| 2 |
|
from crab_logger import Logger |
| 3 |
|
from crab_exceptions import * |
| 4 |
|
from crab_util import * |
| 5 |
+ |
import math |
| 6 |
|
import common |
| 7 |
|
import PsetManipulator |
| 8 |
|
|
| 9 |
< |
import DBSInfo_EDM |
| 10 |
< |
import DataDiscovery_EDM |
| 11 |
< |
import DataLocation_EDM |
| 9 |
> |
import DBSInfo |
| 10 |
> |
import DataDiscovery |
| 11 |
> |
import DataLocation |
| 12 |
|
import Scram |
| 13 |
|
|
| 14 |
|
import os, string, re |
| 15 |
|
|
| 16 |
|
class Cmssw(JobType): |
| 17 |
< |
def __init__(self, cfg_params): |
| 17 |
> |
def __init__(self, cfg_params, ncjobs): |
| 18 |
|
JobType.__init__(self, 'CMSSW') |
| 19 |
|
common.logger.debug(3,'CMSSW::__init__') |
| 20 |
|
|
| 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) |
| 45 |
|
|
| 46 |
|
### collect Data cards |
| 47 |
|
try: |
| 48 |
< |
# self.owner = cfg_params['CMSSW.owner'] |
| 49 |
< |
# log.debug(6, "CMSSW::CMSSW(): owner = "+self.owner) |
| 50 |
< |
# self.dataset = cfg_params['CMSSW.dataset'] |
| 51 |
< |
self.datasetPath = cfg_params['CMSSW.datasetpath'] |
| 52 |
< |
log.debug(6, "CMSSW::CMSSW(): datasetPath = "+self.datasetPath) |
| 48 |
> |
tmp = cfg_params['CMSSW.datasetpath'] |
| 49 |
> |
log.debug(6, "CMSSW::CMSSW(): datasetPath = "+tmp) |
| 50 |
> |
if string.lower(tmp)=='none': |
| 51 |
> |
self.datasetPath = None |
| 52 |
> |
self.selectNoInput = 1 |
| 53 |
> |
else: |
| 54 |
> |
self.datasetPath = tmp |
| 55 |
> |
self.selectNoInput = 0 |
| 56 |
|
except KeyError: |
| 50 |
– |
# msg = "Error: owner and/or dataset not defined " |
| 57 |
|
msg = "Error: datasetpath not defined " |
| 58 |
|
raise CrabException(msg) |
| 59 |
|
|
| 60 |
|
# ML monitoring |
| 61 |
|
# split dataset path style: /PreProdR3Minbias/SIM/GEN-SIM |
| 62 |
< |
datasetpath_split = self.datasetPath.split("/") |
| 63 |
< |
self.setParam_('dataset', datasetpath_split[1]) |
| 64 |
< |
self.setParam_('owner', datasetpath_split[-1]) |
| 65 |
< |
|
| 66 |
< |
|
| 62 |
> |
if not self.datasetPath: |
| 63 |
> |
self.setParam_('dataset', 'None') |
| 64 |
> |
self.setParam_('owner', 'None') |
| 65 |
> |
else: |
| 66 |
> |
datasetpath_split = self.datasetPath.split("/") |
| 67 |
> |
self.setParam_('dataset', datasetpath_split[1]) |
| 68 |
> |
self.setParam_('owner', datasetpath_split[-1]) |
| 69 |
|
|
| 70 |
+ |
self.setTaskid_() |
| 71 |
+ |
self.setParam_('taskId', self.cfg_params['taskId']) |
| 72 |
|
|
| 73 |
|
self.dataTiers = [] |
| 64 |
– |
# try: |
| 65 |
– |
# tmpDataTiers = string.split(cfg_params['CMSSW.data_tier'],',') |
| 66 |
– |
# for tmp in tmpDataTiers: |
| 67 |
– |
# tmp=string.strip(tmp) |
| 68 |
– |
# self.dataTiers.append(tmp) |
| 69 |
– |
# pass |
| 70 |
– |
# pass |
| 71 |
– |
# except KeyError: |
| 72 |
– |
# pass |
| 73 |
– |
# log.debug(6, "Cmssw::Cmssw(): dataTiers = "+str(self.dataTiers)) |
| 74 |
|
|
| 75 |
|
## now the application |
| 76 |
|
try: |
| 116 |
|
|
| 117 |
|
# script_exe file as additional file in inputSandbox |
| 118 |
|
try: |
| 119 |
< |
self.scriptExe = cfg_params['USER.script_exe'] |
| 120 |
< |
self.additional_inbox_files.append(self.scriptExe) |
| 119 |
> |
self.scriptExe = cfg_params['USER.script_exe'] |
| 120 |
> |
self.additional_inbox_files.append(self.scriptExe) |
| 121 |
> |
if self.scriptExe != '': |
| 122 |
> |
if not os.path.isfile(self.scriptExe): |
| 123 |
> |
msg ="WARNING. file "+self.scriptExe+" not found" |
| 124 |
> |
raise CrabException(msg) |
| 125 |
|
except KeyError: |
| 126 |
|
pass |
| 123 |
– |
if self.scriptExe != '': |
| 124 |
– |
if os.path.isfile(self.scriptExe): |
| 125 |
– |
pass |
| 126 |
– |
else: |
| 127 |
– |
log.message("WARNING. file "+self.scriptExe+" not found") |
| 128 |
– |
sys.exit() |
| 127 |
|
|
| 128 |
|
## additional input files |
| 129 |
|
try: |
| 130 |
< |
tmpAddFiles = string.split(cfg_params['CMSSW.additional_input_files'],',') |
| 130 |
> |
tmpAddFiles = string.split(cfg_params['USER.additional_input_files'],',') |
| 131 |
|
for tmp in tmpAddFiles: |
| 132 |
|
if not os.path.exists(tmp): |
| 133 |
|
raise CrabException("Additional input file not found: "+tmp) |
| 134 |
< |
tmp=string.strip(tmp) |
| 137 |
< |
self.additional_inbox_files.append(tmp) |
| 134 |
> |
self.additional_inbox_files.append(string.strip(tmp)) |
| 135 |
|
pass |
| 136 |
|
pass |
| 137 |
|
except KeyError: |
| 138 |
|
pass |
| 139 |
|
|
| 140 |
+ |
# files per job |
| 141 |
|
try: |
| 142 |
< |
self.filesPerJob = int(cfg_params['CMSSW.files_per_jobs']) #Daniele |
| 142 |
> |
if (cfg_params['CMSSW.files_per_jobs']): |
| 143 |
> |
raise CrabException("files_per_jobs no longer supported. Quitting.") |
| 144 |
|
except KeyError: |
| 145 |
< |
self.filesPerJob = 1 |
| 145 |
> |
pass |
| 146 |
|
|
| 147 |
< |
## Max event will be total_number_of_events ??? Daniele |
| 147 |
> |
## Events per job |
| 148 |
> |
try: |
| 149 |
> |
self.eventsPerJob =int( cfg_params['CMSSW.events_per_job']) |
| 150 |
> |
self.selectEventsPerJob = 1 |
| 151 |
> |
except KeyError: |
| 152 |
> |
self.eventsPerJob = -1 |
| 153 |
> |
self.selectEventsPerJob = 0 |
| 154 |
> |
|
| 155 |
> |
## number of jobs |
| 156 |
|
try: |
| 157 |
< |
self.maxEv = cfg_params['CMSSW.event_per_job'] |
| 157 |
> |
self.theNumberOfJobs =int( cfg_params['CMSSW.number_of_jobs']) |
| 158 |
> |
self.selectNumberOfJobs = 1 |
| 159 |
|
except KeyError: |
| 160 |
< |
self.maxEv = "-1" |
| 161 |
< |
## |
| 160 |
> |
self.theNumberOfJobs = 0 |
| 161 |
> |
self.selectNumberOfJobs = 0 |
| 162 |
> |
|
| 163 |
|
try: |
| 164 |
|
self.total_number_of_events = int(cfg_params['CMSSW.total_number_of_events']) |
| 165 |
+ |
self.selectTotalNumberEvents = 1 |
| 166 |
|
except KeyError: |
| 167 |
< |
msg = 'Must define total_number_of_events' |
| 167 |
> |
self.total_number_of_events = 0 |
| 168 |
> |
self.selectTotalNumberEvents = 0 |
| 169 |
> |
|
| 170 |
> |
if ( (self.selectTotalNumberEvents + self.selectEventsPerJob + self.selectNumberOfJobs) != 2 ): |
| 171 |
> |
msg = 'Must define exactly two of total_number_of_events, events_per_job, or number_of_jobs.' |
| 172 |
|
raise CrabException(msg) |
| 173 |
< |
|
| 174 |
< |
CEBlackList = [] |
| 173 |
> |
|
| 174 |
> |
## source seed for pythia |
| 175 |
|
try: |
| 176 |
< |
tmpBad = string.split(cfg_params['EDG.ce_black_list'],',') |
| 163 |
< |
for tmp in tmpBad: |
| 164 |
< |
tmp=string.strip(tmp) |
| 165 |
< |
CEBlackList.append(tmp) |
| 176 |
> |
self.sourceSeed = int(cfg_params['CMSSW.pythia_seed']) |
| 177 |
|
except KeyError: |
| 178 |
< |
pass |
| 178 |
> |
self.sourceSeed = None |
| 179 |
> |
common.logger.debug(5,"No seed given") |
| 180 |
|
|
| 181 |
< |
self.reCEBlackList=[] |
| 182 |
< |
for bad in CEBlackList: |
| 171 |
< |
self.reCEBlackList.append(re.compile( bad )) |
| 172 |
< |
|
| 173 |
< |
common.logger.debug(5,'CEBlackList: '+str(CEBlackList)) |
| 174 |
< |
|
| 175 |
< |
CEWhiteList = [] |
| 176 |
< |
try: |
| 177 |
< |
tmpGood = string.split(cfg_params['EDG.ce_white_list'],',') |
| 178 |
< |
for tmp in tmpGood: |
| 179 |
< |
tmp=string.strip(tmp) |
| 180 |
< |
CEWhiteList.append(tmp) |
| 181 |
> |
try: |
| 182 |
> |
self.sourceSeedVtx = int(cfg_params['CMSSW.vtx_seed']) |
| 183 |
|
except KeyError: |
| 184 |
< |
pass |
| 185 |
< |
|
| 184 |
< |
#print 'CEWhiteList: ',CEWhiteList |
| 185 |
< |
self.reCEWhiteList=[] |
| 186 |
< |
for Good in CEWhiteList: |
| 187 |
< |
self.reCEWhiteList.append(re.compile( Good )) |
| 188 |
< |
|
| 189 |
< |
common.logger.debug(5,'CEWhiteList: '+str(CEWhiteList)) |
| 184 |
> |
self.sourceSeedVtx = None |
| 185 |
> |
common.logger.debug(5,"No vertex seed given") |
| 186 |
|
|
| 187 |
|
self.PsetEdit = PsetManipulator.PsetManipulator(self.pset) #Daniele Pset |
| 188 |
|
|
| 190 |
|
## Initialize the variables that are extracted from DBS/DLS and needed in other places of the code |
| 191 |
|
self.maxEvents=0 # max events available ( --> check the requested nb. of evts in Creator.py) |
| 192 |
|
self.DBSPaths={} # all dbs paths requested ( --> input to the site local discovery script) |
| 193 |
+ |
self.jobDestination=[] # Site destination(s) for each job (list of lists) |
| 194 |
|
## Perform the data location and discovery (based on DBS/DLS) |
| 195 |
< |
self.DataDiscoveryAndLocation(cfg_params) |
| 195 |
> |
## SL: Don't if NONE is specified as input (pythia use case) |
| 196 |
> |
blockSites = {} |
| 197 |
> |
if self.datasetPath: |
| 198 |
> |
blockSites = self.DataDiscoveryAndLocation(cfg_params) |
| 199 |
|
#DBSDLS-end |
| 200 |
|
|
| 201 |
|
self.tgzNameWithPath = self.getTarBall(self.executable) |
| 202 |
+ |
|
| 203 |
+ |
## Select Splitting |
| 204 |
+ |
if self.selectNoInput: self.jobSplittingNoInput() |
| 205 |
+ |
else: self.jobSplittingByBlocks(blockSites) |
| 206 |
|
|
| 207 |
< |
self.jobSplitting() #Daniele job Splitting |
| 208 |
< |
self.PsetEdit.maxEvent(self.maxEv) #Daniele |
| 209 |
< |
self.PsetEdit.inputModule("INPUT") #Daniele |
| 210 |
< |
self.PsetEdit.psetWriter(self.configFilename()) |
| 211 |
< |
|
| 212 |
< |
|
| 207 |
> |
# modify Pset |
| 208 |
> |
try: |
| 209 |
> |
if (self.datasetPath): # standard job |
| 210 |
> |
# allow to processa a fraction of events in a file |
| 211 |
> |
self.PsetEdit.inputModule("INPUT") |
| 212 |
> |
self.PsetEdit.maxEvent("INPUTMAXEVENTS") |
| 213 |
> |
self.PsetEdit.skipEvent("INPUTSKIPEVENTS") |
| 214 |
> |
|
| 215 |
> |
else: # pythia like job |
| 216 |
> |
self.PsetEdit.maxEvent(self.eventsPerJob) |
| 217 |
> |
if (self.sourceSeed) : |
| 218 |
> |
self.PsetEdit.pythiaSeed("INPUT") |
| 219 |
> |
if (self.sourceSeedVtx) : |
| 220 |
> |
self.PsetEdit.pythiaSeedVtx("INPUTVTX") |
| 221 |
> |
self.PsetEdit.psetWriter(self.configFilename()) |
| 222 |
> |
except: |
| 223 |
> |
msg='Error while manipuliating ParameterSet: exiting...' |
| 224 |
> |
raise CrabException(msg) |
| 225 |
|
|
| 226 |
|
def DataDiscoveryAndLocation(self, cfg_params): |
| 227 |
|
|
| 228 |
|
common.logger.debug(10,"CMSSW::DataDiscoveryAndLocation()") |
| 229 |
|
|
| 214 |
– |
#datasetPath = "/"+self.owner+"/"+self.dataTiers[0]+"/"+self.dataset |
| 215 |
– |
|
| 230 |
|
datasetPath=self.datasetPath |
| 231 |
|
|
| 232 |
|
## TODO |
| 234 |
|
dataTiers = dataTiersList.split(',') |
| 235 |
|
|
| 236 |
|
## Contact the DBS |
| 237 |
+ |
common.logger.message("Contacting DBS...") |
| 238 |
|
try: |
| 239 |
< |
self.pubdata=DataDiscovery_EDM.DataDiscovery_EDM(datasetPath, dataTiers, cfg_params) |
| 239 |
> |
self.pubdata=DataDiscovery.DataDiscovery(datasetPath, dataTiers, cfg_params) |
| 240 |
|
self.pubdata.fetchDBSInfo() |
| 241 |
|
|
| 242 |
< |
except DataDiscovery_EDM.NotExistingDatasetError, ex : |
| 242 |
> |
except DataDiscovery.NotExistingDatasetError, ex : |
| 243 |
|
msg = 'ERROR ***: failed Data Discovery in DBS : %s'%ex.getErrorMessage() |
| 244 |
|
raise CrabException(msg) |
| 245 |
|
|
| 246 |
< |
except DataDiscovery_EDM.NoDataTierinProvenanceError, ex : |
| 246 |
> |
except DataDiscovery.NoDataTierinProvenanceError, ex : |
| 247 |
|
msg = 'ERROR ***: failed Data Discovery in DBS : %s'%ex.getErrorMessage() |
| 248 |
|
raise CrabException(msg) |
| 249 |
< |
except DataDiscovery_EDM.DataDiscoveryError, ex: |
| 249 |
> |
except DataDiscovery.DataDiscoveryError, ex: |
| 250 |
|
msg = 'ERROR ***: failed Data Discovery in DBS %s'%ex.getErrorMessage() |
| 251 |
|
raise CrabException(msg) |
| 252 |
|
|
| 254 |
|
## self.DBSPaths=self.pubdata.getDBSPaths() |
| 255 |
|
common.logger.message("Required data are :"+self.datasetPath) |
| 256 |
|
|
| 257 |
< |
filesbyblock=self.pubdata.getFiles() |
| 258 |
< |
self.AllInputFiles=filesbyblock.values() |
| 259 |
< |
self.files = self.AllInputFiles |
| 260 |
< |
|
| 261 |
< |
## TEMP |
| 262 |
< |
# self.filesTmp = filesbyblock.values() |
| 248 |
< |
# self.files = [] |
| 249 |
< |
# locPath='rfio:cmsbose2.bo.infn.it:/flatfiles/SE00/cms/fanfani/ProdTest/' |
| 250 |
< |
# locPath='' |
| 251 |
< |
# tmp = [] |
| 252 |
< |
# for file in self.filesTmp[0]: |
| 253 |
< |
# tmp.append(locPath+file) |
| 254 |
< |
# self.files.append(tmp) |
| 255 |
< |
## END TEMP |
| 257 |
> |
self.filesbyblock=self.pubdata.getFiles() |
| 258 |
> |
self.eventsbyblock=self.pubdata.getEventsPerBlock() |
| 259 |
> |
self.eventsbyfile=self.pubdata.getEventsPerFile() |
| 260 |
> |
# print str(self.filesbyblock) |
| 261 |
> |
# print 'self.eventsbyfile',len(self.eventsbyfile) |
| 262 |
> |
# print str(self.eventsbyfile) |
| 263 |
|
|
| 264 |
|
## get max number of events |
| 258 |
– |
#common.logger.debug(10,"number of events for primary fileblocks %i"%self.pubdata.getMaxEvents()) |
| 265 |
|
self.maxEvents=self.pubdata.getMaxEvents() ## self.maxEvents used in Creator.py |
| 266 |
|
common.logger.message("\nThe number of available events is %s"%self.maxEvents) |
| 267 |
|
|
| 268 |
+ |
common.logger.message("Contacting DLS...") |
| 269 |
|
## Contact the DLS and build a list of sites hosting the fileblocks |
| 270 |
|
try: |
| 271 |
< |
dataloc=DataLocation_EDM.DataLocation_EDM(filesbyblock.keys(),cfg_params) |
| 271 |
> |
dataloc=DataLocation.DataLocation(self.filesbyblock.keys(),cfg_params) |
| 272 |
|
dataloc.fetchDLSInfo() |
| 273 |
< |
except DataLocation_EDM.DataLocationError , ex: |
| 273 |
> |
except DataLocation.DataLocationError , ex: |
| 274 |
|
msg = 'ERROR ***: failed Data Location in DLS \n %s '%ex.getErrorMessage() |
| 275 |
|
raise CrabException(msg) |
| 276 |
|
|
| 270 |
– |
allsites=dataloc.getSites() |
| 271 |
– |
common.logger.debug(5,"sites are %s"%allsites) |
| 272 |
– |
sites=self.checkBlackList(allsites) |
| 273 |
– |
common.logger.debug(5,"sites are (after black list) %s"%sites) |
| 274 |
– |
sites=self.checkWhiteList(sites) |
| 275 |
– |
common.logger.debug(5,"sites are (after white list) %s"%sites) |
| 277 |
|
|
| 278 |
< |
if len(sites)==0: |
| 279 |
< |
msg = 'No sites hosting all the needed data! Exiting... ' |
| 280 |
< |
raise CrabException(msg) |
| 278 |
> |
sites = dataloc.getSites() |
| 279 |
> |
allSites = [] |
| 280 |
> |
listSites = sites.values() |
| 281 |
> |
for list in listSites: |
| 282 |
> |
for oneSite in list: |
| 283 |
> |
allSites.append(oneSite) |
| 284 |
> |
allSites = self.uniquelist(allSites) |
| 285 |
|
|
| 286 |
< |
common.logger.message("List of Sites hosting the data : "+str(sites)) |
| 287 |
< |
common.logger.debug(6, "List of Sites: "+str(sites)) |
| 288 |
< |
common.analisys_common_info['sites']=sites ## used in SchedulerEdg.py in createSchScript |
| 284 |
< |
self.setParam_('TargetCE', ','.join(sites)) |
| 285 |
< |
return |
| 286 |
> |
common.logger.message("Sites ("+str(len(allSites))+") hosting part/all of dataset: "+str(allSites)) |
| 287 |
> |
common.logger.debug(6, "List of Sites: "+str(allSites)) |
| 288 |
> |
return sites |
| 289 |
|
|
| 290 |
< |
def jobSplitting(self): |
| 290 |
> |
def jobSplittingByBlocks(self, blockSites): |
| 291 |
|
""" |
| 292 |
< |
first implemntation for job splitting |
| 293 |
< |
""" |
| 294 |
< |
# print 'eventi totali '+str(self.maxEvents) |
| 295 |
< |
# print 'eventi totali richiesti dallo user '+str(self.total_number_of_events) |
| 296 |
< |
#print 'files per job '+str(self.filesPerJob) |
| 297 |
< |
common.logger.message('Required '+str(self.filesPerJob)+' files per job ') |
| 298 |
< |
common.logger.message('Required '+str(self.total_number_of_events)+' events in total ') |
| 299 |
< |
|
| 300 |
< |
## TODO: SL need to have (from DBS) a detailed list of how many events per each file |
| 301 |
< |
n_tot_files = (len(self.files[0])) |
| 302 |
< |
## SL: this is wrong if the files have different number of events |
| 303 |
< |
evPerFile = int(self.maxEvents)/n_tot_files |
| 304 |
< |
|
| 305 |
< |
common.logger.debug(5,'Events per File '+str(evPerFile)) |
| 306 |
< |
|
| 307 |
< |
## if asked to process all events, do it |
| 308 |
< |
if self.total_number_of_events == -1: |
| 309 |
< |
self.total_number_of_events=self.maxEvents |
| 310 |
< |
self.total_number_of_jobs = int(n_tot_files)*1/int(self.filesPerJob) |
| 311 |
< |
common.logger.message(str(self.total_number_of_jobs)+' jobs will be created for all available events '+str(self.total_number_of_events)+' events') |
| 312 |
< |
|
| 292 |
> |
Perform job splitting. Jobs run over an integer number of files |
| 293 |
> |
and no more than one block. |
| 294 |
> |
ARGUMENT: blockSites: dictionary with blocks as keys and list of host sites as values |
| 295 |
> |
REQUIRES: self.selectTotalNumberEvents, self.selectEventsPerJob, self.selectNumberofJobs, |
| 296 |
> |
self.total_number_of_events, self.eventsPerJob, self.theNumberOfJobs, |
| 297 |
> |
self.maxEvents, self.filesbyblock |
| 298 |
> |
SETS: self.jobDestination - Site destination(s) for each job (a list of lists) |
| 299 |
> |
self.total_number_of_jobs - Total # of jobs |
| 300 |
> |
self.list_of_args - File(s) job will run on (a list of lists) |
| 301 |
> |
""" |
| 302 |
> |
|
| 303 |
> |
# ---- Handle the possible job splitting configurations ---- # |
| 304 |
> |
if (self.selectTotalNumberEvents): |
| 305 |
> |
totalEventsRequested = self.total_number_of_events |
| 306 |
> |
if (self.selectEventsPerJob): |
| 307 |
> |
eventsPerJobRequested = self.eventsPerJob |
| 308 |
> |
if (self.selectNumberOfJobs): |
| 309 |
> |
totalEventsRequested = self.theNumberOfJobs * self.eventsPerJob |
| 310 |
> |
|
| 311 |
> |
# If user requested all the events in the dataset |
| 312 |
> |
if (totalEventsRequested == -1): |
| 313 |
> |
eventsRemaining=self.maxEvents |
| 314 |
> |
# If user requested more events than are in the dataset |
| 315 |
> |
elif (totalEventsRequested > self.maxEvents): |
| 316 |
> |
eventsRemaining = self.maxEvents |
| 317 |
> |
common.logger.message("Requested "+str(self.total_number_of_events)+ " events, but only "+str(self.maxEvents)+" events are available.") |
| 318 |
> |
# If user requested less events than are in the dataset |
| 319 |
|
else: |
| 320 |
< |
self.total_number_of_files = int(self.total_number_of_events/evPerFile) |
| 312 |
< |
## SL: if ask for less event than what is computed to be available on a |
| 313 |
< |
## file, process the first file anyhow. |
| 314 |
< |
if self.total_number_of_files == 0: |
| 315 |
< |
self.total_number_of_files = self.total_number_of_files + 1 |
| 320 |
> |
eventsRemaining = totalEventsRequested |
| 321 |
|
|
| 322 |
< |
common.logger.debug(5,'N files '+str(self.total_number_of_files)) |
| 322 |
> |
# If user requested more events per job than are in the dataset |
| 323 |
> |
if (self.selectEventsPerJob and eventsPerJobRequested > self.maxEvents): |
| 324 |
> |
eventsPerJobRequested = self.maxEvents |
| 325 |
|
|
| 326 |
< |
check = 0 |
| 327 |
< |
|
| 321 |
< |
## Compute the number of jobs |
| 322 |
< |
#self.total_number_of_jobs = int(n_tot_files)*1/int(self.filesPerJob) |
| 323 |
< |
self.total_number_of_jobs = int(self.total_number_of_files/self.filesPerJob) |
| 324 |
< |
common.logger.debug(5,'N jobs '+str(self.total_number_of_jobs)) |
| 326 |
> |
# For user info at end |
| 327 |
> |
totalEventCount = 0 |
| 328 |
|
|
| 329 |
< |
## is there any remainder? |
| 330 |
< |
check = int(self.total_number_of_files) - (int(self.total_number_of_jobs)*self.filesPerJob) |
| 329 |
> |
if (self.selectTotalNumberEvents and self.selectNumberOfJobs): |
| 330 |
> |
eventsPerJobRequested = int(eventsRemaining/self.theNumberOfJobs) |
| 331 |
|
|
| 332 |
< |
common.logger.debug(5,'Check '+str(check)) |
| 332 |
> |
if (self.selectNumberOfJobs): |
| 333 |
> |
common.logger.message("May not create the exact number_of_jobs requested.") |
| 334 |
|
|
| 335 |
< |
if check > 0: |
| 336 |
< |
self.total_number_of_jobs = self.total_number_of_jobs + 1 |
| 337 |
< |
common.logger.message('Warning: last job will be created with '+str(check)+' files') |
| 335 |
> |
if ( self.ncjobs == 'all' ) : |
| 336 |
> |
totalNumberOfJobs = 999999999 |
| 337 |
> |
else : |
| 338 |
> |
totalNumberOfJobs = self.ncjobs |
| 339 |
> |
|
| 340 |
|
|
| 341 |
< |
common.logger.message(str(self.total_number_of_jobs)+' jobs will be created for a total of '+str((self.total_number_of_jobs-1)*self.filesPerJob*evPerFile + check*evPerFile)+' events') |
| 342 |
< |
pass |
| 341 |
> |
blocks = blockSites.keys() |
| 342 |
> |
blockCount = 0 |
| 343 |
> |
# Backup variable in case self.maxEvents counted events in a non-included block |
| 344 |
> |
numBlocksInDataset = len(blocks) |
| 345 |
|
|
| 346 |
+ |
jobCount = 0 |
| 347 |
|
list_of_lists = [] |
| 339 |
– |
for i in xrange(0, int(n_tot_files), self.filesPerJob): |
| 340 |
– |
list_of_lists.append(self.files[0][i: i+self.filesPerJob]) |
| 348 |
|
|
| 349 |
< |
self.list_of_files = list_of_lists |
| 350 |
< |
|
| 349 |
> |
# ---- Iterate over the blocks in the dataset until ---- # |
| 350 |
> |
# ---- we've met the requested total # of events ---- # |
| 351 |
> |
while ( (eventsRemaining > 0) and (blockCount < numBlocksInDataset) and (jobCount < totalNumberOfJobs)): |
| 352 |
> |
block = blocks[blockCount] |
| 353 |
> |
|
| 354 |
> |
|
| 355 |
> |
evInBlock = self.eventsbyblock[block] |
| 356 |
> |
common.logger.debug(5,'Events in Block File '+str(evInBlock)) |
| 357 |
> |
|
| 358 |
> |
#Correct - switch to this when DBS up |
| 359 |
> |
#numEventsInBlock = self.eventsbyblock[block] |
| 360 |
> |
numEventsInBlock = evInBlock |
| 361 |
> |
|
| 362 |
> |
files = self.filesbyblock[block] |
| 363 |
> |
numFilesInBlock = len(files) |
| 364 |
> |
if (numFilesInBlock <= 0): |
| 365 |
> |
continue |
| 366 |
> |
fileCount = 0 |
| 367 |
> |
|
| 368 |
> |
# ---- New block => New job ---- # |
| 369 |
> |
parString = "\\{" |
| 370 |
> |
# counter for number of events in files currently worked on |
| 371 |
> |
filesEventCount = 0 |
| 372 |
> |
# flag if next while loop should touch new file |
| 373 |
> |
newFile = 1 |
| 374 |
> |
# job event counter |
| 375 |
> |
jobSkipEventCount = 0 |
| 376 |
> |
|
| 377 |
> |
# ---- Iterate over the files in the block until we've met the requested ---- # |
| 378 |
> |
# ---- total # of events or we've gone over all the files in this block ---- # |
| 379 |
> |
while ( (eventsRemaining > 0) and (fileCount < numFilesInBlock) and (jobCount < totalNumberOfJobs) ): |
| 380 |
> |
file = files[fileCount] |
| 381 |
> |
if newFile : |
| 382 |
> |
try: |
| 383 |
> |
numEventsInFile = self.eventsbyfile[file] |
| 384 |
> |
common.logger.debug(6, "File "+str(file)+" has "+str(numEventsInFile)+" events") |
| 385 |
> |
# increase filesEventCount |
| 386 |
> |
filesEventCount += numEventsInFile |
| 387 |
> |
# Add file to current job |
| 388 |
> |
parString += '\\\"' + file + '\\\"\,' |
| 389 |
> |
newFile = 0 |
| 390 |
> |
except KeyError: |
| 391 |
> |
common.logger.message("File "+str(file)+" has unknown numbe of events: skipping") |
| 392 |
> |
|
| 393 |
> |
|
| 394 |
> |
# if less events in file remain than eventsPerJobRequested |
| 395 |
> |
if ( filesEventCount - jobSkipEventCount < eventsPerJobRequested ) : |
| 396 |
> |
# if last file in block |
| 397 |
> |
if ( fileCount == numFilesInBlock ) : |
| 398 |
> |
# end job using last file, use remaining events in block |
| 399 |
> |
# close job and touch new file |
| 400 |
> |
fullString = parString[:-2] |
| 401 |
> |
fullString += '\\}' |
| 402 |
> |
list_of_lists.append([fullString,str(-1),str(jobSkipEventCount)]) |
| 403 |
> |
common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(filesEventCount - jobSkipEventCount)+" events (last file in block).") |
| 404 |
> |
self.jobDestination.append(blockSites[block]) |
| 405 |
> |
common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount])) |
| 406 |
> |
# reset counter |
| 407 |
> |
jobCount = jobCount + 1 |
| 408 |
> |
totalEventCount = totalEventCount + eventsPerJobRequested |
| 409 |
> |
eventsRemaining = eventsRemaining - eventsPerJobRequested |
| 410 |
> |
jobSkipEventCount = 0 |
| 411 |
> |
# reset file |
| 412 |
> |
parString = "\\{" |
| 413 |
> |
filesEventCount = 0 |
| 414 |
> |
newFile = 1 |
| 415 |
> |
fileCount += 1 |
| 416 |
> |
else : |
| 417 |
> |
# go to next file |
| 418 |
> |
newFile = 1 |
| 419 |
> |
fileCount += 1 |
| 420 |
> |
# if events in file equal to eventsPerJobRequested |
| 421 |
> |
elif ( filesEventCount - jobSkipEventCount == eventsPerJobRequested ) : |
| 422 |
> |
# close job and touch new file |
| 423 |
> |
fullString = parString[:-2] |
| 424 |
> |
fullString += '\\}' |
| 425 |
> |
list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)]) |
| 426 |
> |
common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.") |
| 427 |
> |
self.jobDestination.append(blockSites[block]) |
| 428 |
> |
common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount])) |
| 429 |
> |
# reset counter |
| 430 |
> |
jobCount = jobCount + 1 |
| 431 |
> |
totalEventCount = totalEventCount + eventsPerJobRequested |
| 432 |
> |
eventsRemaining = eventsRemaining - eventsPerJobRequested |
| 433 |
> |
jobSkipEventCount = 0 |
| 434 |
> |
# reset file |
| 435 |
> |
parString = "\\{" |
| 436 |
> |
filesEventCount = 0 |
| 437 |
> |
newFile = 1 |
| 438 |
> |
fileCount += 1 |
| 439 |
> |
|
| 440 |
> |
# if more events in file remain than eventsPerJobRequested |
| 441 |
> |
else : |
| 442 |
> |
# close job but don't touch new file |
| 443 |
> |
fullString = parString[:-2] |
| 444 |
> |
fullString += '\\}' |
| 445 |
> |
list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)]) |
| 446 |
> |
common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.") |
| 447 |
> |
self.jobDestination.append(blockSites[block]) |
| 448 |
> |
common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount])) |
| 449 |
> |
# increase counter |
| 450 |
> |
jobCount = jobCount + 1 |
| 451 |
> |
totalEventCount = totalEventCount + eventsPerJobRequested |
| 452 |
> |
eventsRemaining = eventsRemaining - eventsPerJobRequested |
| 453 |
> |
# calculate skip events for last file |
| 454 |
> |
# use filesEventCount (contains several files), jobSkipEventCount and eventsPerJobRequest |
| 455 |
> |
jobSkipEventCount = eventsPerJobRequested - (filesEventCount - jobSkipEventCount - self.eventsbyfile[file]) |
| 456 |
> |
# remove all but the last file |
| 457 |
> |
filesEventCount = self.eventsbyfile[file] |
| 458 |
> |
parString = "\\{" |
| 459 |
> |
parString += '\\\"' + file + '\\\"\,' |
| 460 |
> |
pass # END if |
| 461 |
> |
pass # END while (iterate over files in the block) |
| 462 |
> |
pass # END while (iterate over blocks in the dataset) |
| 463 |
> |
self.ncjobs = self.total_number_of_jobs = jobCount |
| 464 |
> |
if (eventsRemaining > 0 and jobCount < totalNumberOfJobs ): |
| 465 |
> |
common.logger.message("Could not run on all requested events because some blocks not hosted at allowed sites.") |
| 466 |
> |
common.logger.message("\n"+str(jobCount)+" job(s) can run on "+str(totalEventCount)+" events.\n") |
| 467 |
> |
|
| 468 |
> |
self.list_of_args = list_of_lists |
| 469 |
> |
return |
| 470 |
> |
|
| 471 |
> |
def jobSplittingNoInput(self): |
| 472 |
> |
""" |
| 473 |
> |
Perform job splitting based on number of event per job |
| 474 |
> |
""" |
| 475 |
> |
common.logger.debug(5,'Splitting per events') |
| 476 |
> |
common.logger.message('Required '+str(self.eventsPerJob)+' events per job ') |
| 477 |
> |
common.logger.message('Required '+str(self.theNumberOfJobs)+' jobs in total ') |
| 478 |
> |
common.logger.message('Required '+str(self.total_number_of_events)+' events in total ') |
| 479 |
> |
|
| 480 |
> |
if (self.total_number_of_events < 0): |
| 481 |
> |
msg='Cannot split jobs per Events with "-1" as total number of events' |
| 482 |
> |
raise CrabException(msg) |
| 483 |
> |
|
| 484 |
> |
if (self.selectEventsPerJob): |
| 485 |
> |
self.total_number_of_jobs = int(self.total_number_of_events/self.eventsPerJob) |
| 486 |
> |
elif (self.selectNumberOfJobs) : |
| 487 |
> |
self.total_number_of_jobs = self.theNumberOfJobs |
| 488 |
> |
self.eventsPerJob = int(self.total_number_of_events/self.total_number_of_jobs) |
| 489 |
> |
|
| 490 |
> |
common.logger.debug(5,'N jobs '+str(self.total_number_of_jobs)) |
| 491 |
> |
|
| 492 |
> |
# is there any remainder? |
| 493 |
> |
check = int(self.total_number_of_events) - (int(self.total_number_of_jobs)*self.eventsPerJob) |
| 494 |
> |
|
| 495 |
> |
common.logger.debug(5,'Check '+str(check)) |
| 496 |
> |
|
| 497 |
> |
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') |
| 498 |
> |
if check > 0: |
| 499 |
> |
common.logger.message('Warning: asked '+str(self.total_number_of_events)+' but can do only '+str(int(self.total_number_of_jobs)*self.eventsPerJob)) |
| 500 |
> |
|
| 501 |
> |
# argument is seed number.$i |
| 502 |
> |
self.list_of_args = [] |
| 503 |
> |
for i in range(self.total_number_of_jobs): |
| 504 |
> |
## Since there is no input, any site is good |
| 505 |
> |
self.jobDestination.append(["Any"]) |
| 506 |
> |
if (self.sourceSeed): |
| 507 |
> |
if (self.sourceSeedVtx): |
| 508 |
> |
## pythia + vtx random seed |
| 509 |
> |
self.list_of_args.append([ |
| 510 |
> |
str(self.sourceSeed)+str(i), |
| 511 |
> |
str(self.sourceSeedVtx)+str(i) |
| 512 |
> |
]) |
| 513 |
> |
else: |
| 514 |
> |
## only pythia random seed |
| 515 |
> |
self.list_of_args.append([(str(self.sourceSeed)+str(i))]) |
| 516 |
> |
else: |
| 517 |
> |
## no random seed |
| 518 |
> |
self.list_of_args.append([str(i)]) |
| 519 |
> |
#print self.list_of_args |
| 520 |
> |
|
| 521 |
|
return |
| 522 |
|
|
| 523 |
|
def split(self, jobParams): |
| 525 |
|
common.jobDB.load() |
| 526 |
|
#### Fabio |
| 527 |
|
njobs = self.total_number_of_jobs |
| 528 |
< |
filelist = self.list_of_files |
| 528 |
> |
arglist = self.list_of_args |
| 529 |
|
# create the empty structure |
| 530 |
|
for i in range(njobs): |
| 531 |
|
jobParams.append("") |
| 532 |
|
|
| 533 |
|
for job in range(njobs): |
| 534 |
< |
jobParams[job] = filelist[job] |
| 534 |
> |
jobParams[job] = arglist[job] |
| 535 |
> |
# print str(arglist[job]) |
| 536 |
> |
# print jobParams[job] |
| 537 |
|
common.jobDB.setArguments(job, jobParams[job]) |
| 538 |
+ |
common.logger.debug(5,"Job "+str(job)+" Destination: "+str(self.jobDestination[job])) |
| 539 |
+ |
common.jobDB.setDestination(job, self.jobDestination[job]) |
| 540 |
|
|
| 541 |
|
common.jobDB.save() |
| 542 |
|
return |
| 543 |
|
|
| 544 |
|
def getJobTypeArguments(self, nj, sched): |
| 545 |
< |
params = common.jobDB.arguments(nj) |
| 546 |
< |
#print params |
| 547 |
< |
parString = "\\{" |
| 548 |
< |
|
| 368 |
< |
for i in range(len(params) - 1): |
| 369 |
< |
parString += '\\\"' + params[i] + '\\\"\,' |
| 370 |
< |
|
| 371 |
< |
parString += '\\\"' + params[len(params) - 1] + '\\\"\\}' |
| 372 |
< |
return parString |
| 545 |
> |
result = '' |
| 546 |
> |
for i in common.jobDB.arguments(nj): |
| 547 |
> |
result=result+str(i)+" " |
| 548 |
> |
return result |
| 549 |
|
|
| 550 |
|
def numberOfJobs(self): |
| 551 |
|
# Fabio |
| 376 |
– |
|
| 552 |
|
return self.total_number_of_jobs |
| 378 |
– |
|
| 379 |
– |
|
| 380 |
– |
|
| 381 |
– |
def checkBlackList(self, allSites): |
| 382 |
– |
if len(self.reCEBlackList)==0: return allSites |
| 383 |
– |
sites = [] |
| 384 |
– |
for site in allSites: |
| 385 |
– |
common.logger.debug(10,'Site '+site) |
| 386 |
– |
good=1 |
| 387 |
– |
for re in self.reCEBlackList: |
| 388 |
– |
if re.search(site): |
| 389 |
– |
common.logger.message('CE in black list, skipping site '+site) |
| 390 |
– |
good=0 |
| 391 |
– |
pass |
| 392 |
– |
if good: sites.append(site) |
| 393 |
– |
if len(sites) == 0: |
| 394 |
– |
common.logger.debug(3,"No sites found after BlackList") |
| 395 |
– |
return sites |
| 396 |
– |
|
| 397 |
– |
def checkWhiteList(self, allSites): |
| 398 |
– |
|
| 399 |
– |
if len(self.reCEWhiteList)==0: return allSites |
| 400 |
– |
sites = [] |
| 401 |
– |
for site in allSites: |
| 402 |
– |
good=0 |
| 403 |
– |
for re in self.reCEWhiteList: |
| 404 |
– |
if re.search(site): |
| 405 |
– |
common.logger.debug(5,'CE in white list, adding site '+site) |
| 406 |
– |
good=1 |
| 407 |
– |
if not good: continue |
| 408 |
– |
sites.append(site) |
| 409 |
– |
if len(sites) == 0: |
| 410 |
– |
common.logger.message("No sites found after WhiteList\n") |
| 411 |
– |
else: |
| 412 |
– |
common.logger.debug(5,"Selected sites via WhiteList are "+str(sites)+"\n") |
| 413 |
– |
return sites |
| 553 |
|
|
| 554 |
|
def getTarBall(self, exe): |
| 555 |
|
""" |
| 655 |
|
txt += ' echo "JOB_EXIT_STATUS = 10016"\n' |
| 656 |
|
txt += ' echo "JobExitCode=10016" | tee -a $RUNTIME_AREA/$repo\n' |
| 657 |
|
txt += ' dumpStatus $RUNTIME_AREA/$repo\n' |
| 658 |
+ |
txt += ' rm -f $RUNTIME_AREA/$repo \n' |
| 659 |
+ |
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n' |
| 660 |
+ |
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n' |
| 661 |
|
txt += ' exit 1\n' |
| 662 |
|
txt += ' fi\n' |
| 663 |
|
txt += '\n' |
| 677 |
|
txt += ' echo "JOB_EXIT_STATUS = 10034"\n' |
| 678 |
|
txt += ' echo "JobExitCode=10034" | tee -a $RUNTIME_AREA/$repo\n' |
| 679 |
|
txt += ' dumpStatus $RUNTIME_AREA/$repo\n' |
| 680 |
+ |
txt += ' rm -f $RUNTIME_AREA/$repo \n' |
| 681 |
+ |
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n' |
| 682 |
+ |
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n' |
| 683 |
|
## OLI_Daniele |
| 684 |
|
txt += ' if [ $middleware == OSG ]; then \n' |
| 685 |
|
txt += ' echo "Remove working directory: $WORKING_DIR"\n' |
| 690 |
|
txt += ' echo "JOB_EXIT_STATUS = 10018"\n' |
| 691 |
|
txt += ' echo "JobExitCode=10018" | tee -a $RUNTIME_AREA/$repo\n' |
| 692 |
|
txt += ' dumpStatus $RUNTIME_AREA/$repo\n' |
| 693 |
+ |
txt += ' rm -f $RUNTIME_AREA/$repo \n' |
| 694 |
+ |
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n' |
| 695 |
+ |
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n' |
| 696 |
|
txt += ' fi\n' |
| 697 |
|
txt += ' fi \n' |
| 698 |
|
txt += ' exit 1 \n' |
| 706 |
|
txt += "\n" |
| 707 |
|
txt += "## number of arguments (first argument always jobnumber)\n" |
| 708 |
|
txt += "\n" |
| 709 |
< |
txt += "narg=$#\n" |
| 710 |
< |
txt += "if [ $narg -lt 2 ]\n" |
| 709 |
> |
# txt += "narg=$#\n" |
| 710 |
> |
txt += "if [ $nargs -lt 2 ]\n" |
| 711 |
|
txt += "then\n" |
| 712 |
< |
txt += " echo 'SET_EXE_ENV 1 ==> ERROR Too few arguments' +$narg+ \n" |
| 712 |
> |
txt += " echo 'SET_EXE_ENV 1 ==> ERROR Too few arguments' +$nargs+ \n" |
| 713 |
|
txt += ' echo "JOB_EXIT_STATUS = 50113"\n' |
| 714 |
|
txt += ' echo "JobExitCode=50113" | tee -a $RUNTIME_AREA/$repo\n' |
| 715 |
|
txt += ' dumpStatus $RUNTIME_AREA/$repo\n' |
| 716 |
+ |
txt += ' rm -f $RUNTIME_AREA/$repo \n' |
| 717 |
+ |
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n' |
| 718 |
+ |
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n' |
| 719 |
|
## OLI_Daniele |
| 720 |
|
txt += ' if [ $middleware == OSG ]; then \n' |
| 721 |
|
txt += ' echo "Remove working directory: $WORKING_DIR"\n' |
| 726 |
|
txt += ' echo "JOB_EXIT_STATUS = 50114"\n' |
| 727 |
|
txt += ' echo "JobExitCode=50114" | tee -a $RUNTIME_AREA/$repo\n' |
| 728 |
|
txt += ' dumpStatus $RUNTIME_AREA/$repo\n' |
| 729 |
+ |
txt += ' rm -f $RUNTIME_AREA/$repo \n' |
| 730 |
+ |
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n' |
| 731 |
+ |
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n' |
| 732 |
|
txt += ' fi\n' |
| 733 |
|
txt += ' fi \n' |
| 734 |
|
txt += " exit 1\n" |
| 739 |
|
job = common.job_list[nj] |
| 740 |
|
pset = os.path.basename(job.configFilename()) |
| 741 |
|
txt += '\n' |
| 742 |
< |
txt += 'InputFiles=$2\n' |
| 743 |
< |
txt += 'echo "<$InputFiles>"\n' |
| 744 |
< |
#txt += 'echo sed "s#{\'INPUT\'}#$InputFiles#" $RUNTIME_AREA/'+pset+' \n' |
| 745 |
< |
txt += 'sed "s#{\'INPUT\'}#$InputFiles#" $RUNTIME_AREA/'+pset+' > pset.cfg\n' |
| 746 |
< |
#txt += 'sed "s#{\'INPUT\'}#${InputFiles}#" $RUNTIME_AREA/'+pset+' > pset1.cfg\n' |
| 742 |
> |
if (self.datasetPath): # standard job |
| 743 |
> |
#txt += 'InputFiles=$2\n' |
| 744 |
> |
txt += 'InputFiles=${args[1]}\n' |
| 745 |
> |
txt += 'MaxEvents=${args[2]}\n' |
| 746 |
> |
txt += 'SkipEvents=${args[3]}\n' |
| 747 |
> |
txt += 'echo "Inputfiles:<$InputFiles>"\n' |
| 748 |
> |
txt += 'sed "s#{\'INPUT\'}#$InputFiles#" $RUNTIME_AREA/'+pset+' > pset_tmp_1.cfg\n' |
| 749 |
> |
txt += 'echo "MaxEvents:<$MaxEvents>"\n' |
| 750 |
> |
txt += 'sed "s#INPUTMAXEVENTS#$MaxEvents#" $RUNTIME_AREA/ pset_tmp_1.cfg > pset_tmp_2.cfg\n' |
| 751 |
> |
txt += 'echo "SkipEvents:<$SkipEvents>"\n' |
| 752 |
> |
txt += 'sed "s#INPUTSKIPEVENTS#$SkipEvents#" $RUNTIME_AREA/ pset_tmp_2.cfg > pset.cfg\n' |
| 753 |
> |
else: # pythia like job |
| 754 |
> |
if (self.sourceSeed): |
| 755 |
> |
# txt += 'Seed=$2\n' |
| 756 |
> |
txt += 'Seed=${args[1]}\n' |
| 757 |
> |
txt += 'echo "Seed: <$Seed>"\n' |
| 758 |
> |
txt += 'sed "s#\<INPUT\>#$Seed#" $RUNTIME_AREA/'+pset+' > tmp.cfg\n' |
| 759 |
> |
if (self.sourceSeedVtx): |
| 760 |
> |
# txt += 'VtxSeed=$3\n' |
| 761 |
> |
txt += 'VtxSeed=${args[2]}\n' |
| 762 |
> |
txt += 'echo "VtxSeed: <$VtxSeed>"\n' |
| 763 |
> |
txt += 'sed "s#INPUTVTX#$VtxSeed#" tmp.cfg > pset.cfg\n' |
| 764 |
> |
else: |
| 765 |
> |
txt += 'mv tmp.cfg pset.cfg\n' |
| 766 |
> |
else: |
| 767 |
> |
txt += '# Copy untouched pset\n' |
| 768 |
> |
txt += 'cp $RUNTIME_AREA/'+pset+' pset.cfg\n' |
| 769 |
> |
|
| 770 |
|
|
| 771 |
|
if len(self.additional_inbox_files) > 0: |
| 772 |
|
for file in self.additional_inbox_files: |
| 773 |
< |
txt += 'if [ -e $RUNTIME_AREA/'+file+' ] ; then\n' |
| 774 |
< |
txt += ' cp $RUNTIME_AREA/'+file+' .\n' |
| 775 |
< |
txt += ' chmod +x '+file+'\n' |
| 773 |
> |
relFile = file.split("/")[-1] |
| 774 |
> |
txt += 'if [ -e $RUNTIME_AREA/'+relFile+' ] ; then\n' |
| 775 |
> |
txt += ' cp $RUNTIME_AREA/'+relFile+' .\n' |
| 776 |
> |
txt += ' chmod +x '+relFile+'\n' |
| 777 |
|
txt += 'fi\n' |
| 778 |
|
pass |
| 779 |
|
|
| 810 |
|
txt += ' cd $RUNTIME_AREA\n' |
| 811 |
|
txt += ' /bin/rm -rf $WORKING_DIR\n' |
| 812 |
|
txt += ' if [ -d $WORKING_DIR ] ;then\n' |
| 813 |
< |
txt += ' echo "SET_EXE 50999 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after Untarring .tgz file failed"\n' |
| 814 |
< |
txt += ' echo "JOB_EXIT_STATUS = 50999"\n' |
| 815 |
< |
txt += ' echo "JobExitCode=50999" | tee -a $RUNTIME_AREA/$repo\n' |
| 816 |
< |
txt += ' dumpStatus $RUNTIME_AREA/$repo\n' |
| 813 |
> |
txt += ' echo "SET_EXE 50999 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after Untarring .tgz file failed"\n' |
| 814 |
> |
txt += ' echo "JOB_EXIT_STATUS = 50999"\n' |
| 815 |
> |
txt += ' echo "JobExitCode=50999" | tee -a $RUNTIME_AREA/$repo\n' |
| 816 |
> |
txt += ' dumpStatus $RUNTIME_AREA/$repo\n' |
| 817 |
> |
txt += ' rm -f $RUNTIME_AREA/$repo \n' |
| 818 |
> |
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n' |
| 819 |
> |
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n' |
| 820 |
|
txt += ' fi\n' |
| 821 |
|
txt += ' fi \n' |
| 822 |
|
txt += ' \n' |
| 894 |
|
txt += '\n' |
| 895 |
|
txt += '# check output file\n' |
| 896 |
|
txt += 'ls '+fileWithSuffix+'\n' |
| 897 |
< |
txt += 'exe_result=$?\n' |
| 898 |
< |
txt += 'if [ $exe_result -ne 0 ] ; then\n' |
| 899 |
< |
txt += ' echo "ERROR: No output file to manage"\n' |
| 900 |
< |
txt += ' echo "JOB_EXIT_STATUS = $exe_result"\n' |
| 901 |
< |
txt += ' echo "JobExitCode=60302" | tee -a $RUNTIME_AREA/$repo\n' |
| 902 |
< |
txt += ' dumpStatus $RUNTIME_AREA/$repo\n' |
| 897 |
> |
txt += 'ls_result=$?\n' |
| 898 |
> |
#txt += 'exe_result=$?\n' |
| 899 |
> |
txt += 'if [ $ls_result -ne 0 ] ; then\n' |
| 900 |
> |
txt += ' echo "ERROR: Problem with output file"\n' |
| 901 |
> |
#txt += ' echo "JOB_EXIT_STATUS = $exe_result"\n' |
| 902 |
> |
#txt += ' echo "JobExitCode=60302" | tee -a $RUNTIME_AREA/$repo\n' |
| 903 |
> |
#txt += ' dumpStatus $RUNTIME_AREA/$repo\n' |
| 904 |
|
### OLI_DANIELE |
| 905 |
|
if common.scheduler.boss_scheduler_name == 'condor_g': |
| 906 |
|
txt += ' if [ $middleware == OSG ]; then \n' |
| 914 |
|
txt += 'cd $RUNTIME_AREA\n' |
| 915 |
|
file_list=file_list[:-1] |
| 916 |
|
txt += 'file_list="'+file_list+'"\n' |
| 917 |
+ |
txt += 'cd $RUNTIME_AREA\n' |
| 918 |
|
### OLI_DANIELE |
| 919 |
|
txt += 'if [ $middleware == OSG ]; then\n' |
| 920 |
|
txt += ' cd $RUNTIME_AREA\n' |
| 925 |
|
txt += ' echo "JOB_EXIT_STATUS = 60999"\n' |
| 926 |
|
txt += ' echo "JobExitCode=60999" | tee -a $RUNTIME_AREA/$repo\n' |
| 927 |
|
txt += ' dumpStatus $RUNTIME_AREA/$repo\n' |
| 928 |
+ |
txt += ' rm -f $RUNTIME_AREA/$repo \n' |
| 929 |
+ |
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n' |
| 930 |
+ |
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n' |
| 931 |
|
txt += ' fi\n' |
| 932 |
|
txt += 'fi\n' |
| 933 |
|
txt += '\n' |
| 945 |
|
# add "_txt" |
| 946 |
|
if len(p)>1: |
| 947 |
|
ext = p[len(p)-1] |
| 762 |
– |
#result = name + '_' + str(txt) + "." + ext |
| 948 |
|
result = name + '_' + txt + "." + ext |
| 949 |
|
else: |
| 765 |
– |
#result = name + '_' + str(txt) |
| 950 |
|
result = name + '_' + txt |
| 951 |
|
|
| 952 |
|
return result |
| 956 |
|
return job requirements to add to jdl files |
| 957 |
|
""" |
| 958 |
|
req = '' |
| 959 |
< |
if common.analisys_common_info['sites']: |
| 960 |
< |
if common.analisys_common_info['sw_version']: |
| 961 |
< |
req='Member("VO-cms-' + \ |
| 962 |
< |
common.analisys_common_info['sw_version'] + \ |
| 963 |
< |
'", other.GlueHostApplicationSoftwareRunTimeEnvironment)' |
| 964 |
< |
if len(common.analisys_common_info['sites'])>0: |
| 965 |
< |
req = req + ' && (' |
| 782 |
< |
for i in range(len(common.analisys_common_info['sites'])): |
| 783 |
< |
req = req + 'other.GlueCEInfoHostName == "' \ |
| 784 |
< |
+ common.analisys_common_info['sites'][i] + '"' |
| 785 |
< |
if ( i < (int(len(common.analisys_common_info['sites']) - 1)) ): |
| 786 |
< |
req = req + ' || ' |
| 787 |
< |
req = req + ')' |
| 788 |
< |
#print "req = ", req |
| 959 |
> |
if common.analisys_common_info['sw_version']: |
| 960 |
> |
req='Member("VO-cms-' + \ |
| 961 |
> |
common.analisys_common_info['sw_version'] + \ |
| 962 |
> |
'", other.GlueHostApplicationSoftwareRunTimeEnvironment)' |
| 963 |
> |
|
| 964 |
> |
req = req + ' && (other.GlueHostNetworkAdapterOutboundIP)' |
| 965 |
> |
|
| 966 |
|
return req |
| 967 |
|
|
| 968 |
|
def configFilename(self): |
| 980 |
|
txt += ' if [ -f $GRID3_APP_DIR/cmssoft/cmsset_default.sh ] ;then\n' |
| 981 |
|
txt += ' # Use $GRID3_APP_DIR/cmssoft/cmsset_default.sh to setup cms software\n' |
| 982 |
|
txt += ' source $GRID3_APP_DIR/cmssoft/cmsset_default.sh '+self.version+'\n' |
| 983 |
< |
txt += ' elif [ -f $OSG_APP/cmssoft/cmsset_default.sh ] ;then\n' |
| 984 |
< |
txt += ' # Use $OSG_APP/cmssoft/cmsset_default.sh to setup cms software\n' |
| 985 |
< |
txt += ' source $OSG_APP/cmssoft/cmsset_default.sh '+self.version+'\n' |
| 983 |
> |
txt += ' elif [ -f $OSG_APP/cmssoft/cms/cmsset_default.sh ] ;then\n' |
| 984 |
> |
txt += ' # Use $OSG_APP/cmssoft/cms/cmsset_default.sh to setup cms software\n' |
| 985 |
> |
txt += ' source $OSG_APP/cmssoft/cms/cmsset_default.sh '+self.version+'\n' |
| 986 |
|
txt += ' else\n' |
| 987 |
< |
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' |
| 987 |
> |
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' |
| 988 |
|
txt += ' echo "JOB_EXIT_STATUS = 10020"\n' |
| 989 |
|
txt += ' echo "JobExitCode=10020" | tee -a $RUNTIME_AREA/$repo\n' |
| 990 |
|
txt += ' dumpStatus $RUNTIME_AREA/$repo\n' |
| 991 |
+ |
txt += ' rm -f $RUNTIME_AREA/$repo \n' |
| 992 |
+ |
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n' |
| 993 |
+ |
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n' |
| 994 |
|
txt += ' exit 1\n' |
| 995 |
|
txt += '\n' |
| 996 |
|
txt += ' echo "Remove working directory: $WORKING_DIR"\n' |
| 997 |
|
txt += ' cd $RUNTIME_AREA\n' |
| 998 |
|
txt += ' /bin/rm -rf $WORKING_DIR\n' |
| 999 |
|
txt += ' if [ -d $WORKING_DIR ] ;then\n' |
| 1000 |
< |
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' |
| 1000 |
> |
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' |
| 1001 |
|
txt += ' echo "JOB_EXIT_STATUS = 10017"\n' |
| 1002 |
|
txt += ' echo "JobExitCode=10017" | tee -a $RUNTIME_AREA/$repo\n' |
| 1003 |
|
txt += ' dumpStatus $RUNTIME_AREA/$repo\n' |
| 1004 |
+ |
txt += ' rm -f $RUNTIME_AREA/$repo \n' |
| 1005 |
+ |
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n' |
| 1006 |
+ |
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n' |
| 1007 |
|
txt += ' fi\n' |
| 1008 |
|
txt += '\n' |
| 1009 |
|
txt += ' exit 1\n' |
| 1027 |
|
txt += ' echo "JOB_EXIT_STATUS = 10031" \n' |
| 1028 |
|
txt += ' echo "JobExitCode=10031" | tee -a $RUNTIME_AREA/$repo\n' |
| 1029 |
|
txt += ' dumpStatus $RUNTIME_AREA/$repo\n' |
| 1030 |
+ |
txt += ' rm -f $RUNTIME_AREA/$repo \n' |
| 1031 |
+ |
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n' |
| 1032 |
+ |
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n' |
| 1033 |
|
txt += ' exit 1\n' |
| 1034 |
|
txt += ' else\n' |
| 1035 |
|
txt += ' echo "Sourcing environment... "\n' |
| 1038 |
|
txt += ' echo "JOB_EXIT_STATUS = 10020"\n' |
| 1039 |
|
txt += ' echo "JobExitCode=10020" | 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 "sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n' |
| 1051 |
|
txt += ' echo "JOB_EXIT_STATUS = 10032"\n' |
| 1052 |
|
txt += ' echo "JobExitCode=10032" | tee -a $RUNTIME_AREA/$repo\n' |
| 1053 |
|
txt += ' dumpStatus $RUNTIME_AREA/$repo\n' |
| 1054 |
+ |
txt += ' rm -f $RUNTIME_AREA/$repo \n' |
| 1055 |
+ |
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n' |
| 1056 |
+ |
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n' |
| 1057 |
|
txt += ' exit 1\n' |
| 1058 |
|
txt += ' fi\n' |
| 1059 |
|
txt += ' fi\n' |
| 1070 |
|
txt += ' echo "JOB_EXIT_STATUS = 10033"\n' |
| 1071 |
|
txt += ' echo "JobExitCode=10033" | tee -a $RUNTIME_AREA/$repo\n' |
| 1072 |
|
txt += ' dumpStatus $RUNTIME_AREA/$repo\n' |
| 1073 |
+ |
txt += ' rm -f $RUNTIME_AREA/$repo \n' |
| 1074 |
+ |
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n' |
| 1075 |
+ |
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n' |
| 1076 |
|
txt += ' exit 1\n' |
| 1077 |
|
txt += ' fi\n' |
| 1078 |
|
txt += ' echo "SET_CMS_ENV 0 ==> setup cms environment ok"\n' |
| 1084 |
|
|
| 1085 |
|
def getParams(self): |
| 1086 |
|
return self._params |
| 1087 |
+ |
|
| 1088 |
+ |
def setTaskid_(self): |
| 1089 |
+ |
self._taskId = self.cfg_params['taskId'] |
| 1090 |
+ |
|
| 1091 |
+ |
def getTaskid(self): |
| 1092 |
+ |
return self._taskId |
| 1093 |
+ |
|
| 1094 |
+ |
####################################################################### |
| 1095 |
+ |
def uniquelist(self, old): |
| 1096 |
+ |
""" |
| 1097 |
+ |
remove duplicates from a list |
| 1098 |
+ |
""" |
| 1099 |
+ |
nd={} |
| 1100 |
+ |
for e in old: |
| 1101 |
+ |
nd[e]=0 |
| 1102 |
+ |
return nd.keys() |
