| 3 |
|
import time, glob |
| 4 |
|
from Actor import * |
| 5 |
|
from crab_util import * |
| 6 |
– |
from crab_logger import Logger |
| 6 |
|
from crab_exceptions import * |
| 7 |
< |
from FwkJobRep.ReportParser import readJobReport |
| 7 |
> |
from ProdCommon.FwkJobRep.ReportParser import readJobReport |
| 8 |
> |
from ProdCommon.FwkJobRep.ReportState import checkSuccess |
| 9 |
|
from ProdCommon.MCPayloads.WorkflowSpec import WorkflowSpec |
| 10 |
|
from ProdCommon.DataMgmt.DBS.DBSWriter import DBSWriter |
| 11 |
|
from ProdCommon.DataMgmt.DBS.DBSErrors import DBSWriterError, formatEx,DBSReaderError |
| 23 |
|
- publishes output data on DBS and DLS |
| 24 |
|
""" |
| 25 |
|
|
| 26 |
+ |
self.cfg_params=cfg_params |
| 27 |
+ |
|
| 28 |
|
try: |
| 29 |
< |
self.processedData = cfg_params['USER.publish_data_name'] |
| 28 |
< |
except KeyError: |
| 29 |
> |
if not cfg_params.has_key('USER.publish_data_name') |
| 30 |
|
raise CrabException('Cannot publish output data, because you did not specify USER.publish_data_name parameter in the crab.cfg file') |
| 31 |
< |
try: |
| 32 |
< |
if (int(cfg_params['USER.copy_data']) != 1): raise KeyError |
| 33 |
< |
except KeyError: |
| 34 |
< |
raise CrabException('You can not publish data because you did not selected *** copy_data = 1 *** in the crab.cfg file') |
| 35 |
< |
try: |
| 36 |
< |
self.pset = cfg_params['CMSSW.pset'] |
| 37 |
< |
except KeyError: |
| 38 |
< |
raise CrabException('Cannot publish output data, because you did not specify the psetname in [CMSSW] of your crab.cfg file') |
| 39 |
< |
try: |
| 40 |
< |
self.globalDBS=cfg_params['CMSSW.dbs_url'] |
| 41 |
< |
except KeyError: |
| 42 |
< |
self.globalDBS="http://cmsdbsprod.cern.ch/cms_dbs_prod_global/servlet/DBSServlet" |
| 43 |
< |
try: |
| 44 |
< |
self.DBSURL=cfg_params['USER.dbs_url_for_publication'] |
| 45 |
< |
common.logger.message('<dbs_url_for_publication> = '+self.DBSURL) |
| 46 |
< |
if (self.DBSURL == "http://cmsdbsprod.cern.ch/cms_dbs_prod_global/servlet/DBSServlet") or (self.DBSURL == "https://cmsdbsprod.cern.ch:8443/cms_dbs_prod_global_writer/servlet/DBSServlet"): |
| 47 |
< |
msg = "You can not publish your data in the globalDBS = " + self.DBSURL + "\n" |
| 48 |
< |
msg = msg + "Please write your local one in the [USER] section 'dbs_url_for_publication'" |
| 49 |
< |
raise CrabException(msg) |
| 50 |
< |
except KeyError: |
| 51 |
< |
msg = "Error. The [USER] section does not have 'dbs_url_for_publication'" |
| 52 |
< |
msg = msg + " entry, necessary to publish the data" |
| 31 |
> |
self.userprocessedData = cfg_params['USER.publish_data_name'] |
| 32 |
> |
self.processedData = None |
| 33 |
> |
|
| 34 |
> |
if (not cfg_params.has_key('USER.copy_data') or int(cfg_params['USER.copy_data']) != 1 ) or \ |
| 35 |
> |
(not cfg_params.has_key('USER.publish_data') or int(cfg_params['USER.publish_data']) != 1 ): |
| 36 |
> |
msg = 'You can not publish data because you did not selected \n' |
| 37 |
> |
msg += '\t*** copy_data = 1 or publish_data = 1 *** in the crab.cfg file' |
| 38 |
> |
|
| 39 |
> |
# try: |
| 40 |
> |
# self.pset = cfg_params['CMSSW.pset'] |
| 41 |
> |
# except KeyError: |
| 42 |
> |
# raise CrabException('Cannot publish output data, because you did not specify the psetname in [CMSSW] of your crab.cfg file') |
| 43 |
> |
|
| 44 |
> |
self.globalDBS=cfg_params.get('CMSSW.dbs_url',"http://cmsdbsprod.cern.ch/cms_dbs_prod_global/servlet/DBSServlet") |
| 45 |
> |
|
| 46 |
> |
if not cfg_params.has_key('USER.dbs_url_for_publication'): |
| 47 |
> |
msg = "Warning. The [USER] section does not have 'dbs_url_for_publication'" |
| 48 |
> |
msg = msg + " entry, necessary to publish the data.\n" |
| 49 |
> |
msg = msg + "Use the command **crab -publish -USER.dbs_url_for_publication=dbs_url_for_publication*** \nwhere dbs_url_for_publication is your local dbs instance." |
| 50 |
> |
raise CrabException(msg) |
| 51 |
> |
|
| 52 |
> |
self.DBSURL=cfg_params['USER.dbs_url_for_publication'] |
| 53 |
> |
common.logger.info('<dbs_url_for_publication> = '+self.DBSURL) |
| 54 |
> |
if (self.DBSURL == "http://cmsdbsprod.cern.ch/cms_dbs_prod_global/servlet/DBSServlet") or (self.DBSURL == "https://cmsdbsprod.cern.ch:8443/cms_dbs_prod_global_writer/servlet/DBSServlet"): |
| 55 |
> |
msg = "You can not publish your data in the globalDBS = " + self.DBSURL + "\n" |
| 56 |
> |
msg = msg + "Please write your local one in the [USER] section 'dbs_url_for_publication'" |
| 57 |
|
raise CrabException(msg) |
| 58 |
|
|
| 59 |
|
self.content=file(self.pset).read() |
| 60 |
|
self.resDir = common.work_space.resDir() |
| 61 |
+ |
|
| 62 |
+ |
self.dataset_to_import=[] |
| 63 |
+ |
|
| 64 |
|
self.datasetpath=cfg_params['CMSSW.datasetpath'] |
| 65 |
+ |
if (self.datasetpath.upper() != 'NONE'): |
| 66 |
+ |
self.dataset_to_import.append(self.datasetpath) |
| 67 |
+ |
|
| 68 |
+ |
### Added PU dataset |
| 69 |
+ |
tmp = cfg_params.get('CMSSW.dataset_pu',None) |
| 70 |
+ |
if tmp : |
| 71 |
+ |
datasets = tmp.split(',') |
| 72 |
+ |
for dataset in datasets: |
| 73 |
+ |
dataset=string.strip(dataset) |
| 74 |
+ |
self.dataset_to_import.append(dataset) |
| 75 |
+ |
### |
| 76 |
+ |
|
| 77 |
+ |
self.skipOcheck=cfg_params.get('CMSSW.publish_zero_event',0) |
| 78 |
+ |
|
| 79 |
|
self.SEName='' |
| 80 |
|
self.CMSSW_VERSION='' |
| 81 |
|
self.exit_status='' |
| 90 |
|
dbsWriter = DBSWriter(self.DBSURL,level='ERROR') |
| 91 |
|
|
| 92 |
|
try: |
| 93 |
+ |
#dbsWriter.importDatasetWithoutParentage(globalDBS, self.datasetpath, self.DBSURL) |
| 94 |
|
dbsWriter.importDataset(globalDBS, self.datasetpath, self.DBSURL) |
| 95 |
|
except DBSWriterError, ex: |
| 96 |
|
msg = "Error importing dataset to be processed into local DBS\n" |
| 97 |
|
msg += "Source Dataset: %s\n" % datasetpath |
| 98 |
|
msg += "Source DBS: %s\n" % globalDBS |
| 99 |
|
msg += "Destination DBS: %s\n" % self.DBSURL |
| 100 |
< |
common.logger.message(msg) |
| 100 |
> |
common.logger.info(msg) |
| 101 |
> |
common.logger.debug(str(ex)) |
| 102 |
|
return 1 |
| 103 |
|
return 0 |
| 104 |
|
|
| 111 |
|
except IndexError: |
| 112 |
|
self.exit_status = '1' |
| 113 |
|
msg = "Error: Problem with "+file+" file" |
| 114 |
< |
common.logger.message(msg) |
| 114 |
> |
common.logger.info(msg) |
| 115 |
|
return self.exit_status |
| 116 |
< |
|
| 117 |
< |
if (self.datasetpath != 'None'): |
| 118 |
< |
common.logger.message("--->>> Importing parent dataset in the dbs") |
| 119 |
< |
status_import=self.importParentDataset(self.globalDBS, self.datasetpath) |
| 120 |
< |
if (status_import == 1): |
| 121 |
< |
common.logger.message('Problem with parent import from the global DBS '+self.globalDBS+ 'to the local one '+self.DBSURL) |
| 122 |
< |
self.exit_status='1' |
| 123 |
< |
return self.exit_status |
| 124 |
< |
common.logger.message("Parent import ok") |
| 116 |
> |
|
| 117 |
> |
if (len(self.dataset_to_import) != 0): |
| 118 |
> |
for dataset in self.dataset_to_import: |
| 119 |
> |
common.logger.info("--->>> Importing parent dataset in the dbs: " +dataset) |
| 120 |
> |
status_import=self.importParentDataset(self.globalDBS, dataset) |
| 121 |
> |
if (status_import == 1): |
| 122 |
> |
common.logger.info('Problem with parent '+ dataset +' import from the global DBS '+self.globalDBS+ 'to the local one '+self.DBSURL) |
| 123 |
> |
self.exit_status='1' |
| 124 |
> |
return self.exit_status |
| 125 |
> |
else: |
| 126 |
> |
common.logger.info('Import ok of dataset '+dataset) |
| 127 |
|
|
| 128 |
|
#// DBS to contact |
| 129 |
|
dbswriter = DBSWriter(self.DBSURL) |
| 133 |
|
except IndexError: |
| 134 |
|
self.exit_status = '1' |
| 135 |
|
msg = "Error: No file to publish in xml file"+file+" file" |
| 136 |
< |
common.logger.message(msg) |
| 136 |
> |
common.logger.info(msg) |
| 137 |
|
return self.exit_status |
| 138 |
|
|
| 139 |
|
datasets=fileinfo.dataset |
| 140 |
< |
common.logger.debug(6,"FileInfo = " + str(fileinfo)) |
| 141 |
< |
common.logger.debug(6,"DatasetInfo = " + str(datasets)) |
| 140 |
> |
common.logger.log(10-1,"FileInfo = " + str(fileinfo)) |
| 141 |
> |
common.logger.log(10-1,"DatasetInfo = " + str(datasets)) |
| 142 |
> |
if len(datasets)<=0: |
| 143 |
> |
self.exit_status = '1' |
| 144 |
> |
msg = "Error: No info about dataset in the xml file "+file |
| 145 |
> |
common.logger.info(msg) |
| 146 |
> |
return self.exit_status |
| 147 |
|
for dataset in datasets: |
| 148 |
|
#### for production data |
| 149 |
+ |
self.processedData = dataset['ProcessedDataset'] |
| 150 |
|
if (dataset['PrimaryDataset'] == 'null'): |
| 151 |
< |
dataset['PrimaryDataset'] = dataset['ProcessedDataset'] |
| 152 |
< |
|
| 151 |
> |
#dataset['PrimaryDataset'] = dataset['ProcessedDataset'] |
| 152 |
> |
dataset['PrimaryDataset'] = self.userprocessedData |
| 153 |
> |
#else: # add parentage from input dataset |
| 154 |
> |
elif self.datasetpath.upper() != 'NONE': |
| 155 |
> |
dataset['ParentDataset']= self.datasetpath |
| 156 |
> |
|
| 157 |
|
dataset['PSetContent']=self.content |
| 158 |
|
cfgMeta = {'name' : self.pset , 'Type' : 'user' , 'annotation': 'user cfg', 'version' : 'private version'} # add real name of user cfg |
| 159 |
< |
common.logger.message("PrimaryDataset = %s"%dataset['PrimaryDataset']) |
| 160 |
< |
common.logger.message("ProcessedDataset = %s"%dataset['ProcessedDataset']) |
| 161 |
< |
common.logger.message("<User Dataset Name> = /"+dataset['PrimaryDataset']+"/"+dataset['ProcessedDataset']+"/USER") |
| 159 |
> |
common.logger.info("PrimaryDataset = %s"%dataset['PrimaryDataset']) |
| 160 |
> |
common.logger.info("ProcessedDataset = %s"%dataset['ProcessedDataset']) |
| 161 |
> |
common.logger.info("<User Dataset Name> = /"+dataset['PrimaryDataset']+"/"+dataset['ProcessedDataset']+"/USER") |
| 162 |
> |
self.dataset_to_check="/"+dataset['PrimaryDataset']+"/"+dataset['ProcessedDataset']+"/USER" |
| 163 |
|
|
| 164 |
< |
common.logger.debug(6,"--->>> Inserting primary: %s processed : %s"%(dataset['PrimaryDataset'],dataset['ProcessedDataset'])) |
| 164 |
> |
common.logger.log(10-1,"--->>> Inserting primary: %s processed : %s"%(dataset['PrimaryDataset'],dataset['ProcessedDataset'])) |
| 165 |
|
|
| 166 |
|
primary = DBSWriterObjects.createPrimaryDataset( dataset, dbswriter.dbs) |
| 167 |
< |
common.logger.debug(6,"Primary: %s "%primary) |
| 167 |
> |
common.logger.log(10-1,"Primary: %s "%primary) |
| 168 |
|
|
| 169 |
|
algo = DBSWriterObjects.createAlgorithm(dataset, cfgMeta, dbswriter.dbs) |
| 170 |
< |
common.logger.debug(6,"Algo: %s "%algo) |
| 170 |
> |
common.logger.log(10-1,"Algo: %s "%algo) |
| 171 |
|
|
| 172 |
|
processed = DBSWriterObjects.createProcessedDataset(primary, algo, dataset, dbswriter.dbs) |
| 173 |
< |
common.logger.debug(6,"Processed: %s "%processed) |
| 173 |
> |
common.logger.log(10-1,"Processed: %s "%processed) |
| 174 |
|
|
| 175 |
< |
common.logger.debug(6,"Inserted primary %s processed %s"%(primary,processed)) |
| 175 |
> |
common.logger.log(10-1,"Inserted primary %s processed %s"%(primary,processed)) |
| 176 |
|
|
| 177 |
< |
common.logger.debug(6,"exit_status = %s "%self.exit_status) |
| 177 |
> |
common.logger.log(10-1,"exit_status = %s "%self.exit_status) |
| 178 |
|
return self.exit_status |
| 179 |
|
|
| 180 |
|
def publishAJobReport(self,file,procdataset): |
| 199 |
|
elif (file['LFN'] == ''): |
| 200 |
|
self.noLFN.append(file['PFN']) |
| 201 |
|
else: |
| 202 |
< |
if int(file['TotalEvents']) != 0 : |
| 203 |
< |
file.lumisections = {} |
| 202 |
> |
if self.skipOcheck==0: |
| 203 |
> |
if int(file['TotalEvents']) != 0: |
| 204 |
> |
#file.lumisections = {} |
| 205 |
> |
# lumi info are now in run hash |
| 206 |
> |
file.runs = {} |
| 207 |
> |
for ds in file.dataset: |
| 208 |
> |
### Fede for production |
| 209 |
> |
if (ds['PrimaryDataset'] == 'null'): |
| 210 |
> |
#ds['PrimaryDataset']=procdataset |
| 211 |
> |
ds['PrimaryDataset']=self.userprocessedData |
| 212 |
> |
filestopublish.append(file) |
| 213 |
> |
else: |
| 214 |
> |
self.noEventsFiles.append(file['LFN']) |
| 215 |
> |
else: |
| 216 |
> |
file.runs = {} |
| 217 |
|
for ds in file.dataset: |
| 168 |
– |
ds['ProcessedDataset']=procdataset |
| 218 |
|
### Fede for production |
| 219 |
|
if (ds['PrimaryDataset'] == 'null'): |
| 220 |
< |
ds['PrimaryDataset']=procdataset |
| 220 |
> |
#ds['PrimaryDataset']=procdataset |
| 221 |
> |
ds['PrimaryDataset']=self.userprocessedData |
| 222 |
|
filestopublish.append(file) |
| 223 |
< |
else: |
| 174 |
< |
self.noEventsFiles.append(file['LFN']) |
| 223 |
> |
|
| 224 |
|
jobReport.files = filestopublish |
| 225 |
|
### if all files of FJR have number of events = 0 |
| 226 |
|
if (len(filestopublish) == 0): |
| 232 |
|
Blocks=None |
| 233 |
|
try: |
| 234 |
|
Blocks=dbswriter.insertFiles(jobReport) |
| 235 |
< |
common.logger.message("Blocks = %s"%Blocks) |
| 235 |
> |
common.logger.info("Inserting file in blocks = %s"%Blocks) |
| 236 |
|
except DBSWriterError, ex: |
| 237 |
< |
common.logger.message("Insert file error: %s"%ex) |
| 237 |
> |
common.logger.info("Insert file error: %s"%ex) |
| 238 |
|
return Blocks |
| 239 |
|
|
| 240 |
|
def run(self): |
| 243 |
|
""" |
| 244 |
|
|
| 245 |
|
file_list = glob.glob(self.resDir+"crab_fjr*.xml") |
| 246 |
< |
common.logger.debug(6, "file_list = "+str(file_list)) |
| 247 |
< |
common.logger.debug(6, "len(file_list) = "+str(len(file_list))) |
| 246 |
> |
## Select only those fjr that are succesfull |
| 247 |
> |
good_list=[] |
| 248 |
> |
for fjr in file_list: |
| 249 |
> |
reports = readJobReport(fjr) |
| 250 |
> |
if len(reports)>0: |
| 251 |
> |
if reports[0].status == "Success": |
| 252 |
> |
good_list.append(fjr) |
| 253 |
> |
file_list=good_list |
| 254 |
> |
## |
| 255 |
> |
common.logger.log(10-1, "file_list = "+str(file_list)) |
| 256 |
> |
common.logger.log(10-1, "len(file_list) = "+str(len(file_list))) |
| 257 |
|
|
| 258 |
|
if (len(file_list)>0): |
| 259 |
|
BlocksList=[] |
| 260 |
< |
common.logger.message("--->>> Start dataset publication") |
| 260 |
> |
common.logger.info("--->>> Start dataset publication") |
| 261 |
|
self.exit_status=self.publishDataset(file_list[0]) |
| 262 |
|
if (self.exit_status == '1'): |
| 263 |
|
return self.exit_status |
| 264 |
< |
common.logger.message("--->>> End dataset publication") |
| 264 |
> |
common.logger.info("--->>> End dataset publication") |
| 265 |
|
|
| 266 |
|
|
| 267 |
< |
common.logger.message("--->>> Start files publication") |
| 267 |
> |
common.logger.info("--->>> Start files publication") |
| 268 |
|
for file in file_list: |
| 269 |
< |
common.logger.message("file = "+file) |
| 269 |
> |
common.logger.debug( "file = "+file) |
| 270 |
|
Blocks=self.publishAJobReport(file,self.processedData) |
| 271 |
|
if Blocks: |
| 272 |
< |
[BlocksList.append(x) for x in Blocks] |
| 272 |
> |
for x in Blocks: # do not allow multiple entries of the same block |
| 273 |
> |
if x not in BlocksList: |
| 274 |
> |
BlocksList.append(x) |
| 275 |
|
|
| 276 |
|
# close the blocks |
| 277 |
< |
common.logger.debug(6, "BlocksList = %s"%BlocksList) |
| 277 |
> |
common.logger.log(10-1, "BlocksList = %s"%BlocksList) |
| 278 |
|
# dbswriter = DBSWriter(self.DBSURL,level='ERROR') |
| 279 |
|
dbswriter = DBSWriter(self.DBSURL) |
| 280 |
|
|
| 281 |
|
for BlockName in BlocksList: |
| 282 |
|
try: |
| 283 |
|
closeBlock=dbswriter.manageFileBlock(BlockName,maxFiles= 1) |
| 284 |
< |
common.logger.debug(6, "closeBlock %s"%closeBlock) |
| 284 |
> |
common.logger.log(10-1, "closeBlock %s"%closeBlock) |
| 285 |
|
#dbswriter.dbs.closeBlock(BlockName) |
| 286 |
|
except DBSWriterError, ex: |
| 287 |
< |
common.logger.message("Close block error %s"%ex) |
| 287 |
> |
common.logger.info("Close block error %s"%ex) |
| 288 |
|
|
| 289 |
|
if (len(self.noEventsFiles)>0): |
| 290 |
< |
common.logger.message("--->>> WARNING: "+str(len(self.noEventsFiles))+" files not published because they contain 0 events are:") |
| 290 |
> |
common.logger.info("--->>> WARNING: "+str(len(self.noEventsFiles))+" files not published because they contain 0 events are:") |
| 291 |
|
for lfn in self.noEventsFiles: |
| 292 |
< |
common.logger.message("------ LFN: %s"%lfn) |
| 292 |
> |
common.logger.info("------ LFN: %s"%lfn) |
| 293 |
|
if (len(self.noLFN)>0): |
| 294 |
< |
common.logger.message("--->>> WARNING: there are "+str(len(self.noLFN))+" files not published because they have empty LFN") |
| 294 |
> |
common.logger.info("--->>> WARNING: there are "+str(len(self.noLFN))+" files not published because they have empty LFN") |
| 295 |
|
for pfn in self.noLFN: |
| 296 |
< |
common.logger.message("------ pfn: %s"%pfn) |
| 296 |
> |
common.logger.info("------ pfn: %s"%pfn) |
| 297 |
|
if (len(self.problemFiles)>0): |
| 298 |
< |
common.logger.message("--->>> WARNING: "+str(len(self.problemFiles))+" files not published because they had problem with copy to SE") |
| 298 |
> |
common.logger.info("--->>> WARNING: "+str(len(self.problemFiles))+" files not published because they had problem with copy to SE") |
| 299 |
|
for lfn in self.problemFiles: |
| 300 |
< |
common.logger.message("------ LFN: %s"%lfn) |
| 301 |
< |
common.logger.message("--->>> End files publication") |
| 302 |
< |
common.logger.message("--->>> To check data publication please use: InspectDBS2.py --DBSURL=<dbs_url_for_publication> --datasetPath=<User Dataset Name>") |
| 300 |
> |
common.logger.info("------ LFN: %s"%lfn) |
| 301 |
> |
common.logger.info("--->>> End files publication") |
| 302 |
> |
|
| 303 |
> |
self.cfg_params['USER.dataset_to_check']=self.dataset_to_check |
| 304 |
> |
from InspectDBS import InspectDBS |
| 305 |
> |
check=InspectDBS(self.cfg_params) |
| 306 |
> |
check.checkPublication() |
| 307 |
|
return self.exit_status |
| 308 |
|
|
| 309 |
|
else: |
| 310 |
< |
common.logger.message("--->>> "+self.resDir+" empty: no file to publish on DBS") |
| 310 |
> |
common.logger.info("--->>> "+self.resDir+" empty: no file to publish on DBS") |
| 311 |
|
self.exit_status = '1' |
| 312 |
|
return self.exit_status |
| 313 |
|
|
