3 |
|
from crab_exceptions import * |
4 |
|
from crab_util import * |
5 |
|
import common |
6 |
+ |
import PsetManipulator |
7 |
|
|
8 |
< |
import DataDiscovery |
9 |
< |
import DataLocation |
8 |
> |
import DBSInfo_EDM |
9 |
> |
#from DataDiscovery_EDM import DataDiscovery_EDM |
10 |
> |
import DataDiscovery_EDM |
11 |
> |
#from DataLocation_EDM import DataLocation_EDM |
12 |
> |
import DataLocation_EDM |
13 |
|
import Scram |
14 |
|
|
15 |
|
import os, string, re |
20 |
|
common.logger.debug(3,'CMSSW::__init__') |
21 |
|
|
22 |
|
self.analisys_common_info = {} |
23 |
+ |
# Marco. |
24 |
+ |
self._params = {} |
25 |
+ |
self.cfg_params = cfg_params |
26 |
|
|
27 |
|
log = common.logger |
28 |
|
|
33 |
|
self.executable = '' |
34 |
|
self.tgz_name = 'default.tgz' |
35 |
|
|
36 |
+ |
|
37 |
|
self.version = self.scram.getSWVersion() |
38 |
|
common.analisys_common_info['sw_version'] = self.version |
39 |
+ |
### FEDE |
40 |
+ |
common.analisys_common_info['copy_input_data'] = 0 |
41 |
+ |
common.analisys_common_info['events_management'] = 1 |
42 |
|
|
43 |
|
### collect Data cards |
44 |
|
try: |
45 |
< |
self.owner = cfg_params['CMSSW.owner'] |
46 |
< |
log.debug(6, "CMSSW::CMSSW(): owner = "+self.owner) |
47 |
< |
self.dataset = cfg_params['CMSSW.dataset'] |
48 |
< |
log.debug(6, "CMSSW::CMSSW(): dataset = "+self.dataset) |
45 |
> |
# self.owner = cfg_params['CMSSW.owner'] |
46 |
> |
# log.debug(6, "CMSSW::CMSSW(): owner = "+self.owner) |
47 |
> |
# self.dataset = cfg_params['CMSSW.dataset'] |
48 |
> |
self.datasetPath = cfg_params['CMSSW.datasetpath'] |
49 |
> |
log.debug(6, "CMSSW::CMSSW(): datasetPath = "+self.datasetPath) |
50 |
|
except KeyError: |
51 |
< |
msg = "Error: owner and/or dataset not defined " |
51 |
> |
# msg = "Error: owner and/or dataset not defined " |
52 |
> |
msg = "Error: datasetpath not defined " |
53 |
|
raise CrabException(msg) |
41 |
– |
|
54 |
|
self.dataTiers = [] |
55 |
< |
try: |
56 |
< |
tmpDataTiers = string.split(cfg_params['CMSSW.data_tier'],',') |
57 |
< |
for tmp in tmpDataTiers: |
58 |
< |
tmp=string.strip(tmp) |
59 |
< |
self.dataTiers.append(tmp) |
60 |
< |
pass |
61 |
< |
pass |
62 |
< |
except KeyError: |
63 |
< |
pass |
64 |
< |
log.debug(6, "Cmssw::Cmssw(): dataTiers = "+str(self.dataTiers)) |
55 |
> |
# try: |
56 |
> |
# tmpDataTiers = string.split(cfg_params['CMSSW.data_tier'],',') |
57 |
> |
# for tmp in tmpDataTiers: |
58 |
> |
# tmp=string.strip(tmp) |
59 |
> |
# self.dataTiers.append(tmp) |
60 |
> |
# pass |
61 |
> |
# pass |
62 |
> |
# except KeyError: |
63 |
> |
# pass |
64 |
> |
# log.debug(6, "Cmssw::Cmssw(): dataTiers = "+str(self.dataTiers)) |
65 |
|
|
66 |
|
## now the application |
67 |
|
try: |
95 |
|
tmp=string.strip(tmp) |
96 |
|
self.output_file.append(tmp) |
97 |
|
pass |
86 |
– |
|
98 |
|
else: |
99 |
|
log.message("No output file defined: only stdout/err will be available") |
100 |
|
pass |
105 |
|
|
106 |
|
# script_exe file as additional file in inputSandbox |
107 |
|
try: |
108 |
< |
self.scriptExe = cfg_params['CMSSW.script_exe'] |
108 |
> |
self.scriptExe = cfg_params['USER.script_exe'] |
109 |
|
self.additional_inbox_files.append(self.scriptExe) |
110 |
|
except KeyError: |
111 |
|
pass |
120 |
|
try: |
121 |
|
tmpAddFiles = string.split(cfg_params['CMSSW.additional_input_files'],',') |
122 |
|
for tmp in tmpAddFiles: |
123 |
+ |
if not os.path.exists(tmp): |
124 |
+ |
raise CrabException("Additional input file not found: "+tmp) |
125 |
|
tmp=string.strip(tmp) |
126 |
|
self.additional_inbox_files.append(tmp) |
127 |
|
pass |
130 |
|
pass |
131 |
|
|
132 |
|
try: |
133 |
+ |
self.filesPerJob = int(cfg_params['CMSSW.files_per_jobs']) #Daniele |
134 |
+ |
except KeyError: |
135 |
+ |
self.filesPerJob = 1 |
136 |
+ |
|
137 |
+ |
## Max event will be total_number_of_events ??? Daniele |
138 |
+ |
try: |
139 |
+ |
self.maxEv = cfg_params['CMSSW.event_per_job'] |
140 |
+ |
except KeyError: |
141 |
+ |
self.maxEv = "-1" |
142 |
+ |
## |
143 |
+ |
try: |
144 |
|
self.total_number_of_events = int(cfg_params['CMSSW.total_number_of_events']) |
145 |
|
except KeyError: |
146 |
< |
msg = 'Must define total_number_of_events and job_number_of_events' |
146 |
> |
msg = 'Must define total_number_of_events' |
147 |
|
raise CrabException(msg) |
124 |
– |
|
125 |
– |
#Marco: FirstEvent is nolonger used inside PSet |
126 |
– |
# try: |
127 |
– |
# self.first = int(cfg_params['CMSSW.first_event']) |
128 |
– |
# except KeyError: |
129 |
– |
# self.first = 0 |
130 |
– |
# pass |
131 |
– |
# log.debug(6, "Orca::Orca(): total number of events = "+`self.total_number_of_events`) |
132 |
– |
#log.debug(6, "Orca::Orca(): events per job = "+`self.job_number_of_events`) |
133 |
– |
# log.debug(6, "Orca::Orca(): first event = "+`self.first`) |
148 |
|
|
149 |
|
CEBlackList = [] |
150 |
|
try: |
164 |
|
CEWhiteList = [] |
165 |
|
try: |
166 |
|
tmpGood = string.split(cfg_params['EDG.ce_white_list'],',') |
153 |
– |
#tmpGood = ['cern'] |
167 |
|
for tmp in tmpGood: |
168 |
|
tmp=string.strip(tmp) |
156 |
– |
#if (tmp == 'cnaf'): tmp = 'webserver' ########## warning: temp. patch |
169 |
|
CEWhiteList.append(tmp) |
170 |
|
except KeyError: |
171 |
|
pass |
177 |
|
|
178 |
|
common.logger.debug(5,'CEWhiteList: '+str(CEWhiteList)) |
179 |
|
|
180 |
+ |
self.PsetEdit = PsetManipulator.PsetManipulator(self.pset) #Daniele Pset |
181 |
+ |
|
182 |
|
#DBSDLS-start |
183 |
|
## Initialize the variables that are extracted from DBS/DLS and needed in other places of the code |
184 |
|
self.maxEvents=0 # max events available ( --> check the requested nb. of evts in Creator.py) |
189 |
|
|
190 |
|
self.tgzNameWithPath = self.getTarBall(self.executable) |
191 |
|
|
192 |
+ |
self.jobSplitting() #Daniele job Splitting |
193 |
+ |
self.PsetEdit.maxEvent(self.maxEv) #Daniele |
194 |
+ |
self.PsetEdit.inputModule("INPUT") #Daniele |
195 |
+ |
self.PsetEdit.psetWriter(self.configFilename()) |
196 |
+ |
|
197 |
|
def DataDiscoveryAndLocation(self, cfg_params): |
198 |
|
|
199 |
< |
fun = "CMSSW::DataDiscoveryAndLocation()" |
199 |
> |
common.logger.debug(10,"CMSSW::DataDiscoveryAndLocation()") |
200 |
> |
|
201 |
> |
#datasetPath = "/"+self.owner+"/"+self.dataTiers[0]+"/"+self.dataset |
202 |
> |
|
203 |
> |
datasetPath=self.datasetPath |
204 |
> |
|
205 |
> |
## TODO |
206 |
> |
dataTiersList = "" |
207 |
> |
dataTiers = dataTiersList.split(',') |
208 |
|
|
209 |
|
## Contact the DBS |
210 |
|
try: |
211 |
< |
self.pubdata=DataDiscovery.DataDiscovery(self.owner, |
185 |
< |
self.dataset, |
186 |
< |
self.dataTiers, |
187 |
< |
cfg_params) |
211 |
> |
self.pubdata=DataDiscovery_EDM.DataDiscovery_EDM(datasetPath, dataTiers, dataTiers) |
212 |
|
self.pubdata.fetchDBSInfo() |
213 |
|
|
214 |
< |
except DataDiscovery.NotExistingDatasetError, ex : |
214 |
> |
except DataDiscovery_EDM.NotExistingDatasetError, ex : |
215 |
|
msg = 'ERROR ***: failed Data Discovery in DBS : %s'%ex.getErrorMessage() |
216 |
|
raise CrabException(msg) |
217 |
|
|
218 |
< |
except DataDiscovery.NoDataTierinProvenanceError, ex : |
218 |
> |
except DataDiscovery_EDM.NoDataTierinProvenanceError, ex : |
219 |
|
msg = 'ERROR ***: failed Data Discovery in DBS : %s'%ex.getErrorMessage() |
220 |
|
raise CrabException(msg) |
221 |
< |
except DataDiscovery.DataDiscoveryError, ex: |
221 |
> |
except DataDiscovery_EDM.DataDiscoveryError, ex: |
222 |
|
msg = 'ERROR ***: failed Data Discovery in DBS %s'%ex.getErrorMessage() |
223 |
|
raise CrabException(msg) |
224 |
|
|
225 |
|
## get list of all required data in the form of dbs paths (dbs path = /dataset/datatier/owner) |
226 |
< |
self.DBSPaths=self.pubdata.getDBSPaths() |
227 |
< |
common.logger.message("Required data are : ") |
228 |
< |
for path in self.DBSPaths: |
229 |
< |
common.logger.message(" --> "+path ) |
226 |
> |
## self.DBSPaths=self.pubdata.getDBSPaths() |
227 |
> |
common.logger.message("Required data are :"+self.datasetPath) |
228 |
> |
|
229 |
> |
filesbyblock=self.pubdata.getFiles() |
230 |
> |
self.AllInputFiles=filesbyblock.values() |
231 |
> |
self.files = self.AllInputFiles |
232 |
> |
|
233 |
> |
## TEMP |
234 |
> |
# self.filesTmp = filesbyblock.values() |
235 |
> |
# self.files = [] |
236 |
> |
# locPath='rfio:cmsbose2.bo.infn.it:/flatfiles/SE00/cms/fanfani/ProdTest/' |
237 |
> |
# locPath='' |
238 |
> |
# tmp = [] |
239 |
> |
# for file in self.filesTmp[0]: |
240 |
> |
# tmp.append(locPath+file) |
241 |
> |
# self.files.append(tmp) |
242 |
> |
## END TEMP |
243 |
|
|
244 |
|
## get max number of events |
245 |
< |
common.logger.debug(10,"number of events for primary fileblocks %i"%self.pubdata.getMaxEvents()) |
245 |
> |
#common.logger.debug(10,"number of events for primary fileblocks %i"%self.pubdata.getMaxEvents()) |
246 |
|
self.maxEvents=self.pubdata.getMaxEvents() ## self.maxEvents used in Creator.py |
247 |
|
common.logger.message("\nThe number of available events is %s"%self.maxEvents) |
248 |
|
|
212 |
– |
## get fileblocks corresponding to the required data |
213 |
– |
fb=self.pubdata.getFileBlocks() |
214 |
– |
common.logger.debug(5,"fileblocks are %s"%fb) |
215 |
– |
|
249 |
|
## Contact the DLS and build a list of sites hosting the fileblocks |
250 |
|
try: |
251 |
< |
dataloc=DataLocation.DataLocation(self.pubdata.getFileBlocks(),cfg_params) |
251 |
> |
dataloc=DataLocation_EDM.DataLocation_EDM(filesbyblock.keys(),cfg_params) |
252 |
|
dataloc.fetchDLSInfo() |
253 |
< |
except DataLocation.DataLocationError , ex: |
253 |
> |
except DataLocation_EDM.DataLocationError , ex: |
254 |
|
msg = 'ERROR ***: failed Data Location in DLS \n %s '%ex.getErrorMessage() |
255 |
|
raise CrabException(msg) |
256 |
|
|
264 |
|
if len(sites)==0: |
265 |
|
msg = 'No sites hosting all the needed data! Exiting... ' |
266 |
|
raise CrabException(msg) |
267 |
+ |
|
268 |
|
common.logger.message("List of Sites hosting the data : "+str(sites)) |
269 |
|
common.logger.debug(6, "List of Sites: "+str(sites)) |
270 |
|
common.analisys_common_info['sites']=sites ## used in SchedulerEdg.py in createSchScript |
271 |
|
return |
272 |
+ |
|
273 |
+ |
def jobSplitting(self): |
274 |
+ |
""" |
275 |
+ |
first implemntation for job splitting |
276 |
+ |
""" |
277 |
+ |
# print 'eventi totali '+str(self.maxEvents) |
278 |
+ |
# print 'eventi totali richiesti dallo user '+str(self.total_number_of_events) |
279 |
+ |
#print 'files per job '+str(self.filesPerJob) |
280 |
+ |
common.logger.message('Required '+str(self.filesPerJob)+' files per job ') |
281 |
+ |
common.logger.message('Required '+str(self.total_number_of_events)+' events in total ') |
282 |
+ |
|
283 |
+ |
## TODO: SL need to have (from DBS) a detailed list of how many events per each file |
284 |
+ |
n_tot_files = (len(self.files[0])) |
285 |
+ |
## SL: this is wrong if the files have different number of events |
286 |
+ |
evPerFile = int(self.maxEvents)/n_tot_files |
287 |
+ |
|
288 |
+ |
common.logger.debug(5,'Events per File '+str(evPerFile)) |
289 |
+ |
|
290 |
+ |
## if asked to process all events, do it |
291 |
+ |
if self.total_number_of_events == -1: |
292 |
+ |
self.total_number_of_events=self.maxEvents |
293 |
+ |
self.total_number_of_jobs = int(n_tot_files)*1/int(self.filesPerJob) |
294 |
+ |
common.logger.message(str(self.total_number_of_jobs)+' jobs will be created for all available events '+str(self.total_number_of_events)+' events') |
295 |
+ |
|
296 |
+ |
else: |
297 |
+ |
self.total_number_of_files = int(self.total_number_of_events/evPerFile) |
298 |
+ |
## SL: if ask for less event than what is computed to be available on a |
299 |
+ |
## file, process the first file anyhow. |
300 |
+ |
if self.total_number_of_files == 0: |
301 |
+ |
self.total_number_of_files = self.total_number_of_files + 1 |
302 |
+ |
|
303 |
+ |
common.logger.debug(5,'N files '+str(self.total_number_of_files)) |
304 |
+ |
|
305 |
+ |
check = 0 |
306 |
+ |
|
307 |
+ |
## Compute the number of jobs |
308 |
+ |
#self.total_number_of_jobs = int(n_tot_files)*1/int(self.filesPerJob) |
309 |
+ |
self.total_number_of_jobs = int(self.total_number_of_files/self.filesPerJob) |
310 |
+ |
common.logger.debug(5,'N jobs '+str(self.total_number_of_jobs)) |
311 |
+ |
|
312 |
+ |
## is there any remainder? |
313 |
+ |
check = int(self.total_number_of_files) - (int(self.total_number_of_jobs)*self.filesPerJob) |
314 |
+ |
|
315 |
+ |
common.logger.debug(5,'Check '+str(check)) |
316 |
+ |
|
317 |
+ |
if check > 0: |
318 |
+ |
self.total_number_of_jobs = self.total_number_of_jobs + 1 |
319 |
+ |
common.logger.message('Warning: last job will be created with '+str(check)+' files') |
320 |
+ |
|
321 |
+ |
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') |
322 |
+ |
pass |
323 |
+ |
|
324 |
+ |
list_of_lists = [] |
325 |
+ |
for i in xrange(0, int(n_tot_files), self.filesPerJob): |
326 |
+ |
list_of_lists.append(self.files[0][i: i+self.filesPerJob]) |
327 |
+ |
|
328 |
+ |
self.list_of_files = list_of_lists |
329 |
+ |
|
330 |
+ |
return |
331 |
+ |
|
332 |
+ |
def split(self, jobParams): |
333 |
+ |
|
334 |
+ |
common.jobDB.load() |
335 |
+ |
#### Fabio |
336 |
+ |
njobs = self.total_number_of_jobs |
337 |
+ |
filelist = self.list_of_files |
338 |
+ |
# create the empty structure |
339 |
+ |
for i in range(njobs): |
340 |
+ |
jobParams.append("") |
341 |
+ |
|
342 |
+ |
for job in range(njobs): |
343 |
+ |
jobParams[job] = filelist[job] |
344 |
+ |
common.jobDB.setArguments(job, jobParams[job]) |
345 |
+ |
|
346 |
+ |
common.jobDB.save() |
347 |
+ |
return |
348 |
+ |
|
349 |
+ |
def getJobTypeArguments(self, nj, sched): |
350 |
+ |
params = common.jobDB.arguments(nj) |
351 |
+ |
#print params |
352 |
+ |
parString = "\\{" |
353 |
+ |
|
354 |
+ |
for i in range(len(params) - 1): |
355 |
+ |
parString += '\\\"' + params[i] + '\\\"\,' |
356 |
|
|
357 |
+ |
parString += '\\\"' + params[len(params) - 1] + '\\\"\\}' |
358 |
+ |
return parString |
359 |
+ |
|
360 |
+ |
def numberOfJobs(self): |
361 |
+ |
# Fabio |
362 |
+ |
|
363 |
+ |
return self.total_number_of_jobs |
364 |
+ |
|
365 |
+ |
|
366 |
+ |
|
367 |
|
def checkBlackList(self, allSites): |
368 |
|
if len(self.reCEBlackList)==0: return allSites |
369 |
|
sites = [] |
380 |
|
common.logger.debug(3,"No sites found after BlackList") |
381 |
|
return sites |
382 |
|
|
383 |
< |
def checkWhiteList(self, allsites): |
383 |
> |
def checkWhiteList(self, allSites): |
384 |
|
|
385 |
< |
if len(self.reCEWhiteList)==0: return pubDBUrls |
385 |
> |
if len(self.reCEWhiteList)==0: return allSites |
386 |
|
sites = [] |
387 |
< |
for site in allsites: |
260 |
< |
#print 'connecting to the URL ',url |
387 |
> |
for site in allSites: |
388 |
|
good=0 |
389 |
|
for re in self.reCEWhiteList: |
390 |
|
if re.search(site): |
454 |
|
if os.path.exists(lib): |
455 |
|
filesToBeTarred.append(libDir) |
456 |
|
|
457 |
+ |
## Now check if module dir is present |
458 |
+ |
moduleDir = 'module' |
459 |
+ |
if os.path.isdir(swArea+'/'+moduleDir): |
460 |
+ |
filesToBeTarred.append(moduleDir) |
461 |
+ |
|
462 |
|
## Now check if the Data dir is present |
463 |
|
dataDir = 'src/Data/' |
464 |
|
if os.path.isdir(swArea+'/'+dataDir): |
486 |
|
the execution environment for the job 'nj'. |
487 |
|
""" |
488 |
|
# Prepare JobType-independent part |
489 |
< |
txt = self.wsSetupCMSEnvironment_() |
489 |
> |
txt = '' |
490 |
> |
|
491 |
> |
## OLI_Daniele at this level middleware already known |
492 |
> |
|
493 |
> |
txt += 'if [ $middleware == LCG ]; then \n' |
494 |
> |
txt += self.wsSetupCMSLCGEnvironment_() |
495 |
> |
txt += 'elif [ $middleware == OSG ]; then\n' |
496 |
> |
txt += ' time=`date -u +"%s"`\n' |
497 |
> |
txt += ' WORKING_DIR=$OSG_WN_TMP/cms_$time\n' |
498 |
> |
txt += ' echo "Creating working directory: $WORKING_DIR"\n' |
499 |
> |
txt += ' /bin/mkdir -p $WORKING_DIR\n' |
500 |
> |
txt += ' if [ ! -d $WORKING_DIR ] ;then\n' |
501 |
> |
txt += ' echo "OSG WORKING DIR ==> $WORKING_DIR could not be created on on WN `hostname`"\n' |
502 |
> |
|
503 |
> |
txt += ' echo "JOB_EXIT_STATUS = 1"\n' |
504 |
> |
txt += ' exit 1\n' |
505 |
> |
txt += ' fi\n' |
506 |
> |
txt += '\n' |
507 |
> |
txt += ' echo "Change to working directory: $WORKING_DIR"\n' |
508 |
> |
txt += ' cd $WORKING_DIR\n' |
509 |
> |
txt += self.wsSetupCMSOSGEnvironment_() |
510 |
> |
txt += 'fi\n' |
511 |
|
|
512 |
|
# Prepare JobType-specific part |
513 |
|
scram = self.scram.commandName() |
517 |
|
txt += 'status=$?\n' |
518 |
|
txt += 'if [ $status != 0 ] ; then\n' |
519 |
|
txt += ' echo "SET_EXE_ENV 1 ==>ERROR CMSSW '+self.version+' not found on `hostname`" \n' |
520 |
< |
txt += ' echo "JOB_EXIT_STATUS = 5"\n' |
521 |
< |
txt += ' echo "SanityCheckCode = 5" | tee -a $RUNTIME_AREA/$repo\n' |
520 |
> |
txt += ' echo "JOB_EXIT_STATUS = 10034"\n' |
521 |
> |
txt += ' echo "SanityCheckCode = 10034" | tee -a $RUNTIME_AREA/$repo\n' |
522 |
|
txt += ' dumpStatus $RUNTIME_AREA/$repo\n' |
523 |
< |
txt += ' exit 5 \n' |
523 |
> |
## OLI_Daniele |
524 |
> |
txt += ' if [ $middleware == OSG ]; then \n' |
525 |
> |
txt += ' echo "Remove working directory: $WORKING_DIR"\n' |
526 |
> |
txt += ' cd $RUNTIME_AREA\n' |
527 |
> |
txt += ' /bin/rm -rf $WORKING_DIR\n' |
528 |
> |
txt += ' if [ -d $WORKING_DIR ] ;then\n' |
529 |
> |
txt += ' echo "OSG WORKING DIR ==> $WORKING_DIR could not be deleted on on WN `hostname`"\n' |
530 |
> |
txt += ' fi\n' |
531 |
> |
txt += ' fi \n' |
532 |
> |
txt += ' exit 1 \n' |
533 |
|
txt += 'fi \n' |
534 |
|
txt += 'echo "CMSSW_VERSION = '+self.version+'"\n' |
535 |
|
txt += 'cd '+self.version+'\n' |
543 |
|
# txt += "## ARGUMNETS: $3 Max Event for this job\n" |
544 |
|
txt += "\n" |
545 |
|
txt += "narg=$#\n" |
546 |
< |
txt += "if [ $narg -lt 1 ]\n" |
546 |
> |
txt += "if [ $narg -lt 2 ]\n" |
547 |
|
txt += "then\n" |
548 |
|
txt += " echo 'SET_EXE_ENV 1 ==> ERROR Too few arguments' +$narg+ \n" |
549 |
< |
txt += ' echo "JOB_EXIT_STATUS = 1"\n' |
550 |
< |
txt += ' echo "SanityCheckCode = 1" | tee -a $RUNTIME_AREA/$repo\n' |
549 |
> |
txt += ' echo "JOB_EXIT_STATUS = 50113"\n' |
550 |
> |
txt += ' echo "SanityCheckCode = 50113" | tee -a $RUNTIME_AREA/$repo\n' |
551 |
|
txt += ' dumpStatus $RUNTIME_AREA/$repo\n' |
552 |
+ |
## OLI_Daniele |
553 |
+ |
txt += ' if [ $middleware == OSG ]; then \n' |
554 |
+ |
txt += ' echo "Remove working directory: $WORKING_DIR"\n' |
555 |
+ |
txt += ' cd $RUNTIME_AREA\n' |
556 |
+ |
txt += ' /bin/rm -rf $WORKING_DIR\n' |
557 |
+ |
txt += ' if [ -d $WORKING_DIR ] ;then\n' |
558 |
+ |
txt += ' echo "OSG WORKING DIR ==> $WORKING_DIR could not be deleted on on WN `hostname`"\n' |
559 |
+ |
txt += ' fi\n' |
560 |
+ |
txt += ' fi \n' |
561 |
|
txt += " exit 1\n" |
562 |
|
txt += "fi\n" |
563 |
|
txt += "\n" |
564 |
|
txt += "NJob=$1\n" |
565 |
< |
# txt += "FirstEvent=$2\n" |
566 |
< |
# txt += "MaxEvents=$3\n" |
565 |
> |
txt += "InputFiles=$2\n" |
566 |
> |
txt += "echo \"<$InputFiles>\"\n" |
567 |
> |
# txt += "Args = ` cat $2 | sed -e \'s/\\\\//g\' -e \'s/\"/\\x27/g\' `" |
568 |
> |
|
569 |
> |
### OLI_DANIELE |
570 |
> |
txt += 'if [ $middleware == LCG ]; then \n' |
571 |
> |
txt += ' echo "MonitorJobID=`echo ${NJob}_$EDG_WL_JOBID`" | tee -a $RUNTIME_AREA/$repo\n' |
572 |
> |
txt += ' echo "SyncGridJobId=`echo $EDG_WL_JOBID`" | tee -a $RUNTIME_AREA/$repo\n' |
573 |
> |
txt += ' echo "SyncCE=`edg-brokerinfo getCE`" | tee -a $RUNTIME_AREA/$repo\n' |
574 |
> |
txt += 'elif [ $middleware == OSG ]; then\n' |
575 |
> |
|
576 |
> |
# OLI: added monitoring for dashbord, use hash of crab.cfg |
577 |
> |
if common.scheduler.boss_scheduler_name == 'condor_g': |
578 |
> |
# create hash of cfg file |
579 |
> |
hash = makeCksum(common.work_space.cfgFileName()) |
580 |
> |
txt += ' echo "MonitorJobID=`echo ${NJob}_'+hash+'_$GLOBUS_GRAM_JOB_CONTACT`" | tee -a $RUNTIME_AREA/$repo\n' |
581 |
> |
txt += ' echo "SyncGridJobId=`echo $GLOBUS_GRAM_JOB_CONTACT`" | tee -a $RUNTIME_AREA/$repo\n' |
582 |
> |
txt += ' echo "SyncCE=`echo $hostname`" | tee -a $RUNTIME_AREA/$repo\n' |
583 |
> |
else : |
584 |
> |
txt += ' echo "MonitorJobID=`echo ${NJob}_$EDG_WL_JOBID`" | tee -a $RUNTIME_AREA/$repo\n' |
585 |
> |
txt += ' echo "SyncGridJobId=`echo $EDG_WL_JOBID`" | tee -a $RUNTIME_AREA/$repo\n' |
586 |
> |
txt += ' echo "SyncCE=`$EDG_WL_LOG_DESTINATION`" | tee -a $RUNTIME_AREA/$repo\n' |
587 |
> |
|
588 |
> |
txt += 'fi\n' |
589 |
> |
txt += 'dumpStatus $RUNTIME_AREA/$repo\n' |
590 |
|
|
591 |
|
# Prepare job-specific part |
592 |
|
job = common.job_list[nj] |
593 |
|
pset = os.path.basename(job.configFilename()) |
594 |
|
txt += '\n' |
595 |
< |
txt += 'cp $RUNTIME_AREA/'+pset+' pset.cfg\n' |
596 |
< |
# txt += 'if [ -e $RUNTIME_AREA/orcarc_$CE ] ; then\n' |
597 |
< |
# txt += ' cat $RUNTIME_AREA/orcarc_$CE .orcarc >> .orcarc_tmp\n' |
404 |
< |
# txt += ' mv .orcarc_tmp .orcarc\n' |
405 |
< |
# txt += 'fi\n' |
406 |
< |
# txt += 'if [ -e $RUNTIME_AREA/init_$CE.sh ] ; then\n' |
407 |
< |
# txt += ' cp $RUNTIME_AREA/init_$CE.sh init.sh\n' |
408 |
< |
# txt += 'fi\n' |
595 |
> |
#txt += 'echo sed "s#{\'INPUT\'}#$InputFiles#" $RUNTIME_AREA/'+pset+' \n' |
596 |
> |
txt += 'sed "s#{\'INPUT\'}#$InputFiles#" $RUNTIME_AREA/'+pset+' > pset.cfg\n' |
597 |
> |
#txt += 'sed "s#{\'INPUT\'}#${InputFiles}#" $RUNTIME_AREA/'+pset+' > pset1.cfg\n' |
598 |
|
|
599 |
|
if len(self.additional_inbox_files) > 0: |
600 |
|
for file in self.additional_inbox_files: |
604 |
|
txt += 'fi\n' |
605 |
|
pass |
606 |
|
|
418 |
– |
# txt += '\n' |
419 |
– |
# txt += 'chmod +x ./init.sh\n' |
420 |
– |
# txt += './init.sh\n' |
421 |
– |
# txt += 'exitStatus=$?\n' |
422 |
– |
# txt += 'if [ $exitStatus != 0 ] ; then\n' |
423 |
– |
# txt += ' echo "SET_EXE_ENV 1 ==> ERROR StageIn init script failed"\n' |
424 |
– |
# txt += ' echo "JOB_EXIT_STATUS = $exitStatus" \n' |
425 |
– |
# txt += ' echo "SanityCheckCode = $exitStatus" | tee -a $RUNTIME_AREA/$repo\n' |
426 |
– |
# txt += ' dumpStatus $RUNTIME_AREA/$repo\n' |
427 |
– |
# txt += ' exit $exitStatus\n' |
428 |
– |
# txt += 'fi\n' |
429 |
– |
# txt += "echo 'SET_EXE_ENV 0 ==> job setup ok'\n" |
607 |
|
txt += 'echo "### END JOB SETUP ENVIRONMENT ###"\n\n' |
608 |
|
|
432 |
– |
# txt += 'echo "FirstEvent=$FirstEvent" >> .orcarc\n' |
433 |
– |
# txt += 'echo "MaxEvents=$MaxEvents" >> .orcarc\n' |
434 |
– |
# if self.ML: |
435 |
– |
# txt += 'echo "MonalisaJobId=$NJob" >> .orcarc\n' |
436 |
– |
|
609 |
|
txt += '\n' |
610 |
|
txt += 'echo "***** cat pset.cfg *********"\n' |
611 |
|
txt += 'cat pset.cfg\n' |
612 |
|
txt += 'echo "****** end pset.cfg ********"\n' |
613 |
+ |
txt += '\n' |
614 |
+ |
# txt += 'echo "***** cat pset1.cfg *********"\n' |
615 |
+ |
# txt += 'cat pset1.cfg\n' |
616 |
+ |
# txt += 'echo "****** end pset1.cfg ********"\n' |
617 |
+ |
return txt |
618 |
+ |
|
619 |
+ |
def wsBuildExe(self, nj): |
620 |
+ |
""" |
621 |
+ |
Put in the script the commands to build an executable |
622 |
+ |
or a library. |
623 |
+ |
""" |
624 |
+ |
|
625 |
+ |
txt = "" |
626 |
+ |
|
627 |
+ |
if os.path.isfile(self.tgzNameWithPath): |
628 |
+ |
txt += 'echo "tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+'"\n' |
629 |
+ |
txt += 'tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+'\n' |
630 |
+ |
txt += 'untar_status=$? \n' |
631 |
+ |
txt += 'if [ $untar_status -ne 0 ]; then \n' |
632 |
+ |
txt += ' echo "SET_EXE 1 ==> ERROR Untarring .tgz file failed"\n' |
633 |
+ |
txt += ' echo "JOB_EXIT_STATUS = $untar_status" \n' |
634 |
+ |
txt += ' echo "SanityCheckCode = $untar_status" | tee -a $repo\n' |
635 |
+ |
txt += ' if [ $middleware == OSG ]; then \n' |
636 |
+ |
txt += ' echo "Remove working directory: $WORKING_DIR"\n' |
637 |
+ |
txt += ' cd $RUNTIME_AREA\n' |
638 |
+ |
txt += ' /bin/rm -rf $WORKING_DIR\n' |
639 |
+ |
txt += ' if [ -d $WORKING_DIR ] ;then\n' |
640 |
+ |
txt += ' echo "OSG WORKING DIR ==> $WORKING_DIR could not be deleted on on WN `hostname`"\n' |
641 |
+ |
txt += ' fi\n' |
642 |
+ |
txt += ' fi \n' |
643 |
+ |
txt += ' \n' |
644 |
+ |
txt += ' exit $untar_status \n' |
645 |
+ |
txt += 'else \n' |
646 |
+ |
txt += ' echo "Successful untar" \n' |
647 |
+ |
txt += 'fi \n' |
648 |
+ |
pass |
649 |
+ |
|
650 |
|
return txt |
651 |
|
|
652 |
|
def modifySteeringCards(self, nj): |
659 |
|
return self.executable |
660 |
|
|
661 |
|
def executableArgs(self): |
662 |
< |
return "-p pset.cfg" |
662 |
> |
return " -p pset.cfg" |
663 |
|
|
664 |
|
def inputSandbox(self, nj): |
665 |
|
""" |
674 |
|
## config |
675 |
|
inp_box.append(common.job_list[nj].configFilename()) |
676 |
|
## additional input files |
677 |
< |
for file in self.additional_inbox_files: |
678 |
< |
inp_box.append(common.work_space.cwdDir()+file) |
470 |
< |
#print "sono inputSandbox, inp_box = ", inp_box |
677 |
> |
#for file in self.additional_inbox_files: |
678 |
> |
# inp_box.append(common.work_space.cwdDir()+file) |
679 |
|
return inp_box |
680 |
|
|
681 |
|
def outputSandbox(self, nj): |
698 |
|
""" |
699 |
|
Make initial modifications of the user's steering card file. |
700 |
|
""" |
493 |
– |
infile = open(self.pset,'r') |
494 |
– |
|
495 |
– |
outfile = open(common.work_space.jobDir()+self.name()+'.cfg', 'w') |
496 |
– |
|
497 |
– |
outfile.write('\n\n##### The following cards have been created by CRAB: DO NOT TOUCH #####\n') |
498 |
– |
|
499 |
– |
outfile.write('InputCollections=/System/'+self.owner+'/'+self.dataset+'/'+self.dataset+'\n') |
500 |
– |
|
501 |
– |
infile.close() |
502 |
– |
outfile.close() |
701 |
|
return |
702 |
|
|
703 |
|
def wsRenameOutput(self, nj): |
707 |
|
|
708 |
|
txt = '\n' |
709 |
|
file_list = '' |
710 |
+ |
check = len(self.output_file) |
711 |
+ |
i = 0 |
712 |
|
for fileWithSuffix in self.output_file: |
713 |
+ |
i= i + 1 |
714 |
|
output_file_num = self.numberFile_(fileWithSuffix, '$NJob') |
715 |
< |
file_list=file_list+output_file_num+' ' |
715 |
> |
file_list=file_list+output_file_num+'' |
716 |
|
txt += '\n' |
717 |
|
txt += 'ls \n' |
718 |
|
txt += '\n' |
720 |
|
txt += 'exe_result=$?\n' |
721 |
|
txt += 'if [ $exe_result -ne 0 ] ; then\n' |
722 |
|
txt += ' echo "ERROR: No output file to manage"\n' |
723 |
< |
txt += ' echo "JOB_EXIT_STATUS = $exe_result"\n' |
724 |
< |
txt += ' echo "SanityCheckCode = $exe_result" | tee -a $RUNTIME_AREA/$repo\n' |
725 |
< |
txt += ' dumpStatus $RUNTIME_AREA/$repo\n' |
726 |
< |
txt += ' exit $exe_result \n' |
723 |
> |
### OLI_DANIELE |
724 |
> |
txt += ' if [ $middleware == OSG ]; then \n' |
725 |
> |
txt += ' echo "prepare dummy output file"\n' |
726 |
> |
txt += ' cp '+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n' |
727 |
> |
txt += ' fi \n' |
728 |
|
txt += 'else\n' |
729 |
|
txt += ' cp '+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n' |
730 |
|
txt += 'fi\n' |
731 |
< |
txt += 'cd $RUNTIME_AREA\n' |
732 |
< |
|
731 |
> |
if i == check: |
732 |
> |
txt += 'cd $RUNTIME_AREA\n' |
733 |
> |
pass |
734 |
|
pass |
735 |
|
|
736 |
|
file_list=file_list[:-1] |
737 |
|
txt += 'file_list="'+file_list+'"\n' |
738 |
+ |
### OLI_DANIELE |
739 |
+ |
txt += 'if [ $middleware == OSG ]; then\n' |
740 |
+ |
txt += ' cd $RUNTIME_AREA\n' |
741 |
+ |
txt += ' echo "Remove working directory: $WORKING_DIR"\n' |
742 |
+ |
txt += ' /bin/rm -rf $WORKING_DIR\n' |
743 |
+ |
txt += ' if [ -d $WORKING_DIR ] ;then\n' |
744 |
+ |
txt += ' echo "OSG WORKING DIR ==> $WORKING_DIR could not be deleted on on WN `hostname`"\n' |
745 |
+ |
txt += ' fi\n' |
746 |
+ |
txt += 'fi\n' |
747 |
+ |
txt += '\n' |
748 |
|
return txt |
749 |
|
|
750 |
|
def numberFile_(self, file, txt): |
787 |
|
req = req + ')' |
788 |
|
#print "req = ", req |
789 |
|
return req |
790 |
+ |
|
791 |
+ |
def configFilename(self): |
792 |
+ |
""" return the config filename """ |
793 |
+ |
return self.name()+'.cfg' |
794 |
+ |
|
795 |
+ |
### OLI_DANIELE |
796 |
+ |
def wsSetupCMSOSGEnvironment_(self): |
797 |
+ |
""" |
798 |
+ |
Returns part of a job script which is prepares |
799 |
+ |
the execution environment and which is common for all CMS jobs. |
800 |
+ |
""" |
801 |
+ |
txt = '\n' |
802 |
+ |
txt += ' echo "### SETUP CMS OSG ENVIRONMENT ###"\n' |
803 |
+ |
txt += ' if [ -f $GRID3_APP_DIR/cmssoft/cmsset_default.sh ] ;then\n' |
804 |
+ |
txt += ' # Use $GRID3_APP_DIR/cmssoft/cmsset_default.sh to setup cms software\n' |
805 |
+ |
txt += ' source $GRID3_APP_DIR/cmssoft/cmsset_default.sh '+self.version+'\n' |
806 |
+ |
txt += ' elif [ -f $OSG_APP/cmssoft/cmsset_default.sh ] ;then\n' |
807 |
+ |
txt += ' # Use $OSG_APP/cmssoft/cmsset_default.sh to setup cms software\n' |
808 |
+ |
txt += ' source $OSG_APP/cmssoft/cmsset_default.sh '+self.version+'\n' |
809 |
+ |
txt += ' else\n' |
810 |
+ |
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' |
811 |
+ |
txt += ' echo "JOB_EXIT_STATUS = 10020"\n' |
812 |
+ |
txt += ' echo "JobExitCode=10020" | tee -a $RUNTIME_AREA/$repo\n' |
813 |
+ |
txt += ' dumpStatus $RUNTIME_AREA/$repo\n' |
814 |
+ |
txt += ' exit\n' |
815 |
+ |
txt += '\n' |
816 |
+ |
txt += ' echo "Remove working directory: $WORKING_DIR"\n' |
817 |
+ |
txt += ' cd $RUNTIME_AREA\n' |
818 |
+ |
txt += ' /bin/rm -rf $WORKING_DIR\n' |
819 |
+ |
txt += ' if [ -d $WORKING_DIR ] ;then\n' |
820 |
+ |
txt += ' echo "OSG WORKING DIR ==> $WORKING_DIR could not be deleted on on WN `hostname`"\n' |
821 |
+ |
txt += ' fi\n' |
822 |
+ |
txt += '\n' |
823 |
+ |
txt += ' exit\n' |
824 |
+ |
txt += ' fi\n' |
825 |
+ |
txt += '\n' |
826 |
+ |
txt += ' echo "SET_CMS_ENV 0 ==> setup cms environment ok"\n' |
827 |
+ |
txt += ' echo " END SETUP CMS OSG ENVIRONMENT "\n' |
828 |
+ |
|
829 |
+ |
return txt |
830 |
+ |
|
831 |
+ |
### OLI_DANIELE |
832 |
+ |
def wsSetupCMSLCGEnvironment_(self): |
833 |
+ |
""" |
834 |
+ |
Returns part of a job script which is prepares |
835 |
+ |
the execution environment and which is common for all CMS jobs. |
836 |
+ |
""" |
837 |
+ |
txt = ' \n' |
838 |
+ |
txt += ' echo " ### SETUP CMS LCG ENVIRONMENT ### "\n' |
839 |
+ |
txt += ' echo "JOB_EXIT_STATUS = 0"\n' |
840 |
+ |
txt += ' if [ ! $VO_CMS_SW_DIR ] ;then\n' |
841 |
+ |
txt += ' echo "SET_CMS_ENV 10031 ==> ERROR CMS software dir not found on WN `hostname`"\n' |
842 |
+ |
txt += ' echo "JOB_EXIT_STATUS = 10031" \n' |
843 |
+ |
txt += ' echo "JobExitCode=10031" | tee -a $RUNTIME_AREA/$repo\n' |
844 |
+ |
txt += ' dumpStatus $RUNTIME_AREA/$repo\n' |
845 |
+ |
txt += ' exit\n' |
846 |
+ |
txt += ' else\n' |
847 |
+ |
txt += ' echo "Sourcing environment... "\n' |
848 |
+ |
txt += ' if [ ! -s $VO_CMS_SW_DIR/cmsset_default.sh ] ;then\n' |
849 |
+ |
txt += ' echo "SET_CMS_ENV 10020 ==> ERROR cmsset_default.sh file not found into dir $VO_CMS_SW_DIR"\n' |
850 |
+ |
txt += ' echo "JOB_EXIT_STATUS = 10020"\n' |
851 |
+ |
txt += ' echo "JobExitCode=10020" | tee -a $RUNTIME_AREA/$repo\n' |
852 |
+ |
txt += ' dumpStatus $RUNTIME_AREA/$repo\n' |
853 |
+ |
txt += ' exit\n' |
854 |
+ |
txt += ' fi\n' |
855 |
+ |
txt += ' echo "sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n' |
856 |
+ |
txt += ' source $VO_CMS_SW_DIR/cmsset_default.sh\n' |
857 |
+ |
txt += ' result=$?\n' |
858 |
+ |
txt += ' if [ $result -ne 0 ]; then\n' |
859 |
+ |
txt += ' echo "SET_CMS_ENV 10032 ==> ERROR problem sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n' |
860 |
+ |
txt += ' echo "JOB_EXIT_STATUS = 10032"\n' |
861 |
+ |
txt += ' echo "JobExitCode=10032" | tee -a $RUNTIME_AREA/$repo\n' |
862 |
+ |
txt += ' dumpStatus $RUNTIME_AREA/$repo\n' |
863 |
+ |
txt += ' exit\n' |
864 |
+ |
txt += ' fi\n' |
865 |
+ |
txt += ' fi\n' |
866 |
+ |
txt += ' \n' |
867 |
+ |
txt += ' string=`cat /etc/redhat-release`\n' |
868 |
+ |
txt += ' echo $string\n' |
869 |
+ |
txt += ' if [[ $string = *alhalla* ]]; then\n' |
870 |
+ |
txt += ' echo "SCRAM_ARCH= $SCRAM_ARCH"\n' |
871 |
+ |
txt += ' elif [[ $string = *Enterprise* ]] || [[ $string = *cientific* ]]; then\n' |
872 |
+ |
txt += ' export SCRAM_ARCH=slc3_ia32_gcc323\n' |
873 |
+ |
txt += ' echo "SCRAM_ARCH= $SCRAM_ARCH"\n' |
874 |
+ |
txt += ' else\n' |
875 |
+ |
txt += ' echo "SET_CMS_ENV 1 ==> ERROR OS unknown, LCG environment not initialized"\n' |
876 |
+ |
txt += ' echo "JOB_EXIT_STATUS = 10033"\n' |
877 |
+ |
txt += ' echo "JobExitCode=10033" | tee -a $RUNTIME_AREA/$repo\n' |
878 |
+ |
txt += ' dumpStatus $RUNTIME_AREA/$repo\n' |
879 |
+ |
txt += ' exit\n' |
880 |
+ |
txt += ' fi\n' |
881 |
+ |
txt += ' echo "SET_CMS_ENV 0 ==> setup cms environment ok"\n' |
882 |
+ |
txt += ' echo "### END SETUP CMS LCG ENVIRONMENT ###"\n' |
883 |
+ |
return txt |