def getMaxAtomSize(self, effectiveSize=False, getNumEvents=False):
# number of files per job if defined
if not self.isMerging:
nFilesPerJob = self.taskSpec.getNumFilesPerJob()
else:
nFilesPerJob = self.taskSpec.getNumFilesPerMergeJob()
nEventsPerJob = None
if nFilesPerJob == None:
# number of events per job
if not self.isMerging:
nEventsPerJob = self.taskSpec.getNumEventsPerJob()
else:
nEventsPerJob = self.taskSpec.getNumEventsPerMergeJob()
if nEventsPerJob == None:
nFilesPerJob = 1
# grouping with boundaryID
useBoundary = self.taskSpec.useGroupWithBoundaryID()
# LB
respectLB = self.taskSpec.respectLumiblock()
maxAtomSize = 0
while True:
if not self.isMerging:
maxNumFiles = self.taskSpec.getMaxNumFilesPerJob()
else:
maxNumFiles = self.taskSpec.getMaxNumFilesPerMergeJob()
# get one subchunk
subChunk = self.getSubChunk(None,
nFilesPerJob=nFilesPerJob,
nEventsPerJob=nEventsPerJob,
useBoundary=useBoundary,
respectLB=respectLB,
maxNumFiles=maxNumFiles)
if subChunk == None:
break
# get size
tmpAtomSize = 0
for tmpDatasetSpec, tmpFileSpecList in subChunk:
if (effectiveSize
or getNumEvents) and not tmpDatasetSpec.isMaster():
continue
for tmpFileSpec in tmpFileSpecList:
if effectiveSize:
tmpAtomSize += JediCoreUtils.getEffectiveFileSize(
tmpFileSpec.fsize, tmpFileSpec.startEvent,
tmpFileSpec.endEvent, tmpFileSpec.nEvents)
elif getNumEvents:
tmpAtomSize += tmpFileSpec.getEffectiveNumEvents()
else:
tmpAtomSize += tmpFileSpec.fsize
if maxAtomSize < tmpAtomSize:
maxAtomSize = tmpAtomSize
# reset counters
self.resetUsedCounters()
# return
return maxAtomSize
def getMaxAtomSize(self,effectiveSize=False,getNumEvents=False):
# number of files per job if defined
if not self.isMerging:
nFilesPerJob = self.taskSpec.getNumFilesPerJob()
else:
nFilesPerJob = self.taskSpec.getNumFilesPerMergeJob()
nEventsPerJob = None
if nFilesPerJob == None:
# number of events per job
if not self.isMerging:
nEventsPerJob = self.taskSpec.getNumEventsPerJob()
else:
nEventsPerJob = self.taskSpec.getNumEventsPerMergeJob()
if nEventsPerJob == None:
nFilesPerJob = 1
# grouping with boundaryID
useBoundary = self.taskSpec.useGroupWithBoundaryID()
# LB
respectLB = self.taskSpec.respectLumiblock()
maxAtomSize = 0
while True:
if not self.isMerging:
maxNumFiles = self.taskSpec.getMaxNumFilesPerJob()
else:
maxNumFiles = self.taskSpec.getMaxNumFilesPerMergeJob()
# get one subchunk
subChunk = self.getSubChunk(None,nFilesPerJob=nFilesPerJob,
nEventsPerJob=nEventsPerJob,
useBoundary=useBoundary,
respectLB=respectLB,
maxNumFiles=maxNumFiles)
if subChunk == None:
break
# get size
tmpAtomSize = 0
for tmpDatasetSpec,tmpFileSpecList in subChunk:
if (effectiveSize or getNumEvents) and not tmpDatasetSpec.isMaster():
continue
for tmpFileSpec in tmpFileSpecList:
if effectiveSize:
tmpAtomSize += JediCoreUtils.getEffectiveFileSize(tmpFileSpec.fsize,tmpFileSpec.startEvent,
tmpFileSpec.endEvent,tmpFileSpec.nEvents)
elif getNumEvents:
tmpAtomSize += tmpFileSpec.getEffectiveNumEvents()
else:
tmpAtomSize += tmpFileSpec.fsize
if maxAtomSize < tmpAtomSize:
maxAtomSize = tmpAtomSize
# reset counters
self.resetUsedCounters()
# return
return maxAtomSize
def wrappedMain(self):
while True:
proc = multiprocessing.Process(target=self.target, args=self.args)
proc.start()
pid = proc.pid
while True:
try:
proc.join(20)
if not JediCoreUtils.checkProcess(pid):
timeNow = datetime.datetime.utcnow()
print "{0} {1}: INFO pid={2} not exist".format(
str(timeNow), self.__class__.__name__, pid)
break
except:
timeNow = datetime.datetime.utcnow()
errType, errValue = sys.exc_info()[:2]
print "{0} {1}: INFO failed to check pid={2} with {3} {4}".format(
str(timeNow), self.__class__.__name__, pid, errType,
errValue)
def wrappedMain(self):
while True:
proc = multiprocessing.Process(target=self.target,
args=self.args)
proc.start()
pid = proc.pid
while True:
try:
proc.join(20)
if not JediCoreUtils.checkProcess(pid):
timeNow = datetime.datetime.utcnow()
print "{0} {1}: INFO pid={2} not exist".format(str(timeNow),
self.__class__.__name__,
pid)
break
except:
timeNow = datetime.datetime.utcnow()
errType,errValue = sys.exc_info()[:2]
print "{0} {1}: INFO failed to check pid={2} with {3} {4}".format(str(timeNow),
self.__class__.__name__,
pid,errType,errValue)
def getSubChunk(self,
siteName,
maxNumFiles=None,
maxSize=None,
sizeGradients=0,
sizeIntercepts=0,
nFilesPerJob=None,
multiplicand=1,
walltimeGradient=0,
maxWalltime=0,
nEventsPerJob=None,
useBoundary=None,
sizeGradientsPerInSize=None,
maxOutSize=None,
coreCount=1,
respectLB=False,
corePower=None,
dynNumEvents=False,
maxNumEventRanges=None,
multiplicity=None,
splitByFields=None,
tmpLog=None,
useDirectIO=False):
# check if there are unused files/events
if not self.checkUnused():
return None
# protection against unreasonable values
if nFilesPerJob == 0:
nFilesPerJob = None
if nEventsPerJob == 0:
nEventsPerJob = None
# set default max number of files
if maxNumFiles == None:
maxNumFiles = 50
# set default max number of event ranges
if maxNumEventRanges == None:
maxNumEventRanges = 20
# set default max size
if maxSize == None and nFilesPerJob == None and nEventsPerJob == None:
# 20 GB at most by default
maxSize = 20 * 1024 * 1024 * 1024
# set default output size
minOutSize = self.defaultOutputSize
# set default max number of events
maxNumEvents = None
# ignore negative walltime gradient
if walltimeGradient < 0:
walltimeGradient = 0
# overwrite parameters when nFiles/EventsPerJob is used
if nFilesPerJob != None and not dynNumEvents:
maxNumFiles = nFilesPerJob
if not respectLB:
multiplicand = nFilesPerJob
if nEventsPerJob != None:
maxNumEvents = nEventsPerJob
# split with boundayID
splitWithBoundaryID = False
if useBoundary != None:
splitWithBoundaryID = True
if useBoundary['inSplit'] == 2:
# unset max values to split only with boundaryID
maxNumFiles = None
maxSize = None
maxWalltime = 0
maxNumEvents = None
multiplicand = 1
# get site when splitting per site
if siteName != None:
siteCandidate = self.siteCandidates[siteName]
# use event ratios
useEventRatio = self.useEventRatioForSec()
# start splitting
inputNumFiles = 0
inputNumEvents = 0
fileSize = 0
firstLoop = True
firstMaster = True
inputFileMap = {}
expWalltime = 0
nextStartEvent = None
boundaryID = None
newBoundaryID = False
eventJump = False
nSecFilesMap = {}
nSecEventsMap = {}
numMaster = 0
outSizeMap = {}
lumiBlockNr = None
newLumiBlockNr = False
siteAvailable = True
inputFileSet = set()
fieldStr = None
while (maxNumFiles == None or (not dynNumEvents and inputNumFiles <= maxNumFiles) or \
(dynNumEvents and len(inputFileSet) <= maxNumFiles and inputNumFiles <= maxNumEventRanges)) \
and (maxSize == None or (maxSize != None and fileSize <= maxSize)) \
and (maxWalltime <= 0 or expWalltime <= maxWalltime) \
and (maxNumEvents == None or (maxNumEvents != None and inputNumEvents <= maxNumEvents)) \
and (maxOutSize == None or self.getOutSize(outSizeMap) <= maxOutSize):
# get one file (or one file group for MP) from master
datasetUsage = self.datasetMap[self.masterDataset.datasetID]
if not self.masterDataset.datasetID in outSizeMap:
outSizeMap[self.masterDataset.datasetID] = 0
boundaryIDs = set()
primaryHasEvents = False
for tmpFileSpec in self.masterDataset.Files[
datasetUsage['used']:datasetUsage['used'] + multiplicand]:
# check start event to keep continuity
if (maxNumEvents != None
or dynNumEvents) and tmpFileSpec.startEvent != None:
if nextStartEvent != None and nextStartEvent != tmpFileSpec.startEvent:
eventJump = True
break
# check boundaryID
if splitWithBoundaryID and boundaryID != None and boundaryID != tmpFileSpec.boundaryID \
and useBoundary['inSplit'] != 3:
newBoundaryID = True
break
# check LB
if respectLB and lumiBlockNr != None and lumiBlockNr != tmpFileSpec.lumiBlockNr:
newLumiBlockNr = True
break
# check field
if splitByFields != None:
tmpFieldStr = tmpFileSpec.extractFieldsStr(splitByFields)
if fieldStr == None:
fieldStr = tmpFieldStr
elif tmpFieldStr != fieldStr:
newBoundaryID = True
break
# check for distributed datasets
if self.masterDataset.isDistributed() and siteName != None and \
not siteCandidate.isAvailableFile(tmpFileSpec):
siteAvailable = False
break
if not inputFileMap.has_key(self.masterDataset.datasetID):
inputFileMap[self.masterDataset.datasetID] = []
inputFileMap[self.masterDataset.datasetID].append(tmpFileSpec)
inputFileSet.add(tmpFileSpec.lfn)
datasetUsage['used'] += 1
numMaster += 1
# get effective file size
effectiveFsize = JediCoreUtils.getEffectiveFileSize(
tmpFileSpec.fsize, tmpFileSpec.startEvent,
tmpFileSpec.endEvent, tmpFileSpec.nEvents)
# get num of events
effectiveNumEvents = tmpFileSpec.getEffectiveNumEvents()
# sum
inputNumFiles += 1
if self.taskSpec.outputScaleWithEvents():
fileSize += long(sizeGradients * effectiveNumEvents)
if not dynNumEvents or tmpFileSpec.lfn not in inputFileSet:
if not useDirectIO:
fileSize += long(tmpFileSpec.fsize)
outSizeMap[self.masterDataset.datasetID] += long(
sizeGradients * effectiveNumEvents)
else:
fileSize += long(sizeGradients * effectiveFsize)
if not dynNumEvents or tmpFileSpec.lfn not in inputFileSet:
if not useDirectIO:
fileSize += long(tmpFileSpec.fsize)
outSizeMap[self.masterDataset.datasetID] += long(
sizeGradients * effectiveFsize)
if sizeGradientsPerInSize != None:
fileSize += long(effectiveFsize * sizeGradientsPerInSize)
outSizeMap[self.masterDataset.datasetID] += long(
effectiveFsize * sizeGradientsPerInSize)
# sum offset only for the first master
if firstMaster:
fileSize += sizeIntercepts
# walltime
if self.taskSpec.useHS06():
if firstMaster:
expWalltime += self.taskSpec.baseWalltime
tmpExpWalltime = walltimeGradient * effectiveNumEvents / float(
coreCount)
if not corePower in [None, 0]:
tmpExpWalltime /= corePower
if self.taskSpec.cpuEfficiency == 0:
tmpExpWalltime = 0
else:
tmpExpWalltime /= float(
self.taskSpec.cpuEfficiency) / 100.0
if multiplicity != None:
tmpExpWalltime /= float(multiplicity)
expWalltime += long(tmpExpWalltime)
else:
tmpExpWalltime = walltimeGradient * effectiveFsize / float(
coreCount)
if multiplicity != None:
tmpExpWalltime /= float(multiplicity)
expWalltime += long(tmpExpWalltime)
# the number of events
if (
maxNumEvents != None or useEventRatio
) and tmpFileSpec.startEvent != None and tmpFileSpec.endEvent != None:
primaryHasEvents = True
inputNumEvents += (tmpFileSpec.endEvent -
tmpFileSpec.startEvent + 1)
# set next start event
nextStartEvent = tmpFileSpec.endEvent + 1
if nextStartEvent == tmpFileSpec.nEvents:
nextStartEvent = 0
# boundaryID
if splitWithBoundaryID:
boundaryID = tmpFileSpec.boundaryID
if not boundaryID in boundaryIDs:
boundaryIDs.add(boundaryID)
# LB
if respectLB:
lumiBlockNr = tmpFileSpec.lumiBlockNr
firstMaster = False
# get files from secondaries
firstSecondary = True
for datasetSpec in self.secondaryDatasetList:
if not datasetSpec.datasetID in outSizeMap:
outSizeMap[datasetSpec.datasetID] = 0
if datasetSpec.isNoSplit():
# every job uses dataset without splitting
if firstLoop:
datasetUsage = self.datasetMap[datasetSpec.datasetID]
for tmpFileSpec in datasetSpec.Files:
if not inputFileMap.has_key(datasetSpec.datasetID):
inputFileMap[datasetSpec.datasetID] = []
inputFileMap[datasetSpec.datasetID].append(
tmpFileSpec)
# sum
if not useDirectIO:
fileSize += tmpFileSpec.fsize
if sizeGradientsPerInSize != None:
fileSize += (tmpFileSpec.fsize *
sizeGradientsPerInSize)
outSizeMap[datasetSpec.datasetID] += (
tmpFileSpec.fsize * sizeGradientsPerInSize)
datasetUsage['used'] += 1
else:
if not nSecFilesMap.has_key(datasetSpec.datasetID):
nSecFilesMap[datasetSpec.datasetID] = 0
# get number of files to be used for the secondary
nSecondary = datasetSpec.getNumFilesPerJob()
if nSecondary != None and firstLoop == False:
# read files only in the first bunch when number of files per job is specified
continue
if nSecondary == None:
nSecondary = datasetSpec.getNumMultByRatio(
numMaster) - nSecFilesMap[datasetSpec.datasetID]
if (datasetSpec.getEventRatio() != None
and inputNumEvents > 0) or (
splitWithBoundaryID
and useBoundary['inSplit'] != 3):
# set large number to get all associated secondary files
nSecondary = 10000
datasetUsage = self.datasetMap[datasetSpec.datasetID]
# reset nUsed
if datasetSpec.isReusable(
) and datasetUsage['used'] + nSecondary > len(
datasetSpec.Files):
datasetUsage['used'] = 0
for tmpFileSpec in datasetSpec.Files[
datasetUsage['used']:datasetUsage['used'] +
nSecondary]:
# check boundaryID
if (splitWithBoundaryID or (useBoundary != None and useBoundary['inSplit'] == 3 and datasetSpec.getRatioToMaster() > 1)) \
and boundaryID != None and \
not (boundaryID == tmpFileSpec.boundaryID or tmpFileSpec.boundaryID in boundaryIDs):
break
# check for distributed datasets
if datasetSpec.isDistributed() and siteName != None and \
not siteCandidate.isAvailableFile(tmpFileSpec):
break
# check ratio
if not datasetSpec.datasetID in nSecEventsMap:
nSecEventsMap[datasetSpec.datasetID] = 0
if datasetSpec.getEventRatio(
) != None and inputNumEvents > 0:
if float(
nSecEventsMap[datasetSpec.datasetID]
) / float(inputNumEvents
) >= datasetSpec.getEventRatio():
break
if not inputFileMap.has_key(datasetSpec.datasetID):
inputFileMap[datasetSpec.datasetID] = []
inputFileMap[datasetSpec.datasetID].append(tmpFileSpec)
# sum
if not useDirectIO:
fileSize += tmpFileSpec.fsize
if sizeGradientsPerInSize != None:
fileSize += (tmpFileSpec.fsize *
sizeGradientsPerInSize)
outSizeMap[datasetSpec.datasetID] += (
tmpFileSpec.fsize * sizeGradientsPerInSize)
datasetUsage['used'] += 1
nSecFilesMap[datasetSpec.datasetID] += 1
# the number of events
if firstSecondary and maxNumEvents != None and not primaryHasEvents:
if tmpFileSpec.startEvent != None and tmpFileSpec.endEvent != None:
inputNumEvents += (tmpFileSpec.endEvent -
tmpFileSpec.startEvent + 1)
elif tmpFileSpec.nEvents != None:
inputNumEvents += tmpFileSpec.nEvents
if tmpFileSpec.nEvents != None:
nSecEventsMap[
datasetSpec.datasetID] += tmpFileSpec.nEvents
# use only the first secondary
firstSecondary = False
# unset first loop flag
firstLoop = False
# check if there are unused files/evets
if not self.checkUnused():
break
# break if nFilesPerJob is used as multiplicand
if nFilesPerJob != None and not respectLB:
break
# boundayID is changed
if newBoundaryID:
break
# LB is changed
if newLumiBlockNr:
break
# event jump
if eventJump:
break
# distributed files are unavailable
if not siteAvailable:
break
primaryHasEvents = False
# check master in the next loop
datasetUsage = self.datasetMap[self.masterDataset.datasetID]
newInputNumFiles = inputNumFiles
newInputNumEvents = inputNumEvents
newFileSize = fileSize
newExpWalltime = expWalltime
newNextStartEvent = nextStartEvent
newNumMaster = numMaster
terminateFlag = False
newOutSizeMap = copy.copy(outSizeMap)
newBoundaryIDs = set()
newInputFileSet = copy.copy(inputFileSet)
if not self.masterDataset.datasetID in newOutSizeMap:
newOutSizeMap[self.masterDataset.datasetID] = 0
for tmpFileSpec in self.masterDataset.Files[
datasetUsage['used']:datasetUsage['used'] + multiplicand]:
# check continuity of event
if maxNumEvents != None and tmpFileSpec.startEvent != None and tmpFileSpec.endEvent != None:
primaryHasEvents = True
newInputNumEvents += (tmpFileSpec.endEvent -
tmpFileSpec.startEvent + 1)
# continuity of event is broken
if newNextStartEvent != None and newNextStartEvent != tmpFileSpec.startEvent:
# no files in the next loop
if newInputNumFiles == 0:
terminateFlag = True
break
newNextStartEvent = tmpFileSpec.endEvent + 1
# check boundary
if splitWithBoundaryID and boundaryID != None and boundaryID != tmpFileSpec.boundaryID \
and useBoundary['inSplit'] != 3:
# no files in the next loop
if newInputNumFiles == 0:
terminateFlag = True
break
# check LB
if respectLB and lumiBlockNr != None and lumiBlockNr != tmpFileSpec.lumiBlockNr:
# no files in the next loop
if newInputNumFiles == 0:
terminateFlag = True
break
# check field
if splitByFields != None:
tmpFieldStr = tmpFileSpec.extractFieldsStr(splitByFields)
if tmpFieldStr != fieldStr:
# no files in the next loop
if newInputNumFiles == 0:
terminateFlag = True
break
# check for distributed datasets
if self.masterDataset.isDistributed() and siteName != None and \
not siteCandidate.isAvailableFile(tmpFileSpec):
# no files in the next loop
if newInputNumFiles == 0:
terminateFlag = True
break
# get effective file size
effectiveFsize = JediCoreUtils.getEffectiveFileSize(
tmpFileSpec.fsize, tmpFileSpec.startEvent,
tmpFileSpec.endEvent, tmpFileSpec.nEvents)
# get num of events
effectiveNumEvents = tmpFileSpec.getEffectiveNumEvents()
newInputNumFiles += 1
newNumMaster += 1
newInputFileSet.add(tmpFileSpec.lfn)
if self.taskSpec.outputScaleWithEvents():
newFileSize += long(sizeGradients * effectiveNumEvents)
if not dynNumEvents or tmpFileSpec.lfn not in inputFileSet:
if not useDirectIO:
newFileSize += long(tmpFileSpec.fsize)
newOutSizeMap[self.masterDataset.datasetID] += long(
sizeGradients * effectiveNumEvents)
else:
newFileSize += long(sizeGradients * effectiveFsize)
if not dynNumEvents or tmpFileSpec.lfn not in inputFileSet:
if not useDirectIO:
newFileSize += long(tmpFileSpec.fsize)
newOutSizeMap[self.masterDataset.datasetID] += long(
sizeGradients * effectiveFsize)
if sizeGradientsPerInSize != None:
newFileSize += long(effectiveFsize *
sizeGradientsPerInSize)
newOutSizeMap[self.masterDataset.datasetID] += long(
effectiveFsize * sizeGradientsPerInSize)
if self.taskSpec.useHS06():
tmpExpWalltime = walltimeGradient * effectiveNumEvents / float(
coreCount)
if not corePower in [None, 0]:
tmpExpWalltime /= corePower
if self.taskSpec.cpuEfficiency == 0:
tmpExpWalltime = 0
else:
tmpExpWalltime /= float(
self.taskSpec.cpuEfficiency) / 100.0
if multiplicity != None:
tmpExpWalltime /= float(multiplicity)
newExpWalltime += long(tmpExpWalltime)
else:
tmpExpWalltime = walltimeGradient * effectiveFsize / float(
coreCount)
if multiplicity != None:
tmpExpWalltime /= float(multiplicity)
newExpWalltime += long(tmpExpWalltime)
# boundaryID
if splitWithBoundaryID:
newBoundaryIDs.add(tmpFileSpec.boundaryID)
# check secondaries
firstSecondary = True
for datasetSpec in self.secondaryDatasetList:
if not datasetSpec.datasetID in newOutSizeMap:
newOutSizeMap[datasetSpec.datasetID] = 0
if not datasetSpec.isNoSplit(
) and datasetSpec.getNumFilesPerJob() == None:
# check boundaryID
if splitWithBoundaryID and boundaryID != None and boundaryID != tmpFileSpec.boundaryID \
and useBoundary['inSplit'] != 3:
break
newNumSecondary = datasetSpec.getNumMultByRatio(
newNumMaster) - nSecFilesMap[datasetSpec.datasetID]
datasetUsage = self.datasetMap[datasetSpec.datasetID]
for tmpFileSpec in datasetSpec.Files[
datasetUsage['used']:datasetUsage['used'] +
nSecondary]:
# check boundaryID
if splitWithBoundaryID and boundaryID != None and boundaryID != tmpFileSpec.boundaryID \
and not tmpFileSpec.boundaryID in boundaryIDs and not tmpFileSpec.boundaryID in newBoundaryIDs:
break
if not useDirectIO:
newFileSize += tmpFileSpec.fsize
if sizeGradientsPerInSize != None:
newFileSize += (tmpFileSpec.fsize *
sizeGradientsPerInSize)
newOutSizeMap[datasetSpec.datasetID] += (
tmpFileSpec.fsize * sizeGradientsPerInSize)
# the number of events
if firstSecondary and maxNumEvents != None and not primaryHasEvents:
if tmpFileSpec.startEvent != None and tmpFileSpec.endEvent != None:
newInputNumEvents += (tmpFileSpec.endEvent -
tmpFileSpec.startEvent +
1)
elif tmpFileSpec.nEvents != None:
newInputNumEvents += tmpFileSpec.nEvents
firstSecondary = False
# termination
if terminateFlag:
break
# check
newOutSize = self.getOutSize(newOutSizeMap)
if (maxNumFiles != None and ((not dynNumEvents and newInputNumFiles > maxNumFiles) \
or (dynNumEvents and (len(newInputFileSet) > maxNumFiles or newInputNumFiles > maxNumEventRanges)))) \
or (maxSize != None and newFileSize > maxSize) \
or (maxSize != None and newOutSize < minOutSize and maxSize-minOutSize < newFileSize-newOutSize) \
or (maxWalltime > 0 and newExpWalltime > maxWalltime) \
or (maxNumEvents != None and newInputNumEvents > maxNumEvents) \
or (maxOutSize != None and self.getOutSize(newOutSizeMap) > maxOutSize):
break
# reset nUsed for repeated datasets
for tmpDatasetID, datasetUsage in self.datasetMap.iteritems():
tmpDatasetSpec = datasetUsage['datasetSpec']
if tmpDatasetSpec.isRepeated():
if len(tmpDatasetSpec.Files) > 0:
datasetUsage['used'] %= len(tmpDatasetSpec.Files)
# make copy to return
returnList = []
for tmpDatasetID, inputFileList in inputFileMap.iteritems():
tmpRetList = []
for tmpFileSpec in inputFileList:
# split par site or get atomic subchunk
if siteName != None:
# make copy to individually set locality
newFileSpec = copy.copy(tmpFileSpec)
# set locality
newFileSpec.locality = siteCandidate.getFileLocality(
tmpFileSpec)
if newFileSpec.locality == 'remote':
newFileSpec.sourceName = siteCandidate.remoteSource
# append
tmpRetList.append(newFileSpec)
else:
# getting atomic subchunk
tmpRetList.append(tmpFileSpec)
# add to return map
tmpDatasetSpec = self.getDatasetWithID(tmpDatasetID)
returnList.append((tmpDatasetSpec, tmpRetList))
# return
return returnList
def getSubChunk(self,siteName,maxNumFiles=None,maxSize=None,
sizeGradients=0,sizeIntercepts=0,
nFilesPerJob=None,multiplicand=1,
walltimeGradient=0,maxWalltime=0,
nEventsPerJob=None,useBoundary=None,
sizeGradientsPerInSize=None,
maxOutSize=None,
coreCount=1,
respectLB=False,
corePower=None,
dynNumEvents=False,
maxNumEventRanges=None,
multiplicity=None,
splitByFields=None,
tmpLog=None,
useDirectIO=False,
maxDiskSize=None):
# check if there are unused files/events
if not self.checkUnused():
return None
# protection against unreasonable values
if nFilesPerJob == 0:
nFilesPerJob = None
if nEventsPerJob == 0:
nEventsPerJob = None
# set default max number of files
if maxNumFiles == None:
maxNumFiles = 50
# set default max number of event ranges
if maxNumEventRanges == None:
maxNumEventRanges = 20
# set default max size
if maxSize == None and nFilesPerJob == None and nEventsPerJob == None:
# 20 GB at most by default
maxSize = 20 * 1024 * 1024 * 1024
# set default output size
minOutSize = self.defaultOutputSize
# set default max number of events
maxNumEvents = None
# ignore negative walltime gradient
if walltimeGradient < 0:
walltimeGradient = 0
# overwrite parameters when nFiles/EventsPerJob is used
if nFilesPerJob != None and not dynNumEvents:
maxNumFiles = nFilesPerJob
if not respectLB:
multiplicand = nFilesPerJob
if nEventsPerJob != None:
maxNumEvents = nEventsPerJob
# split with boundayID
splitWithBoundaryID = False
if useBoundary != None:
splitWithBoundaryID = True
if useBoundary['inSplit'] == 2:
# unset max values to split only with boundaryID
maxNumFiles = None
maxSize = None
maxWalltime = 0
maxNumEvents = None
multiplicand = 1
# get site when splitting per site
if siteName != None:
siteCandidate = self.siteCandidates[siteName]
# use event ratios
useEventRatio = self.useEventRatioForSec()
# start splitting
inputNumFiles = 0
inputNumEvents = 0
fileSize = 0
firstLoop = True
firstMaster = True
inputFileMap = {}
expWalltime = 0
nextStartEvent = None
boundaryID = None
newBoundaryID = False
eventJump = False
nSecFilesMap = {}
nSecEventsMap = {}
numMaster = 0
outSizeMap = {}
lumiBlockNr = None
newLumiBlockNr = False
siteAvailable = True
inputFileSet = set()
fieldStr = None
diskSize = 0
while (maxNumFiles == None or (not dynNumEvents and inputNumFiles <= maxNumFiles) or \
(dynNumEvents and len(inputFileSet) <= maxNumFiles and inputNumFiles <= maxNumEventRanges)) \
and (maxSize == None or (maxSize != None and fileSize <= maxSize)) \
and (maxWalltime <= 0 or expWalltime <= maxWalltime) \
and (maxNumEvents == None or (maxNumEvents != None and inputNumEvents <= maxNumEvents)) \
and (maxOutSize == None or self.getOutSize(outSizeMap) <= maxOutSize) \
and (maxDiskSize is None or diskSize <= maxDiskSize):
# get one file (or one file group for MP) from master
datasetUsage = self.datasetMap[self.masterDataset.datasetID]
if not self.masterDataset.datasetID in outSizeMap:
outSizeMap[self.masterDataset.datasetID] = 0
boundaryIDs = set()
primaryHasEvents = False
for tmpFileSpec in self.masterDataset.Files[datasetUsage['used']:datasetUsage['used']+multiplicand]:
# check start event to keep continuity
if (maxNumEvents != None or dynNumEvents) and tmpFileSpec.startEvent != None:
if nextStartEvent != None and nextStartEvent != tmpFileSpec.startEvent:
eventJump = True
break
# check boundaryID
if splitWithBoundaryID and boundaryID != None and boundaryID != tmpFileSpec.boundaryID \
and useBoundary['inSplit'] != 3:
newBoundaryID = True
break
# check LB
if respectLB and lumiBlockNr != None and lumiBlockNr != tmpFileSpec.lumiBlockNr:
newLumiBlockNr = True
break
# check field
if splitByFields != None:
tmpFieldStr = tmpFileSpec.extractFieldsStr(splitByFields)
if fieldStr == None:
fieldStr = tmpFieldStr
elif tmpFieldStr != fieldStr:
newBoundaryID = True
break
# check for distributed datasets
if self.masterDataset.isDistributed() and siteName != None and \
not siteCandidate.isAvailableFile(tmpFileSpec):
siteAvailable = False
break
if not inputFileMap.has_key(self.masterDataset.datasetID):
inputFileMap[self.masterDataset.datasetID] = []
inputFileMap[self.masterDataset.datasetID].append(tmpFileSpec)
inputFileSet.add(tmpFileSpec.lfn)
datasetUsage['used'] += 1
numMaster += 1
# get effective file size
effectiveFsize = JediCoreUtils.getEffectiveFileSize(tmpFileSpec.fsize,tmpFileSpec.startEvent,
tmpFileSpec.endEvent,tmpFileSpec.nEvents)
# get num of events
effectiveNumEvents = tmpFileSpec.getEffectiveNumEvents()
# sum
inputNumFiles += 1
if self.taskSpec.outputScaleWithEvents():
tmpOutSize = long(sizeGradients * effectiveNumEvents)
fileSize += tmpOutSize
diskSize += tmpOutSize
if not dynNumEvents or tmpFileSpec.lfn not in inputFileSet:
fileSize += long(tmpFileSpec.fsize)
if not useDirectIO:
diskSize += long(tmpFileSpec.fsize)
outSizeMap[self.masterDataset.datasetID] += long(sizeGradients * effectiveNumEvents)
else:
tmpOutSize = long(sizeGradients * effectiveFsize)
fileSize += tmpOutSize
diskSize += tmpOutSize
if not dynNumEvents or tmpFileSpec.lfn not in inputFileSet:
fileSize += long(tmpFileSpec.fsize)
if not useDirectIO:
diskSize += long(tmpFileSpec.fsize)
outSizeMap[self.masterDataset.datasetID] += long(sizeGradients * effectiveFsize)
if sizeGradientsPerInSize != None:
tmpOutSize = long(effectiveFsize * sizeGradientsPerInSize)
fileSize += tmpOutSize
diskSize += tmpOutSize
outSizeMap[self.masterDataset.datasetID] += long(effectiveFsize * sizeGradientsPerInSize)
# sum offset only for the first master
if firstMaster:
fileSize += sizeIntercepts
# walltime
if self.taskSpec.useHS06():
if firstMaster:
expWalltime += self.taskSpec.baseWalltime
tmpExpWalltime = walltimeGradient * effectiveNumEvents / float(coreCount)
if not corePower in [None,0]:
tmpExpWalltime /= corePower
if self.taskSpec.cpuEfficiency == 0:
tmpExpWalltime = 0
else:
tmpExpWalltime /= float(self.taskSpec.cpuEfficiency)/100.0
if multiplicity != None:
tmpExpWalltime /= float(multiplicity)
expWalltime += long(tmpExpWalltime)
else:
tmpExpWalltime = walltimeGradient * effectiveFsize / float(coreCount)
if multiplicity != None:
tmpExpWalltime /= float(multiplicity)
expWalltime += long(tmpExpWalltime)
# the number of events
if (maxNumEvents != None or useEventRatio) and tmpFileSpec.startEvent != None and tmpFileSpec.endEvent != None:
primaryHasEvents = True
inputNumEvents += (tmpFileSpec.endEvent - tmpFileSpec.startEvent + 1)
# set next start event
nextStartEvent = tmpFileSpec.endEvent + 1
if nextStartEvent == tmpFileSpec.nEvents:
nextStartEvent = 0
# boundaryID
if splitWithBoundaryID:
boundaryID = tmpFileSpec.boundaryID
if not boundaryID in boundaryIDs:
boundaryIDs.add(boundaryID)
# LB
if respectLB:
lumiBlockNr = tmpFileSpec.lumiBlockNr
firstMaster = False
# get files from secondaries
firstSecondary = True
for datasetSpec in self.secondaryDatasetList:
if not datasetSpec.datasetID in outSizeMap:
outSizeMap[datasetSpec.datasetID] = 0
if datasetSpec.isNoSplit():
# every job uses dataset without splitting
if firstLoop:
datasetUsage = self.datasetMap[datasetSpec.datasetID]
for tmpFileSpec in datasetSpec.Files:
if not inputFileMap.has_key(datasetSpec.datasetID):
inputFileMap[datasetSpec.datasetID] = []
inputFileMap[datasetSpec.datasetID].append(tmpFileSpec)
# sum
fileSize += tmpFileSpec.fsize
if not useDirectIO:
diskSize += tmpFileSpec.fsize
if sizeGradientsPerInSize != None:
tmpOutSize = (tmpFileSpec.fsize * sizeGradientsPerInSize)
fileSize += tmpOutSize
diskSize += tmpOutSize
outSizeMap[datasetSpec.datasetID] += (tmpFileSpec.fsize * sizeGradientsPerInSize)
datasetUsage['used'] += 1
else:
if not nSecFilesMap.has_key(datasetSpec.datasetID):
nSecFilesMap[datasetSpec.datasetID] = 0
# get number of files to be used for the secondary
nSecondary = datasetSpec.getNumFilesPerJob()
if nSecondary != None and firstLoop == False:
# read files only in the first bunch when number of files per job is specified
continue
if nSecondary == None:
nSecondary = datasetSpec.getNumMultByRatio(numMaster) - nSecFilesMap[datasetSpec.datasetID]
if (datasetSpec.getEventRatio() != None and inputNumEvents > 0) or (splitWithBoundaryID and useBoundary['inSplit'] != 3):
# set large number to get all associated secondary files
nSecondary = 10000
datasetUsage = self.datasetMap[datasetSpec.datasetID]
# reset nUsed
if datasetSpec.isReusable() and datasetUsage['used']+nSecondary > len(datasetSpec.Files):
datasetUsage['used'] = 0
for tmpFileSpec in datasetSpec.Files[datasetUsage['used']:datasetUsage['used']+nSecondary]:
# check boundaryID
if (splitWithBoundaryID or (useBoundary != None and useBoundary['inSplit'] == 3 and datasetSpec.getRatioToMaster() > 1)) \
and boundaryID != None and \
not (boundaryID == tmpFileSpec.boundaryID or tmpFileSpec.boundaryID in boundaryIDs):
break
# check for distributed datasets
if datasetSpec.isDistributed() and siteName != None and \
not siteCandidate.isAvailableFile(tmpFileSpec):
break
# check ratio
if not datasetSpec.datasetID in nSecEventsMap:
nSecEventsMap[datasetSpec.datasetID] = 0
if datasetSpec.getEventRatio() != None and inputNumEvents > 0:
if float(nSecEventsMap[datasetSpec.datasetID]) / float(inputNumEvents) >= datasetSpec.getEventRatio():
break
if not inputFileMap.has_key(datasetSpec.datasetID):
inputFileMap[datasetSpec.datasetID] = []
inputFileMap[datasetSpec.datasetID].append(tmpFileSpec)
# sum
fileSize += tmpFileSpec.fsize
if not useDirectIO:
diskSize += tmpFileSpec.fsize
if sizeGradientsPerInSize != None:
tmpOutSize = (tmpFileSpec.fsize * sizeGradientsPerInSize)
fileSize += tmpOutSize
diskSize += tmpOutSize
outSizeMap[datasetSpec.datasetID] += (tmpFileSpec.fsize * sizeGradientsPerInSize)
datasetUsage['used'] += 1
nSecFilesMap[datasetSpec.datasetID] += 1
# the number of events
if firstSecondary and maxNumEvents != None and not primaryHasEvents:
if tmpFileSpec.startEvent != None and tmpFileSpec.endEvent != None:
inputNumEvents += (tmpFileSpec.endEvent - tmpFileSpec.startEvent + 1)
elif tmpFileSpec.nEvents != None:
inputNumEvents += tmpFileSpec.nEvents
if tmpFileSpec.nEvents != None:
nSecEventsMap[datasetSpec.datasetID] += tmpFileSpec.nEvents
# use only the first secondary
firstSecondary = False
# unset first loop flag
firstLoop = False
# check if there are unused files/evets
if not self.checkUnused():
break
# break if nFilesPerJob is used as multiplicand
if nFilesPerJob != None and not respectLB:
break
# boundayID is changed
if newBoundaryID:
break
# LB is changed
if newLumiBlockNr:
break
# event jump
if eventJump:
break
# distributed files are unavailable
if not siteAvailable:
break
primaryHasEvents = False
# check master in the next loop
datasetUsage = self.datasetMap[self.masterDataset.datasetID]
newInputNumFiles = inputNumFiles
newInputNumEvents = inputNumEvents
newFileSize = fileSize
newExpWalltime = expWalltime
newNextStartEvent = nextStartEvent
newNumMaster = numMaster
terminateFlag = False
newOutSizeMap = copy.copy(outSizeMap)
newBoundaryIDs = set()
newInputFileSet = copy.copy(inputFileSet)
newDiskSize = diskSize
if not self.masterDataset.datasetID in newOutSizeMap:
newOutSizeMap[self.masterDataset.datasetID] = 0
for tmpFileSpec in self.masterDataset.Files[datasetUsage['used']:datasetUsage['used']+multiplicand]:
# check continuity of event
if maxNumEvents != None and tmpFileSpec.startEvent != None and tmpFileSpec.endEvent != None:
primaryHasEvents = True
newInputNumEvents += (tmpFileSpec.endEvent - tmpFileSpec.startEvent + 1)
# continuity of event is broken
if newNextStartEvent != None and newNextStartEvent != tmpFileSpec.startEvent:
# no files in the next loop
if newInputNumFiles == 0:
terminateFlag = True
break
newNextStartEvent = tmpFileSpec.endEvent + 1
# check boundary
if splitWithBoundaryID and boundaryID != None and boundaryID != tmpFileSpec.boundaryID \
and useBoundary['inSplit'] != 3:
# no files in the next loop
if newInputNumFiles == 0:
terminateFlag = True
break
# check LB
if respectLB and lumiBlockNr != None and lumiBlockNr != tmpFileSpec.lumiBlockNr:
# no files in the next loop
if newInputNumFiles == 0:
terminateFlag = True
break
# check field
if splitByFields != None:
tmpFieldStr = tmpFileSpec.extractFieldsStr(splitByFields)
if tmpFieldStr != fieldStr:
# no files in the next loop
if newInputNumFiles == 0:
terminateFlag = True
break
# check for distributed datasets
if self.masterDataset.isDistributed() and siteName != None and \
not siteCandidate.isAvailableFile(tmpFileSpec):
# no files in the next loop
if newInputNumFiles == 0:
terminateFlag = True
break
# get effective file size
effectiveFsize = JediCoreUtils.getEffectiveFileSize(tmpFileSpec.fsize,tmpFileSpec.startEvent,
tmpFileSpec.endEvent,tmpFileSpec.nEvents)
# get num of events
effectiveNumEvents = tmpFileSpec.getEffectiveNumEvents()
newInputNumFiles += 1
newNumMaster += 1
newInputFileSet.add(tmpFileSpec.lfn)
if self.taskSpec.outputScaleWithEvents():
tmpOutSize = long(sizeGradients * effectiveNumEvents)
newFileSize += tmpOutSize
newDiskSize += tmpOutSize
if not dynNumEvents or tmpFileSpec.lfn not in inputFileSet:
newFileSize += long(tmpFileSpec.fsize)
if not useDirectIO:
newDiskSize += long(tmpFileSpec.fsize)
newOutSizeMap[self.masterDataset.datasetID] += long(sizeGradients * effectiveNumEvents)
else:
tmpOutSize = long(sizeGradients * effectiveFsize)
newFileSize += tmpOutSize
newDiskSize += tmpOutSize
if not dynNumEvents or tmpFileSpec.lfn not in inputFileSet:
newFileSize += long(tmpFileSpec.fsize)
if not useDirectIO:
newDiskSize += long(tmpFileSpec.fsize)
newOutSizeMap[self.masterDataset.datasetID] += long(sizeGradients * effectiveFsize)
if sizeGradientsPerInSize != None:
tmpOutSize = long(effectiveFsize * sizeGradientsPerInSize)
newFileSize += tmpOutSize
newDiskSize += tmpOutSize
newOutSizeMap[self.masterDataset.datasetID] += long(effectiveFsize * sizeGradientsPerInSize)
if self.taskSpec.useHS06():
tmpExpWalltime = walltimeGradient * effectiveNumEvents / float(coreCount)
if not corePower in [None,0]:
tmpExpWalltime /= corePower
if self.taskSpec.cpuEfficiency == 0:
tmpExpWalltime = 0
else:
tmpExpWalltime /= float(self.taskSpec.cpuEfficiency)/100.0
if multiplicity != None:
tmpExpWalltime /= float(multiplicity)
newExpWalltime += long(tmpExpWalltime)
else:
tmpExpWalltime = walltimeGradient * effectiveFsize / float(coreCount)
if multiplicity != None:
tmpExpWalltime /= float(multiplicity)
newExpWalltime += long(tmpExpWalltime)
# boundaryID
if splitWithBoundaryID:
newBoundaryIDs.add(tmpFileSpec.boundaryID)
# check secondaries
firstSecondary = True
for datasetSpec in self.secondaryDatasetList:
if not datasetSpec.datasetID in newOutSizeMap:
newOutSizeMap[datasetSpec.datasetID] = 0
if not datasetSpec.isNoSplit() and datasetSpec.getNumFilesPerJob() == None:
# check boundaryID
if splitWithBoundaryID and boundaryID != None and boundaryID != tmpFileSpec.boundaryID \
and useBoundary['inSplit'] != 3:
break
newNumSecondary = datasetSpec.getNumMultByRatio(newNumMaster) - nSecFilesMap[datasetSpec.datasetID]
datasetUsage = self.datasetMap[datasetSpec.datasetID]
for tmpFileSpec in datasetSpec.Files[datasetUsage['used']:datasetUsage['used']+nSecondary]:
# check boundaryID
if splitWithBoundaryID and boundaryID != None and boundaryID != tmpFileSpec.boundaryID \
and not tmpFileSpec.boundaryID in boundaryIDs and not tmpFileSpec.boundaryID in newBoundaryIDs:
break
newFileSize += tmpFileSpec.fsize
if not useDirectIO:
newDiskSize += tmpFileSpec.fsize
if sizeGradientsPerInSize != None:
tmpOutSize = (tmpFileSpec.fsize * sizeGradientsPerInSize)
newFileSize += tmpOutSize
newDiskSize += tmpOutSize
newOutSizeMap[datasetSpec.datasetID] += (tmpFileSpec.fsize * sizeGradientsPerInSize)
# the number of events
if firstSecondary and maxNumEvents != None and not primaryHasEvents:
if tmpFileSpec.startEvent != None and tmpFileSpec.endEvent != None:
newInputNumEvents += (tmpFileSpec.endEvent - tmpFileSpec.startEvent + 1)
elif tmpFileSpec.nEvents != None:
newInputNumEvents += tmpFileSpec.nEvents
firstSecondary = False
# termination
if terminateFlag:
break
# check
newOutSize = self.getOutSize(newOutSizeMap)
if (maxNumFiles != None and ((not dynNumEvents and newInputNumFiles > maxNumFiles) \
or (dynNumEvents and (len(newInputFileSet) > maxNumFiles or newInputNumFiles > maxNumEventRanges)))) \
or (maxSize != None and newFileSize > maxSize) \
or (maxSize != None and newOutSize < minOutSize and maxSize-minOutSize < newFileSize-newOutSize) \
or (maxWalltime > 0 and newExpWalltime > maxWalltime) \
or (maxNumEvents != None and newInputNumEvents > maxNumEvents) \
or (maxOutSize != None and self.getOutSize(newOutSizeMap) > maxOutSize) \
or (maxDiskSize is not None and newDiskSize > maxDiskSize):
break
# reset nUsed for repeated datasets
for tmpDatasetID,datasetUsage in self.datasetMap.iteritems():
tmpDatasetSpec = datasetUsage['datasetSpec']
if tmpDatasetSpec.isRepeated():
if len(tmpDatasetSpec.Files) > 0:
datasetUsage['used'] %= len(tmpDatasetSpec.Files)
# make copy to return
returnList = []
for tmpDatasetID,inputFileList in inputFileMap.iteritems():
tmpRetList = []
for tmpFileSpec in inputFileList:
# split par site or get atomic subchunk
if siteName != None:
# make copy to individually set locality
newFileSpec = copy.copy(tmpFileSpec)
# set locality
newFileSpec.locality = siteCandidate.getFileLocality(tmpFileSpec)
if newFileSpec.locality == 'remote':
newFileSpec.sourceName = siteCandidate.remoteSource
# append
tmpRetList.append(newFileSpec)
else:
# getting atomic subchunk
tmpRetList.append(tmpFileSpec)
# add to return map
tmpDatasetSpec = self.getDatasetWithID(tmpDatasetID)
returnList.append((tmpDatasetSpec,tmpRetList))
# return
return returnList
def getSubChunk(self,siteName,maxNumFiles=None,maxSize=None,
sizeGradients=0,sizeIntercepts=0,
nFilesPerJob=None,multiplicand=1,
walltimeGradient=0,maxWalltime=0,
nEventsPerJob=None,useBoundary=None,
sizeGradientsPerInSize=None,
maxOutSize=None,
tmpLog=None):
# check if there are unused files/events
if not self.checkUnused():
return None
# set default max number of files
if maxNumFiles == None:
# 20 files at most by default
maxNumFiles = 20
# set default max size
if maxSize == None:
# 20 GB at most by default
maxSize = 20 * 1024 * 1024 * 1024
# set default output size 2G + 500MB (safety merging)
minOutSize = 2500 * 1024 * 1024
# set default max number of events
maxNumEvents = None
# overwrite parameters when nFiles/EventsPerJob is used
if nFilesPerJob != None:
maxNumFiles = nFilesPerJob
multiplicand = nFilesPerJob
if nEventsPerJob != None:
maxNumEvents = nEventsPerJob
# split with boundayID
splitWithBoundaryID = False
if useBoundary != None:
splitWithBoundaryID = True
if useBoundary['inSplit'] == 2:
# unset max values to split only with boundaryID
maxNumFiles = None
maxSize = None
maxWalltime = 0
maxNumEvents = None
multiplicand = 1
# get site when splitting per site
if siteName != None:
siteCandidate = self.siteCandidates[siteName]
# start splitting
inputNumFiles = 0
inputNumEvents = 0
fileSize = 0
firstLoop = True
firstMaster = True
inputFileMap = {}
expWalltime = 0
nextStartEvent = None
boundaryID = None
newBoundaryID = False
nSecFilesMap = {}
numMaster = 0
outSize = 0
while (maxNumFiles == None or inputNumFiles <= maxNumFiles) \
and (maxSize == None or (maxSize != None and fileSize <= maxSize)) \
and (maxWalltime <= 0 or expWalltime <= maxWalltime) \
and (maxNumEvents == None or (maxNumEvents != None and inputNumEvents <= maxNumEvents)) \
and (maxOutSize == None or outSize <= maxOutSize):
# get one file (or one file group for MP) from master
datasetUsage = self.datasetMap[self.masterDataset.datasetID]
for tmpFileSpec in self.masterDataset.Files[datasetUsage['used']:datasetUsage['used']+multiplicand]:
# check start event to keep continuity
if maxNumEvents != None and tmpFileSpec.startEvent != None:
if nextStartEvent != None and nextStartEvent != tmpFileSpec.startEvent:
break
# check boundaryID
if splitWithBoundaryID and boundaryID != None and boundaryID != tmpFileSpec.boundaryID \
and useBoundary['inSplit'] != 3:
newBoundaryID = True
break
if not inputFileMap.has_key(self.masterDataset.datasetID):
inputFileMap[self.masterDataset.datasetID] = []
inputFileMap[self.masterDataset.datasetID].append(tmpFileSpec)
datasetUsage['used'] += 1
numMaster += 1
# get effective file size
effectiveFsize = JediCoreUtils.getEffectiveFileSize(tmpFileSpec.fsize,tmpFileSpec.startEvent,
tmpFileSpec.endEvent,tmpFileSpec.nEvents)
# sum
inputNumFiles += 1
fileSize += long(tmpFileSpec.fsize + sizeGradients * effectiveFsize)
outSize += long(sizeGradients * effectiveFsize)
if sizeGradientsPerInSize != None:
fileSize += long(effectiveFsize * sizeGradientsPerInSize)
outSize += long(effectiveFsize * sizeGradientsPerInSize)
# sum offset only for the first master
if firstMaster:
fileSize += sizeIntercepts
firstMaster = False
# walltime
expWalltime += long(walltimeGradient * effectiveFsize)
# the number of events
if maxNumEvents != None and tmpFileSpec.startEvent != None and tmpFileSpec.endEvent != None:
inputNumEvents += (tmpFileSpec.endEvent - tmpFileSpec.startEvent + 1)
# set next start event
nextStartEvent = tmpFileSpec.endEvent + 1
if nextStartEvent == tmpFileSpec.nEvents:
nextStartEvent = 0
# boundaryID
if splitWithBoundaryID:
boundaryID = tmpFileSpec.boundaryID
# get files from secondaries
for datasetSpec in self.secondaryDatasetList:
if datasetSpec.isNoSplit():
# every job uses dataset without splitting
if firstLoop:
datasetUsage = self.datasetMap[datasetSpec.datasetID]
for tmpFileSpec in datasetSpec.Files:
if not inputFileMap.has_key(datasetSpec.datasetID):
inputFileMap[datasetSpec.datasetID] = []
inputFileMap[datasetSpec.datasetID].append(tmpFileSpec)
# sum
fileSize += tmpFileSpec.fsize
if sizeGradientsPerInSize != None:
fileSize += (tmpFileSpec.fsize * sizeGradientsPerInSize)
outSize += (tmpFileSpec.fsize * sizeGradientsPerInSize)
datasetUsage['used'] += 1
else:
if not nSecFilesMap.has_key(datasetSpec.datasetID):
nSecFilesMap[datasetSpec.datasetID] = 0
# get number of files to be used for the secondary
nSecondary = datasetSpec.getNumFilesPerJob()
if nSecondary != None and firstLoop == False:
# read files only in the first bunch when number of files per job is specified
continue
if nSecondary == None:
nSecondary = datasetSpec.getNumMultByRatio(numMaster) - nSecFilesMap[datasetSpec.datasetID]
if splitWithBoundaryID and useBoundary['inSplit'] != 3:
# set large number to get all associated secondary files
nSecondary = 10000
datasetUsage = self.datasetMap[datasetSpec.datasetID]
for tmpFileSpec in datasetSpec.Files[datasetUsage['used']:datasetUsage['used']+nSecondary]:
# check boundaryID
if splitWithBoundaryID and boundaryID != None and boundaryID != tmpFileSpec.boundaryID:
break
if not inputFileMap.has_key(datasetSpec.datasetID):
inputFileMap[datasetSpec.datasetID] = []
inputFileMap[datasetSpec.datasetID].append(tmpFileSpec)
# sum
fileSize += tmpFileSpec.fsize
if sizeGradientsPerInSize != None:
fileSize += (tmpFileSpec.fsize * sizeGradientsPerInSize)
outSize += (tmpFileSpec.fsize * sizeGradientsPerInSize)
datasetUsage['used'] += 1
nSecFilesMap[datasetSpec.datasetID] += 1
# unset first loop flag
firstLoop = False
# check if there are unused files/evets
if not self.checkUnused():
break
# break if nFilesPerJob is used
if nFilesPerJob != None:
break
# boundayID is changed
if newBoundaryID:
break
# check master in the next loop
datasetUsage = self.datasetMap[self.masterDataset.datasetID]
newInputNumFiles = inputNumFiles
newInputNumEvents = inputNumEvents
newFileSize = fileSize
newExpWalltime = expWalltime
newNextStartEvent = nextStartEvent
newNumMaster = numMaster
terminateFlag = False
newOutSize = outSize
for tmpFileSpec in self.masterDataset.Files[datasetUsage['used']:datasetUsage['used']+multiplicand]:
# check continuity of event
if maxNumEvents != None and tmpFileSpec.startEvent != None and tmpFileSpec.endEvent != None:
newInputNumEvents += (tmpFileSpec.endEvent - tmpFileSpec.startEvent + 1)
# continuity of event is broken
if newNextStartEvent != None and newNextStartEvent != tmpFileSpec.startEvent:
# no files in the next loop
if newInputNumFiles == 0:
terminateFlag = True
break
newNextStartEvent = tmpFileSpec.endEvent + 1
# check boundary
if splitWithBoundaryID and boundaryID != None and boundaryID != tmpFileSpec.boundaryID:
# no files in the next loop
if newInputNumFiles == 0:
terminateFlag = True
break
# get effective file size
effectiveFsize = JediCoreUtils.getEffectiveFileSize(tmpFileSpec.fsize,tmpFileSpec.startEvent,
tmpFileSpec.endEvent,tmpFileSpec.nEvents)
newInputNumFiles += 1
newNumMaster += 1
newFileSize += long(tmpFileSpec.fsize + sizeGradients * effectiveFsize)
newOutSize += long(sizeGradients * effectiveFsize)
if sizeGradientsPerInSize != None:
newFileSize += long(effectiveFsize * sizeGradientsPerInSize)
newOutSize += long(effectiveFsize * sizeGradientsPerInSize)
newExpWalltime += long(walltimeGradient * effectiveFsize)
# check secondaries
for datasetSpec in self.secondaryDatasetList:
if not datasetSpec.isNoSplit() and datasetSpec.getNumFilesPerJob() == None:
# check boundaryID
if splitWithBoundaryID and boundaryID != None and boundaryID != tmpFileSpec.boundaryID:
break
newNumSecondary = datasetSpec.getNumMultByRatio(newNumMaster) - nSecFilesMap[datasetSpec.datasetID]
datasetUsage = self.datasetMap[datasetSpec.datasetID]
for tmpFileSpec in datasetSpec.Files[datasetUsage['used']:datasetUsage['used']+nSecondary]:
# check boundaryID
if splitWithBoundaryID and boundaryID != None and boundaryID != tmpFileSpec.boundaryID:
break
newFileSize += tmpFileSpec.fsize
if sizeGradientsPerInSize != None:
newFileSize += (tmpFileSpec.fsize * sizeGradientsPerInSize)
# termination
if terminateFlag:
break
# check
if (maxNumFiles != None and newInputNumFiles > maxNumFiles) \
or (maxSize != None and newFileSize > maxSize) \
or (maxSize != None and newOutSize < minOutSize and maxSize-minOutSize < newFileSize-newOutSize) \
or (maxWalltime > 0 and newExpWalltime > maxWalltime) \
or (maxNumEvents != None and newInputNumEvents > maxNumEvents) \
or (maxOutSize != None and newOutSize > maxOutSize):
break
# reset nUsed for repeated datasets
for tmpDatasetID,datasetUsage in self.datasetMap.iteritems():
tmpDatasetSpec = datasetUsage['datasetSpec']
if tmpDatasetSpec.isRepeated():
datasetUsage['used'] %= len(tmpDatasetSpec.Files)
# make copy to return
returnList = []
for tmpDatasetID,inputFileList in inputFileMap.iteritems():
tmpRetList = []
for tmpFileSpec in inputFileList:
# split par site or get atomic subchunk
if siteName != None:
# make copy to individually set locality
newFileSpec = copy.copy(tmpFileSpec)
# set locality
newFileSpec.locality = siteCandidate.getFileLocality(tmpFileSpec)
if newFileSpec.locality == 'remote':
newFileSpec.sourceName = siteCandidate.remoteSource
# append
tmpRetList.append(newFileSpec)
else:
# getting atomic subchunk
tmpRetList.append(tmpFileSpec)
# add to return map
tmpDatasetSpec = self.getDatasetWithID(tmpDatasetID)
returnList.append((tmpDatasetSpec,tmpRetList))
# return
return returnList
