626 |
spiga |
1.57 |
if (self.firstRun):
|
| 627 |
|
|
## pythia first run
|
| 628 |
slacapra |
1.86 |
#self.list_of_args.append([(str(self.firstRun)+str(i))])
|
| 629 |
slacapra |
1.90 |
args.append(str(self.firstRun)+str(i))
|
| 630 |
spiga |
1.57 |
else:
|
| 631 |
|
|
## no first run
|
| 632 |
slacapra |
1.86 |
#self.list_of_args.append([str(i)])
|
| 633 |
slacapra |
1.90 |
args.append(str(i))
|
| 634 |
slacapra |
1.23 |
if (self.sourceSeed):
|
| 635 |
slacapra |
1.90 |
args.append(str(self.sourceSeed)+str(i))
|
| 636 |
slacapra |
1.28 |
if (self.sourceSeedVtx):
|
| 637 |
slacapra |
1.90 |
## + vtx random seed
|
| 638 |
|
|
args.append(str(self.sourceSeedVtx)+str(i))
|
| 639 |
|
|
if (self.sourceSeedG4):
|
| 640 |
|
|
## + G4 random seed
|
| 641 |
|
|
args.append(str(self.sourceSeedG4)+str(i))
|
| 642 |
|
|
if (self.sourceSeedMix):
|
| 643 |
|
|
## + Mix random seed
|
| 644 |
|
|
args.append(str(self.sourceSeedMix)+str(i))
|
| 645 |
|
|
pass
|
| 646 |
|
|
pass
|
| 647 |
|
|
self.list_of_args.append(args)
|
| 648 |
|
|
pass
|
| 649 |
slacapra |
1.86 |
|
| 650 |
slacapra |
1.90 |
# print self.list_of_args
|
| 651 |
gutsche |
1.3 |
|
| 652 |
|
|
return
|
| 653 |
|
|
|
| 654 |
spiga |
1.42 |
|
| 655 |
|
|
def jobSplittingForScript(self):#CarlosDaniele
|
| 656 |
|
|
"""
|
| 657 |
|
|
Perform job splitting based on number of job
|
| 658 |
|
|
"""
|
| 659 |
|
|
common.logger.debug(5,'Splitting per job')
|
| 660 |
|
|
common.logger.message('Required '+str(self.theNumberOfJobs)+' jobs in total ')
|
| 661 |
|
|
|
| 662 |
|
|
self.total_number_of_jobs = self.theNumberOfJobs
|
| 663 |
|
|
|
| 664 |
|
|
common.logger.debug(5,'N jobs '+str(self.total_number_of_jobs))
|
| 665 |
|
|
|
| 666 |
|
|
common.logger.message(str(self.total_number_of_jobs)+' jobs can be created')
|
| 667 |
|
|
|
| 668 |
|
|
# argument is seed number.$i
|
| 669 |
|
|
self.list_of_args = []
|
| 670 |
|
|
for i in range(self.total_number_of_jobs):
|
| 671 |
|
|
## Since there is no input, any site is good
|
| 672 |
|
|
# self.jobDestination.append(["Any"])
|
| 673 |
|
|
self.jobDestination.append([""])
|
| 674 |
|
|
## no random seed
|
| 675 |
|
|
self.list_of_args.append([str(i)])
|
| 676 |
|
|
return
|
| 677 |
|
|
|
| 678 |
gutsche |
1.3 |
def split(self, jobParams):
|
| 679 |
|
|
|
| 680 |
|
|
common.jobDB.load()
|
| 681 |
|
|
#### Fabio
|
| 682 |
|
|
njobs = self.total_number_of_jobs
|
| 683 |
slacapra |
1.9 |
arglist = self.list_of_args
|
| 684 |
gutsche |
1.3 |
# create the empty structure
|
| 685 |
|
|
for i in range(njobs):
|
| 686 |
|
|
jobParams.append("")
|
| 687 |
|
|
|
| 688 |
|
|
for job in range(njobs):
|
| 689 |
slacapra |
1.17 |
jobParams[job] = arglist[job]
|
| 690 |
|
|
# print str(arglist[job])
|
| 691 |
|
|
# print jobParams[job]
|
| 692 |
gutsche |
1.3 |
common.jobDB.setArguments(job, jobParams[job])
|
| 693 |
gutsche |
1.35 |
common.logger.debug(5,"Job "+str(job)+" Destination: "+str(self.jobDestination[job]))
|
| 694 |
|
|
common.jobDB.setDestination(job, self.jobDestination[job])
|
| 695 |
gutsche |
1.3 |
|
| 696 |
|
|
common.jobDB.save()
|
| 697 |
|
|
return
|
| 698 |
|
|
|
| 699 |
|
|
def getJobTypeArguments(self, nj, sched):
|
| 700 |
slacapra |
1.17 |
result = ''
|
| 701 |
|
|
for i in common.jobDB.arguments(nj):
|
| 702 |
|
|
result=result+str(i)+" "
|
| 703 |
|
|
return result
|
| 704 |
gutsche |
1.3 |
|
| 705 |
|
|
def numberOfJobs(self):
|
| 706 |
|
|
# Fabio
|
| 707 |
|
|
return self.total_number_of_jobs
|
| 708 |
|
|
|
| 709 |
slacapra |
1.1 |
def getTarBall(self, exe):
|
| 710 |
|
|
"""
|
| 711 |
|
|
Return the TarBall with lib and exe
|
| 712 |
|
|
"""
|
| 713 |
|
|
|
| 714 |
|
|
# if it exist, just return it
|
| 715 |
corvo |
1.56 |
#
|
| 716 |
|
|
# Marco. Let's start to use relative path for Boss XML files
|
| 717 |
|
|
#
|
| 718 |
|
|
self.tgzNameWithPath = common.work_space.pathForTgz()+'share/'+self.tgz_name
|
| 719 |
slacapra |
1.1 |
if os.path.exists(self.tgzNameWithPath):
|
| 720 |
|
|
return self.tgzNameWithPath
|
| 721 |
|
|
|
| 722 |
|
|
# Prepare a tar gzipped file with user binaries.
|
| 723 |
|
|
self.buildTar_(exe)
|
| 724 |
|
|
|
| 725 |
|
|
return string.strip(self.tgzNameWithPath)
|
| 726 |
|
|
|
| 727 |
|
|
def buildTar_(self, executable):
|
| 728 |
|
|
|
| 729 |
|
|
# First of all declare the user Scram area
|
| 730 |
|
|
swArea = self.scram.getSWArea_()
|
| 731 |
|
|
#print "swArea = ", swArea
|
| 732 |
slacapra |
1.63 |
# swVersion = self.scram.getSWVersion()
|
| 733 |
|
|
# print "swVersion = ", swVersion
|
| 734 |
slacapra |
1.1 |
swReleaseTop = self.scram.getReleaseTop_()
|
| 735 |
|
|
#print "swReleaseTop = ", swReleaseTop
|
| 736 |
|
|
|
| 737 |
|
|
## check if working area is release top
|
| 738 |
|
|
if swReleaseTop == '' or swArea == swReleaseTop:
|
| 739 |
|
|
return
|
| 740 |
|
|
|
| 741 |
slacapra |
1.61 |
import tarfile
|
| 742 |
|
|
try: # create tar ball
|
| 743 |
|
|
tar = tarfile.open(self.tgzNameWithPath, "w:gz")
|
| 744 |
|
|
## First find the executable
|
| 745 |
slacapra |
1.86 |
if (self.executable != ''):
|
| 746 |
slacapra |
1.61 |
exeWithPath = self.scram.findFile_(executable)
|
| 747 |
|
|
if ( not exeWithPath ):
|
| 748 |
|
|
raise CrabException('User executable '+executable+' not found')
|
| 749 |
|
|
|
| 750 |
|
|
## then check if it's private or not
|
| 751 |
|
|
if exeWithPath.find(swReleaseTop) == -1:
|
| 752 |
|
|
# the exe is private, so we must ship
|
| 753 |
|
|
common.logger.debug(5,"Exe "+exeWithPath+" to be tarred")
|
| 754 |
|
|
path = swArea+'/'
|
| 755 |
corvo |
1.85 |
# distinguish case when script is in user project area or given by full path somewhere else
|
| 756 |
|
|
if exeWithPath.find(path) >= 0 :
|
| 757 |
|
|
exe = string.replace(exeWithPath, path,'')
|
| 758 |
|
|
tar.add(path+exe,os.path.basename(executable))
|
| 759 |
|
|
else :
|
| 760 |
|
|
tar.add(exeWithPath,os.path.basename(executable))
|
| 761 |
slacapra |
1.61 |
pass
|
| 762 |
|
|
else:
|
| 763 |
|
|
# the exe is from release, we'll find it on WN
|
| 764 |
|
|
pass
|
| 765 |
|
|
|
| 766 |
|
|
## Now get the libraries: only those in local working area
|
| 767 |
|
|
libDir = 'lib'
|
| 768 |
|
|
lib = swArea+'/' +libDir
|
| 769 |
|
|
common.logger.debug(5,"lib "+lib+" to be tarred")
|
| 770 |
|
|
if os.path.exists(lib):
|
| 771 |
|
|
tar.add(lib,libDir)
|
| 772 |
|
|
|
| 773 |
|
|
## Now check if module dir is present
|
| 774 |
|
|
moduleDir = 'module'
|
| 775 |
|
|
module = swArea + '/' + moduleDir
|
| 776 |
|
|
if os.path.isdir(module):
|
| 777 |
|
|
tar.add(module,moduleDir)
|
| 778 |
|
|
|
| 779 |
|
|
## Now check if any data dir(s) is present
|
| 780 |
|
|
swAreaLen=len(swArea)
|
| 781 |
|
|
for root, dirs, files in os.walk(swArea):
|
| 782 |
|
|
if "data" in dirs:
|
| 783 |
|
|
common.logger.debug(5,"data "+root+"/data"+" to be tarred")
|
| 784 |
|
|
tar.add(root+"/data",root[swAreaLen:]+"/data")
|
| 785 |
|
|
|
| 786 |
|
|
## Add ProdAgent dir to tar
|
| 787 |
|
|
paDir = 'ProdAgentApi'
|
| 788 |
|
|
pa = os.environ['CRABDIR'] + '/' + 'ProdAgentApi'
|
| 789 |
|
|
if os.path.isdir(pa):
|
| 790 |
|
|
tar.add(pa,paDir)
|
| 791 |
fanzago |
1.93 |
|
| 792 |
|
|
### FEDE FOR DBS PUBLICATION
|
| 793 |
|
|
## Add PRODCOMMON dir to tar
|
| 794 |
|
|
prodcommonDir = 'ProdCommon'
|
| 795 |
|
|
prodcommonPath = os.environ['CRABDIR'] + '/' + 'ProdCommon'
|
| 796 |
|
|
if os.path.isdir(prodcommonPath):
|
| 797 |
|
|
tar.add(prodcommonPath,prodcommonDir)
|
| 798 |
|
|
#############################
|
| 799 |
slacapra |
1.61 |
|
| 800 |
|
|
common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
|
| 801 |
|
|
tar.close()
|
| 802 |
|
|
except :
|
| 803 |
|
|
raise CrabException('Could not create tar-ball')
|
| 804 |
gutsche |
1.72 |
|
| 805 |
|
|
## check for tarball size
|
| 806 |
|
|
tarballinfo = os.stat(self.tgzNameWithPath)
|
| 807 |
|
|
if ( tarballinfo.st_size > self.MaxTarBallSize*1024*1024 ) :
|
| 808 |
|
|
raise CrabException('Input sandbox size of ' + str(float(tarballinfo.st_size)/1024.0/1024.0) + ' MB is larger than the allowed ' + str(self.MaxTarBallSize) + ' MB input sandbox limit and not supported by the used GRID submission system. Please make sure that no unnecessary files are in all data directories in your local CMSSW project area as they are automatically packed into the input sandbox.')
|
| 809 |
|
|
|
| 810 |
slacapra |
1.61 |
## create tar-ball with ML stuff
|
| 811 |
corvo |
1.58 |
self.MLtgzfile = common.work_space.pathForTgz()+'share/MLfiles.tgz'
|
| 812 |
slacapra |
1.61 |
try:
|
| 813 |
|
|
tar = tarfile.open(self.MLtgzfile, "w:gz")
|
| 814 |
|
|
path=os.environ['CRABDIR'] + '/python/'
|
| 815 |
|
|
for file in ['report.py', 'DashboardAPI.py', 'Logger.py', 'ProcInfo.py', 'apmon.py', 'parseCrabFjr.py']:
|
| 816 |
|
|
tar.add(path+file,file)
|
| 817 |
|
|
common.logger.debug(5,"Files added to "+self.MLtgzfile+" : "+str(tar.getnames()))
|
| 818 |
|
|
tar.close()
|
| 819 |
|
|
except :
|
| 820 |
corvo |
1.58 |
raise CrabException('Could not create ML files tar-ball')
|
| 821 |
|
|
|
| 822 |
slacapra |
1.1 |
return
|
| 823 |
|
|
|
| 824 |
slacapra |
1.97 |
def additionalInputFileTgz(self):
|
| 825 |
|
|
"""
|
| 826 |
|
|
Put all additional files into a tar ball and return its name
|
| 827 |
|
|
"""
|
| 828 |
|
|
import tarfile
|
| 829 |
|
|
tarName= common.work_space.pathForTgz()+'share/'+self.additional_tgz_name
|
| 830 |
|
|
tar = tarfile.open(tarName, "w:gz")
|
| 831 |
|
|
for file in self.additional_inbox_files:
|
| 832 |
|
|
tar.add(file,string.split(file,'/')[-1])
|
| 833 |
|
|
common.logger.debug(5,"Files added to "+self.additional_tgz_name+" : "+str(tar.getnames()))
|
| 834 |
|
|
tar.close()
|
| 835 |
|
|
return tarName
|
| 836 |
|
|
|
| 837 |
slacapra |
1.1 |
def wsSetupEnvironment(self, nj):
|
| 838 |
|
|
"""
|
| 839 |
|
|
Returns part of a job script which prepares
|
| 840 |
|
|
the execution environment for the job 'nj'.
|
| 841 |
|
|
"""
|
| 842 |
|
|
# Prepare JobType-independent part
|
| 843 |
gutsche |
1.3 |
txt = ''
|
| 844 |
|
|
|
| 845 |
|
|
## OLI_Daniele at this level middleware already known
|
| 846 |
|
|
|
| 847 |
|
|
txt += 'if [ $middleware == LCG ]; then \n'
|
| 848 |
|
|
txt += self.wsSetupCMSLCGEnvironment_()
|
| 849 |
|
|
txt += 'elif [ $middleware == OSG ]; then\n'
|
| 850 |
gutsche |
1.43 |
txt += ' WORKING_DIR=`/bin/mktemp -d $OSG_WN_TMP/cms_XXXXXXXXXXXX`\n'
|
| 851 |
|
|
txt += ' echo "Created working directory: $WORKING_DIR"\n'
|
| 852 |
gutsche |
1.3 |
txt += ' if [ ! -d $WORKING_DIR ] ;then\n'
|
| 853 |
gutsche |
1.7 |
txt += ' echo "SET_CMS_ENV 10016 ==> OSG $WORKING_DIR could not be created on WN `hostname`"\n'
|
| 854 |
slacapra |
1.90 |
txt += ' echo "JOB_EXIT_STATUS = 10016"\n'
|
| 855 |
|
|
txt += ' echo "JobExitCode=10016" | tee -a $RUNTIME_AREA/$repo\n'
|
| 856 |
|
|
txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
|
| 857 |
gutsche |
1.13 |
txt += ' rm -f $RUNTIME_AREA/$repo \n'
|
| 858 |
|
|
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
|
| 859 |
|
|
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
|
| 860 |
gutsche |
1.3 |
txt += ' exit 1\n'
|
| 861 |
|
|
txt += ' fi\n'
|
| 862 |
|
|
txt += '\n'
|
| 863 |
|
|
txt += ' echo "Change to working directory: $WORKING_DIR"\n'
|
| 864 |
|
|
txt += ' cd $WORKING_DIR\n'
|
| 865 |
|
|
txt += self.wsSetupCMSOSGEnvironment_()
|
| 866 |
|
|
txt += 'fi\n'
|
| 867 |
slacapra |
1.1 |
|
| 868 |
|
|
# Prepare JobType-specific part
|
| 869 |
|
|
scram = self.scram.commandName()
|
| 870 |
|
|
txt += '\n\n'
|
| 871 |
|
|
txt += 'echo "### SPECIFIC JOB SETUP ENVIRONMENT ###"\n'
|
| 872 |
slacapra |
1.86 |
txt += 'echo "Setting SCRAM_ARCH='+self.executable_arch+'"\n'
|
| 873 |
|
|
txt += 'export SCRAM_ARCH='+self.executable_arch+'\n'
|
| 874 |
slacapra |
1.1 |
txt += scram+' project CMSSW '+self.version+'\n'
|
| 875 |
|
|
txt += 'status=$?\n'
|
| 876 |
|
|
txt += 'if [ $status != 0 ] ; then\n'
|
| 877 |
gutsche |
1.7 |
txt += ' echo "SET_EXE_ENV 10034 ==>ERROR CMSSW '+self.version+' not found on `hostname`" \n'
|
| 878 |
gutsche |
1.3 |
txt += ' echo "JOB_EXIT_STATUS = 10034"\n'
|
| 879 |
gutsche |
1.7 |
txt += ' echo "JobExitCode=10034" | tee -a $RUNTIME_AREA/$repo\n'
|
| 880 |
slacapra |
1.1 |
txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
|
| 881 |
gutsche |
1.13 |
txt += ' rm -f $RUNTIME_AREA/$repo \n'
|
| 882 |
|
|
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
|
| 883 |
|
|
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
|
| 884 |
gutsche |
1.3 |
## OLI_Daniele
|
| 885 |
|
|
txt += ' if [ $middleware == OSG ]; then \n'
|
| 886 |
|
|
txt += ' echo "Remove working directory: $WORKING_DIR"\n'
|
| 887 |
|
|
txt += ' cd $RUNTIME_AREA\n'
|
| 888 |
|
|
txt += ' /bin/rm -rf $WORKING_DIR\n'
|
| 889 |
|
|
txt += ' if [ -d $WORKING_DIR ] ;then\n'
|
| 890 |
fanzago |
1.96 |
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'
|
| 891 |
|
|
txt += ' echo "JOB_EXIT_STATUS = 10018"\n'
|
| 892 |
|
|
txt += ' echo "JobExitCode=10018" | tee -a $RUNTIME_AREA/$repo\n'
|
| 893 |
|
|
txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
|
| 894 |
gutsche |
1.13 |
txt += ' rm -f $RUNTIME_AREA/$repo \n'
|
| 895 |
|
|
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
|
| 896 |
|
|
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
|
| 897 |
gutsche |
1.3 |
txt += ' fi\n'
|
| 898 |
|
|
txt += ' fi \n'
|
| 899 |
|
|
txt += ' exit 1 \n'
|
| 900 |
slacapra |
1.1 |
txt += 'fi \n'
|
| 901 |
|
|
txt += 'echo "CMSSW_VERSION = '+self.version+'"\n'
|
| 902 |
|
|
txt += 'cd '+self.version+'\n'
|
| 903 |
fanzago |
1.99 |
########## FEDE FOR DBS2 ######################
|
| 904 |
|
|
txt += 'SOFTWARE_DIR=`pwd`\n'
|
| 905 |
|
|
txt += 'echo SOFTWARE_DIR=$SOFTWARE_DIR \n'
|
| 906 |
|
|
###############################################
|
| 907 |
slacapra |
1.1 |
### needed grep for bug in scramv1 ###
|
| 908 |
corvo |
1.58 |
txt += scram+' runtime -sh\n'
|
| 909 |
slacapra |
1.1 |
txt += 'eval `'+scram+' runtime -sh | grep -v SCRAMRT_LSB_JOBNAME`\n'
|
| 910 |
corvo |
1.58 |
txt += 'echo $PATH\n'
|
| 911 |
slacapra |
1.1 |
|
| 912 |
|
|
# Handle the arguments:
|
| 913 |
|
|
txt += "\n"
|
| 914 |
gutsche |
1.7 |
txt += "## number of arguments (first argument always jobnumber)\n"
|
| 915 |
slacapra |
1.1 |
txt += "\n"
|
| 916 |
mkirn |
1.32 |
# txt += "narg=$#\n"
|
| 917 |
|
|
txt += "if [ $nargs -lt 2 ]\n"
|
| 918 |
slacapra |
1.1 |
txt += "then\n"
|
| 919 |
mkirn |
1.33 |
txt += " echo 'SET_EXE_ENV 1 ==> ERROR Too few arguments' +$nargs+ \n"
|
| 920 |
gutsche |
1.3 |
txt += ' echo "JOB_EXIT_STATUS = 50113"\n'
|
| 921 |
gutsche |
1.7 |
txt += ' echo "JobExitCode=50113" | tee -a $RUNTIME_AREA/$repo\n'
|
| 922 |
slacapra |
1.1 |
txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
|
| 923 |
gutsche |
1.13 |
txt += ' rm -f $RUNTIME_AREA/$repo \n'
|
| 924 |
|
|
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
|
| 925 |
|
|
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
|
| 926 |
gutsche |
1.3 |
## OLI_Daniele
|
| 927 |
|
|
txt += ' if [ $middleware == OSG ]; then \n'
|
| 928 |
|
|
txt += ' echo "Remove working directory: $WORKING_DIR"\n'
|
| 929 |
|
|
txt += ' cd $RUNTIME_AREA\n'
|
| 930 |
|
|
txt += ' /bin/rm -rf $WORKING_DIR\n'
|
| 931 |
|
|
txt += ' if [ -d $WORKING_DIR ] ;then\n'
|
| 932 |
fanzago |
1.96 |
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'
|
| 933 |
|
|
txt += ' echo "JOB_EXIT_STATUS = 50114"\n'
|
| 934 |
|
|
txt += ' echo "JobExitCode=50114" | tee -a $RUNTIME_AREA/$repo\n'
|
| 935 |
|
|
txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
|
| 936 |
gutsche |
1.13 |
txt += ' rm -f $RUNTIME_AREA/$repo \n'
|
| 937 |
|
|
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
|
| 938 |
|
|
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
|
| 939 |
gutsche |
1.3 |
txt += ' fi\n'
|
| 940 |
|
|
txt += ' fi \n'
|
| 941 |
slacapra |
1.1 |
txt += " exit 1\n"
|
| 942 |
|
|
txt += "fi\n"
|
| 943 |
|
|
txt += "\n"
|
| 944 |
|
|
|
| 945 |
|
|
# Prepare job-specific part
|
| 946 |
|
|
job = common.job_list[nj]
|
| 947 |
fanzago |
1.93 |
### FEDE FOR DBS OUTPUT PUBLICATION
|
| 948 |
|
|
if (self.datasetPath):
|
| 949 |
|
|
txt += '\n'
|
| 950 |
|
|
txt += 'DatasetPath='+self.datasetPath+'\n'
|
| 951 |
|
|
|
| 952 |
|
|
datasetpath_split = self.datasetPath.split("/")
|
| 953 |
|
|
|
| 954 |
|
|
txt += 'PrimaryDataset='+datasetpath_split[1]+'\n'
|
| 955 |
|
|
txt += 'DataTier='+datasetpath_split[2]+'\n'
|
| 956 |
fanzago |
1.96 |
#txt += 'ProcessedDataset='+datasetpath_split[3]+'\n'
|
| 957 |
|
|
txt += 'ApplicationFamily=cmsRun\n'
|
| 958 |
fanzago |
1.93 |
|
| 959 |
|
|
else:
|
| 960 |
|
|
txt += 'DatasetPath=MCDataTier\n'
|
| 961 |
|
|
txt += 'PrimaryDataset=null\n'
|
| 962 |
|
|
txt += 'DataTier=null\n'
|
| 963 |
fanzago |
1.96 |
#txt += 'ProcessedDataset=null\n'
|
| 964 |
fanzago |
1.93 |
txt += 'ApplicationFamily=MCDataTier\n'
|
| 965 |
spiga |
1.42 |
if self.pset != None: #CarlosDaniele
|
| 966 |
|
|
pset = os.path.basename(job.configFilename())
|
| 967 |
|
|
txt += '\n'
|
| 968 |
spiga |
1.95 |
txt += 'cp $RUNTIME_AREA/'+pset+' .\n'
|
| 969 |
spiga |
1.42 |
if (self.datasetPath): # standard job
|
| 970 |
|
|
#txt += 'InputFiles=$2\n'
|
| 971 |
|
|
txt += 'InputFiles=${args[1]}\n'
|
| 972 |
|
|
txt += 'MaxEvents=${args[2]}\n'
|
| 973 |
|
|
txt += 'SkipEvents=${args[3]}\n'
|
| 974 |
|
|
txt += 'echo "Inputfiles:<$InputFiles>"\n'
|
| 975 |
slacapra |
1.90 |
txt += 'sed "s#{\'INPUT\'}#$InputFiles#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
|
| 976 |
spiga |
1.42 |
txt += 'echo "MaxEvents:<$MaxEvents>"\n'
|
| 977 |
slacapra |
1.90 |
txt += 'sed "s#INPUTMAXEVENTS#$MaxEvents#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
|
| 978 |
spiga |
1.42 |
txt += 'echo "SkipEvents:<$SkipEvents>"\n'
|
| 979 |
slacapra |
1.90 |
txt += 'sed "s#INPUTSKIPEVENTS#$SkipEvents#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
|
| 980 |
spiga |
1.42 |
else: # pythia like job
|
| 981 |
slacapra |
1.90 |
seedIndex=1
|
| 982 |
|
|
if (self.firstRun):
|
| 983 |
|
|
txt += 'FirstRun=${args['+str(seedIndex)+']}\n'
|
| 984 |
spiga |
1.57 |
txt += 'echo "FirstRun: <$FirstRun>"\n'
|
| 985 |
slacapra |
1.90 |
txt += 'sed "s#\<INPUTFIRSTRUN\>#$FirstRun#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
|
| 986 |
|
|
seedIndex=seedIndex+1
|
| 987 |
|
|
|
| 988 |
spiga |
1.57 |
if (self.sourceSeed):
|
| 989 |
slacapra |
1.90 |
txt += 'Seed=${args['+str(seedIndex)+']}\n'
|
| 990 |
|
|
txt += 'sed "s#\<INPUT\>#$Seed#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
|
| 991 |
|
|
seedIndex=seedIndex+1
|
| 992 |
|
|
## the following seeds are not always present
|
| 993 |
spiga |
1.42 |
if (self.sourceSeedVtx):
|
| 994 |
slacapra |
1.90 |
txt += 'VtxSeed=${args['+str(seedIndex)+']}\n'
|
| 995 |
spiga |
1.42 |
txt += 'echo "VtxSeed: <$VtxSeed>"\n'
|
| 996 |
slacapra |
1.90 |
txt += 'sed "s#\<INPUTVTX\>#$VtxSeed#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
|
| 997 |
|
|
seedIndex += 1
|
| 998 |
|
|
if (self.sourceSeedG4):
|
| 999 |
|
|
txt += 'G4Seed=${args['+str(seedIndex)+']}\n'
|
| 1000 |
|
|
txt += 'echo "G4Seed: <$G4Seed>"\n'
|
| 1001 |
|
|
txt += 'sed "s#\<INPUTG4\>#$G4Seed#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
|
| 1002 |
|
|
seedIndex += 1
|
| 1003 |
|
|
if (self.sourceSeedMix):
|
| 1004 |
|
|
txt += 'mixSeed=${args['+str(seedIndex)+']}\n'
|
| 1005 |
|
|
txt += 'echo "MixSeed: <$mixSeed>"\n'
|
| 1006 |
|
|
txt += 'sed "s#\<INPUTMIX\>#$mixSeed#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
|
| 1007 |
|
|
seedIndex += 1
|
| 1008 |
|
|
pass
|
| 1009 |
|
|
pass
|
| 1010 |
|
|
txt += 'mv -f '+pset+' pset.cfg\n'
|
| 1011 |
slacapra |
1.1 |
|
| 1012 |
|
|
if len(self.additional_inbox_files) > 0:
|
| 1013 |
slacapra |
1.97 |
txt += 'if [ -e $RUNTIME_AREA/'+self.additional_tgz_name+' ] ; then\n'
|
| 1014 |
|
|
txt += ' tar xzvf $RUNTIME_AREA/'+self.additional_tgz_name+'\n'
|
| 1015 |
|
|
txt += 'fi\n'
|
| 1016 |
slacapra |
1.1 |
pass
|
| 1017 |
|
|
|
| 1018 |
spiga |
1.42 |
if self.pset != None: #CarlosDaniele
|
| 1019 |
|
|
txt += 'echo "### END JOB SETUP ENVIRONMENT ###"\n\n'
|
| 1020 |
|
|
|
| 1021 |
|
|
txt += '\n'
|
| 1022 |
|
|
txt += 'echo "***** cat pset.cfg *********"\n'
|
| 1023 |
|
|
txt += 'cat pset.cfg\n'
|
| 1024 |
|
|
txt += 'echo "****** end pset.cfg ********"\n'
|
| 1025 |
|
|
txt += '\n'
|
| 1026 |
fanzago |
1.93 |
### FEDE FOR DBS OUTPUT PUBLICATION
|
| 1027 |
fanzago |
1.94 |
txt += 'PSETHASH=`EdmConfigHash < pset.cfg` \n'
|
| 1028 |
|
|
txt += 'echo "PSETHASH = $PSETHASH" \n'
|
| 1029 |
fanzago |
1.93 |
##############
|
| 1030 |
|
|
txt += '\n'
|
| 1031 |
|
|
# txt += 'echo "***** cat pset1.cfg *********"\n'
|
| 1032 |
|
|
# txt += 'cat pset1.cfg\n'
|
| 1033 |
|
|
# txt += 'echo "****** end pset1.cfg ********"\n'
|
| 1034 |
gutsche |
1.3 |
return txt
|
| 1035 |
|
|
|
| 1036 |
slacapra |
1.63 |
def wsBuildExe(self, nj=0):
|
| 1037 |
gutsche |
1.3 |
"""
|
| 1038 |
|
|
Put in the script the commands to build an executable
|
| 1039 |
|
|
or a library.
|
| 1040 |
|
|
"""
|
| 1041 |
|
|
|
| 1042 |
|
|
txt = ""
|
| 1043 |
|
|
|
| 1044 |
|
|
if os.path.isfile(self.tgzNameWithPath):
|
| 1045 |
|
|
txt += 'echo "tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+'"\n'
|
| 1046 |
|
|
txt += 'tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+'\n'
|
| 1047 |
|
|
txt += 'untar_status=$? \n'
|
| 1048 |
|
|
txt += 'if [ $untar_status -ne 0 ]; then \n'
|
| 1049 |
|
|
txt += ' echo "SET_EXE 1 ==> ERROR Untarring .tgz file failed"\n'
|
| 1050 |
|
|
txt += ' echo "JOB_EXIT_STATUS = $untar_status" \n'
|
| 1051 |
gutsche |
1.7 |
txt += ' echo "JobExitCode=$untar_status" | tee -a $RUNTIME_AREA/$repo\n'
|
| 1052 |
gutsche |
1.3 |
txt += ' if [ $middleware == OSG ]; then \n'
|
| 1053 |
|
|
txt += ' echo "Remove working directory: $WORKING_DIR"\n'
|
| 1054 |
|
|
txt += ' cd $RUNTIME_AREA\n'
|
| 1055 |
|
|
txt += ' /bin/rm -rf $WORKING_DIR\n'
|
| 1056 |
|
|
txt += ' if [ -d $WORKING_DIR ] ;then\n'
|
| 1057 |
gutsche |
1.13 |
txt += ' echo "SET_EXE 50999 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after Untarring .tgz file failed"\n'
|
| 1058 |
|
|
txt += ' echo "JOB_EXIT_STATUS = 50999"\n'
|
| 1059 |
|
|
txt += ' echo "JobExitCode=50999" | tee -a $RUNTIME_AREA/$repo\n'
|
| 1060 |
|
|
txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
|
| 1061 |
|
|
txt += ' rm -f $RUNTIME_AREA/$repo \n'
|
| 1062 |
|
|
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
|
| 1063 |
|
|
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
|
| 1064 |
gutsche |
1.3 |
txt += ' fi\n'
|
| 1065 |
|
|
txt += ' fi \n'
|
| 1066 |
|
|
txt += ' \n'
|
| 1067 |
gutsche |
1.7 |
txt += ' exit 1 \n'
|
| 1068 |
gutsche |
1.3 |
txt += 'else \n'
|
| 1069 |
|
|
txt += ' echo "Successful untar" \n'
|
| 1070 |
|
|
txt += 'fi \n'
|
| 1071 |
gutsche |
1.50 |
txt += '\n'
|
| 1072 |
fanzago |
1.93 |
txt += 'echo "Include ProdAgentApi and PRODCOMMON in PYTHONPATH"\n'
|
| 1073 |
gutsche |
1.50 |
txt += 'if [ -z "$PYTHONPATH" ]; then\n'
|
| 1074 |
fanzago |
1.93 |
#### FEDE FOR DBS OUTPUT PUBLICATION
|
| 1075 |
fanzago |
1.99 |
txt += ' export PYTHONPATH=$SOFTWARE_DIR/ProdAgentApi:$SOFTWARE_DIR/ProdCommon\n'
|
| 1076 |
|
|
#txt += ' export PYTHONPATH=`pwd`/ProdAgentApi:`pwd`/ProdCommon\n'
|
| 1077 |
fanzago |
1.93 |
#txt += ' export PYTHONPATH=ProdAgentApi\n'
|
| 1078 |
gutsche |
1.50 |
txt += 'else\n'
|
| 1079 |
fanzago |
1.99 |
txt += ' export PYTHONPATH=$SOFTWARE_DIR/ProdAgentApi:$SOFTWARE_DIR/ProdCommon:${PYTHONPATH}\n'
|
| 1080 |
|
|
#txt += ' export PYTHONPATH=`pwd`/ProdAgentApi:`pwd`/ProdCommon:${PYTHONPATH}\n'
|
| 1081 |
fanzago |
1.93 |
#txt += ' export PYTHONPATH=ProdAgentApi:${PYTHONPATH}\n'
|
| 1082 |
|
|
txt += 'echo "PYTHONPATH=$PYTHONPATH"\n'
|
| 1083 |
|
|
###################
|
| 1084 |
gutsche |
1.50 |
txt += 'fi\n'
|
| 1085 |
|
|
txt += '\n'
|
| 1086 |
|
|
|
| 1087 |
gutsche |
1.3 |
pass
|
| 1088 |
|
|
|
| 1089 |
slacapra |
1.1 |
return txt
|
| 1090 |
|
|
|
| 1091 |
|
|
def modifySteeringCards(self, nj):
|
| 1092 |
|
|
"""
|
| 1093 |
|
|
modify the card provided by the user,
|
| 1094 |
|
|
writing a new card into share dir
|
| 1095 |
|
|
"""
|
| 1096 |
|
|
|
| 1097 |
|
|
def executableName(self):
|
| 1098 |
slacapra |
1.70 |
if self.scriptExe: #CarlosDaniele
|
| 1099 |
spiga |
1.42 |
return "sh "
|
| 1100 |
|
|
else:
|
| 1101 |
|
|
return self.executable
|
| 1102 |
slacapra |
1.1 |
|
| 1103 |
|
|
def executableArgs(self):
|
| 1104 |
slacapra |
1.70 |
if self.scriptExe:#CarlosDaniele
|
| 1105 |
spiga |
1.42 |
return self.scriptExe + " $NJob"
|
| 1106 |
|
|
else:
|
| 1107 |
|
|
return " -p pset.cfg"
|
| 1108 |
slacapra |
1.1 |
|
| 1109 |
|
|
def inputSandbox(self, nj):
|
| 1110 |
|
|
"""
|
| 1111 |
|
|
Returns a list of filenames to be put in JDL input sandbox.
|
| 1112 |
|
|
"""
|
| 1113 |
|
|
inp_box = []
|
| 1114 |
slacapra |
1.53 |
# # dict added to delete duplicate from input sandbox file list
|
| 1115 |
|
|
# seen = {}
|
| 1116 |
slacapra |
1.1 |
## code
|
| 1117 |
|
|
if os.path.isfile(self.tgzNameWithPath):
|
| 1118 |
|
|
inp_box.append(self.tgzNameWithPath)
|
| 1119 |
corvo |
1.58 |
if os.path.isfile(self.MLtgzfile):
|
| 1120 |
|
|
inp_box.append(self.MLtgzfile)
|
| 1121 |
slacapra |
1.1 |
## config
|
| 1122 |
slacapra |
1.70 |
if not self.pset is None:
|
| 1123 |
corvo |
1.56 |
inp_box.append(common.work_space.pathForTgz() + 'job/' + self.configFilename())
|
| 1124 |
slacapra |
1.1 |
## additional input files
|
| 1125 |
slacapra |
1.97 |
tgz = self.additionalInputFileTgz()
|
| 1126 |
|
|
inp_box.append(tgz)
|
| 1127 |
slacapra |
1.1 |
return inp_box
|
| 1128 |
|
|
|
| 1129 |
|
|
def outputSandbox(self, nj):
|
| 1130 |
|
|
"""
|
| 1131 |
|
|
Returns a list of filenames to be put in JDL output sandbox.
|
| 1132 |
|
|
"""
|
| 1133 |
|
|
out_box = []
|
| 1134 |
|
|
|
| 1135 |
|
|
## User Declared output files
|
| 1136 |
slacapra |
1.54 |
for out in (self.output_file+self.output_file_sandbox):
|
| 1137 |
slacapra |
1.1 |
n_out = nj + 1
|
| 1138 |
|
|
out_box.append(self.numberFile_(out,str(n_out)))
|
| 1139 |
|
|
return out_box
|
| 1140 |
|
|
|
| 1141 |
|
|
def prepareSteeringCards(self):
|
| 1142 |
|
|
"""
|
| 1143 |
|
|
Make initial modifications of the user's steering card file.
|
| 1144 |
|
|
"""
|
| 1145 |
|
|
return
|
| 1146 |
|
|
|
| 1147 |
|
|
def wsRenameOutput(self, nj):
|
| 1148 |
|
|
"""
|
| 1149 |
|
|
Returns part of a job script which renames the produced files.
|
| 1150 |
|
|
"""
|
| 1151 |
|
|
|
| 1152 |
|
|
txt = '\n'
|
| 1153 |
gutsche |
1.7 |
txt += '# directory content\n'
|
| 1154 |
|
|
txt += 'ls \n'
|
| 1155 |
slacapra |
1.54 |
|
| 1156 |
|
|
for fileWithSuffix in (self.output_file+self.output_file_sandbox):
|
| 1157 |
slacapra |
1.1 |
output_file_num = self.numberFile_(fileWithSuffix, '$NJob')
|
| 1158 |
|
|
txt += '\n'
|
| 1159 |
gutsche |
1.7 |
txt += '# check output file\n'
|
| 1160 |
slacapra |
1.1 |
txt += 'ls '+fileWithSuffix+'\n'
|
| 1161 |
fanzago |
1.18 |
txt += 'ls_result=$?\n'
|
| 1162 |
|
|
txt += 'if [ $ls_result -ne 0 ] ; then\n'
|
| 1163 |
spiga |
1.88 |
txt += ' exit_status=60302\n'
|
| 1164 |
fanzago |
1.18 |
txt += ' echo "ERROR: Problem with output file"\n'
|
| 1165 |
gutsche |
1.7 |
if common.scheduler.boss_scheduler_name == 'condor_g':
|
| 1166 |
|
|
txt += ' if [ $middleware == OSG ]; then \n'
|
| 1167 |
|
|
txt += ' echo "prepare dummy output file"\n'
|
| 1168 |
|
|
txt += ' echo "Processing of job output failed" > $RUNTIME_AREA/'+output_file_num+'\n'
|
| 1169 |
|
|
txt += ' fi \n'
|
| 1170 |
slacapra |
1.1 |
txt += 'else\n'
|
| 1171 |
fanzago |
1.93 |
### FEDE FOR DBS OUTPUT PUBLICATION
|
| 1172 |
|
|
txt += ' mv '+fileWithSuffix+' $RUNTIME_AREA\n'
|
| 1173 |
|
|
txt += ' cp $RUNTIME_AREA/'+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n'
|
| 1174 |
|
|
#################################
|
| 1175 |
slacapra |
1.1 |
txt += 'fi\n'
|
| 1176 |
|
|
|
| 1177 |
gutsche |
1.7 |
txt += 'cd $RUNTIME_AREA\n'
|
| 1178 |
fanzago |
1.99 |
#### FEDE this is the cleanEnv function
|
| 1179 |
gutsche |
1.3 |
### OLI_DANIELE
|
| 1180 |
fanzago |
1.99 |
#txt += 'if [ $middleware == OSG ]; then\n'
|
| 1181 |
|
|
#txt += ' cd $RUNTIME_AREA\n'
|
| 1182 |
|
|
#txt += ' echo "Remove working directory: $WORKING_DIR"\n'
|
| 1183 |
|
|
#txt += ' /bin/rm -rf $WORKING_DIR\n'
|
| 1184 |
|
|
#txt += ' if [ -d $WORKING_DIR ] ;then\n'
|
| 1185 |
|
|
#txt += ' echo "SET_EXE 60999 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after cleanup of WN"\n'
|
| 1186 |
|
|
#txt += ' echo "JOB_EXIT_STATUS = 60999"\n'
|
| 1187 |
|
|
#txt += ' echo "JobExitCode=60999" | tee -a $RUNTIME_AREA/$repo\n'
|
| 1188 |
|
|
#txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
|
| 1189 |
|
|
#txt += ' rm -f $RUNTIME_AREA/$repo \n'
|
| 1190 |
|
|
#txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
|
| 1191 |
|
|
#txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
|
| 1192 |
|
|
#txt += ' fi\n'
|
| 1193 |
|
|
#txt += 'fi\n'
|
| 1194 |
|
|
#txt += '\n'
|
| 1195 |
slacapra |
1.54 |
|
| 1196 |
|
|
file_list = ''
|
| 1197 |
|
|
## Add to filelist only files to be possibly copied to SE
|
| 1198 |
|
|
for fileWithSuffix in self.output_file:
|
| 1199 |
|
|
output_file_num = self.numberFile_(fileWithSuffix, '$NJob')
|
| 1200 |
|
|
file_list=file_list+output_file_num+' '
|
| 1201 |
|
|
file_list=file_list[:-1]
|
| 1202 |
|
|
txt += 'file_list="'+file_list+'"\n'
|
| 1203 |
|
|
|
| 1204 |
slacapra |
1.1 |
return txt
|
| 1205 |
|
|
|
| 1206 |
|
|
def numberFile_(self, file, txt):
|
| 1207 |
|
|
"""
|
| 1208 |
|
|
append _'txt' before last extension of a file
|
| 1209 |
|
|
"""
|
| 1210 |
|
|
p = string.split(file,".")
|
| 1211 |
|
|
# take away last extension
|
| 1212 |
|
|
name = p[0]
|
| 1213 |
|
|
for x in p[1:-1]:
|
| 1214 |
slacapra |
1.90 |
name=name+"."+x
|
| 1215 |
slacapra |
1.1 |
# add "_txt"
|
| 1216 |
|
|
if len(p)>1:
|
| 1217 |
slacapra |
1.90 |
ext = p[len(p)-1]
|
| 1218 |
|
|
result = name + '_' + txt + "." + ext
|
| 1219 |
slacapra |
1.1 |
else:
|
| 1220 |
slacapra |
1.90 |
result = name + '_' + txt
|
| 1221 |
slacapra |
1.1 |
|
| 1222 |
|
|
return result
|
| 1223 |
|
|
|
| 1224 |
slacapra |
1.63 |
def getRequirements(self, nj=[]):
|
| 1225 |
slacapra |
1.1 |
"""
|
| 1226 |
|
|
return job requirements to add to jdl files
|
| 1227 |
|
|
"""
|
| 1228 |
|
|
req = ''
|
| 1229 |
slacapra |
1.47 |
if self.version:
|
| 1230 |
slacapra |
1.10 |
req='Member("VO-cms-' + \
|
| 1231 |
slacapra |
1.47 |
self.version + \
|
| 1232 |
slacapra |
1.10 |
'", other.GlueHostApplicationSoftwareRunTimeEnvironment)'
|
| 1233 |
slacapra |
1.91 |
# if self.executable_arch:
|
| 1234 |
|
|
# req='Member("VO-cms-' + \
|
| 1235 |
|
|
# self.executable_arch + \
|
| 1236 |
|
|
# '", other.GlueHostApplicationSoftwareRunTimeEnvironment)'
|
| 1237 |
gutsche |
1.35 |
|
| 1238 |
|
|
req = req + ' && (other.GlueHostNetworkAdapterOutboundIP)'
|
| 1239 |
|
|
|
| 1240 |
slacapra |
1.1 |
return req
|
| 1241 |
gutsche |
1.3 |
|
| 1242 |
|
|
def configFilename(self):
|
| 1243 |
|
|
""" return the config filename """
|
| 1244 |
|
|
return self.name()+'.cfg'
|
| 1245 |
|
|
|
| 1246 |
|
|
### OLI_DANIELE
|
| 1247 |
|
|
def wsSetupCMSOSGEnvironment_(self):
|
| 1248 |
|
|
"""
|
| 1249 |
|
|
Returns part of a job script which is prepares
|
| 1250 |
|
|
the execution environment and which is common for all CMS jobs.
|
| 1251 |
|
|
"""
|
| 1252 |
|
|
txt = '\n'
|
| 1253 |
|
|
txt += ' echo "### SETUP CMS OSG ENVIRONMENT ###"\n'
|
| 1254 |
|
|
txt += ' if [ -f $GRID3_APP_DIR/cmssoft/cmsset_default.sh ] ;then\n'
|
| 1255 |
|
|
txt += ' # Use $GRID3_APP_DIR/cmssoft/cmsset_default.sh to setup cms software\n'
|
| 1256 |
spiga |
1.87 |
txt += ' export SCRAM_ARCH='+self.executable_arch+'\n'
|
| 1257 |
gutsche |
1.3 |
txt += ' source $GRID3_APP_DIR/cmssoft/cmsset_default.sh '+self.version+'\n'
|
| 1258 |
mkirn |
1.40 |
txt += ' elif [ -f $OSG_APP/cmssoft/cms/cmsset_default.sh ] ;then\n'
|
| 1259 |
|
|
txt += ' # Use $OSG_APP/cmssoft/cms/cmsset_default.sh to setup cms software\n'
|
| 1260 |
spiga |
1.87 |
txt += ' export SCRAM_ARCH='+self.executable_arch+'\n'
|
| 1261 |
mkirn |
1.40 |
txt += ' source $OSG_APP/cmssoft/cms/cmsset_default.sh '+self.version+'\n'
|
| 1262 |
gutsche |
1.3 |
txt += ' else\n'
|
| 1263 |
mkirn |
1.40 |
txt += ' echo "SET_CMS_ENV 10020 ==> ERROR $GRID3_APP_DIR/cmssoft/cmsset_default.sh and $OSG_APP/cmssoft/cms/cmsset_default.sh file not found"\n'
|
| 1264 |
gutsche |
1.3 |
txt += ' echo "JOB_EXIT_STATUS = 10020"\n'
|
| 1265 |
|
|
txt += ' echo "JobExitCode=10020" | tee -a $RUNTIME_AREA/$repo\n'
|
| 1266 |
|
|
txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
|
| 1267 |
gutsche |
1.13 |
txt += ' rm -f $RUNTIME_AREA/$repo \n'
|
| 1268 |
|
|
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
|
| 1269 |
|
|
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
|
| 1270 |
gutsche |
1.7 |
txt += ' exit 1\n'
|
| 1271 |
gutsche |
1.3 |
txt += '\n'
|
| 1272 |
|
|
txt += ' echo "Remove working directory: $WORKING_DIR"\n'
|
| 1273 |
|
|
txt += ' cd $RUNTIME_AREA\n'
|
| 1274 |
|
|
txt += ' /bin/rm -rf $WORKING_DIR\n'
|
| 1275 |
|
|
txt += ' if [ -d $WORKING_DIR ] ;then\n'
|
| 1276 |
fanzago |
1.96 |
txt += ' echo "SET_CMS_ENV 10017 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after $GRID3_APP_DIR/cmssoft/cmsset_default.sh and $OSG_APP/cmssoft/cms/cmsset_default.sh file not found"\n'
|
| 1277 |
|
|
txt += ' echo "JOB_EXIT_STATUS = 10017"\n'
|
| 1278 |
|
|
txt += ' echo "JobExitCode=10017" | tee -a $RUNTIME_AREA/$repo\n'
|
| 1279 |
|
|
txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
|
| 1280 |
|
|
txt += ' rm -f $RUNTIME_AREA/$repo \n'
|
| 1281 |
|
|
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
|
| 1282 |
|
|
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
|
| 1283 |
gutsche |
1.3 |
txt += ' fi\n'
|
| 1284 |
|
|
txt += '\n'
|
| 1285 |
gutsche |
1.7 |
txt += ' exit 1\n'
|
| 1286 |
gutsche |
1.3 |
txt += ' fi\n'
|
| 1287 |
|
|
txt += '\n'
|
| 1288 |
|
|
txt += ' echo "SET_CMS_ENV 0 ==> setup cms environment ok"\n'
|
| 1289 |
|
|
txt += ' echo " END SETUP CMS OSG ENVIRONMENT "\n'
|
| 1290 |
|
|
|
| 1291 |
|
|
return txt
|
| 1292 |
|
|
|
| 1293 |
|
|
### OLI_DANIELE
|
| 1294 |
|
|
def wsSetupCMSLCGEnvironment_(self):
|
| 1295 |
|
|
"""
|
| 1296 |
|
|
Returns part of a job script which is prepares
|
| 1297 |
|
|
the execution environment and which is common for all CMS jobs.
|
| 1298 |
|
|
"""
|
| 1299 |
|
|
txt = ' \n'
|
| 1300 |
|
|
txt += ' echo " ### SETUP CMS LCG ENVIRONMENT ### "\n'
|
| 1301 |
|
|
txt += ' if [ ! $VO_CMS_SW_DIR ] ;then\n'
|
| 1302 |
|
|
txt += ' echo "SET_CMS_ENV 10031 ==> ERROR CMS software dir not found on WN `hostname`"\n'
|
| 1303 |
|
|
txt += ' echo "JOB_EXIT_STATUS = 10031" \n'
|
| 1304 |
|
|
txt += ' echo "JobExitCode=10031" | tee -a $RUNTIME_AREA/$repo\n'
|
| 1305 |
|
|
txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
|
| 1306 |
gutsche |
1.13 |
txt += ' rm -f $RUNTIME_AREA/$repo \n'
|
| 1307 |
|
|
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
|
| 1308 |
|
|
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
|
| 1309 |
gutsche |
1.7 |
txt += ' exit 1\n'
|
| 1310 |
gutsche |
1.3 |
txt += ' else\n'
|
| 1311 |
|
|
txt += ' echo "Sourcing environment... "\n'
|
| 1312 |
|
|
txt += ' if [ ! -s $VO_CMS_SW_DIR/cmsset_default.sh ] ;then\n'
|
| 1313 |
|
|
txt += ' echo "SET_CMS_ENV 10020 ==> ERROR cmsset_default.sh file not found into dir $VO_CMS_SW_DIR"\n'
|
| 1314 |
|
|
txt += ' echo "JOB_EXIT_STATUS = 10020"\n'
|
| 1315 |
|
|
txt += ' echo "JobExitCode=10020" | tee -a $RUNTIME_AREA/$repo\n'
|
| 1316 |
|
|
txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
|
| 1317 |
gutsche |
1.13 |
txt += ' rm -f $RUNTIME_AREA/$repo \n'
|
| 1318 |
|
|
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
|
| 1319 |
|
|
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
|
| 1320 |
gutsche |
1.7 |
txt += ' exit 1\n'
|
| 1321 |
gutsche |
1.3 |
txt += ' fi\n'
|
| 1322 |
|
|
txt += ' echo "sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
|
| 1323 |
|
|
txt += ' source $VO_CMS_SW_DIR/cmsset_default.sh\n'
|
| 1324 |
|
|
txt += ' result=$?\n'
|
| 1325 |
|
|
txt += ' if [ $result -ne 0 ]; then\n'
|
| 1326 |
|
|
txt += ' echo "SET_CMS_ENV 10032 ==> ERROR problem sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
|
| 1327 |
|
|
txt += ' echo "JOB_EXIT_STATUS = 10032"\n'
|
| 1328 |
|
|
txt += ' echo "JobExitCode=10032" | tee -a $RUNTIME_AREA/$repo\n'
|
| 1329 |
|
|
txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
|
| 1330 |
gutsche |
1.13 |
txt += ' rm -f $RUNTIME_AREA/$repo \n'
|
| 1331 |
|
|
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
|
| 1332 |
|
|
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
|
| 1333 |
gutsche |
1.7 |
txt += ' exit 1\n'
|
| 1334 |
gutsche |
1.3 |
txt += ' fi\n'
|
| 1335 |
|
|
txt += ' fi\n'
|
| 1336 |
|
|
txt += ' \n'
|
| 1337 |
|
|
txt += ' echo "SET_CMS_ENV 0 ==> setup cms environment ok"\n'
|
| 1338 |
|
|
txt += ' echo "### END SETUP CMS LCG ENVIRONMENT ###"\n'
|
| 1339 |
|
|
return txt
|
| 1340 |
gutsche |
1.5 |
|
| 1341 |
fanzago |
1.93 |
### FEDE FOR DBS OUTPUT PUBLICATION
|
| 1342 |
|
|
def modifyReport(self, nj):
|
| 1343 |
|
|
"""
|
| 1344 |
|
|
insert the part of the script that modifies the FrameworkJob Report
|
| 1345 |
|
|
"""
|
| 1346 |
fanzago |
1.94 |
|
| 1347 |
fanzago |
1.93 |
txt = ''
|
| 1348 |
|
|
txt += 'echo "Modify Job Report" \n'
|
| 1349 |
fanzago |
1.99 |
#txt += 'chmod a+x $RUNTIME_AREA/'+self.version+'/ProdAgentApi/FwkJobRep/ModifyJobReport.py\n'
|
| 1350 |
|
|
################ FEDE FOR DBS2 #############################################
|
| 1351 |
|
|
txt += 'chmod a+x $SOFTWARE_DIR/ProdAgentApi/FwkJobRep/ModifyJobReport.py\n'
|
| 1352 |
|
|
#############################################################################
|
| 1353 |
fanzago |
1.94 |
try:
|
| 1354 |
|
|
publish_data = int(self.cfg_params['USER.publish_data'])
|
| 1355 |
|
|
except KeyError:
|
| 1356 |
|
|
publish_data = 0
|
| 1357 |
|
|
|
| 1358 |
fanzago |
1.93 |
txt += 'if [ -z "$SE" ]; then\n'
|
| 1359 |
|
|
txt += ' SE="" \n'
|
| 1360 |
|
|
txt += 'fi \n'
|
| 1361 |
|
|
txt += 'if [ -z "$SE_PATH" ]; then\n'
|
| 1362 |
|
|
txt += ' SE_PATH="" \n'
|
| 1363 |
|
|
txt += 'fi \n'
|
| 1364 |
|
|
txt += 'echo "SE = $SE"\n'
|
| 1365 |
|
|
txt += 'echo "SE_PATH = $SE_PATH"\n'
|
| 1366 |
fanzago |
1.94 |
|
| 1367 |
|
|
if (publish_data == 1):
|
| 1368 |
|
|
#processedDataset = self.cfg_params['USER.processed_datasetname']
|
| 1369 |
|
|
processedDataset = self.cfg_params['USER.publish_data_name']
|
| 1370 |
|
|
txt += 'ProcessedDataset='+processedDataset+'\n'
|
| 1371 |
|
|
#### LFN=/store/user/<user>/processedDataset_PSETHASH
|
| 1372 |
|
|
txt += 'if [ "$SE_PATH" == "" ]; then\n'
|
| 1373 |
fanzago |
1.101 |
#### FEDE: added slash in LFN ##############
|
| 1374 |
|
|
txt += ' FOR_LFN=/copy_problems/ \n'
|
| 1375 |
fanzago |
1.94 |
txt += 'else \n'
|
| 1376 |
|
|
txt += ' tmp=`echo $SE_PATH | awk -F \'store\' \'{print$2}\'` \n'
|
| 1377 |
fanzago |
1.101 |
##### FEDE TO BE CHANGED, BECAUSE STORE IS HARDCODED!!!! ########
|
| 1378 |
fanzago |
1.94 |
txt += ' FOR_LFN=/store$tmp \n'
|
| 1379 |
|
|
txt += 'fi \n'
|
| 1380 |
|
|
txt += 'echo "ProcessedDataset = $ProcessedDataset"\n'
|
| 1381 |
|
|
txt += 'echo "FOR_LFN = $FOR_LFN" \n'
|
| 1382 |
|
|
else:
|
| 1383 |
|
|
txt += 'ProcessedDataset=no_data_to_publish \n'
|
| 1384 |
fanzago |
1.101 |
#### FEDE: added slash in LFN ##############
|
| 1385 |
|
|
txt += 'FOR_LFN=/local/ \n'
|
| 1386 |
fanzago |
1.94 |
txt += 'echo "ProcessedDataset = $ProcessedDataset"\n'
|
| 1387 |
|
|
txt += 'echo "FOR_LFN = $FOR_LFN" \n'
|
| 1388 |
|
|
txt += 'echo "CMSSW_VERSION = $CMSSW_VERSION"\n\n'
|
| 1389 |
fanzago |
1.99 |
#txt += 'echo "$RUNTIME_AREA/'+self.version+'/ProdAgentApi/FwkJobRep/ModifyJobReport.py crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH"\n'
|
| 1390 |
|
|
txt += 'echo "$SOFTWARE_DIR/ProdAgentApi/FwkJobRep/ModifyJobReport.py crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH"\n'
|
| 1391 |
|
|
txt += '$SOFTWARE_DIR/ProdAgentApi/FwkJobRep/ModifyJobReport.py crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH\n'
|
| 1392 |
|
|
#txt += '$RUNTIME_AREA/'+self.version+'/ProdAgentApi/FwkJobRep/ModifyJobReport.py crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH\n'
|
| 1393 |
fanzago |
1.93 |
txt += 'modifyReport_result=$?\n'
|
| 1394 |
|
|
txt += 'echo modifyReport_result = $modifyReport_result\n'
|
| 1395 |
fanzago |
1.100 |
txt += 'if [ $modifyReport_result -ne 0 ]; then\n'
|
| 1396 |
fanzago |
1.93 |
txt += ' exit_status=1\n'
|
| 1397 |
|
|
txt += ' echo "ERROR: Problem with ModifyJobReport"\n'
|
| 1398 |
|
|
txt += 'else\n'
|
| 1399 |
|
|
txt += ' mv NewFrameworkJobReport.xml crab_fjr_$NJob.xml\n'
|
| 1400 |
|
|
txt += 'fi\n'
|
| 1401 |
|
|
return txt
|
| 1402 |
fanzago |
1.99 |
|
| 1403 |
|
|
def cleanEnv(self):
|
| 1404 |
|
|
### OLI_DANIELE
|
| 1405 |
|
|
txt = ''
|
| 1406 |
|
|
txt += 'if [ $middleware == OSG ]; then\n'
|
| 1407 |
|
|
txt += ' cd $RUNTIME_AREA\n'
|
| 1408 |
|
|
txt += ' echo "Remove working directory: $WORKING_DIR"\n'
|
| 1409 |
|
|
txt += ' /bin/rm -rf $WORKING_DIR\n'
|
| 1410 |
|
|
txt += ' if [ -d $WORKING_DIR ] ;then\n'
|
| 1411 |
|
|
txt += ' echo "SET_EXE 60999 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after cleanup of WN"\n'
|
| 1412 |
|
|
txt += ' echo "JOB_EXIT_STATUS = 60999"\n'
|
| 1413 |
|
|
txt += ' echo "JobExitCode=60999" | tee -a $RUNTIME_AREA/$repo\n'
|
| 1414 |
|
|
txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
|
| 1415 |
|
|
txt += ' rm -f $RUNTIME_AREA/$repo \n'
|
| 1416 |
|
|
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
|
| 1417 |
|
|
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
|
| 1418 |
|
|
txt += ' fi\n'
|
| 1419 |
|
|
txt += 'fi\n'
|
| 1420 |
|
|
txt += '\n'
|
| 1421 |
|
|
return txt
|
| 1422 |
fanzago |
1.93 |
|
| 1423 |
gutsche |
1.5 |
def setParam_(self, param, value):
|
| 1424 |
|
|
self._params[param] = value
|
| 1425 |
|
|
|
| 1426 |
|
|
def getParams(self):
|
| 1427 |
|
|
return self._params
|
| 1428 |
gutsche |
1.8 |
|
| 1429 |
|
|
def setTaskid_(self):
|
| 1430 |
|
|
self._taskId = self.cfg_params['taskId']
|
| 1431 |
|
|
|
| 1432 |
|
|
def getTaskid(self):
|
| 1433 |
|
|
return self._taskId
|
| 1434 |
gutsche |
1.35 |
|
| 1435 |
|
|
def uniquelist(self, old):
|
| 1436 |
|
|
"""
|
| 1437 |
|
|
remove duplicates from a list
|
| 1438 |
|
|
"""
|
| 1439 |
|
|
nd={}
|
| 1440 |
|
|
for e in old:
|
| 1441 |
|
|
nd[e]=0
|
| 1442 |
|
|
return nd.keys()
|
