| 10 |
|
import os, string, glob |
| 11 |
|
|
| 12 |
|
class Cmssw(JobType): |
| 13 |
< |
def __init__(self, cfg_params, ncjobs): |
| 13 |
> |
def __init__(self, cfg_params, ncjobs,skip_blocks, isNew): |
| 14 |
|
JobType.__init__(self, 'CMSSW') |
| 15 |
|
common.logger.debug(3,'CMSSW::__init__') |
| 16 |
< |
|
| 16 |
> |
self.skip_blocks = skip_blocks |
| 17 |
> |
|
| 18 |
|
self.argsList = [] |
| 19 |
|
|
| 20 |
|
self._params = {} |
| 36 |
|
self.executable_arch = self.scram.getArch() |
| 37 |
|
self.tgz_name = 'default.tgz' |
| 38 |
|
self.scriptName = 'CMSSW.sh' |
| 39 |
< |
self.pset = '' #scrip use case Da |
| 40 |
< |
self.datasetPath = '' #scrip use case Da |
| 39 |
> |
self.pset = '' |
| 40 |
> |
self.datasetPath = '' |
| 41 |
|
|
| 42 |
|
# set FJR file name |
| 43 |
|
self.fjrFileName = 'crab_fjr.xml' |
| 44 |
|
|
| 45 |
|
self.version = self.scram.getSWVersion() |
| 46 |
|
version_array = self.version.split('_') |
| 47 |
< |
self.major_version = 0 |
| 48 |
< |
self.minor_version = 0 |
| 47 |
> |
self.CMSSW_major = 0 |
| 48 |
> |
self.CMSSW_minor = 0 |
| 49 |
> |
self.CMSSW_patch = 0 |
| 50 |
|
try: |
| 51 |
< |
self.major_version = int(version_array[1]) |
| 52 |
< |
self.minor_version = int(version_array[2]) |
| 51 |
> |
self.CMSSW_major = int(version_array[1]) |
| 52 |
> |
self.CMSSW_minor = int(version_array[2]) |
| 53 |
> |
self.CMSSW_patch = int(version_array[3]) |
| 54 |
|
except: |
| 55 |
< |
msg = "Cannot parse CMSSW version string: " + "_".join(version_array) + " for major and minor release number!" |
| 55 |
> |
msg = "Cannot parse CMSSW version string: " + self.version + " for major and minor release number!" |
| 56 |
|
raise CrabException(msg) |
| 57 |
|
|
| 55 |
– |
|
| 56 |
– |
# |
| 57 |
– |
# Try to block creation in case of arch/version mismatch |
| 58 |
– |
# |
| 59 |
– |
|
| 60 |
– |
# a = string.split(self.version, "_") |
| 61 |
– |
# |
| 62 |
– |
# if int(a[1]) == 1 and (int(a[2]) < 5 and self.executable_arch.find('slc4') == 0): |
| 63 |
– |
# msg = "Warning: You are using %s version of CMSSW with %s architecture. \n--> Did you compile your libraries with SLC3? Otherwise you can find some problems running on SLC4 Grid nodes.\n"%(self.version, self.executable_arch) |
| 64 |
– |
# common.logger.message(msg) |
| 65 |
– |
# if int(a[1]) == 1 and (int(a[2]) >= 5 and self.executable_arch.find('slc3') == 0): |
| 66 |
– |
# msg = "Error: CMS does not support %s with %s architecture"%(self.version, self.executable_arch) |
| 67 |
– |
# raise CrabException(msg) |
| 68 |
– |
# |
| 69 |
– |
|
| 70 |
– |
|
| 58 |
|
### collect Data cards |
| 59 |
|
|
| 60 |
|
if not cfg_params.has_key('CMSSW.datasetpath'): |
| 61 |
|
msg = "Error: datasetpath not defined " |
| 62 |
|
raise CrabException(msg) |
| 63 |
+ |
|
| 64 |
+ |
### Temporary: added to remove input file control in the case of PU |
| 65 |
+ |
if not cfg_params.has_key('USER.dataset_pu'): |
| 66 |
+ |
self.dataset_pu = 'NONE' |
| 67 |
+ |
else: |
| 68 |
+ |
self.dataset_pu = cfg_params['USER.dataset_pu'] |
| 69 |
+ |
#### |
| 70 |
+ |
|
| 71 |
|
tmp = cfg_params['CMSSW.datasetpath'] |
| 72 |
|
log.debug(6, "CMSSW::CMSSW(): datasetPath = "+tmp) |
| 73 |
|
if string.lower(tmp)=='none': |
| 78 |
|
self.selectNoInput = 0 |
| 79 |
|
|
| 80 |
|
self.dataTiers = [] |
| 81 |
< |
|
| 81 |
> |
self.debugWrap = '' |
| 82 |
> |
self.debug_wrapper = cfg_params.get('USER.debug_wrapper',False) |
| 83 |
> |
if self.debug_wrapper: self.debugWrap='--debug' |
| 84 |
|
## now the application |
| 85 |
|
self.executable = cfg_params.get('CMSSW.executable','cmsRun') |
| 86 |
|
log.debug(6, "CMSSW::CMSSW(): executable = "+self.executable) |
| 103 |
|
self.output_file_sandbox.append(self.fjrFileName) |
| 104 |
|
|
| 105 |
|
# other output files to be returned via sandbox or copied to SE |
| 106 |
+ |
outfileflag = False |
| 107 |
|
self.output_file = [] |
| 108 |
|
tmp = cfg_params.get('CMSSW.output_file',None) |
| 109 |
|
if tmp : |
| 110 |
< |
tmpOutFiles = string.split(tmp,',') |
| 111 |
< |
log.debug(7, 'cmssw::cmssw(): output files '+str(tmpOutFiles)) |
| 112 |
< |
for tmp in tmpOutFiles: |
| 113 |
< |
tmp=string.strip(tmp) |
| 116 |
< |
self.output_file.append(tmp) |
| 117 |
< |
pass |
| 118 |
< |
else: |
| 119 |
< |
log.message("No output file defined: only stdout/err and the CRAB Framework Job Report will be available\n") |
| 120 |
< |
pass |
| 110 |
> |
self.output_file = [x.strip() for x in tmp.split(',')] |
| 111 |
> |
outfileflag = True #output found |
| 112 |
> |
#else: |
| 113 |
> |
# log.message("No output file defined: only stdout/err and the CRAB Framework Job Report will be available\n") |
| 114 |
|
|
| 115 |
|
# script_exe file as additional file in inputSandbox |
| 116 |
|
self.scriptExe = cfg_params.get('USER.script_exe',None) |
| 120 |
|
raise CrabException(msg) |
| 121 |
|
self.additional_inbox_files.append(string.strip(self.scriptExe)) |
| 122 |
|
|
| 130 |
– |
#CarlosDaniele |
| 123 |
|
if self.datasetPath == None and self.pset == None and self.scriptExe == '' : |
| 124 |
|
msg ="Error. script_exe not defined" |
| 125 |
|
raise CrabException(msg) |
| 126 |
|
|
| 127 |
+ |
# use parent files... |
| 128 |
+ |
self.useParent = self.cfg_params.get('CMSSW.use_parent',False) |
| 129 |
+ |
|
| 130 |
|
## additional input files |
| 131 |
|
if cfg_params.has_key('USER.additional_input_files'): |
| 132 |
|
tmpAddFiles = string.split(cfg_params['USER.additional_input_files'],',') |
| 145 |
|
if not os.path.exists(file): |
| 146 |
|
raise CrabException("Additional input file not found: "+file) |
| 147 |
|
pass |
| 153 |
– |
# fname = string.split(file, '/')[-1] |
| 154 |
– |
# storedFile = common.work_space.pathForTgz()+'share/'+fname |
| 155 |
– |
# shutil.copyfile(file, storedFile) |
| 148 |
|
self.additional_inbox_files.append(string.strip(file)) |
| 149 |
|
pass |
| 150 |
|
pass |
| 170 |
|
if cfg_params.has_key('CMSSW.total_number_of_events'): |
| 171 |
|
self.total_number_of_events = int(cfg_params['CMSSW.total_number_of_events']) |
| 172 |
|
self.selectTotalNumberEvents = 1 |
| 173 |
+ |
if self.selectNumberOfJobs == 1: |
| 174 |
+ |
if (self.total_number_of_events != -1) and int(self.total_number_of_events) < int(self.theNumberOfJobs): |
| 175 |
+ |
msg = 'Must specify at least one event per job. total_number_of_events > number_of_jobs ' |
| 176 |
+ |
raise CrabException(msg) |
| 177 |
|
else: |
| 178 |
|
self.total_number_of_events = 0 |
| 179 |
|
self.selectTotalNumberEvents = 0 |
| 180 |
|
|
| 181 |
< |
if self.pset != None: #CarlosDaniele |
| 181 |
> |
if self.pset != None: |
| 182 |
|
if ( (self.selectTotalNumberEvents + self.selectEventsPerJob + self.selectNumberOfJobs) != 2 ): |
| 183 |
|
msg = 'Must define exactly two of total_number_of_events, events_per_job, or number_of_jobs.' |
| 184 |
|
raise CrabException(msg) |
| 208 |
|
if self.sourceSeed: |
| 209 |
|
print "pythia_seed is a deprecated parameter. Use preserve_seeds or increment_seeds in the future.\n","Added to increment_seeds." |
| 210 |
|
self.incrementSeeds.append('sourceSeed') |
| 211 |
+ |
self.incrementSeeds.append('theSource') |
| 212 |
|
|
| 213 |
|
self.sourceSeedVtx = cfg_params.get('CMSSW.vtx_seed',None) |
| 214 |
|
if self.sourceSeedVtx: |
| 227 |
|
|
| 228 |
|
self.firstRun = cfg_params.get('CMSSW.first_run',None) |
| 229 |
|
|
| 233 |
– |
if self.pset != None: #CarlosDaniele |
| 234 |
– |
import PsetManipulator as pp |
| 235 |
– |
PsetEdit = pp.PsetManipulator(self.pset) #Daniele Pset |
| 230 |
|
|
| 231 |
|
# Copy/return |
| 238 |
– |
|
| 232 |
|
self.copy_data = int(cfg_params.get('USER.copy_data',0)) |
| 233 |
|
self.return_data = int(cfg_params.get('USER.return_data',0)) |
| 234 |
|
|
| 244 |
|
blockSites = self.DataDiscoveryAndLocation(cfg_params) |
| 245 |
|
#DBSDLS-end |
| 246 |
|
|
| 254 |
– |
|
| 247 |
|
## Select Splitting |
| 248 |
|
if self.selectNoInput: |
| 249 |
< |
if self.pset == None: #CarlosDaniele |
| 249 |
> |
if self.pset == None: |
| 250 |
|
self.jobSplittingForScript() |
| 251 |
|
else: |
| 252 |
|
self.jobSplittingNoInput() |
| 253 |
|
else: |
| 254 |
|
self.jobSplittingByBlocks(blockSites) |
| 255 |
|
|
| 256 |
< |
# modify Pset |
| 257 |
< |
if self.pset != None: #CarlosDaniele |
| 258 |
< |
try: |
| 259 |
< |
# Add FrameworkJobReport to parameter-set, set max events. |
| 260 |
< |
# Reset later for data jobs by writeCFG which does all modifications |
| 261 |
< |
PsetEdit.addCrabFJR(self.fjrFileName) # FUTURE: Job report addition not needed by CMSSW>1.5 |
| 262 |
< |
PsetEdit.maxEvent(self.eventsPerJob) |
| 263 |
< |
PsetEdit.psetWriter(self.configFilename()) |
| 264 |
< |
except: |
| 265 |
< |
msg='Error while manipuliating ParameterSet: exiting...' |
| 266 |
< |
raise CrabException(msg) |
| 267 |
< |
self.tgzNameWithPath = self.getTarBall(self.executable) |
| 256 |
> |
# modify Pset only the first time |
| 257 |
> |
if isNew: |
| 258 |
> |
if self.pset != None: |
| 259 |
> |
import PsetManipulator as pp |
| 260 |
> |
PsetEdit = pp.PsetManipulator(self.pset) |
| 261 |
> |
try: |
| 262 |
> |
# Add FrameworkJobReport to parameter-set, set max events. |
| 263 |
> |
# Reset later for data jobs by writeCFG which does all modifications |
| 264 |
> |
PsetEdit.addCrabFJR(self.fjrFileName) # FUTURE: Job report addition not needed by CMSSW>1.5 |
| 265 |
> |
PsetEdit.maxEvent(self.eventsPerJob) |
| 266 |
> |
PsetEdit.psetWriter(self.configFilename()) |
| 267 |
> |
## If present, add TFileService to output files |
| 268 |
> |
if not int(cfg_params.get('CMSSW.skip_TFileService_output',0)): |
| 269 |
> |
tfsOutput = PsetEdit.getTFileService() |
| 270 |
> |
if tfsOutput: |
| 271 |
> |
if tfsOutput in self.output_file: |
| 272 |
> |
common.logger.debug(5,"Output from TFileService "+tfsOutput+" already in output files") |
| 273 |
> |
else: |
| 274 |
> |
outfileflag = True #output found |
| 275 |
> |
self.output_file.append(tfsOutput) |
| 276 |
> |
common.logger.message("Adding "+tfsOutput+" to output files (from TFileService)") |
| 277 |
> |
pass |
| 278 |
> |
pass |
| 279 |
> |
## If present and requested, add PoolOutputModule to output files |
| 280 |
> |
if int(cfg_params.get('CMSSW.get_edm_output',0)): |
| 281 |
> |
edmOutput = PsetEdit.getPoolOutputModule() |
| 282 |
> |
if edmOutput: |
| 283 |
> |
if edmOutput in self.output_file: |
| 284 |
> |
common.logger.debug(5,"Output from PoolOutputModule "+edmOutput+" already in output files") |
| 285 |
> |
else: |
| 286 |
> |
self.output_file.append(edmOutput) |
| 287 |
> |
common.logger.message("Adding "+edmOutput+" to output files (from PoolOutputModule)") |
| 288 |
> |
pass |
| 289 |
> |
pass |
| 290 |
> |
except CrabException: |
| 291 |
> |
msg='Error while manipulating ParameterSet: exiting...' |
| 292 |
> |
raise CrabException(msg) |
| 293 |
> |
## Prepare inputSandbox TarBall (only the first time) |
| 294 |
> |
self.tgzNameWithPath = self.getTarBall(self.executable) |
| 295 |
|
|
| 296 |
|
def DataDiscoveryAndLocation(self, cfg_params): |
| 297 |
|
|
| 304 |
|
## Contact the DBS |
| 305 |
|
common.logger.message("Contacting Data Discovery Services ...") |
| 306 |
|
try: |
| 307 |
< |
self.pubdata=DataDiscovery.DataDiscovery(datasetPath, cfg_params) |
| 307 |
> |
self.pubdata=DataDiscovery.DataDiscovery(datasetPath, cfg_params,self.skip_blocks) |
| 308 |
|
self.pubdata.fetchDBSInfo() |
| 309 |
|
|
| 310 |
|
except DataDiscovery.NotExistingDatasetError, ex : |
| 320 |
|
self.filesbyblock=self.pubdata.getFiles() |
| 321 |
|
self.eventsbyblock=self.pubdata.getEventsPerBlock() |
| 322 |
|
self.eventsbyfile=self.pubdata.getEventsPerFile() |
| 323 |
+ |
self.parentFiles=self.pubdata.getParent() |
| 324 |
|
|
| 325 |
|
## get max number of events |
| 326 |
< |
self.maxEvents=self.pubdata.getMaxEvents() ## self.maxEvents used in Creator.py |
| 326 |
> |
self.maxEvents=self.pubdata.getMaxEvents() |
| 327 |
|
|
| 328 |
|
## Contact the DLS and build a list of sites hosting the fileblocks |
| 329 |
|
try: |
| 347 |
|
|
| 348 |
|
return sites |
| 349 |
|
|
| 330 |
– |
# to Be Removed DS -- BL |
| 331 |
– |
# def setArgsList(self, argsList): |
| 332 |
– |
# self.argsList = argsList |
| 333 |
– |
|
| 350 |
|
def jobSplittingByBlocks(self, blockSites): |
| 351 |
|
""" |
| 352 |
|
Perform job splitting. Jobs run over an integer number of files |
| 437 |
|
|
| 438 |
|
# ---- Iterate over the files in the block until we've met the requested ---- # |
| 439 |
|
# ---- total # of events or we've gone over all the files in this block ---- # |
| 440 |
+ |
pString='' |
| 441 |
|
while ( (eventsRemaining > 0) and (fileCount < numFilesInBlock) and (jobCount < totalNumberOfJobs) ): |
| 442 |
|
file = files[fileCount] |
| 443 |
+ |
if self.useParent: |
| 444 |
+ |
parent = self.parentFiles[file] |
| 445 |
+ |
for f in parent : |
| 446 |
+ |
pString += '\\\"' + f + '\\\"\,' |
| 447 |
+ |
common.logger.debug(6, "File "+str(file)+" has the following parents: "+str(parent)) |
| 448 |
+ |
common.logger.write("File "+str(file)+" has the following parents: "+str(parent)) |
| 449 |
|
if newFile : |
| 450 |
|
try: |
| 451 |
|
numEventsInFile = self.eventsbyfile[file] |
| 466 |
|
# end job using last file, use remaining events in block |
| 467 |
|
# close job and touch new file |
| 468 |
|
fullString = parString[:-2] |
| 469 |
< |
list_of_lists.append([fullString,str(-1),str(jobSkipEventCount)]) |
| 469 |
> |
if self.useParent: |
| 470 |
> |
fullParentString = pString[:-2] |
| 471 |
> |
list_of_lists.append([fullString,fullParentString,str(-1),str(jobSkipEventCount)]) |
| 472 |
> |
else: |
| 473 |
> |
list_of_lists.append([fullString,str(-1),str(jobSkipEventCount)]) |
| 474 |
|
common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(filesEventCount - jobSkipEventCount)+" events (last file in block).") |
| 475 |
|
self.jobDestination.append(blockSites[block]) |
| 476 |
|
common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount])) |
| 482 |
|
eventsRemaining = eventsRemaining - filesEventCount + jobSkipEventCount |
| 483 |
|
jobSkipEventCount = 0 |
| 484 |
|
# reset file |
| 485 |
+ |
pString = "" |
| 486 |
|
parString = "" |
| 487 |
|
filesEventCount = 0 |
| 488 |
|
newFile = 1 |
| 495 |
|
elif ( filesEventCount - jobSkipEventCount == eventsPerJobRequested ) : |
| 496 |
|
# close job and touch new file |
| 497 |
|
fullString = parString[:-2] |
| 498 |
< |
list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)]) |
| 498 |
> |
if self.useParent: |
| 499 |
> |
fullParentString = pString[:-2] |
| 500 |
> |
list_of_lists.append([fullString,fullParentString,str(eventsPerJobRequested),str(jobSkipEventCount)]) |
| 501 |
> |
else: |
| 502 |
> |
list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)]) |
| 503 |
|
common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.") |
| 504 |
|
self.jobDestination.append(blockSites[block]) |
| 505 |
|
common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount])) |
| 510 |
|
eventsRemaining = eventsRemaining - eventsPerJobRequested |
| 511 |
|
jobSkipEventCount = 0 |
| 512 |
|
# reset file |
| 513 |
+ |
pString = "" |
| 514 |
|
parString = "" |
| 515 |
|
filesEventCount = 0 |
| 516 |
|
newFile = 1 |
| 520 |
|
else : |
| 521 |
|
# close job but don't touch new file |
| 522 |
|
fullString = parString[:-2] |
| 523 |
< |
list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)]) |
| 523 |
> |
if self.useParent: |
| 524 |
> |
fullParentString = pString[:-2] |
| 525 |
> |
list_of_lists.append([fullString,fullParentString,str(eventsPerJobRequested),str(jobSkipEventCount)]) |
| 526 |
> |
else: |
| 527 |
> |
list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)]) |
| 528 |
|
common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.") |
| 529 |
|
self.jobDestination.append(blockSites[block]) |
| 530 |
|
common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount])) |
| 538 |
|
jobSkipEventCount = eventsPerJobRequested - (filesEventCount - jobSkipEventCount - self.eventsbyfile[file]) |
| 539 |
|
# remove all but the last file |
| 540 |
|
filesEventCount = self.eventsbyfile[file] |
| 541 |
+ |
if self.useParent: |
| 542 |
+ |
for f in parent : pString += '\\\"' + f + '\\\"\,' |
| 543 |
|
parString = '\\\"' + file + '\\\"\,' |
| 544 |
|
pass # END if |
| 545 |
|
pass # END while (iterate over files in the block) |
| 648 |
|
return |
| 649 |
|
|
| 650 |
|
|
| 651 |
< |
def jobSplittingForScript(self):#CarlosDaniele |
| 651 |
> |
def jobSplittingForScript(self): |
| 652 |
|
""" |
| 653 |
|
Perform job splitting based on number of job |
| 654 |
|
""" |
| 664 |
|
# argument is seed number.$i |
| 665 |
|
self.list_of_args = [] |
| 666 |
|
for i in range(self.total_number_of_jobs): |
| 628 |
– |
## Since there is no input, any site is good |
| 629 |
– |
# self.jobDestination.append(["Any"]) |
| 667 |
|
self.jobDestination.append([""]) |
| 631 |
– |
## no random seed |
| 668 |
|
self.list_of_args.append([str(i)]) |
| 669 |
|
return |
| 670 |
|
|
| 671 |
< |
def split(self, jobParams): |
| 671 |
> |
def split(self, jobParams,firstJobID): |
| 672 |
|
|
| 637 |
– |
#### Fabio |
| 673 |
|
njobs = self.total_number_of_jobs |
| 674 |
|
arglist = self.list_of_args |
| 675 |
|
# create the empty structure |
| 678 |
|
|
| 679 |
|
listID=[] |
| 680 |
|
listField=[] |
| 681 |
< |
for job in range(njobs): |
| 682 |
< |
jobParams[job] = arglist[job] |
| 681 |
> |
for id in range(njobs): |
| 682 |
> |
job = id + int(firstJobID) |
| 683 |
> |
jobParams[id] = arglist[id] |
| 684 |
|
listID.append(job+1) |
| 685 |
|
job_ToSave ={} |
| 686 |
|
concString = ' ' |
| 687 |
|
argu='' |
| 688 |
< |
if len(jobParams[job]): |
| 689 |
< |
argu += concString.join(jobParams[job] ) |
| 690 |
< |
job_ToSave['arguments']= str(job+1)+' '+argu## new BL--DS |
| 691 |
< |
job_ToSave['dlsDestination']= self.jobDestination[job]## new BL--DS |
| 656 |
< |
#common._db.updateJob_(job,job_ToSave)## new BL--DS |
| 688 |
> |
if len(jobParams[id]): |
| 689 |
> |
argu += concString.join(jobParams[id] ) |
| 690 |
> |
job_ToSave['arguments']= str(job+1)+' '+argu |
| 691 |
> |
job_ToSave['dlsDestination']= self.jobDestination[id] |
| 692 |
|
listField.append(job_ToSave) |
| 693 |
|
msg="Job "+str(job)+" Arguments: "+str(job+1)+" "+argu+"\n" \ |
| 694 |
< |
+" Destination: "+str(self.jobDestination[job]) |
| 694 |
> |
+" Destination: "+str(self.jobDestination[id]) |
| 695 |
|
common.logger.debug(5,msg) |
| 696 |
< |
#common.logger.debug(5,"Job "+str(job)+" Destination: "+str(self.jobDestination[job])) |
| 662 |
< |
common._db.updateJob_(listID,listField)## new BL--DS |
| 663 |
< |
## Pay Attention Here....DS--BL |
| 696 |
> |
common._db.updateJob_(listID,listField) |
| 697 |
|
self.argsList = (len(jobParams[0])+1) |
| 698 |
|
|
| 699 |
|
return |
| 700 |
|
|
| 701 |
|
def numberOfJobs(self): |
| 669 |
– |
# Fabio |
| 702 |
|
return self.total_number_of_jobs |
| 703 |
|
|
| 704 |
|
def getTarBall(self, exe): |
| 705 |
|
""" |
| 706 |
|
Return the TarBall with lib and exe |
| 707 |
|
""" |
| 676 |
– |
|
| 677 |
– |
# if it exist, just return it |
| 678 |
– |
# |
| 679 |
– |
# Marco. Let's start to use relative path for Boss XML files |
| 680 |
– |
# |
| 708 |
|
self.tgzNameWithPath = common.work_space.pathForTgz()+'share/'+self.tgz_name |
| 709 |
|
if os.path.exists(self.tgzNameWithPath): |
| 710 |
|
return self.tgzNameWithPath |
| 718 |
|
|
| 719 |
|
# First of all declare the user Scram area |
| 720 |
|
swArea = self.scram.getSWArea_() |
| 694 |
– |
#print "swArea = ", swArea |
| 695 |
– |
# swVersion = self.scram.getSWVersion() |
| 696 |
– |
# print "swVersion = ", swVersion |
| 721 |
|
swReleaseTop = self.scram.getReleaseTop_() |
| 698 |
– |
#print "swReleaseTop = ", swReleaseTop |
| 722 |
|
|
| 723 |
|
## check if working area is release top |
| 724 |
|
if swReleaseTop == '' or swArea == swReleaseTop: |
| 764 |
|
tar.add(module,moduleDir) |
| 765 |
|
|
| 766 |
|
## Now check if any data dir(s) is present |
| 744 |
– |
swAreaLen=len(swArea) |
| 767 |
|
self.dataExist = False |
| 768 |
< |
for root, dirs, files in os.walk(swArea): |
| 769 |
< |
if "data" in dirs: |
| 770 |
< |
self.dataExist=True |
| 771 |
< |
common.logger.debug(5,"data "+root+"/data"+" to be tarred") |
| 772 |
< |
tar.add(root+"/data",root[swAreaLen:]+"/data") |
| 768 |
> |
todo_list = [(i, i) for i in os.listdir(swArea+"/src")] |
| 769 |
> |
while len(todo_list): |
| 770 |
> |
entry, name = todo_list.pop() |
| 771 |
> |
if name.startswith('crab_0_') or name.startswith('.') or name == 'CVS': |
| 772 |
> |
continue |
| 773 |
> |
if os.path.isdir(swArea+"/src/"+entry): |
| 774 |
> |
entryPath = entry + '/' |
| 775 |
> |
todo_list += [(entryPath + i, i) for i in os.listdir(swArea+"/src/"+entry)] |
| 776 |
> |
if name == 'data': |
| 777 |
> |
self.dataExist=True |
| 778 |
> |
common.logger.debug(5,"data "+entry+" to be tarred") |
| 779 |
> |
tar.add(swArea+"/src/"+entry,"src/"+entry) |
| 780 |
> |
pass |
| 781 |
> |
pass |
| 782 |
|
|
| 783 |
|
### CMSSW ParameterSet |
| 784 |
|
if not self.pset is None: |
| 788 |
|
|
| 789 |
|
|
| 790 |
|
## Add ProdCommon dir to tar |
| 791 |
< |
prodcommonDir = 'ProdCommon' |
| 792 |
< |
prodcommonPath = os.environ['CRABDIR'] + '/' + 'ProdCommon' |
| 793 |
< |
if os.path.isdir(prodcommonPath): |
| 794 |
< |
tar.add(prodcommonPath,prodcommonDir) |
| 791 |
> |
prodcommonDir = './' |
| 792 |
> |
prodcommonPath = os.environ['CRABDIR'] + '/' + 'external/' |
| 793 |
> |
neededStuff = ['ProdCommon/__init__.py','ProdCommon/FwkJobRep', 'ProdCommon/CMSConfigTools','ProdCommon/Core','ProdCommon/MCPayloads', 'IMProv'] |
| 794 |
> |
for file in neededStuff: |
| 795 |
> |
tar.add(prodcommonPath+file,prodcommonDir+file) |
| 796 |
|
common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames())) |
| 797 |
|
|
| 798 |
|
##### ML stuff |
| 803 |
|
common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames())) |
| 804 |
|
|
| 805 |
|
##### Utils |
| 806 |
< |
Utils_file_list=['parseCrabFjr.py','writeCfg.py', 'JobReportErrorCode.py'] |
| 806 |
> |
Utils_file_list=['parseCrabFjr.py','writeCfg.py', 'fillCrabFjr.py'] |
| 807 |
|
for file in Utils_file_list: |
| 808 |
|
tar.add(path+file,file) |
| 809 |
|
common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames())) |
| 814 |
|
common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames())) |
| 815 |
|
|
| 816 |
|
tar.close() |
| 817 |
< |
except : |
| 818 |
< |
raise CrabException('Could not create tar-ball') |
| 817 |
> |
except IOError: |
| 818 |
> |
raise CrabException('Could not create tar-ball '+self.tgzNameWithPath) |
| 819 |
> |
except tarfile.TarError: |
| 820 |
> |
raise CrabException('Could not create tar-ball '+self.tgzNameWithPath) |
| 821 |
|
|
| 822 |
|
## check for tarball size |
| 823 |
|
tarballinfo = os.stat(self.tgzNameWithPath) |
| 831 |
|
Returns part of a job script which prepares |
| 832 |
|
the execution environment for the job 'nj'. |
| 833 |
|
""" |
| 834 |
+ |
if (self.CMSSW_major >= 2 and self.CMSSW_minor >= 1) or (self.CMSSW_major >= 3): |
| 835 |
+ |
psetName = 'pset.py' |
| 836 |
+ |
else: |
| 837 |
+ |
psetName = 'pset.cfg' |
| 838 |
|
# Prepare JobType-independent part |
| 839 |
|
txt = '\n#Written by cms_cmssw::wsSetupEnvironment\n' |
| 840 |
|
txt += 'echo ">>> setup environment"\n' |
| 853 |
|
txt += ' cd $WORKING_DIR\n' |
| 854 |
|
txt += ' echo ">>> current directory (WORKING_DIR): $WORKING_DIR"\n' |
| 855 |
|
txt += self.wsSetupCMSOSGEnvironment_() |
| 818 |
– |
#txt += ' echo "### Set SCRAM ARCH to ' + self.executable_arch + ' ###"\n' |
| 819 |
– |
#txt += ' export SCRAM_ARCH='+self.executable_arch+'\n' |
| 856 |
|
txt += 'fi\n' |
| 857 |
|
|
| 858 |
|
# Prepare JobType-specific part |
| 891 |
|
|
| 892 |
|
# Prepare job-specific part |
| 893 |
|
job = common.job_list[nj] |
| 858 |
– |
### FEDE FOR DBS OUTPUT PUBLICATION |
| 894 |
|
if (self.datasetPath): |
| 895 |
|
txt += '\n' |
| 896 |
|
txt += 'DatasetPath='+self.datasetPath+'\n' |
| 912 |
|
txt += 'cp $RUNTIME_AREA/'+pset+' .\n' |
| 913 |
|
if (self.datasetPath): # standard job |
| 914 |
|
txt += 'InputFiles=${args[1]}; export InputFiles\n' |
| 915 |
< |
txt += 'MaxEvents=${args[2]}; export MaxEvents\n' |
| 916 |
< |
txt += 'SkipEvents=${args[3]}; export SkipEvents\n' |
| 915 |
> |
if (self.useParent): |
| 916 |
> |
txt += 'ParentFiles=${args[2]}; export ParentFiles\n' |
| 917 |
> |
txt += 'MaxEvents=${args[3]}; export MaxEvents\n' |
| 918 |
> |
txt += 'SkipEvents=${args[4]}; export SkipEvents\n' |
| 919 |
> |
else: |
| 920 |
> |
txt += 'MaxEvents=${args[2]}; export MaxEvents\n' |
| 921 |
> |
txt += 'SkipEvents=${args[3]}; export SkipEvents\n' |
| 922 |
|
txt += 'echo "Inputfiles:<$InputFiles>"\n' |
| 923 |
+ |
if (self.useParent): txt += 'echo "ParentFiles:<$ParentFiles>"\n' |
| 924 |
|
txt += 'echo "MaxEvents:<$MaxEvents>"\n' |
| 925 |
|
txt += 'echo "SkipEvents:<$SkipEvents>"\n' |
| 926 |
|
else: # pythia like job |
| 932 |
|
txt += 'FirstRun=${args[1]}; export FirstRun\n' |
| 933 |
|
txt += 'echo "FirstRun: <$FirstRun>"\n' |
| 934 |
|
|
| 935 |
< |
txt += 'mv -f '+pset+' pset.cfg\n' |
| 935 |
> |
txt += 'mv -f ' + pset + ' ' + psetName + '\n' |
| 936 |
|
|
| 937 |
|
|
| 938 |
|
if self.pset != None: |
| 939 |
+ |
# FUTURE: Can simply for 2_1_x and higher |
| 940 |
|
txt += '\n' |
| 941 |
< |
txt += 'echo "***** cat pset.cfg *********"\n' |
| 942 |
< |
txt += 'cat pset.cfg\n' |
| 943 |
< |
txt += 'echo "****** end pset.cfg ********"\n' |
| 944 |
< |
txt += '\n' |
| 945 |
< |
txt += 'PSETHASH=`EdmConfigHash < pset.cfg` \n' |
| 941 |
> |
if self.debug_wrapper==True: |
| 942 |
> |
txt += 'echo "***** cat ' + psetName + ' *********"\n' |
| 943 |
> |
txt += 'cat ' + psetName + '\n' |
| 944 |
> |
txt += 'echo "****** end ' + psetName + ' ********"\n' |
| 945 |
> |
txt += '\n' |
| 946 |
> |
txt += 'PSETHASH=`edmConfigHash < ' + psetName + '` \n' |
| 947 |
|
txt += 'echo "PSETHASH = $PSETHASH" \n' |
| 948 |
|
txt += '\n' |
| 949 |
|
return txt |
| 959 |
|
if os.path.isfile(self.tgzNameWithPath): |
| 960 |
|
txt += 'echo ">>> tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+' :" \n' |
| 961 |
|
txt += 'tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+'\n' |
| 962 |
< |
txt += 'ls -Al \n' |
| 962 |
> |
if self.debug_wrapper: |
| 963 |
> |
txt += 'ls -Al \n' |
| 964 |
|
txt += 'untar_status=$? \n' |
| 965 |
|
txt += 'if [ $untar_status -ne 0 ]; then \n' |
| 966 |
|
txt += ' echo "ERROR ==> Untarring .tgz file failed"\n' |
| 970 |
|
txt += ' echo "Successful untar" \n' |
| 971 |
|
txt += 'fi \n' |
| 972 |
|
txt += '\n' |
| 973 |
< |
txt += 'echo ">>> Include ProdCommon in PYTHONPATH:"\n' |
| 973 |
> |
txt += 'echo ">>> Include $RUNTIME_AREA in PYTHONPATH:"\n' |
| 974 |
|
txt += 'if [ -z "$PYTHONPATH" ]; then\n' |
| 975 |
< |
txt += ' export PYTHONPATH=$RUNTIME_AREA/ProdCommon\n' |
| 975 |
> |
txt += ' export PYTHONPATH=$RUNTIME_AREA/\n' |
| 976 |
|
txt += 'else\n' |
| 977 |
< |
txt += ' export PYTHONPATH=$RUNTIME_AREA/ProdCommon:${PYTHONPATH}\n' |
| 977 |
> |
txt += ' export PYTHONPATH=$RUNTIME_AREA/:${PYTHONPATH}\n' |
| 978 |
|
txt += 'echo "PYTHONPATH=$PYTHONPATH"\n' |
| 979 |
|
txt += 'fi\n' |
| 980 |
|
txt += '\n' |
| 995 |
|
txt += 'rm -r lib/ module/ \n' |
| 996 |
|
txt += 'mv $RUNTIME_AREA/lib/ . \n' |
| 997 |
|
txt += 'mv $RUNTIME_AREA/module/ . \n' |
| 998 |
< |
if self.dataExist == True: txt += 'mv $RUNTIME_AREA/src/ . \n' |
| 998 |
> |
if self.dataExist == True: |
| 999 |
> |
txt += 'rm -r src/ \n' |
| 1000 |
> |
txt += 'mv $RUNTIME_AREA/src/ . \n' |
| 1001 |
|
if len(self.additional_inbox_files)>0: |
| 1002 |
|
for file in self.additional_inbox_files: |
| 1003 |
< |
txt += 'mv $RUNTIME_AREA/'+file+' . \n' |
| 1004 |
< |
txt += 'mv $RUNTIME_AREA/ProdCommon/ . \n' |
| 1003 |
> |
txt += 'mv $RUNTIME_AREA/'+os.path.basename(file)+' . \n' |
| 1004 |
> |
# txt += 'mv $RUNTIME_AREA/ProdCommon/ . \n' |
| 1005 |
> |
# txt += 'mv $RUNTIME_AREA/IMProv/ . \n' |
| 1006 |
|
|
| 1007 |
+ |
txt += 'echo ">>> Include $RUNTIME_AREA in PYTHONPATH:"\n' |
| 1008 |
|
txt += 'if [ -z "$PYTHONPATH" ]; then\n' |
| 1009 |
< |
txt += ' export PYTHONPATH=$SOFTWARE_DIR/ProdCommon\n' |
| 1009 |
> |
txt += ' export PYTHONPATH=$RUNTIME_AREA/\n' |
| 1010 |
|
txt += 'else\n' |
| 1011 |
< |
txt += ' export PYTHONPATH=$SOFTWARE_DIR/ProdCommon:${PYTHONPATH}\n' |
| 1011 |
> |
txt += ' export PYTHONPATH=$RUNTIME_AREA/:${PYTHONPATH}\n' |
| 1012 |
|
txt += 'echo "PYTHONPATH=$PYTHONPATH"\n' |
| 1013 |
|
txt += 'fi\n' |
| 1014 |
|
txt += '\n' |
| 1015 |
|
|
| 1016 |
|
return txt |
| 1017 |
|
|
| 970 |
– |
def modifySteeringCards(self, nj): |
| 971 |
– |
""" |
| 972 |
– |
modify the card provided by the user, |
| 973 |
– |
writing a new card into share dir |
| 974 |
– |
""" |
| 1018 |
|
|
| 1019 |
|
def executableName(self): |
| 1020 |
< |
if self.scriptExe: #CarlosDaniele |
| 1020 |
> |
if self.scriptExe: |
| 1021 |
|
return "sh " |
| 1022 |
|
else: |
| 1023 |
|
return self.executable |
| 1030 |
|
ex_args = "" |
| 1031 |
|
# FUTURE: This tests the CMSSW version. Can remove code as versions deprecated |
| 1032 |
|
# Framework job report |
| 1033 |
< |
if (self.major_version >= 1 and self.minor_version >= 5) or (self.major_version >= 2): |
| 1033 |
> |
if (self.CMSSW_major >= 1 and self.CMSSW_minor >= 5) or (self.CMSSW_major >= 2): |
| 1034 |
|
ex_args += " -j $RUNTIME_AREA/crab_fjr_$NJob.xml" |
| 1035 |
< |
# Type of cfg file |
| 1036 |
< |
if self.major_version >= 2 : |
| 1035 |
> |
# Type of config file |
| 1036 |
> |
if self.CMSSW_major >= 2 : |
| 1037 |
|
ex_args += " -p pset.py" |
| 1038 |
|
else: |
| 1039 |
|
ex_args += " -p pset.cfg" |
| 1044 |
|
Returns a list of filenames to be put in JDL input sandbox. |
| 1045 |
|
""" |
| 1046 |
|
inp_box = [] |
| 1004 |
– |
# # dict added to delete duplicate from input sandbox file list |
| 1005 |
– |
# seen = {} |
| 1006 |
– |
## code |
| 1047 |
|
if os.path.isfile(self.tgzNameWithPath): |
| 1048 |
|
inp_box.append(self.tgzNameWithPath) |
| 1049 |
|
wrapper = os.path.basename(str(common._db.queryTask('scriptName'))) |
| 1059 |
|
## User Declared output files |
| 1060 |
|
for out in (self.output_file+self.output_file_sandbox): |
| 1061 |
|
n_out = nj + 1 |
| 1062 |
< |
out_box.append(self.numberFile_(out,str(n_out))) |
| 1062 |
> |
out_box.append(numberFile(out,str(n_out))) |
| 1063 |
|
return out_box |
| 1064 |
|
|
| 1025 |
– |
def prepareSteeringCards(self): |
| 1026 |
– |
""" |
| 1027 |
– |
Make initial modifications of the user's steering card file. |
| 1028 |
– |
""" |
| 1029 |
– |
return |
| 1065 |
|
|
| 1066 |
|
def wsRenameOutput(self, nj): |
| 1067 |
|
""" |
| 1071 |
|
txt = '\n#Written by cms_cmssw::wsRenameOutput\n' |
| 1072 |
|
txt += 'echo ">>> current directory (SOFTWARE_DIR): $SOFTWARE_DIR" \n' |
| 1073 |
|
txt += 'echo ">>> current directory content:"\n' |
| 1074 |
< |
txt += 'ls \n' |
| 1074 |
> |
if self.debug_wrapper: |
| 1075 |
> |
txt += 'ls -Al\n' |
| 1076 |
|
txt += '\n' |
| 1077 |
|
|
| 1078 |
|
for fileWithSuffix in (self.output_file): |
| 1079 |
< |
output_file_num = self.numberFile_(fileWithSuffix, '$NJob') |
| 1079 |
> |
output_file_num = numberFile(fileWithSuffix, '$NJob') |
| 1080 |
|
txt += '\n' |
| 1081 |
|
txt += '# check output file\n' |
| 1082 |
|
txt += 'if [ -e ./'+fileWithSuffix+' ] ; then\n' |
| 1097 |
|
txt += 'fi\n' |
| 1098 |
|
file_list = [] |
| 1099 |
|
for fileWithSuffix in (self.output_file): |
| 1100 |
< |
file_list.append(self.numberFile_(fileWithSuffix, '$NJob')) |
| 1100 |
> |
file_list.append(numberFile(fileWithSuffix, '$NJob')) |
| 1101 |
|
|
| 1102 |
|
txt += 'file_list="'+string.join(file_list,' ')+'"\n' |
| 1103 |
|
txt += '\n' |
| 1104 |
|
txt += 'echo ">>> current directory (SOFTWARE_DIR): $SOFTWARE_DIR" \n' |
| 1105 |
|
txt += 'echo ">>> current directory content:"\n' |
| 1106 |
< |
txt += 'ls \n' |
| 1106 |
> |
if self.debug_wrapper: |
| 1107 |
> |
txt += 'ls -Al\n' |
| 1108 |
|
txt += '\n' |
| 1109 |
|
txt += 'cd $RUNTIME_AREA\n' |
| 1110 |
|
txt += 'echo ">>> current directory (RUNTIME_AREA): $RUNTIME_AREA"\n' |
| 1111 |
|
return txt |
| 1112 |
|
|
| 1076 |
– |
def numberFile_(self, file, txt): |
| 1077 |
– |
""" |
| 1078 |
– |
append _'txt' before last extension of a file |
| 1079 |
– |
""" |
| 1080 |
– |
p = string.split(file,".") |
| 1081 |
– |
# take away last extension |
| 1082 |
– |
name = p[0] |
| 1083 |
– |
for x in p[1:-1]: |
| 1084 |
– |
name=name+"."+x |
| 1085 |
– |
# add "_txt" |
| 1086 |
– |
if len(p)>1: |
| 1087 |
– |
ext = p[len(p)-1] |
| 1088 |
– |
result = name + '_' + txt + "." + ext |
| 1089 |
– |
else: |
| 1090 |
– |
result = name + '_' + txt |
| 1091 |
– |
|
| 1092 |
– |
return result |
| 1093 |
– |
|
| 1113 |
|
def getRequirements(self, nj=[]): |
| 1114 |
|
""" |
| 1115 |
|
return job requirements to add to jdl files |
| 1119 |
|
req='Member("VO-cms-' + \ |
| 1120 |
|
self.version + \ |
| 1121 |
|
'", other.GlueHostApplicationSoftwareRunTimeEnvironment)' |
| 1122 |
< |
## SL add requirement for OS version only if SL4 |
| 1104 |
< |
#reSL4 = re.compile( r'slc4' ) |
| 1105 |
< |
if self.executable_arch: # and reSL4.search(self.executable_arch): |
| 1122 |
> |
if self.executable_arch: |
| 1123 |
|
req+=' && Member("VO-cms-' + \ |
| 1124 |
|
self.executable_arch + \ |
| 1125 |
|
'", other.GlueHostApplicationSoftwareRunTimeEnvironment)' |
| 1133 |
|
def configFilename(self): |
| 1134 |
|
""" return the config filename """ |
| 1135 |
|
# FUTURE: Can remove cfg mode for CMSSW >= 2_1_x |
| 1136 |
< |
if (self.major_version >= 2 and self.minor_version >= 1) or (self.major_version >= 3): |
| 1136 |
> |
if (self.CMSSW_major >= 2 and self.CMSSW_minor >= 1) or (self.CMSSW_major >= 3): |
| 1137 |
|
return self.name()+'.py' |
| 1138 |
|
else: |
| 1139 |
|
return self.name()+'.cfg' |
| 1162 |
|
|
| 1163 |
|
return txt |
| 1164 |
|
|
| 1148 |
– |
### OLI_DANIELE |
| 1165 |
|
def wsSetupCMSLCGEnvironment_(self): |
| 1166 |
|
""" |
| 1167 |
|
Returns part of a job script which is prepares |
| 1196 |
|
txt += ' echo "==> setup cms environment ok"\n' |
| 1197 |
|
return txt |
| 1198 |
|
|
| 1183 |
– |
### FEDE FOR DBS OUTPUT PUBLICATION |
| 1199 |
|
def modifyReport(self, nj): |
| 1200 |
|
""" |
| 1201 |
|
insert the part of the script that modifies the FrameworkJob Report |
| 1202 |
|
""" |
| 1188 |
– |
|
| 1203 |
|
txt = '\n#Written by cms_cmssw::modifyReport\n' |
| 1204 |
|
publish_data = int(self.cfg_params.get('USER.publish_data',0)) |
| 1205 |
|
if (publish_data == 1): |
| 1215 |
|
txt += 'fi\n' |
| 1216 |
|
|
| 1217 |
|
txt += 'echo ">>> Modify Job Report:" \n' |
| 1218 |
< |
txt += 'chmod a+x $SOFTWARE_DIR/ProdCommon/ProdCommon/FwkJobRep/ModifyJobReport.py\n' |
| 1218 |
> |
txt += 'chmod a+x $RUNTIME_AREA/ProdCommon/FwkJobRep/ModifyJobReport.py\n' |
| 1219 |
|
txt += 'ProcessedDataset='+processedDataset+'\n' |
| 1220 |
|
txt += 'echo "ProcessedDataset = $ProcessedDataset"\n' |
| 1221 |
|
txt += 'echo "SE = $SE"\n' |
| 1222 |
|
txt += 'echo "SE_PATH = $SE_PATH"\n' |
| 1223 |
|
txt += 'echo "FOR_LFN = $FOR_LFN" \n' |
| 1224 |
|
txt += 'echo "CMSSW_VERSION = $CMSSW_VERSION"\n\n' |
| 1225 |
< |
txt += 'echo "$SOFTWARE_DIR/ProdCommon/ProdCommon/FwkJobRep/ModifyJobReport.py $RUNTIME_AREA/crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH"\n' |
| 1226 |
< |
txt += '$SOFTWARE_DIR/ProdCommon/ProdCommon/FwkJobRep/ModifyJobReport.py $RUNTIME_AREA/crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH\n' |
| 1225 |
> |
txt += 'echo "$RUNTIME_AREA/ProdCommon/FwkJobRep/ModifyJobReport.py $RUNTIME_AREA/crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH"\n' |
| 1226 |
> |
txt += '$RUNTIME_AREA/ProdCommon/FwkJobRep/ModifyJobReport.py $RUNTIME_AREA/crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH\n' |
| 1227 |
|
txt += 'modifyReport_result=$?\n' |
| 1228 |
|
txt += 'if [ $modifyReport_result -ne 0 ]; then\n' |
| 1229 |
|
txt += ' modifyReport_result=70500\n' |
| 1235 |
|
txt += 'fi\n' |
| 1236 |
|
return txt |
| 1237 |
|
|
| 1238 |
+ |
def wsParseFJR(self): |
| 1239 |
+ |
""" |
| 1240 |
+ |
Parse the FrameworkJobReport to obtain useful infos |
| 1241 |
+ |
""" |
| 1242 |
+ |
txt = '\n#Written by cms_cmssw::wsParseFJR\n' |
| 1243 |
+ |
txt += 'echo ">>> Parse FrameworkJobReport crab_fjr.xml"\n' |
| 1244 |
+ |
txt += 'if [ -s $RUNTIME_AREA/crab_fjr_$NJob.xml ]; then\n' |
| 1245 |
+ |
txt += ' if [ -s $RUNTIME_AREA/parseCrabFjr.py ]; then\n' |
| 1246 |
+ |
txt += ' cmd_out=`python $RUNTIME_AREA/parseCrabFjr.py --input $RUNTIME_AREA/crab_fjr_$NJob.xml --dashboard $MonitorID,$MonitorJobID '+self.debugWrap+'`\n' |
| 1247 |
+ |
if self.debug_wrapper : |
| 1248 |
+ |
txt += ' echo "Result of parsing the FrameworkJobReport crab_fjr.xml: $cmd_out"\n' |
| 1249 |
+ |
txt += ' executable_exit_status=`python $RUNTIME_AREA/parseCrabFjr.py --input $RUNTIME_AREA/crab_fjr_$NJob.xml --exitcode`\n' |
| 1250 |
+ |
txt += ' if [ $executable_exit_status -eq 50115 ];then\n' |
| 1251 |
+ |
txt += ' echo ">>> crab_fjr.xml contents: "\n' |
| 1252 |
+ |
txt += ' cat $RUNTIME_AREA/crab_fjr_NJob.xml\n' |
| 1253 |
+ |
txt += ' echo "Wrong FrameworkJobReport --> does not contain useful info. ExitStatus: $executable_exit_status"\n' |
| 1254 |
+ |
txt += ' elif [ $executable_exit_status -eq -999 ];then\n' |
| 1255 |
+ |
txt += ' echo "ExitStatus from FrameworkJobReport not available. not available. Using exit code of executable from command line."\n' |
| 1256 |
+ |
txt += ' else\n' |
| 1257 |
+ |
txt += ' echo "Extracted ExitStatus from FrameworkJobReport parsing output: $executable_exit_status"\n' |
| 1258 |
+ |
txt += ' fi\n' |
| 1259 |
+ |
txt += ' else\n' |
| 1260 |
+ |
txt += ' echo "CRAB python script to parse CRAB FrameworkJobReport crab_fjr.xml is not available, using exit code of executable from command line."\n' |
| 1261 |
+ |
txt += ' fi\n' |
| 1262 |
+ |
#### Patch to check input data reading for CMSSW16x Hopefully we-ll remove it asap |
| 1263 |
+ |
|
| 1264 |
+ |
if (self.datasetPath and self.dataset_pu == 'NONE'): |
| 1265 |
+ |
# VERIFY PROCESSED DATA |
| 1266 |
+ |
txt += ' if [ $executable_exit_status -eq 0 ];then\n' |
| 1267 |
+ |
txt += ' echo ">>> Verify list of processed files:"\n' |
| 1268 |
+ |
txt += ' echo $InputFiles |tr -d \'\\\\\' |tr \',\' \'\\n\'|tr -d \'"\' > input-files.txt\n' |
| 1269 |
+ |
txt += ' python $RUNTIME_AREA/parseCrabFjr.py --input $RUNTIME_AREA/crab_fjr_$NJob.xml --lfn > processed-files.txt\n' |
| 1270 |
+ |
txt += ' cat input-files.txt | sort | uniq > tmp.txt\n' |
| 1271 |
+ |
txt += ' mv tmp.txt input-files.txt\n' |
| 1272 |
+ |
txt += ' echo "cat input-files.txt"\n' |
| 1273 |
+ |
txt += ' echo "----------------------"\n' |
| 1274 |
+ |
txt += ' cat input-files.txt\n' |
| 1275 |
+ |
txt += ' cat processed-files.txt | sort | uniq > tmp.txt\n' |
| 1276 |
+ |
txt += ' mv tmp.txt processed-files.txt\n' |
| 1277 |
+ |
txt += ' echo "----------------------"\n' |
| 1278 |
+ |
txt += ' echo "cat processed-files.txt"\n' |
| 1279 |
+ |
txt += ' echo "----------------------"\n' |
| 1280 |
+ |
txt += ' cat processed-files.txt\n' |
| 1281 |
+ |
txt += ' echo "----------------------"\n' |
| 1282 |
+ |
txt += ' diff -q input-files.txt processed-files.txt\n' |
| 1283 |
+ |
txt += ' fileverify_status=$?\n' |
| 1284 |
+ |
txt += ' if [ $fileverify_status -ne 0 ]; then\n' |
| 1285 |
+ |
txt += ' executable_exit_status=30001\n' |
| 1286 |
+ |
txt += ' echo "ERROR ==> not all input files processed"\n' |
| 1287 |
+ |
txt += ' echo " ==> list of processed files from crab_fjr.xml differs from list in pset.cfg"\n' |
| 1288 |
+ |
txt += ' echo " ==> diff input-files.txt processed-files.txt"\n' |
| 1289 |
+ |
txt += ' fi\n' |
| 1290 |
+ |
txt += ' fi\n' |
| 1291 |
+ |
txt += '\n' |
| 1292 |
+ |
txt += 'else\n' |
| 1293 |
+ |
txt += ' echo "CRAB FrameworkJobReport crab_fjr.xml is not available, using exit code of executable from command line."\n' |
| 1294 |
+ |
txt += 'fi\n' |
| 1295 |
+ |
txt += '\n' |
| 1296 |
+ |
txt += 'echo "ExeExitCode=$executable_exit_status" | tee -a $RUNTIME_AREA/$repo\n' |
| 1297 |
+ |
txt += 'echo "EXECUTABLE_EXIT_STATUS = $executable_exit_status"\n' |
| 1298 |
+ |
txt += 'job_exit_code=$executable_exit_status\n' |
| 1299 |
+ |
|
| 1300 |
+ |
return txt |
| 1301 |
+ |
|
| 1302 |
|
def setParam_(self, param, value): |
| 1303 |
|
self._params[param] = value |
| 1304 |
|
|
| 1325 |
|
stderr = 'CMSSW_$NJob.stderr' |
| 1326 |
|
if (self.return_data == 1): |
| 1327 |
|
for file in (self.output_file+self.output_file_sandbox): |
| 1328 |
< |
listOutFiles.append(self.numberFile_(file, '$NJob')) |
| 1328 |
> |
listOutFiles.append(numberFile(file, '$NJob')) |
| 1329 |
|
listOutFiles.append(stdout) |
| 1330 |
|
listOutFiles.append(stderr) |
| 1331 |
|
else: |
| 1332 |
|
for file in (self.output_file_sandbox): |
| 1333 |
< |
listOutFiles.append(self.numberFile_(file, '$NJob')) |
| 1333 |
> |
listOutFiles.append(numberFile(file, '$NJob')) |
| 1334 |
|
listOutFiles.append(stdout) |
| 1335 |
|
listOutFiles.append(stderr) |
| 1336 |
|
txt += 'echo "output files: '+string.join(listOutFiles,' ')+'"\n' |
