| 2 |
|
from crab_logger import Logger |
| 3 |
|
from crab_exceptions import * |
| 4 |
|
from crab_util import * |
| 5 |
+ |
from BlackWhiteListParser import BlackWhiteListParser |
| 6 |
|
import common |
| 7 |
|
import Scram |
| 8 |
|
|
| 16 |
|
self._params = {} |
| 17 |
|
self.cfg_params = cfg_params |
| 18 |
|
|
| 19 |
+ |
# init BlackWhiteListParser |
| 20 |
+ |
self.blackWhiteListParser = BlackWhiteListParser(cfg_params) |
| 21 |
+ |
|
| 22 |
|
try: |
| 23 |
|
self.MaxTarBallSize = float(self.cfg_params['EDG.maxtarballsize']) |
| 24 |
|
except KeyError: |
| 44 |
|
self.fjrFileName = 'crab_fjr.xml' |
| 45 |
|
|
| 46 |
|
self.version = self.scram.getSWVersion() |
| 47 |
+ |
|
| 48 |
+ |
# |
| 49 |
+ |
# Try to block creation in case of arch/version mismatch |
| 50 |
+ |
# |
| 51 |
+ |
|
| 52 |
+ |
a = string.split(self.version, "_") |
| 53 |
+ |
|
| 54 |
+ |
if int(a[1]) == 1 and (int(a[2]) < 5 and self.executable_arch.find('slc4') == 0): |
| 55 |
+ |
msg = "Error: CMS does not support %s with %s architecture"%(self.version, self.executable_arch) |
| 56 |
+ |
raise CrabException(msg) |
| 57 |
+ |
if int(a[1]) == 1 and (int(a[2]) >= 5 and self.executable_arch.find('slc3') == 0): |
| 58 |
+ |
msg = "Error: CMS does not support %s with %s architecture"%(self.version, self.executable_arch) |
| 59 |
+ |
raise CrabException(msg) |
| 60 |
+ |
|
| 61 |
|
common.taskDB.setDict('codeVersion',self.version) |
| 62 |
|
self.setParam_('application', self.version) |
| 63 |
|
|
| 91 |
|
try: |
| 92 |
|
datasetpath_split = self.datasetPath.split("/") |
| 93 |
|
# standard style |
| 94 |
+ |
self.setParam_('datasetFull', self.datasetPath) |
| 95 |
|
if self.use_dbs_1 == 1 : |
| 96 |
|
self.setParam_('dataset', datasetpath_split[1]) |
| 97 |
|
self.setParam_('owner', datasetpath_split[-1]) |
| 319 |
|
if (self.sourceSeedVtx) : |
| 320 |
|
PsetEdit.vtxSeed("INPUTVTX") |
| 321 |
|
if (self.sourceSeedG4) : |
| 322 |
< |
self.PsetEdit.g4Seed("INPUTG4") |
| 322 |
> |
PsetEdit.g4Seed("INPUTG4") |
| 323 |
|
if (self.sourceSeedMix) : |
| 324 |
< |
self.PsetEdit.mixSeed("INPUTMIX") |
| 324 |
> |
PsetEdit.mixSeed("INPUTMIX") |
| 325 |
|
# add FrameworkJobReport to parameter-set |
| 326 |
|
PsetEdit.addCrabFJR(self.fjrFileName) |
| 327 |
|
PsetEdit.psetWriter(self.configFilename()) |
| 587 |
|
for block in blocks: |
| 588 |
|
if block in jobsOfBlock.keys() : |
| 589 |
|
blockCounter += 1 |
| 590 |
< |
screenOutput += "Block %5i: jobs %20s: sites: %s\n" % (blockCounter,spanRanges(jobsOfBlock[block]),','.join(blockSites[block])) |
| 590 |
> |
screenOutput += "Block %5i: jobs %20s: sites: %s\n" % (blockCounter,spanRanges(jobsOfBlock[block]),','.join(self.blackWhiteListParser.checkWhiteList(self.blackWhiteListParser.checkBlackList(blockSites[block],block),block))) |
| 591 |
|
|
| 592 |
< |
# common.logger.message(screenOutput) |
| 592 |
> |
common.logger.message(screenOutput) |
| 593 |
|
|
| 594 |
|
self.list_of_args = list_of_lists |
| 595 |
|
return |
| 857 |
|
|
| 858 |
|
## OLI_Daniele at this level middleware already known |
| 859 |
|
|
| 841 |
– |
txt += 'echo "### Firtst set SCRAM ARCH and BUILD_ARCH ###"\n' |
| 842 |
– |
txt += 'echo "Setting SCRAM_ARCH='+self.executable_arch+'"\n' |
| 843 |
– |
txt += 'export SCRAM_ARCH='+self.executable_arch+'\n' |
| 844 |
– |
txt += 'export BUILD_ARCH='+self.executable_arch+'\n' |
| 860 |
|
txt += 'if [ $middleware == LCG ]; then \n' |
| 861 |
+ |
txt += ' echo "### First set SCRAM ARCH and BUILD_ARCH to ' + self.executable_arch + ' ###"\n' |
| 862 |
+ |
txt += ' export SCRAM_ARCH='+self.executable_arch+'\n' |
| 863 |
+ |
txt += ' export BUILD_ARCH='+self.executable_arch+'\n' |
| 864 |
|
txt += self.wsSetupCMSLCGEnvironment_() |
| 865 |
|
txt += 'elif [ $middleware == OSG ]; then\n' |
| 866 |
|
txt += ' WORKING_DIR=`/bin/mktemp -d $OSG_WN_TMP/cms_XXXXXXXXXXXX`\n' |
| 879 |
|
txt += ' echo "Change to working directory: $WORKING_DIR"\n' |
| 880 |
|
txt += ' cd $WORKING_DIR\n' |
| 881 |
|
txt += self.wsSetupCMSOSGEnvironment_() |
| 882 |
+ |
txt += ' echo "### Set SCRAM ARCH to ' + self.executable_arch + ' ###"\n' |
| 883 |
+ |
txt += ' export SCRAM_ARCH='+self.executable_arch+'\n' |
| 884 |
|
txt += 'fi\n' |
| 885 |
|
|
| 886 |
|
# Prepare JobType-specific part |
| 1119 |
|
def executableArgs(self): |
| 1120 |
|
if self.scriptExe:#CarlosDaniele |
| 1121 |
|
return self.scriptExe + " $NJob" |
| 1122 |
< |
else: |
| 1123 |
< |
return " -p pset.cfg" |
| 1122 |
> |
else: |
| 1123 |
> |
# if >= CMSSW_1_5_X, add -e |
| 1124 |
> |
version_array = self.scram.getSWVersion().split('_') |
| 1125 |
> |
major = 0 |
| 1126 |
> |
minor = 0 |
| 1127 |
> |
try: |
| 1128 |
> |
major = int(version_array[1]) |
| 1129 |
> |
minor = int(version_array[2]) |
| 1130 |
> |
except: |
| 1131 |
> |
msg = "Cannot parse CMSSW version string: " + "_".join(version_array) + " for major and minor release number!" |
| 1132 |
> |
raise CrabException(msg) |
| 1133 |
> |
if major >= 1 and minor >= 5 : |
| 1134 |
> |
return " -e -p pset.cfg" |
| 1135 |
> |
else: |
| 1136 |
> |
return " -p pset.cfg" |
| 1137 |
|
|
| 1138 |
|
def inputSandbox(self, nj): |
| 1139 |
|
""" |
| 1189 |
|
# txt += 'ls '+fileWithSuffix+'\n' |
| 1190 |
|
# txt += 'ls_result=$?\n' |
| 1191 |
|
txt += 'if [ -e ./'+fileWithSuffix+' ] ; then\n' |
| 1192 |
< |
txt += ' mv '+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n' |
| 1192 |
> |
###### FEDE FOR OUTPUT DATA PUBLICATION ######## |
| 1193 |
> |
txt += ' mv '+fileWithSuffix+' $RUNTIME_AREA\n' |
| 1194 |
> |
txt += ' cp $RUNTIME_AREA/'+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n' |
| 1195 |
> |
################################################ |
| 1196 |
|
txt += 'else\n' |
| 1197 |
< |
txt += ' exit_status=60302\n' |
| 1198 |
< |
txt += ' echo "ERROR: Problem with output file '+fileWithSuffix+'"\n' |
| 1197 |
> |
txt += ' exit_status=60302\n' |
| 1198 |
> |
txt += ' echo "ERROR: Problem with output file '+fileWithSuffix+'"\n' |
| 1199 |
> |
############# FEDE ADDED CHECK FOR OUTPUT ############# |
| 1200 |
> |
## MATTY's FIX: the exit option was interrupting the execution |
| 1201 |
> |
if fileWithSuffix in self.output_file: |
| 1202 |
> |
txt += ' echo "JOB_EXIT_STATUS = $exit_status"\n' |
| 1203 |
> |
txt += ' output_exit_status=$exit_status\n' |
| 1204 |
> |
txt += ' # exit $exit_status\n' |
| 1205 |
> |
####################################################### |
| 1206 |
|
if common.scheduler.boss_scheduler_name == 'condor_g': |
| 1207 |
|
txt += ' if [ $middleware == OSG ]; then \n' |
| 1208 |
|
txt += ' echo "prepare dummy output file"\n' |
| 1212 |
|
file_list = [] |
| 1213 |
|
for fileWithSuffix in (self.output_file): |
| 1214 |
|
file_list.append(self.numberFile_(fileWithSuffix, '$NJob')) |
| 1215 |
+ |
|
| 1216 |
|
txt += 'file_list="'+string.join(file_list,' ')+'"\n' |
| 1173 |
– |
|
| 1217 |
|
txt += 'cd $RUNTIME_AREA\n' |
| 1175 |
– |
#### FEDE this is the cleanEnv function |
| 1176 |
– |
### OLI_DANIELE |
| 1177 |
– |
#txt += 'if [ $middleware == OSG ]; then\n' |
| 1178 |
– |
#txt += ' cd $RUNTIME_AREA\n' |
| 1179 |
– |
#txt += ' echo "Remove working directory: $WORKING_DIR"\n' |
| 1180 |
– |
#txt += ' /bin/rm -rf $WORKING_DIR\n' |
| 1181 |
– |
#txt += ' if [ -d $WORKING_DIR ] ;then\n' |
| 1182 |
– |
#txt += ' echo "SET_EXE 60999 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after cleanup of WN"\n' |
| 1183 |
– |
#txt += ' echo "JOB_EXIT_STATUS = 60999"\n' |
| 1184 |
– |
#txt += ' echo "JobExitCode=60999" | tee -a $RUNTIME_AREA/$repo\n' |
| 1185 |
– |
#txt += ' dumpStatus $RUNTIME_AREA/$repo\n' |
| 1186 |
– |
#txt += ' rm -f $RUNTIME_AREA/$repo \n' |
| 1187 |
– |
#txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n' |
| 1188 |
– |
#txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n' |
| 1189 |
– |
#txt += ' fi\n' |
| 1190 |
– |
#txt += 'fi\n' |
| 1191 |
– |
#txt += '\n' |
| 1192 |
– |
|
| 1193 |
– |
|
| 1218 |
|
return txt |
| 1219 |
|
|
| 1220 |
|
def numberFile_(self, file, txt): |
| 1361 |
|
""" |
| 1362 |
|
|
| 1363 |
|
txt = '' |
| 1340 |
– |
txt += 'echo "Modify Job Report" \n' |
| 1341 |
– |
#txt += 'chmod a+x $RUNTIME_AREA/'+self.version+'/ProdAgentApi/FwkJobRep/ModifyJobReport.py\n' |
| 1342 |
– |
################ FEDE FOR DBS2 ############################################# |
| 1343 |
– |
txt += 'chmod a+x $SOFTWARE_DIR/ProdAgentApi/FwkJobRep/ModifyJobReport.py\n' |
| 1344 |
– |
############################################################################# |
| 1364 |
|
try: |
| 1365 |
|
publish_data = int(self.cfg_params['USER.publish_data']) |
| 1366 |
|
except KeyError: |
| 1367 |
|
publish_data = 0 |
| 1368 |
+ |
if (publish_data == 1): |
| 1369 |
+ |
txt += 'echo "Modify Job Report" \n' |
| 1370 |
+ |
#txt += 'chmod a+x $RUNTIME_AREA/'+self.version+'/ProdAgentApi/FwkJobRep/ModifyJobReport.py\n' |
| 1371 |
+ |
################ FEDE FOR DBS2 ############################################# |
| 1372 |
+ |
txt += 'chmod a+x $SOFTWARE_DIR/ProdAgentApi/FwkJobRep/ModifyJobReport.py\n' |
| 1373 |
+ |
############################################################################# |
| 1374 |
+ |
#try: |
| 1375 |
+ |
# publish_data = int(self.cfg_params['USER.publish_data']) |
| 1376 |
+ |
#except KeyError: |
| 1377 |
+ |
# publish_data = 0 |
| 1378 |
|
|
| 1379 |
< |
txt += 'if [ -z "$SE" ]; then\n' |
| 1380 |
< |
txt += ' SE="" \n' |
| 1381 |
< |
txt += 'fi \n' |
| 1382 |
< |
txt += 'if [ -z "$SE_PATH" ]; then\n' |
| 1383 |
< |
txt += ' SE_PATH="" \n' |
| 1384 |
< |
txt += 'fi \n' |
| 1385 |
< |
txt += 'echo "SE = $SE"\n' |
| 1386 |
< |
txt += 'echo "SE_PATH = $SE_PATH"\n' |
| 1379 |
> |
txt += 'if [ -z "$SE" ]; then\n' |
| 1380 |
> |
txt += ' SE="" \n' |
| 1381 |
> |
txt += 'fi \n' |
| 1382 |
> |
txt += 'if [ -z "$SE_PATH" ]; then\n' |
| 1383 |
> |
txt += ' SE_PATH="" \n' |
| 1384 |
> |
txt += 'fi \n' |
| 1385 |
> |
txt += 'echo "SE = $SE"\n' |
| 1386 |
> |
txt += 'echo "SE_PATH = $SE_PATH"\n' |
| 1387 |
|
|
| 1388 |
< |
if (publish_data == 1): |
| 1388 |
> |
#if (publish_data == 1): |
| 1389 |
|
#processedDataset = self.cfg_params['USER.processed_datasetname'] |
| 1390 |
|
processedDataset = self.cfg_params['USER.publish_data_name'] |
| 1391 |
|
txt += 'ProcessedDataset='+processedDataset+'\n' |
| 1415 |
|
txt += ' mv NewFrameworkJobReport.xml crab_fjr_$NJob.xml\n' |
| 1416 |
|
txt += 'fi\n' |
| 1417 |
|
else: |
| 1418 |
< |
txt += 'ProcessedDataset=no_data_to_publish \n' |
| 1418 |
> |
txt += 'echo "no data publication required"\n' |
| 1419 |
> |
#txt += 'ProcessedDataset=no_data_to_publish \n' |
| 1420 |
|
#### FEDE: added slash in LFN ############## |
| 1421 |
< |
txt += 'FOR_LFN=/local/ \n' |
| 1422 |
< |
txt += 'echo "ProcessedDataset = $ProcessedDataset"\n' |
| 1423 |
< |
txt += 'echo "FOR_LFN = $FOR_LFN" \n' |
| 1421 |
> |
#txt += 'FOR_LFN=/local/ \n' |
| 1422 |
> |
#txt += 'echo "ProcessedDataset = $ProcessedDataset"\n' |
| 1423 |
> |
#txt += 'echo "FOR_LFN = $FOR_LFN" \n' |
| 1424 |
|
return txt |
| 1425 |
|
|
| 1426 |
|
def cleanEnv(self): |
| 1463 |
|
for e in old: |
| 1464 |
|
nd[e]=0 |
| 1465 |
|
return nd.keys() |
| 1466 |
+ |
|
| 1467 |
+ |
|
| 1468 |
+ |
def checkOut(self, limit): |
| 1469 |
+ |
""" |
| 1470 |
+ |
check the dimension of the output files |
| 1471 |
+ |
""" |
| 1472 |
+ |
txt = 'echo "*****************************************"\n' |
| 1473 |
+ |
txt += 'echo "** Starting output sandbox limit check **"\n' |
| 1474 |
+ |
txt += 'echo "*****************************************"\n' |
| 1475 |
+ |
allOutFiles = "" |
| 1476 |
+ |
listOutFiles = [] |
| 1477 |
+ |
for fileOut in (self.output_file+self.output_file_sandbox): |
| 1478 |
+ |
if fileOut.find('crab_fjr') == -1: |
| 1479 |
+ |
allOutFiles = allOutFiles + " " + self.numberFile_(fileOut, '$NJob') |
| 1480 |
+ |
listOutFiles.append(self.numberFile_(fileOut, '$NJob')) |
| 1481 |
+ |
txt += 'echo "OUTPUT files: '+str(allOutFiles)+'";\n' |
| 1482 |
+ |
txt += 'ls -gGhrta;\n' |
| 1483 |
+ |
txt += 'sum=0;\n' |
| 1484 |
+ |
txt += 'for file in '+str(allOutFiles)+' ; do\n' |
| 1485 |
+ |
txt += ' if [ -e $file ]; then\n' |
| 1486 |
+ |
txt += ' tt=`ls -gGrta $file | awk \'{ print $3 }\'`\n' |
| 1487 |
+ |
txt += ' sum=`expr $sum + $tt`\n' |
| 1488 |
+ |
txt += ' else\n' |
| 1489 |
+ |
txt += ' echo "WARNING: output file $file not found!"\n' |
| 1490 |
+ |
txt += ' fi\n' |
| 1491 |
+ |
txt += 'done\n' |
| 1492 |
+ |
txt += 'echo "Total Output dimension: $sum";\n' |
| 1493 |
+ |
txt += 'limit='+str(limit)+';\n' |
| 1494 |
+ |
txt += 'echo "OUTPUT FILES LIMIT SET TO: $limit";\n' |
| 1495 |
+ |
txt += 'if [ $limit -lt $sum ]; then\n' |
| 1496 |
+ |
txt += ' echo "WARNING: output files have to big size - something will be lost;"\n' |
| 1497 |
+ |
txt += ' echo " checking the output file sizes..."\n' |
| 1498 |
+ |
""" |
| 1499 |
+ |
txt += ' dim=0;\n' |
| 1500 |
+ |
txt += ' exclude=0;\n' |
| 1501 |
+ |
txt += ' for files in '+str(allOutFiles)+' ; do\n' |
| 1502 |
+ |
txt += ' sumTemp=0;\n' |
| 1503 |
+ |
txt += ' for file2 in '+str(allOutFiles)+' ; do\n' |
| 1504 |
+ |
txt += ' if [ $file != $file2 ]; then\n' |
| 1505 |
+ |
txt += ' tt=`ls -gGrta $file2 | awk \'{ print $3 }\';`\n' |
| 1506 |
+ |
txt += ' sumTemp=`expr $sumTemp + $tt`;\n' |
| 1507 |
+ |
txt += ' fi\n' |
| 1508 |
+ |
txt += ' done\n' |
| 1509 |
+ |
txt += ' if [ $sumTemp -lt $limit ]; then\n' |
| 1510 |
+ |
txt += ' if [ $dim -lt $sumTemp ]; then\n' |
| 1511 |
+ |
txt += ' dim=$sumTemp;\n' |
| 1512 |
+ |
txt += ' exclude=$file;\n' |
| 1513 |
+ |
txt += ' fi\n' |
| 1514 |
+ |
txt += ' fi\n' |
| 1515 |
+ |
txt += ' done\n' |
| 1516 |
+ |
txt += ' echo "Dimension calculated: $dim"; echo "File to exclude: $exclude";\n' |
| 1517 |
+ |
""" |
| 1518 |
+ |
txt += ' tot=0;\n' |
| 1519 |
+ |
txt += ' for file2 in '+str(allOutFiles)+' ; do\n' |
| 1520 |
+ |
txt += ' tt=`ls -gGrta $file2 | awk \'{ print $3 }\';`\n' |
| 1521 |
+ |
txt += ' tot=`expr $tot + $tt`;\n' |
| 1522 |
+ |
txt += ' if [ $limit -lt $tot ]; then\n' |
| 1523 |
+ |
txt += ' tot=`expr $tot - $tt`;\n' |
| 1524 |
+ |
txt += ' fileLast=$file;\n' |
| 1525 |
+ |
txt += ' break;\n' |
| 1526 |
+ |
txt += ' fi\n' |
| 1527 |
+ |
txt += ' done\n' |
| 1528 |
+ |
txt += ' echo "Dimension calculated: $tot"; echo "First file to exclude: $file";\n' |
| 1529 |
+ |
txt += ' flag=0;\n' |
| 1530 |
+ |
txt += ' for filess in '+str(allOutFiles)+' ; do\n' |
| 1531 |
+ |
txt += ' if [ $fileLast = $filess ]; then\n' |
| 1532 |
+ |
txt += ' flag=1;\n' |
| 1533 |
+ |
txt += ' fi\n' |
| 1534 |
+ |
txt += ' if [ $flag -eq 1 ]; then\n' |
| 1535 |
+ |
txt += ' rm -f $filess;\n' |
| 1536 |
+ |
txt += ' fi\n' |
| 1537 |
+ |
txt += ' done\n' |
| 1538 |
+ |
txt += ' ls -agGhrt;\n' |
| 1539 |
+ |
txt += ' echo "WARNING: output files are too big in dimension: can not put in the output_sandbox.";\n' |
| 1540 |
+ |
txt += ' echo "JOB_EXIT_STATUS = 70000";\n' |
| 1541 |
+ |
txt += ' exit_status=70000;\n' |
| 1542 |
+ |
txt += 'else' |
| 1543 |
+ |
txt += ' echo "Total Output dimension $sum is fine.";\n' |
| 1544 |
+ |
txt += 'fi\n' |
| 1545 |
+ |
txt += 'echo "*****************************************"\n' |
| 1546 |
+ |
txt += 'echo "*** Ending output sandbox limit check ***"\n' |
| 1547 |
+ |
txt += 'echo "*****************************************"\n' |
| 1548 |
+ |
return txt |
