def createAlgorithmForInsert(datasetInfo):
"""
_createPartialAlgorithm_
Create an Algorithm instance that uses the minimal info needed
to insert a file
"""
exeName = datasetInfo['ApplicationName']
appVersion = datasetInfo['ApplicationVersion']
appFamily = datasetInfo["ApplicationFamily"]
#
# Repacker jobs have no PsetContent/PSetHash
#
psetContent = datasetInfo.get('PSetContent', None)
if psetContent == None:
psetContent = "PSET_CONTENT_NOT_AVAILABLE"
psetHash = datasetInfo.get('PSetHash', None)
if psetHash == None:
psetHash = "NO_PSET_HASH"
else:
if psetHash.find(";"):
# no need for fake hash in new schema
psetHash = psetHash.split(";")[0]
psetHash = psetHash.replace("hash=", "")
psetInstance = DbsQueryableParameterSet(Hash=psetHash)
algorithmInstance = DbsAlgorithm(ExecutableName=exeName,
ApplicationVersion=appVersion,
ApplicationFamily=appFamily,
ParameterSetID=psetInstance)
return algorithmInstance
def createAlgorithm(apiRef,
appName,
appVer,
appFam,
PSetHash=None,
PSetContent=None):
"""
_createAlgorithm_
Create a new DBS Algorithm, explicitly passing in the arguments.
We don't use configs.
The insert tag tell you whether or not to actually write this to DBS
"""
# Take care of PSetHash
if not PSetHash:
PSetHash = "NO_PSET_HASH2"
elif PSetHash.find(";"):
# no need for fake hash in new schema
PSetHash = PSetHash.split(";")[0]
PSetHash = PSetHash.replace("hash=", "")
# Create PSetHash. dbsApi tries to base64encode the value of PSetContent
# which blows up if it's None
if not PSetContent:
PSetContent = ""
psetInstance = DbsQueryableParameterSet(Hash=PSetHash, Content=PSetContent)
algorithmInstance = DbsAlgorithm(ExecutableName=appName,
ApplicationVersion=appVer,
ApplicationFamily=appFam,
ParameterSetID=psetInstance)
if apiRef:
try:
apiRef.insertAlgorithm(algorithmInstance)
except DbsException as ex:
msg = "Error in DBSInterface.createAlgorithm(%s)\n" % appVer
msg += formatEx(ex)
logging.error(msg)
raise DBSInterfaceError(msg)
return algorithmInstance
def createMergeAlgorithm(datasetInfo, apiRef=None):
"""
_createMergeAlgorithm_
Create a DbsAlgorithm for a merge dataset
"""
exeName = datasetInfo['ApplicationName']
version = datasetInfo['ApplicationVersion']
family = datasetInfo.get('ApplicationFamily', None)
if (family == None) or not (family):
family = datasetInfo['OutputModuleName']
mergeAlgo = DbsAlgorithm(
ExecutableName=exeName,
ApplicationVersion=version,
ApplicationFamily=family,
)
if apiRef != None:
apiRef.insertAlgorithm(mergeAlgo)
return mergeAlgo
from DBSAPI.dbsFileBlock import DbsFileBlock
from DBSAPI.dbsFile import DbsFile
from DBSAPI.dbsLumiSection import DbsLumiSection
from DBSAPI.dbsQueryableParameterSet import DbsQueryableParameterSet
from DBSAPI.dbsPrimaryDataset import DbsPrimaryDataset
from DBSAPI.dbsProcessedDataset import DbsProcessedDataset
from DBSAPI.dbsOptions import DbsOptionParser
optManager = DbsOptionParser()
(opts, args) = optManager.getOpt()
api = DbsApi(opts.__dict__)
try:
merge_algo = DbsAlgorithm(
ExecutableName="EdmFastMerge",
ApplicationVersion="v101",
ApplicationFamily="Merge",
)
path = "/test_primary_001/TestProcessedDS001/SIM"
merge_proc = api.insertMergedDataset(path, "ThisISMergedDataset001",
merge_algo)
# File will go into THIS Block
block = DbsFileBlock(StorageElement=['test1', 'test3'],
Name="/test_primary_001/TestProcessedDS001/SIM#12345")
merged_file = DbsFile(
Checksum='00000',
LogicalFileName='MERGEDFILE_001',
NumberOfEvents=10000,
FileSize=1000000,
pri = DbsPrimaryDataset (Name = primary, Type="test")
apiObj.run(pri,"", excep = True)
f.write("\n***********************insertPrimaryDataset API tests***************************")
apiObj = DbsUnitTestApi(api.insertAlgorithm,f)
apiObj.setVerboseLevel(opts.verbose)
f.write("\n\n***********************insertAlgorithm API tests***************************")
algo1 = DbsAlgorithm (ExecutableName="TestExe01",
ApplicationVersion= "TestVersion01" + mytime,
ApplicationFamily="AppFamily01",
ParameterSetID=DbsQueryableParameterSet(Hash="001234565798685",
Name="MyFirstParam01",
Version="V001",
Type="test",
Annotation="This is test",
Content="int a= {}, b={c=1, d=33}, f={}, x, y, x"
)
)
apiObj.run(algo1, excep = False)
algo2 = DbsAlgorithm (ExecutableName="TestExe011",
ApplicationVersion= "TestVersion011" + mytime,
ApplicationFamily="AppFamily011",
ParameterSetID=DbsQueryableParameterSet(Hash="001234565798685",
Name="MyFirstParam01",
Annotation="This is test",
Content="int a= {}, b={c=1, d=33}, f={}, x, y, x"
mytime += str(i)
fileList = []
#Insert Primary
apiObj = DbsUnitTestApi(api.insertPrimaryDataset, f)
primary = 'StressTestPrimary' + mytime
pri1 = DbsPrimaryDataset (Name = primary, Type='MC')
apiObj.run(pri1, excep = False)
#Insert Algorithm
apiObj = DbsUnitTestApi(api.insertAlgorithm,f)
algo1 = DbsAlgorithm (ExecutableName="StressTestExe01",
ApplicationVersion= "StressTestVersion01" + mytime,
ApplicationFamily="StressTestAppFamily01",
ParameterSetID=DbsQueryableParameterSet(Hash="001234565798685",
Name="StressTestParam01",
Version="V001",
Type="test",
Annotation="This is a stress test param",
Content="int a= {}, b={c=1, d=33}, f={}, x, y, x"
)
)
apiObj.run(algo1, excep = False)
#Insert Tier
apiObj = DbsUnitTestApi(api.insertTier, f)
tierName1 = "GEN"
tierName2 = "SIM"
apiObj.run(tierName1, excep = False)
apiObj.run(tierName2, excep = False)
tierList = [tierName1, tierName2]
def createAlgorithm(datasetInfo, configMetadata=None, apiRef=None):
"""
_createAlgorithm_
Create an algorithm assuming that datasetInfo is a
ProdCommon.MCPayloads.DatasetInfo like dictionary
"""
exeName = datasetInfo['ApplicationName']
appVersion = datasetInfo['ApplicationVersion']
appFamily = datasetInfo["ApplicationFamily"]
#
# HACK: Problem with large PSets (is this still relevant ?)
#
# Repacker jobs have no PSetContent/PSetHash
#
psetContent = datasetInfo.get('PSetContent', None)
if psetContent == None:
psetContent = "PSET_CONTENT_NOT_AVAILABLE"
psetHash = datasetInfo.get('PSetHash', None)
if psetHash == None:
psetHash = "NO_PSET_HASH"
else:
if psetHash.find(";"):
# no need for fake hash in new schema
psetHash = psetHash.split(";")[0]
psetHash = psetHash.replace("hash=", "")
## No more hacks
#msg = ">>>>>>>>>>>>>>>>>>>>>>>>>>>>\n"
#msg += "TEST HACK USED FOR PSetContent\n"
#msg += ">>>>>>>>>>>>>>>>>>>>>>>>>>>>"
#logging.warning(msg)
#print msg
#psetContent = "This is not a PSet"
#
# HACK: 100 char limit on cfg file name
if configMetadata != None:
cfgName = configMetadata['name']
if len(cfgName) > 100:
msg = ">>>>>>>>>>>>>>>>>>>>>>>>>>>>\n"
msg += "TEST HACK USED FOR Config File Name"
msg += ">>>>>>>>>>>>>>>>>>>>>>>>>>>>"
logging.warning(msg)
print msg
configMetadata['name'] = cfgName[-99]
psetInstance = DbsQueryableParameterSet(
Hash=psetHash,
Name=configMetadata['name'],
Version=configMetadata['version'],
Type=configMetadata['Type'],
Annotation=configMetadata['annotation'],
Content=psetContent,
)
algorithmInstance = DbsAlgorithm(ExecutableName=exeName,
ApplicationVersion=appVersion,
ApplicationFamily=appFamily,
ParameterSetID=psetInstance)
else:
psetInstance = DbsQueryableParameterSet(Hash=psetHash)
algorithmInstance = DbsAlgorithm(ExecutableName=exeName,
ApplicationVersion=appVersion,
ApplicationFamily=appFamily,
ParameterSetID=psetInstance)
if apiRef != None:
apiRef.insertAlgorithm(algorithmInstance)
return algorithmInstance
from DBSAPI.dbsPrimaryDataset import DbsPrimaryDataset
from DBSAPI.dbsProcessedDataset import DbsProcessedDataset
from DBSAPI.dbsOptions import DbsOptionParser
optManager = DbsOptionParser()
(opts,args) = optManager.getOpt()
api = DbsApi(opts.__dict__)
mytime = time.strftime("_%Y%m%d_%Hh%Mm%Ss",time.localtime())
primary = DbsPrimaryDataset (Name = "test_primary_001" + mytime, Type="test")
algo = DbsAlgorithm (
ExecutableName="TestExe01" + mytime,
ApplicationVersion= "TestVersion01" + mytime,
ApplicationFamily="AppFamily01" + mytime,
ParameterSetID=DbsQueryableParameterSet(
Hash="001234565798685",
)
)
proc = DbsProcessedDataset (
PrimaryDataset=primary,
Name="TestProcessedDS001" + mytime,
PhysicsGroup="BPositive",
Status="Valid",
TierList=['SIM', 'GEN'],
AlgoList=[algo],
)
procChild = DbsProcessedDataset (
PrimaryDataset=primary,
newBlockName,
storage_element_list=[seName])
blockRef = dbsApi.listBlocks(dataset=datasetPath,
block_name=newBlockName)[0]
print blockRef
newFiles = []
for newFileLFN in badFiles[newBlockName]:
localFile = DBSBufferFile(lfn=newFileLFN)
localFile.load(parentage=1)
(primaryDS, procDS, tier) = datasetPath[1:].split("/", 3)
primary = DbsPrimaryDataset(Name=primaryDS, Type="mc")
algo = DbsAlgorithm(ExecutableName=localFile["appName"],
ApplicationVersion=localFile["appVer"],
ApplicationFamily=localFile["appFam"],
ParameterSetID=psetInstance)
processed = DbsProcessedDataset(PrimaryDataset=primary,
AlgoList=[algo],
Name=procDS,
TierList=[tier],
ParentList=[],
PhysicsGroup="NoGroup",
Status="VALID",
GlobalTag="")
newFiles.append(
DBSInterface.createDBSFileFromBufferFile(localFile, processed))
dbsApi.insertFiles(datasetPath, newFiles, blockRef)
dbsApi.closeBlock(block=newBlockName)
optManager = DbsOptionParser()
(opts, args) = optManager.getOpt()
api = DbsApi(opts.__dict__)
import pdb
##Primary Dataset
primary = DbsPrimaryDataset(Name="test_primary_001", Type="TEST")
api.insertPrimaryDataset(primary)
#Algorithm used by Parent and Child Datasets for our test
algo = DbsAlgorithm(ExecutableName="TestExe01",
ApplicationVersion="TestVersion01",
ApplicationFamily="AppFamily01",
ParameterSetID=DbsQueryableParameterSet(
Hash="001234565798685",
Name="MyFirstParam01",
Version="V001",
Type="test",
Annotation="This is test",
Content="int a= {}, b={c=1, d=33}, f={}, x, y, x"))
api.insertAlgorithm(algo)
# Parent Dataset
parent_procds = DbsProcessedDataset(
PrimaryDataset=primary,
Name="TestProcessedDS001-Parent",
PhysicsGroup="BPositive",
Status="VALID",
TierList=['GEN', 'SIM'],
AlgoList=[algo],
fileSizeG = 563226500
fileStatusG = validStatus
fileValidStatusG = validStatus
fileTypeG = 'STREAMER'
qim_name1 = "Tracker_Global"
qim_name2 = "TIB_Local"
qim_name3 = "TIB_Power"
qim_int = "TIB_Percentage"
primObj = DbsPrimaryDataset(Name=primName, Type=primType)
algoObj1 = DbsAlgorithm(ExecutableName=algoExe1,
ApplicationVersion=algoVer1,
ApplicationFamily=algoFam1,
ParameterSetID=DbsQueryableParameterSet(
Hash=psHash1,
Name=psName1,
Version=psVer1,
Type=psType1,
Annotation=psAnno1,
Content=psCon1))
algoObj2 = DbsAlgorithm(ExecutableName=algoExe2,
ApplicationVersion=algoVer2,
ApplicationFamily=algoFam2,
ParameterSetID=DbsQueryableParameterSet(
Hash=psHash2,
Name=psName2,
Version=psVer2,
Type=psType2,
Annotation=psAnno2,
Content=psCon2))
