| 5 |
|
from crab_util import * |
| 6 |
|
from crab_logger import Logger |
| 7 |
|
from crab_exceptions import * |
| 8 |
< |
from FwkJobRep.ReportParser import readJobReport |
| 8 |
> |
from ProdCommon.FwkJobRep.ReportParser import readJobReport |
| 9 |
> |
from ProdCommon.FwkJobRep.ReportState import checkSuccess |
| 10 |
|
from ProdCommon.MCPayloads.WorkflowSpec import WorkflowSpec |
| 11 |
|
from ProdCommon.DataMgmt.DBS.DBSWriter import DBSWriter |
| 12 |
|
from ProdCommon.DataMgmt.DBS.DBSErrors import DBSWriterError, formatEx,DBSReaderError |
| 25 |
|
""" |
| 26 |
|
|
| 27 |
|
try: |
| 28 |
< |
self.processedData = cfg_params['USER.publish_data_name'] |
| 28 |
> |
self.userprocessedData = cfg_params['USER.publish_data_name'] |
| 29 |
> |
self.processedData = None |
| 30 |
|
except KeyError: |
| 31 |
|
raise CrabException('Cannot publish output data, because you did not specify USER.publish_data_name parameter in the crab.cfg file') |
| 32 |
+ |
|
| 33 |
|
try: |
| 34 |
|
if (int(cfg_params['USER.copy_data']) != 1): raise KeyError |
| 35 |
|
except KeyError: |
| 44 |
|
self.globalDBS="http://cmsdbsprod.cern.ch/cms_dbs_prod_global/servlet/DBSServlet" |
| 45 |
|
try: |
| 46 |
|
self.DBSURL=cfg_params['USER.dbs_url_for_publication'] |
| 47 |
< |
common.logger.message('dbs url = '+self.DBSURL) |
| 48 |
< |
if (self.DBSURL == "http://cmsdbsprod.cern.ch/cms_dbs_prod_global/servlet/DBSServlet"): |
| 47 |
> |
common.logger.message('<dbs_url_for_publication> = '+self.DBSURL) |
| 48 |
> |
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"): |
| 49 |
|
msg = "You can not publish your data in the globalDBS = " + self.DBSURL + "\n" |
| 50 |
|
msg = msg + "Please write your local one in the [USER] section 'dbs_url_for_publication'" |
| 51 |
|
raise CrabException(msg) |
| 52 |
|
except KeyError: |
| 53 |
< |
msg = "Error. The [USER] section does not have 'dbs_url_for_publication'" |
| 54 |
< |
msg = msg + " entry, necessary to publish the data" |
| 53 |
> |
msg = "Warning. The [USER] section does not have 'dbs_url_for_publication'" |
| 54 |
> |
msg = msg + " entry, necessary to publish the data.\n" |
| 55 |
> |
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." |
| 56 |
|
raise CrabException(msg) |
| 57 |
|
|
| 58 |
|
self.content=file(self.pset).read() |
| 59 |
|
self.resDir = common.work_space.resDir() |
| 60 |
+ |
|
| 61 |
+ |
self.dataset_to_import=[] |
| 62 |
+ |
|
| 63 |
|
self.datasetpath=cfg_params['CMSSW.datasetpath'] |
| 64 |
+ |
if (self.datasetpath.upper() != 'NONE'): |
| 65 |
+ |
self.dataset_to_import.append(self.datasetpath) |
| 66 |
+ |
|
| 67 |
+ |
### Added PU dataset |
| 68 |
+ |
tmp = cfg_params.get('CMSSW.dataset_pu',None) |
| 69 |
+ |
if tmp : |
| 70 |
+ |
datasets = tmp.split(',') |
| 71 |
+ |
for dataset in datasets: |
| 72 |
+ |
dataset=string.strip(dataset) |
| 73 |
+ |
self.dataset_to_import.append(dataset) |
| 74 |
+ |
### |
| 75 |
+ |
|
| 76 |
|
self.SEName='' |
| 77 |
|
self.CMSSW_VERSION='' |
| 78 |
|
self.exit_status='' |
| 87 |
|
dbsWriter = DBSWriter(self.DBSURL,level='ERROR') |
| 88 |
|
|
| 89 |
|
try: |
| 90 |
< |
dbsWriter.importDataset(globalDBS, self.datasetpath, self.DBSURL) |
| 90 |
> |
dbsWriter.importDatasetWithoutParentage(globalDBS, self.datasetpath, self.DBSURL) |
| 91 |
|
except DBSWriterError, ex: |
| 92 |
|
msg = "Error importing dataset to be processed into local DBS\n" |
| 93 |
|
msg += "Source Dataset: %s\n" % datasetpath |
| 108 |
|
msg = "Error: Problem with "+file+" file" |
| 109 |
|
common.logger.message(msg) |
| 110 |
|
return self.exit_status |
| 111 |
< |
|
| 112 |
< |
if (self.datasetpath != 'None'): |
| 113 |
< |
common.logger.message("--->>> Importing parent dataset in the dbs") |
| 114 |
< |
status_import=self.importParentDataset(self.globalDBS, self.datasetpath) |
| 115 |
< |
if (status_import == 1): |
| 116 |
< |
common.logger.message('Problem with parent import from the global DBS '+self.globalDBS+ 'to the local one '+self.DBSURL) |
| 117 |
< |
self.exit_status='1' |
| 118 |
< |
return self.exit_status |
| 119 |
< |
common.logger.message("Parent import ok") |
| 111 |
> |
|
| 112 |
> |
if (len(self.dataset_to_import) != 0): |
| 113 |
> |
for dataset in self.dataset_to_import: |
| 114 |
> |
common.logger.message("--->>> Importing parent dataset in the dbs: " +dataset) |
| 115 |
> |
status_import=self.importParentDataset(self.globalDBS, dataset) |
| 116 |
> |
if (status_import == 1): |
| 117 |
> |
common.logger.message('Problem with parent '+ dataset +' import from the global DBS '+self.globalDBS+ 'to the local one '+self.DBSURL) |
| 118 |
> |
self.exit_status='1' |
| 119 |
> |
return self.exit_status |
| 120 |
> |
else: |
| 121 |
> |
common.logger.message('Import ok of dataset '+dataset) |
| 122 |
|
|
| 123 |
|
#// DBS to contact |
| 124 |
|
dbswriter = DBSWriter(self.DBSURL) |
| 134 |
|
datasets=fileinfo.dataset |
| 135 |
|
common.logger.debug(6,"FileInfo = " + str(fileinfo)) |
| 136 |
|
common.logger.debug(6,"DatasetInfo = " + str(datasets)) |
| 137 |
+ |
if len(datasets)<=0: |
| 138 |
+ |
self.exit_status = '1' |
| 139 |
+ |
msg = "Error: No info about dataset in the xml file "+file |
| 140 |
+ |
common.logger.message(msg) |
| 141 |
+ |
return self.exit_status |
| 142 |
|
for dataset in datasets: |
| 143 |
+ |
#### for production data |
| 144 |
+ |
self.processedData = dataset['ProcessedDataset'] |
| 145 |
+ |
if (dataset['PrimaryDataset'] == 'null'): |
| 146 |
+ |
#dataset['PrimaryDataset'] = dataset['ProcessedDataset'] |
| 147 |
+ |
dataset['PrimaryDataset'] = self.userprocessedData |
| 148 |
+ |
#else: # add parentage from input dataset |
| 149 |
+ |
elif self.datasetpath.upper() != 'NONE': |
| 150 |
+ |
dataset['ParentDataset']= self.datasetpath |
| 151 |
+ |
|
| 152 |
|
dataset['PSetContent']=self.content |
| 118 |
– |
#cfgMeta = {'name' : 'usercfg' , 'Type' : 'user' , 'annotation': 'user cfg', 'version' : 'private version'} # add real name of user cfg |
| 153 |
|
cfgMeta = {'name' : self.pset , 'Type' : 'user' , 'annotation': 'user cfg', 'version' : 'private version'} # add real name of user cfg |
| 154 |
|
common.logger.message("PrimaryDataset = %s"%dataset['PrimaryDataset']) |
| 155 |
|
common.logger.message("ProcessedDataset = %s"%dataset['ProcessedDataset']) |
| 156 |
< |
common.logger.message("--->>> Inserting primary: %s processed : %s"%(dataset['PrimaryDataset'],dataset['ProcessedDataset'])) |
| 156 |
> |
common.logger.message("<User Dataset Name> = /"+dataset['PrimaryDataset']+"/"+dataset['ProcessedDataset']+"/USER") |
| 157 |
> |
|
| 158 |
> |
common.logger.debug(6,"--->>> Inserting primary: %s processed : %s"%(dataset['PrimaryDataset'],dataset['ProcessedDataset'])) |
| 159 |
|
|
| 160 |
|
primary = DBSWriterObjects.createPrimaryDataset( dataset, dbswriter.dbs) |
| 161 |
|
common.logger.debug(6,"Primary: %s "%primary) |
| 166 |
|
processed = DBSWriterObjects.createProcessedDataset(primary, algo, dataset, dbswriter.dbs) |
| 167 |
|
common.logger.debug(6,"Processed: %s "%processed) |
| 168 |
|
|
| 169 |
< |
common.logger.message("Inserted primary %s processed %s"%(primary,processed)) |
| 169 |
> |
common.logger.debug(6,"Inserted primary %s processed %s"%(primary,processed)) |
| 170 |
|
|
| 171 |
|
common.logger.debug(6,"exit_status = %s "%self.exit_status) |
| 172 |
|
return self.exit_status |
| 194 |
|
self.noLFN.append(file['PFN']) |
| 195 |
|
else: |
| 196 |
|
if int(file['TotalEvents']) != 0 : |
| 197 |
< |
file.lumisections = {} |
| 197 |
> |
#file.lumisections = {} |
| 198 |
> |
# lumi info are now in run hash |
| 199 |
> |
file.runs = {} |
| 200 |
|
for ds in file.dataset: |
| 201 |
< |
ds['ProcessedDataset']=procdataset |
| 201 |
> |
### Fede for production |
| 202 |
> |
if (ds['PrimaryDataset'] == 'null'): |
| 203 |
> |
#ds['PrimaryDataset']=procdataset |
| 204 |
> |
ds['PrimaryDataset']=self.userprocessedData |
| 205 |
|
filestopublish.append(file) |
| 206 |
|
else: |
| 207 |
|
self.noEventsFiles.append(file['LFN']) |
| 216 |
|
Blocks=None |
| 217 |
|
try: |
| 218 |
|
Blocks=dbswriter.insertFiles(jobReport) |
| 219 |
< |
common.logger.message("Blocks = %s"%Blocks) |
| 219 |
> |
common.logger.message("Inserting file in blocks = %s"%Blocks) |
| 220 |
|
except DBSWriterError, ex: |
| 221 |
< |
common.logger.message("Insert file error: %s"%ex) |
| 221 |
> |
common.logger.error("Insert file error: %s"%ex) |
| 222 |
|
return Blocks |
| 223 |
|
|
| 224 |
|
def run(self): |
| 227 |
|
""" |
| 228 |
|
|
| 229 |
|
file_list = glob.glob(self.resDir+"crab_fjr*.xml") |
| 230 |
+ |
## Select only those fjr that are succesfull |
| 231 |
+ |
good_list=[] |
| 232 |
+ |
for fjr in file_list: |
| 233 |
+ |
reports = readJobReport(fjr) |
| 234 |
+ |
if len(reports)>0: |
| 235 |
+ |
if reports[0].status == "Success": |
| 236 |
+ |
good_list.append(fjr) |
| 237 |
+ |
file_list=good_list |
| 238 |
+ |
## |
| 239 |
|
common.logger.debug(6, "file_list = "+str(file_list)) |
| 240 |
|
common.logger.debug(6, "len(file_list) = "+str(len(file_list))) |
| 241 |
|
|
| 250 |
|
|
| 251 |
|
common.logger.message("--->>> Start files publication") |
| 252 |
|
for file in file_list: |
| 253 |
< |
common.logger.message("file = "+file) |
| 253 |
> |
common.logger.debug(1, "file = "+file) |
| 254 |
|
Blocks=self.publishAJobReport(file,self.processedData) |
| 255 |
|
if Blocks: |
| 256 |
< |
[BlocksList.append(x) for x in Blocks] |
| 256 |
> |
for x in Blocks: # do not allow multiple entries of the same block |
| 257 |
> |
if x not in BlocksList: |
| 258 |
> |
BlocksList.append(x) |
| 259 |
|
|
| 260 |
|
# close the blocks |
| 261 |
< |
common.logger.message("BlocksList = %s"%BlocksList) |
| 261 |
> |
common.logger.debug(6, "BlocksList = %s"%BlocksList) |
| 262 |
|
# dbswriter = DBSWriter(self.DBSURL,level='ERROR') |
| 263 |
|
dbswriter = DBSWriter(self.DBSURL) |
| 264 |
|
|
| 265 |
|
for BlockName in BlocksList: |
| 266 |
|
try: |
| 267 |
|
closeBlock=dbswriter.manageFileBlock(BlockName,maxFiles= 1) |
| 268 |
< |
common.logger.message("closeBlock %s"%closeBlock) |
| 268 |
> |
common.logger.debug(6, "closeBlock %s"%closeBlock) |
| 269 |
|
#dbswriter.dbs.closeBlock(BlockName) |
| 270 |
|
except DBSWriterError, ex: |
| 271 |
|
common.logger.message("Close block error %s"%ex) |
| 272 |
|
|
| 221 |
– |
common.logger.message("--->>> End files publication") |
| 273 |
|
if (len(self.noEventsFiles)>0): |
| 274 |
|
common.logger.message("--->>> WARNING: "+str(len(self.noEventsFiles))+" files not published because they contain 0 events are:") |
| 275 |
|
for lfn in self.noEventsFiles: |
| 282 |
|
common.logger.message("--->>> WARNING: "+str(len(self.problemFiles))+" files not published because they had problem with copy to SE") |
| 283 |
|
for lfn in self.problemFiles: |
| 284 |
|
common.logger.message("------ LFN: %s"%lfn) |
| 285 |
+ |
common.logger.message("--->>> End files publication") |
| 286 |
+ |
common.logger.message("--->>> To check data publication please use: InspectDBS2.py --DBSURL=<dbs_url_for_publication> --datasetPath=<User Dataset Name>") |
| 287 |
|
return self.exit_status |
| 288 |
|
|
| 289 |
|
else: |
