import os
from DBSAPI.dbsException import *
from DBSAPI.dbsApiException import *
from DBSAPI.dbsOptions import DbsOptionParser
from DBSAPI.dbsApi import DbsApi
try:
optManager = DbsOptionParser()
(opts,args) = optManager.getOpt()
args = {}
# These dummy values are required to create the DbsApi object so let them here
args['url']='http://cmsdbsprod.cern.ch/cms_dbs_prod_global/servlet/DBSServlet'
args['version']='DBS_2_0_9'
args['mode']='POST'
api = DbsApi(args)
if len(sys.argv) < 3:
print "USAGE: python %s <srcURL> <dstURL> block_name" % sys.argv[0]
sys.exit(1)
srcURL = sys.argv[1]
dstURL = sys.argv[2]
block = sys.argv[3]
api.dbsMigrateBlock(srcURL, dstURL, block)
except DbsApiException, ex:
print "Caught API Exception %s: %s " % (ex.getClassName(), ex.getErrorMessage() )
if ex.getErrorCode() not in (None, ""):
print "DBS Exception Error Code: ", ex.getErrorCode()
print "Done"
class DBSInterface:
"""
_DBSInterface_
So once upon a time there was a guy named Anzar, and he wrote
a thing called DBSWriter for a project called ProdAgent,
which unfortunately sucked because, let's be honest, everything
in ProdAgent was thrown together just to make stuff work.
So eventually WMAgent came along and I had to rewrite the whole
thing.
"""
def __init__(self, config):
"""
Use Configuration object from config
"""
# Config should have DBSInterface element
self.config = config.DBSInterface
args = {'url': self.config.DBSUrl,
'level': 'ERROR',
"user" :'NORMAL',
'version': self.config.DBSVersion}
self.version = self.config.DBSVersion
self.globalDBSUrl = None
self.committedRuns = collections.deque(maxlen = 1000)
self.maxFilesToCommit = self.config.MaxFilesToCommit
self.doGlobalMigration = getattr(self.config, 'doGlobalMigration', True)
if getattr(self.config, 'globalDBSUrl', None) != None:
globalArgs = {'url': self.config.globalDBSUrl,
'level': 'ERROR',
"user" :'NORMAL',
'version': self.config.globalDBSVersion}
self.globalDBSUrl = self.config.globalDBSUrl
else:
self.doGlobalMigration = False
try:
self.dbs = DbsApi(args)
if self.globalDBSUrl:
self.globalDBS = DbsApi(globalArgs)
except DbsException as ex:
msg = "Error in DBSWriterError with DbsApi\n"
msg += "%s\n" % formatEx(ex)
logging.error(msg)
raise DBSInterfaceError(msg)
return
def runDBSBuffer(self, algo, dataset, blocks, override = False):
"""
_runDBSBuffer_
Run the entire DBSBuffer chain
"""
# First create the dataset
processed = self.insertDatasetAlgo(algo = algo, dataset = dataset,
override = override)
# Next create blocks
affBlocks = self.createAndInsertBlocks(dataset = dataset,
procDataset = processed,
blocks = blocks)
return affBlocks
def createAndInsertBlocks(self, dataset, procDataset, blocks):
"""
_createBlocks_
Create all the blocks we use, and insert the
files into them.
"""
affectedBlocks = []
for block in blocks:
# Create each block one at a time and insert its files
# First create the block
createUncheckedBlock(apiRef = self.dbs, name = block['Name'],
datasetPath = dataset['Path'],
seName = block['location'])
block['Path'] = dataset['Path']
block['StorageElementList'] = block['location']
# Now assemble the files
readyFiles = []
for f in block['newFiles']:
for run in f.getRuns():
if not run.run in self.committedRuns:
insertDBSRunsFromRun(apiRef = self.dbs, dSRun = run)
self.committedRuns.append(run.run)
dbsfile = createDBSFileFromBufferFile(bufferFile = f,
procDataset = procDataset)
readyFiles.append(dbsfile)
block['readyFiles'] = readyFiles
flag = False
if block['open'] == 'Pending':
logging.info("3Found block to close in DBSInterface.createAndInsertBlocks: %s" % (block['Name']))
flag = True
finBlock = self.insertFilesAndCloseBlocks(block = block, close = flag)
affectedBlocks.append(finBlock)
affBlocks = self.migrateClosedBlocks(blocks = affectedBlocks)
return affBlocks
def insertDatasetAlgo(self, algo, dataset, override = False):
"""
_insertDatasetAlgo_
Insert a dataset-algo combination
The override re-inserts it if DBSBuffer thinks it's already there
"""
dbsRef = None
if override or not algo['InDBS']:
# Then put the algo in DBS by referencing local DBS
dbsRef = self.dbs
# Create a DBS Algo
dbsAlgo = createAlgorithm(apiRef = dbsRef,
appName = algo['ApplicationName'],
appVer = algo['ApplicationVersion'],
appFam = algo['ApplicationFamily'],
PSetHash = algo['PSetHash'],
PSetContent = algo['PSetContent'])
if dataset['PrimaryDataset'].lower() == 'bogus':
# Do not commit bogus datasets!
return None
dbsRef = self.dbs
if dataset.get('DASInDBS', None):
# Then this whole thing is already in DBS
dbsRef = None
primary = createPrimaryDataset(primaryName = dataset['PrimaryDataset'],
primaryDatasetType = self.config.primaryDatasetType,
apiRef = dbsRef)
processed = createProcessedDataset(apiRef = dbsRef,
algorithm = dbsAlgo,
primary = primary,
processedName = dataset['ProcessedDataset'],
dataTier = dataset['DataTier'],
status = dataset['status'],
globalTag = dataset['globalTag'],
parent = dataset['parent'])
return processed
def insertFilesAndCloseBlocks(self, block, close = False):
"""
_insertFilesAndCloseBlocks_
Insert files into blocks and close them.
This does all the actual work of closing a block, first
inserting the files, then actually closing if you have
toggled the 'close' flag
"""
# Insert all the files added to the block in this round
if len(block.get('readyFiles', [])) > 0:
insertFiles(apiRef = self.dbs, datasetPath = block['Path'],
files = block['readyFiles'], block = block,
maxFiles = self.maxFilesToCommit)
# Reset the block files
block['insertedFiles'].extend(block['readyFiles'])
block['readyFiles'] = []
# Close the block if requested
if close:
logging.info("Calling close block...")
closeBlock(apiRef = self.dbs, block = block)
block['OpenForWriting'] = '0'
block['open'] = 0
return block
def closeAndMigrateBlocksByName(self, blockNames):
"""
_closeAndMigrateBlocksByName_
This is basically for the timeout in DBSUploadPoller
It allows you to close and migrate a block by name only
It expects a list of names really, although it can
take just one.
"""
if type(blockNames) != list:
blockNames = [blockNames]
blocksToClose = []
for name in blockNames:
blockList = listBlocks(apiRef = self.dbs,
blockName = name)
if len(blockList) != 1:
msg = "Error: We can't load blocks with this name\n"
msg += str(name)
msg += "\nRetrieved %i blocks" % (len(blockList))
logging.error(msg)
raise DBSInterfaceError(msg)
block = blockList[0]
block['open'] = 'Pending'
b2 = self.insertFilesAndCloseBlocks(block = block,
close = True)
blocksToClose.append(b2)
if self.doGlobalMigration:
self.migrateClosedBlocks(blocks = blocksToClose)
return blocksToClose
def migrateClosedBlocks(self, blocks):
"""
_migrateClosedBlocks_
One at a time, migrate closed blocks
This checks to see if blocks are closed.
If they are, it migrates them.
"""
if not self.doGlobalMigration:
logging.debug("Skipping migration due to doGlobalMigration tag.")
return blocks
if type(blocks) != list:
blocks = [blocks]
for block in blocks:
if block.get('OpenForWriting', 1) != '0':
# logging.error("Attempt to migrate open block!")
# Block is not done
# Ignore this, because we send all blocks here
continue
try:
# Migrate each block
logging.info("About to migrate block %s" % (block['Name']))
self.dbs.dbsMigrateBlock(srcURL = self.config.DBSUrl,
dstURL = self.globalDBSUrl,
block_name = block['Name'],
srcVersion = self.version,
dstVersion = self.config.globalDBSVersion)
block['open'] = 'InGlobalDBS'
except DbsException as ex:
msg = "Error in DBSInterface.migrateClosedBlocks()\n"
msg += "%s\n" % formatEx(ex)
msg += str(traceback.format_exc())
raise DBSInterfaceError(msg)
return blocks
def getAPIRef(self, globalRef = False):
"""
Get a DBSAPI ref to either the local or global API
"""
if globalRef:
return self.globalDBS
return self.dbs
#!/usr/bin/env python2.6
from DBSAPI.dbsApi import DbsApi
import sys
#blockToMigrate = '/T_t-channel_TuneZ2star_8TeV-powheg-tauola/Summer12_DR53X-PU_S10_START53_V7A-v3/AODSIM#b6861908-0f9a-11e2-826d-00221959e72f'
blockToMigrate = sys.argv[0]
localDbs = 'https://cmsdbsprod.cern.ch:8443/cms_dbs_ph_analysis_02_writer/servlet/DBSServlet'
globalDbs = 'http://cmsdbsprod.cern.ch/cms_dbs_prod_global/servlet/DBSServlet'
dbsArgs = { 'url' : localDbs,
'level' : 'ERROR',
'user' : 'DEBUG',
'version':'DBS_2_0_8' }
# /T_t-channel_TuneZ2star_8TeV-powheg-tauola/Summer12_DR53X-PU_S10_START53_V7A-v3/AODSIM#b6861908-0f9a-11e2-826d-00221959e72f
# /MET/Run2012C-EcalRecover_11Dec2012-v1/AOD#b38eb352-4304-11e2-a593-003048f02c8a
# /ZJetToMuMu_Pt-0to15_TuneZ2star_8TeV_pythia6/Summer12_DR53X-PU_S10_START53_V7A-v1/AODSIM#d6a384ec-154d-11e2-a364-00221959e72f
dbsApi = DbsApi(dbsArgs)
blockList = list(set(sys.argv[1:]))
for blockToMigrate in blockList:
print "Beginning to migrate %s" % blockToMigrate
newBlock = dbsApi.dbsMigrateBlock(globalDbs, localDbs, blockToMigrate)
print newBlock
print "Migration done"
class DBSWriter:
"""
_DBSWriter_
General API for writing data to DBS
"""
def __init__(self, url, **contact):
args = {"url": url, "level": 'ERROR'}
args.update(contact)
try:
self.dbs = DbsApi(args)
self.args = args
self.version = args.get('version', None)
self.globalDBSUrl = args.get('globalDBSUrl', None)
self.globalVersion = args.get('globalVersion', None)
if self.globalDBSUrl:
globalArgs = {'url': url, 'level': 'ERROR'}
globalArgs.update(contact)
self.globalDBS = DbsApi(globalArgs)
except DbsException as ex:
msg = "Error in DBSWriterError with DbsApi\n"
msg += "%s\n" % formatEx(ex)
raise DBSWriterError(msg)
self.reader = DBSReader(**args)
def createDatasets(self, workflowSpec):
"""
_createDatasets_
Create All the output datasets found in the workflow spec instance
provided
"""
try:
workflowSpec.payload.operate(
_CreateDatasetOperator(self.dbs, workflowSpec))
except DbsException as ex:
msg = "Error in DBSWriter.createDatasets\n"
msg += "For Workflow: %s\n" % workflowSpec.workflowName()
msg += "%s\n" % formatEx(ex)
raise DBSWriterError(msg)
return
def insertFilesForDBSBuffer(self,
files,
procDataset,
algos,
jobType="NotMerge",
insertDetectorData=False,
maxFiles=100,
maxSize=99999999,
timeOut=None,
fileCommitLength=5):
"""
_insertFiles_
list of files inserted in DBS
"""
#TODO: Whats the purpose of insertDetectorData
if len(files) < 1:
return
affectedBlocks = []
insertFiles = []
addedRuns = []
seName = None
#Get the algos in insertable form
# logging.error("About to input algos")
# logging.error(algos)
ialgos = [
DBSWriterObjects.createAlgorithmForInsert(dict(algo))
for algo in algos
]
#print ialgos
for outFile in files:
# //
# // Convert each file into a DBS File object
#//
lumiList = []
#Somehing similar should be the real deal when multiple runs/lumi could be returned from wmbs file
for runlumiinfo in outFile.getRuns():
lrun = long(runlumiinfo.run)
run = DbsRun(
RunNumber=lrun,
NumberOfEvents=0,
NumberOfLumiSections=0,
TotalLuminosity=0,
StoreNumber=0,
StartOfRun=0,
EndOfRun=0,
)
#Only added if not added by another file in this loop, why waste a call to DBS
if lrun not in addedRuns:
self.dbs.insertRun(run)
addedRuns.append(
lrun) #save it so we do not try to add it again to DBS
logging.debug("run %s added to DBS " % str(lrun))
for alsn in runlumiinfo:
lumi = DbsLumiSection(
LumiSectionNumber=long(alsn),
StartEventNumber=0,
EndEventNumber=0,
LumiStartTime=0,
LumiEndTime=0,
RunNumber=lrun,
)
lumiList.append(lumi)
logging.debug("lumi list created for the file")
dbsfile = DbsFile(
#Checksum = str(outFile['cksum']),
NumberOfEvents=outFile['events'],
LogicalFileName=outFile['lfn'],
FileSize=int(outFile['size']),
Status="VALID",
ValidationStatus='VALID',
FileType='EDM',
Dataset=procDataset,
TierList=DBSWriterObjects.makeTierList(
procDataset['Path'].split('/')[3]),
AlgoList=ialgos,
LumiList=lumiList,
ParentList=outFile.getParentLFNs(),
#BranchHash = outFile['BranchHash'],
)
#Set checksums by hand
#dbsfile['Checksum'] = 0 #Set a default?
for entry in outFile['checksums'].keys():
#This should be a dictionary with a cktype key and cksum value
if entry.lower() == 'cksum':
dbsfile['Checksum'] = str(outFile['checksums'][entry])
elif entry.lower() == 'adler32':
dbsfile['Adler32'] = str(outFile['checksums'][entry])
elif entry.lower() == 'md5':
dbsfile['Md5'] = str(outFile['checksums'][entry])
#This check comes from ProdAgent, not sure if its required
if len(outFile["locations"]) > 0:
seName = list(outFile["locations"])[0]
logging.debug("SEname associated to file is: %s" % seName)
else:
msg = "Error in DBSWriter.insertFiles\n"
msg += "No SEname associated to file"
#print "FAKING seName for now"
#seName="cmssrm.fnal.gov"
raise DBSWriterError(msg)
insertFiles.append(dbsfile)
# //Processing Jobs:
# // Insert the lists of sorted files into the appropriate
#// fileblocks
sumSize = 0
sumFiles = 0
tmpFiles = []
blockList = []
#First, get the block. See if the block already exists
try:
fileBlock = DBSWriterObjects.getDBSFileBlock(
self.dbs, procDataset, seName)
fileBlock['files'] = []
#if not fileBlock in affectedBlocks:
# affectedBlocks.append(fileBlock)
except DbsException as ex:
msg = "Error in DBSWriter.insertFilesForDBSBuffer\n"
msg += "Cannot retrieve FileBlock for dataset:\n"
msg += " %s\n" % procDataset['Path']
msg += "%s\n" % formatEx(ex)
raise DBSWriterError(msg)
filesToCommit = []
for file in insertFiles:
# First see if the block is full
if self.manageFileBlock(fileBlock=fileBlock,
maxFiles=maxFiles,
maxSize=maxSize,
timeOut=timeOut,
algos=ialgos,
filesToCommit=filesToCommit,
procDataset=procDataset):
fileBlock['OpenForWriting'] = 0
if not fileBlock in affectedBlocks:
affectedBlocks.append(fileBlock)
# Then we need a new block
try:
fileBlock = DBSWriterObjects.getDBSFileBlock(
self.dbs, procDataset, seName)
fileBlock['files'] = []
except DbsException as ex:
msg = "Error in DBSWriter.insertFilesForDBSBuffer\n"
msg += "Cannot retrieve FileBlock for dataset:\n"
msg += " %s\n" % procDataset['Path']
msg += "%s\n" % formatEx(ex)
raise DBSWriterError(msg)
fileBlock['files'].append(file['LogicalFileName'])
filesToCommit.append(file)
if len(filesToCommit) >= fileCommitLength:
# Only commit the files if there are more of them then the maximum length
try:
self.dbs.insertFiles(procDataset, filesToCommit, fileBlock)
filesToCommit = []
logging.debug("Inserted files: %s to FileBlock: %s" \
% ( ([ x['LogicalFileName'] for x in insertFiles ]),fileBlock['Name']))
except DbsException as ex:
msg = "Error in DBSWriter.insertFiles\n"
msg += "Cannot insert processed files:\n"
msg += " %s\n" % (
[x['LogicalFileName'] for x in insertFiles], )
msg += "%s\n" % formatEx(ex)
raise DBSWriterError(msg)
if len(filesToCommit) > 0:
try:
self.dbs.insertFiles(procDataset, filesToCommit, fileBlock)
filesToCommit = []
logging.debug("Inserted files: %s to FileBlock: %s" \
% ( ([ x['LogicalFileName'] for x in insertFiles ]),fileBlock['Name']))
except DbsException as ex:
msg = "Error in DBSWriter.insertFiles\n"
msg += "Cannot insert processed files:\n"
msg += " %s\n" % ([x['LogicalFileName']
for x in insertFiles], )
msg += "%s\n" % formatEx(ex)
raise DBSWriterError(msg)
if not fileBlock in affectedBlocks:
affectedBlocks.append(fileBlock)
## Do bulk inserts now for DBS
#filesToCommit = []
#count = 0
#count2 = 0
#for file in insertFiles:
# count += 1
# #Try and close the box
# logging.error("Should have a file")
# logging.error(len(filesToCommit))
# count2 += len(filesToCommit)
# if self.manageFileBlock(fileBlock = fileBlock, maxFiles = maxFiles,
# maxSize = maxSize, timeOut = timeOut, algos = ialgos,
# filesToCommit = filesToCommit, procDataset = procDataset):
# fileBlock['OpenForWriting'] = '0'
# if not fileBlock in affectedBlocks:
# affectedBlocks.append(fileBlock)
#
#
#
# # Then we need a new block
# try:
# fileBlock = DBSWriterObjects.getDBSFileBlock(
# self.dbs,
# procDataset,
# seName)
# fileBlock['files'] = []
# except DbsException, ex:
# msg = "Error in DBSWriter.insertFilesForDBSBuffer\n"
# msg += "Cannot retrieve FileBlock for dataset:\n"
# msg += " %s\n" % procDataset['Path']
# msg += "%s\n" % formatEx(ex)
# raise DBSWriterError(msg)
# #At this point, we should commit the block as is
# fileBlock['files'].append(file['LogicalFileName'])
# if jobType == "MergeSpecial":
# for file in fileList:
# file['Block'] = fileBlock
# msg="calling: self.dbs.insertMergedFile(%s, %s)" % (str(file['ParentList']),str(file))
# logging.debug(msg)
# try:
# #
# #
# # NOTE To Anzar From Anzar (File cloning as in DBS API can be done here and then I can use Bulk insert on Merged files as well)
# self.dbs.insertMergedFile(file['ParentList'],
# file)
#
# except DbsException, ex:
# msg = "Error in DBSWriter.insertFiles\n"
# msg += "Cannot insert merged file:\n"
# msg += " %s\n" % file['LogicalFileName']
# msg += "%s\n" % formatEx(ex)
# raise DBSWriterError(msg)
# logging.debug("Inserted merged file: %s to FileBlock: %s"%(file['LogicalFileName'],fileBlock['Name']))
# else:
# filesToCommit.append(file)
# if len(filesToCommit) >= fileCommitLength:
# # Only commit the files if there are more of them then the maximum length
# try:
# logging.error("About to commit %i files" %(len(filesToCommit)))
# count2 += len(filesToCommit)
# self.dbs.insertFiles(procDataset, filesToCommit, fileBlock)
# filesToCommit = []
# logging.debug("Inserted files: %s to FileBlock: %s" \
# % ( ([ x['LogicalFileName'] for x in insertFiles ]),fileBlock['Name']))
#
# except DbsException, ex:
# msg = "Error in DBSWriter.insertFiles\n"
# msg += "Cannot insert processed files:\n"
# msg += " %s\n" % ([ x['LogicalFileName'] for x in insertFiles ],)
# msg += "%s\n" % formatEx(ex)
# raise DBSWriterError(msg)
#
#
#
#
## If we still have files to commit, commit them
#logging.error("Got to the end of the loop")
#logging.error(len(filesToCommit))
#logging.error(count2)
#if len(filesToCommit) > 0:
# try:
# logging.error("About to insert some files")
# self.dbs.insertFiles(procDataset, filesToCommit, fileBlock)
# filesToCommit = []
# logging.debug("Inserted files: %s to FileBlock: %s" \
# % ( ([ x['LogicalFileName'] for x in insertFiles ]),fileBlock['Name']))
#
# except DbsException, ex:
# msg = "Error in DBSWriter.insertFiles\n"
# msg += "Cannot insert processed files:\n"
# msg += " %s\n" % ([ x['LogicalFileName'] for x in insertFiles ],)
# msg += "%s\n" % formatEx(ex)
# raise DBSWriterError(msg)
if not fileBlock in affectedBlocks:
affectedBlocks.append(fileBlock)
return list(affectedBlocks)
def insertFiles(self, fwkJobRep, insertDetectorData=False):
"""
_insertFiles_
Process the files in the FwkJobReport instance and insert
them into the associated datasets
A list of affected fileblock names is returned both for merged
and unmerged fileblocks. Only merged blocks will have to be managed.
#for merged file
#blocks to facilitate management of those blocks.
#This list is not populated for processing jobs since we dont really
#care about the processing job blocks.
"""
insertLists = {}
orderedHashes = []
affectedBlocks = set()
if len(fwkJobRep.files) <= 0:
msg = "Error in DBSWriter.insertFiles\n"
msg += "No files found in FrameWorkJobReport for:\n"
msg += "==> JobSpecId: %s" % fwkJobRep.jobSpecId
msg += " Workflow: %s" % fwkJobRep.workflowSpecId
raise DBSWriterError(msg)
for outFile in fwkJobRep.sortFiles():
# //
# // Convert each file into a DBS File object
#//
seName = None
if "SEName" in outFile:
if outFile['SEName']:
seName = outFile['SEName']
logging.debug("SEname associated to file is: %s" % seName)
## remove the fallback to site se-name if no SE is associated to File
## because it's likely that there is some stage out problem if there
## is no SEName associated to the file.
# if not seName:
# if fwkJobRep.siteDetails.has_key("se-name"):
# seName = fwkJobRep.siteDetails['se-name']
# seName = str(seName)
# logging.debug("site SEname: %s"%seName)
if not seName:
msg = "Error in DBSWriter.insertFiles\n"
msg += "No SEname associated to files in FrameWorkJobReport for "
# msg += "No SEname found in FrameWorkJobReport for "
msg += "==> JobSpecId: %s" % fwkJobRep.jobSpecId
msg += " Workflow: %s" % fwkJobRep.workflowSpecId
raise DBSWriterError(msg)
try:
if (insertDetectorData):
dbsFiles = DBSWriterObjects.createDBSFiles(
outFile, fwkJobRep.jobType, self.dbs)
else:
dbsFiles = DBSWriterObjects.createDBSFiles(
outFile, fwkJobRep.jobType)
except DbsException as ex:
msg = "Error in DBSWriter.insertFiles:\n"
msg += "Error creating DbsFile instances for file:\n"
msg += "%s\n" % outFile['LFN']
msg += "%s\n" % formatEx(ex)
raise DBSWriterError(msg)
if len(dbsFiles) <= 0:
msg = "No DbsFile instances created. Not enough info in the FrameWorkJobReport for"
msg += "==> JobSpecId: %s" % fwkJobRep.jobSpecId
msg += " Workflow: %s" % fwkJobRep.workflowSpecId
raise DBSWriterError(msg)
for f in dbsFiles:
datasetName = makeDBSDSName(f)
hashName = "%s-%s" % (seName, datasetName)
if hashName not in insertLists:
insertLists[hashName] = _InsertFileList(
seName, datasetName)
insertLists[hashName].append(f)
if not orderedHashes.count(hashName):
orderedHashes.append(hashName)
# //Processing Jobs:
# // Insert the lists of sorted files into the appropriate
#// fileblocks
for hash in orderedHashes:
fileList = insertLists[hash]
procDataset = fileList[0]['Dataset']
try:
fileBlock = DBSWriterObjects.getDBSFileBlock(
self.dbs, procDataset, fileList.seName)
except DbsException as ex:
msg = "Error in DBSWriter.insertFiles\n"
msg += "Cannot retrieve FileBlock for dataset:\n"
msg += " %s\n" % procDataset
msg += "In Storage Element:\n %s\n" % fileList.seName
msg += "%s\n" % formatEx(ex)
raise DBSWriterError(msg)
if fwkJobRep.jobType == "Merge":
# //
# // Merge files
#//
for mergedFile in fileList:
mergedFile['Block'] = fileBlock
affectedBlocks.add(fileBlock['Name'])
msg = "calling: self.dbs.insertMergedFile(%s, %s)" % (str(
mergedFile['ParentList']), str(mergedFile))
logging.debug(msg)
try:
self.dbs.insertMergedFile(mergedFile['ParentList'],
mergedFile)
except DbsException as ex:
msg = "Error in DBSWriter.insertFiles\n"
msg += "Cannot insert merged file:\n"
msg += " %s\n" % mergedFile['LogicalFileName']
msg += "%s\n" % formatEx(ex)
raise DBSWriterError(msg)
logging.debug(
"Inserted merged file: %s to FileBlock: %s" %
(mergedFile['LogicalFileName'], fileBlock['Name']))
else:
# //
# // Processing files
#//
affectedBlocks.add(fileBlock['Name'])
msg = "calling: self.dbs.insertFiles(%s, %s, %s)" % (
str(procDataset), str(list(fileList)), str(fileBlock))
logging.debug(msg)
try:
self.dbs.insertFiles(procDataset, list(fileList),
fileBlock)
except DbsException as ex:
msg = "Error in DBSWriter.insertFiles\n"
msg += "Cannot insert processed files:\n"
msg += " %s\n" % ([x['LogicalFileName']
for x in fileList], )
msg += "%s\n" % formatEx(ex)
raise DBSWriterError(msg)
logging.debug("Inserted files: %s to FileBlock: %s" %
(([x['LogicalFileName']
for x in fileList]), fileBlock['Name']))
return list(affectedBlocks)
def manageFileBlock(self,
fileBlock,
maxFiles=100,
maxSize=None,
timeOut=None,
algos=[],
filesToCommit=[],
procDataset=None):
"""
_manageFileBlock_
Check to see wether the fileblock with the provided name
is closeable based on number of files or total size.
If the block equals or exceeds wither the maxFiles or maxSize
parameters, close the block and return True, else do nothing and
return False
"""
# //
# // Check that the block exists, and is open before we close it
#//
fileblockName = fileBlock['Name']
blockInstance = self.dbs.listBlocks(block_name=fileblockName)
if len(blockInstance) > 1:
msg = "Multiple Blocks matching name: %s\n" % fileblockName
msg += "Unable to manage file block..."
raise DBSWriterError(msg)
if len(blockInstance) == 0:
msg = "Block name %s not found\n" % fileblockName
msg += "Cant manage a non-existent fileblock"
raise DBSWriterError(msg)
blockInstance = blockInstance[0]
isClosed = blockInstance.get('OpenForWriting', '1')
if isClosed != '1':
msg = "Block %s already closed" % fileblockName
logging.warning(msg)
# Now we need to commit files
if len(filesToCommit) > 0:
try:
self.dbs.insertFiles(procDataset, filesToCommit, fileBlock)
filesToCommit = []
except DbsException as ex:
msg = "Error in DBSWriter.insertFiles\n"
msg += "Cannot insert processed files:\n"
raise DBSWriterError(msg)
# Attempting to migrate to global
if self.globalDBSUrl:
self.dbs.dbsMigrateBlock(
srcURL=self.args['url'],
dstURL=self.globalDBSUrl,
block_name=fileblockName,
srcVersion=self.version,
dstVersion=self.globalVersion,
)
#for algo in algos:
# self.globalDBS.insertAlgoInPD(dataset = get_path(fileblockName.split('#')[0]),
# algorithm = algo)
logging.info(
"Migrated block %s to global due to pre-closed status" %
(fileblockName))
else:
logging.error(
"Should've migrated block %s because it was already closed, but didn't"
% (fileblockName))
return True
# //
# // We have an open block, sum number of files and file sizes
#//
#fileCount = int(blockInstance.get('NumberOfFiles', 0))
fileCount = len(fileBlock['files'])
totalSize = float(blockInstance.get('BlockSize', 0))
msg = "Fileblock: %s\n ==> Size: %s Files: %s\n" % (
fileblockName, totalSize, fileCount)
logging.warning(msg)
# //
# // Test close block conditions
#//
closeBlock = False
if timeOut:
if int(time.time()) - int(blockInstance['CreationDate']) > timeOut:
closeBlock = True
msg = "Closing Block based on timeOut: %s" % fileblockName
logging.debug(msg)
if fileCount >= maxFiles:
closeBlock = True
msg = "Closing Block Based on files: %s" % fileblockName
logging.debug(msg)
if maxSize != None:
if totalSize >= maxSize:
closeBlock = True
msg = "Closing Block Based on size: %s" % fileblockName
logging.debug(msg)
if closeBlock:
# Now we need to commit files
if len(filesToCommit) > 0:
try:
self.dbs.insertFiles(procDataset, filesToCommit, fileBlock)
filesToCommit = []
#logging.debug("Inserted files: %s to FileBlock: %s" \
# % ( ([ x['LogicalFileName'] for x in insertFiles ]),fileBlock['Name']))
except DbsException as ex:
msg = "Error in DBSWriter.insertFiles\n"
msg += "Cannot insert processed files:\n"
#msg += " %s\n" % ([ x['LogicalFileName'] for x in insertFiles ],)
msg += "%s\n" % formatEx(ex)
raise DBSWriterError(msg)
# //
# // Close the block
#//
self.dbs.closeBlock(
DBSWriterObjects.createDBSFileBlock(fileblockName))
if self.globalDBSUrl:
self.dbs.dbsMigrateBlock(srcURL=self.args['url'],
dstURL=self.globalDBSUrl,
block_name=fileblockName,
srcVersion=self.version,
dstVersion=self.globalVersion)
for algo in algos:
pass
#self.globalDBS.insertAlgoInPD(dataset = get_path(fileblockName.split('#')[0]),
# algorithm = algo)
logging.info("Migrated block %s to global" % (fileblockName))
else:
logging.error("Should've migrated block %s, but didn't" %
(fileblockName))
return closeBlock
def migrateDatasetBlocks(self, inputDBSUrl, datasetPath, blocks):
"""
_migrateDatasetBlocks_
Migrate the list of fileblocks provided by blocks, belonging
to the dataset specified by the dataset path to this DBS instance
from the inputDBSUrl provided
- *inputDBSUrl* : URL for connection to input DBS
- *datasetPath* : Name of dataset in input DBS (must exist in input
DBS)
- *blocks* : list of block names to be migrated (must exist
in input DBS)
"""
if len(blocks) == 0:
msg = "FileBlocks not provided.\n"
msg += "You must provide the name of at least one fileblock\n"
msg += "to be migrated"
raise DBSWriterError(msg)
# //
# // Hook onto input DBSUrl and verify that the dataset & blocks
#// exist
reader = DBSReader(inputDBSUrl)
inputBlocks = reader.listFileBlocks(datasetPath)
for block in blocks:
# //
# // Test block exists at source
#//
if block not in inputBlocks:
msg = "Block name:\n ==> %s\n" % block
msg += "Not found in input dataset:\n ==> %s\n" % datasetPath
msg += "In DBS Instance:\n ==> %s\n" % inputDBSUrl
raise DBSWriterError(msg)
# //
# // Test block does not exist in target
#//
if self.reader.blockExists(block):
# //
# // block exists
#// If block is closed dont attempt transfer
if not self.reader.blockIsOpen(block):
msg = "Block already exists in target DBS and is closed:\n"
msg += " ==> %s\n" % block
msg += "Skipping Migration of that block"
logging.warning(msg)
continue
try:
xferData = reader.dbs.listDatasetContents(datasetPath, block)
except DbsException as ex:
msg = "Error in DBSWriter.migrateDatasetBlocks\n"
msg += "Could not read content of dataset:\n ==> %s\n" % (
datasetPath, )
msg += "Block name:\n ==> %s\n" % block
msg += "%s\n" % formatEx(ex)
raise DBSWriterError(msg)
xferData = _remapBlockParentage(datasetPath, xferData)
try:
self.dbs.insertDatasetContents(xferData)
except DbsException as ex:
msg = "Error in DBSWriter.migrateDatasetBlocks\n"
msg += "Could not write content of dataset:\n ==> %s\n" % (
datasetPath, )
msg += "Block name:\n ==> %s\n" % block
msg += "%s\n" % formatEx(ex)
raise DBSWriterError(msg)
del xferData
return
def importDatasetWithExistingParents(self,
sourceDBS,
sourceDatasetPath,
targetDBS,
onlyClosed=True):
"""
_importDataset_
Import a dataset into the local scope DBS.
It complains if the parent dataset ar not there!!
- *sourceDBS* : URL for input DBS instance
- *sourceDatasetPath* : Dataset Path to be imported
- *targetDBS* : URL for DBS to have dataset imported to
"""
reader = DBSReader(sourceDBS)
inputBlocks = reader.getFileBlocksInfo(sourceDatasetPath,
onlyClosed,
locations=False)
for inputBlock in inputBlocks:
block = inputBlock['Name']
# //
# // Test block does not exist in target
#//
if self.reader.blockExists(block):
# //
# // block exists
#// If block is closed dont attempt transfer
if not str(inputBlock['OpenForWriting']) != '1':
msg = "Block already exists in target DBS and is closed:\n"
msg += " ==> %s\n" % block
msg += "Skipping Import of that block"
logging.warning(msg)
locations = reader.listFileBlockLocation(block)
# only empty file blocks can have no location
if not locations and str(
inputBlock['NumberOfFiles']) != "0":
msg = "Error in DBSWriter.importDatasetWithExistingParents\n"
msg += "Block has no locations defined: %s" % block
raise DBSWriterError(msg)
logging.info("Update block locations to:")
for sename in locations:
self.dbs.addReplicaToBlock(block, sename)
logging.info(sename)
continue
try:
xferData = reader.dbs.listDatasetContents(
sourceDatasetPath, block)
except DbsException as ex:
msg = "Error in DBSWriter.importDatasetWithExistingParents\n"
msg += "Could not read content of dataset:\n ==> %s\n" % (
sourceDatasetPath, )
msg += "Block name:\n ==> %s\n" % block
msg += "%s\n" % formatEx(ex)
raise DBSWriterError(msg)
try:
self.dbs.insertDatasetContents(xferData)
except DbsException as ex:
msg = "Error in DBSWriter.importDatasetWithExistingParents\n"
msg += "Could not write content of dataset:\n ==> %s\n" % (
sourceDatasetPath, )
msg += "Block name:\n ==> %s\n" % block
msg += "%s\n" % formatEx(ex)
raise DBSWriterError(msg)
del xferData
locations = reader.listFileBlockLocation(block)
# only empty file blocks can have no location
if not locations and str(inputBlock['NumberOfFiles']) != "0":
msg = "Error in DBSWriter.importDatasetWithExistingParents\n"
msg += "Block has no locations defined: %s" % block
raise DBSWriterError(msg)
for sename in locations:
self.dbs.addReplicaToBlock(block, sename)
return
def importDataset(self,
sourceDBS,
sourceDatasetPath,
targetDBS,
onlyClosed=True):
"""
_importDataset_
Import a dataset into the local scope DBS with full parentage hirerarchy
(at least not slow because branches info is dropped)
- *sourceDBS* : URL for input DBS instance
- *sourceDatasetPath* : Dataset Path to be imported
- *targetDBS* : URL for DBS to have dataset imported to
"""
reader = DBSReader(sourceDBS)
inputBlocks = reader.getFileBlocksInfo(sourceDatasetPath,
onlyClosed,
locations=False)
blkCounter = 0
for inputBlock in inputBlocks:
block = inputBlock['Name']
# //
# // Test block does not exist in target
#//
blkCounter = blkCounter + 1
msg = "Importing block %s of %s: %s " % (blkCounter,
len(inputBlocks), block)
logging.debug(msg)
if self.reader.blockExists(block):
# //
# // block exists
#// If block is closed dont attempt transfer
if str(inputBlock['OpenForWriting']) != '1':
msg = "Block already exists in target DBS and is closed:\n"
msg += " ==> %s\n" % block
msg += "Skipping Import of that block"
logging.warning(msg)
locations = reader.listFileBlockLocation(block)
# only empty file blocks can have no location
if not locations and str(
inputBlock['NumberOfFiles']) != "0":
msg = "Error in DBSWriter.importDataset\n"
msg += "Block has no locations defined: %s" % block
raise DBSWriterError(msg)
logging.info("Update block locations to:")
for sename in locations:
self.dbs.addReplicaToBlock(block, sename)
logging.info(sename)
continue
try:
self.dbs.migrateDatasetContents(sourceDBS,
targetDBS,
sourceDatasetPath,
block_name=block,
noParentsReadOnly=False)
except DbsException as ex:
msg = "Error in DBSWriter.importDataset\n"
msg += "Could not write content of dataset:\n ==> %s\n" % (
sourceDatasetPath, )
msg += "Block name:\n ==> %s\n" % block
msg += "%s\n" % formatEx(ex)
raise DBSWriterError(msg)
locations = reader.listFileBlockLocation(block)
# only empty file blocks can have no location
if not locations and str(inputBlock['NumberOfFiles']) != "0":
msg = "Error in DBSWriter.importDataset\n"
msg += "Block has no locations defined: %s" % block
raise DBSWriterError(msg)
for sename in locations:
self.dbs.addReplicaToBlock(block, sename)
return
def importDatasetWithoutParentage(self,
sourceDBS,
sourceDatasetPath,
targetDBS,
onlyClosed=True):
"""
_importDataset_
Import a dataset into the local scope DBS with one level parentage,
however it has severe limitation on its use due to the "ReadOnly" concept.
- *sourceDBS* : URL for input DBS instance
- *sourceDatasetPath* : Dataset Path to be imported
- *targetDBS* : URL for DBS to have dataset imported to
"""
reader = DBSReader(sourceDBS)
inputBlocks = reader.getFileBlocksInfo(sourceDatasetPath,
onlyClosed,
locations=False)
for inputBlock in inputBlocks:
block = inputBlock['Name']
# //
# // Test block does not exist in target
#//
if self.reader.blockExists(block):
# //
# // block exists
#// If block is closed dont attempt transfer
if str(inputBlock['OpenForWriting']) != '1':
msg = "Block already exists in target DBS and is closed:\n"
msg += " ==> %s\n" % block
msg += "Skipping Import of that block"
logging.warning(msg)
locations = reader.listFileBlockLocation(block)
# only empty file blocks can have no location
if not locations and str(
inputBlock['NumberOfFiles']) != "0":
msg = "Error in DBSWriter.importDatasetWithoutParentage\n"
msg += "Block has no locations defined: %s" % block
raise DBSWriterError(msg)
logging.info("Update block locations to:")
for sename in locations:
self.dbs.addReplicaToBlock(block, sename)
logging.info(sename)
continue
try:
self.dbs.migrateDatasetContents(sourceDBS,
targetDBS,
sourceDatasetPath,
block_name=block,
noParentsReadOnly=True)
except DbsException as ex:
msg = "Error in DBSWriter.importDatasetWithoutParentage\n"
msg += "Could not write content of dataset:\n ==> %s\n" % (
sourceDatasetPath, )
msg += "Block name:\n ==> %s\n" % block
msg += "%s\n" % formatEx(ex)
raise DBSWriterError(msg)
locations = reader.listFileBlockLocation(block)
# only empty file blocks can have no location
if not locations and str(inputBlock['NumberOfFiles']) != "0":
msg = "Error in DBSWriter.importDatasetWithoutParentage\n"
msg += "Block has no locations defined: %s" % block
raise DBSWriterError(msg)
for sename in locations:
self.dbs.addReplicaToBlock(block, sename)
return
def getOutputDatasetsWithPSet(payloadNode):
"""
_getOutputDatasetsWithPSet_
Extract all the information about output datasets from the
payloadNode object provided, including the {{}} format PSet cfg
Returns a list of DatasetInfo objects including App details
from the node.
"""
result = []
for item in payloadNode._OutputDatasets:
resultEntry = DatasetInfo()
resultEntry.update(item)
resultEntry["ApplicationName"] = payloadNode.application[
'Executable']
resultEntry["ApplicationProject"] = payloadNode.application[
'Project']
resultEntry["ApplicationVersion"] = payloadNode.application[
'Version']
resultEntry["ApplicationFamily"] = item.get(
"OutputModuleName", "AppFamily")
try:
config = payloadNode.cfgInterface
psetStr = config.originalContent()
resultEntry['PSetContent'] = psetStr
except Exception as ex:
resultEntry['PSetContent'] = None
result.append(resultEntry)
return _sortDatasets(result)
class DBSWriter:
"""
_DBSWriter_
General API for writing data to DBS
"""
def __init__(self, url, **contact):
args = { "url" : url, "level" : 'ERROR'}
args.update(contact)
try:
self.dbs = DbsApi(args)
self.args = args
self.version = args.get('version', None)
self.globalDBSUrl = args.get('globalDBSUrl', None)
self.globalVersion = args.get('globalVersion', None)
if self.globalDBSUrl:
globalArgs = {'url': url, 'level': 'ERROR'}
globalArgs.update(contact)
self.globalDBS = DbsApi(globalArgs)
except DbsException as ex:
msg = "Error in DBSWriterError with DbsApi\n"
msg += "%s\n" % formatEx(ex)
raise DBSWriterError(msg)
self.reader = DBSReader(**args)
def createDatasets(self, workflowSpec):
"""
_createDatasets_
Create All the output datasets found in the workflow spec instance
provided
"""
try:
workflowSpec.payload.operate(
_CreateDatasetOperator(self.dbs, workflowSpec)
)
except DbsException as ex:
msg = "Error in DBSWriter.createDatasets\n"
msg += "For Workflow: %s\n" % workflowSpec.workflowName()
msg += "%s\n" % formatEx(ex)
raise DBSWriterError(msg)
return
def insertFilesForDBSBuffer(self, files, procDataset, algos,
jobType = "NotMerge", insertDetectorData = False,
maxFiles = 100, maxSize = 99999999, timeOut = None,
fileCommitLength = 5):
"""
_insertFiles_
list of files inserted in DBS
"""
#TODO: Whats the purpose of insertDetectorData
if len(files) < 1:
return
affectedBlocks = []
insertFiles = []
addedRuns=[]
pnn = None
#Get the algos in insertable form
# logging.error("About to input algos")
# logging.error(algos)
ialgos = [DBSWriterObjects.createAlgorithmForInsert(dict(algo)) for algo in algos ]
#print ialgos
for outFile in files:
# //
# // Convert each file into a DBS File object
#//
lumiList = []
#Somehing similar should be the real deal when multiple runs/lumi could be returned from wmbs file
for runlumiinfo in outFile.getRuns():
lrun=long(runlumiinfo.run)
run = DbsRun(
RunNumber = lrun,
NumberOfEvents = 0,
NumberOfLumiSections = 0,
TotalLuminosity = 0,
StoreNumber = 0,
StartOfRun = 0,
EndOfRun = 0,
)
#Only added if not added by another file in this loop, why waste a call to DBS
if lrun not in addedRuns:
self.dbs.insertRun(run)
addedRuns.append(lrun) #save it so we do not try to add it again to DBS
logging.debug("run %s added to DBS " % str(lrun))
for alsn in runlumiinfo:
lumi = DbsLumiSection(
LumiSectionNumber = long(alsn),
StartEventNumber = 0,
EndEventNumber = 0,
LumiStartTime = 0,
LumiEndTime = 0,
RunNumber = lrun,
)
lumiList.append(lumi)
logging.debug("lumi list created for the file")
dbsfile = DbsFile(
#Checksum = str(outFile['cksum']),
NumberOfEvents = outFile['events'],
LogicalFileName = outFile['lfn'],
FileSize = int(outFile['size']),
Status = "VALID",
ValidationStatus = 'VALID',
FileType = 'EDM',
Dataset = procDataset,
TierList = DBSWriterObjects.makeTierList(procDataset['Path'].split('/')[3]),
AlgoList = ialgos,
LumiList = lumiList,
ParentList = outFile.getParentLFNs(),
#BranchHash = outFile['BranchHash'],
)
#Set checksums by hand
#dbsfile['Checksum'] = 0 #Set a default?
for entry in outFile['checksums'].keys():
#This should be a dictionary with a cktype key and cksum value
if entry.lower() == 'cksum':
dbsfile['Checksum'] = str(outFile['checksums'][entry])
elif entry.lower() == 'adler32':
dbsfile['Adler32'] = str(outFile['checksums'][entry])
elif entry.lower() == 'md5':
dbsfile['Md5'] = str(outFile['checksums'][entry])
#This check comes from ProdAgent, not sure if its required
if len(outFile["locations"]) > 0:
pnn = list(outFile["locations"])[0]
logging.debug("PNN associated to file is: %s"%pnn)
else:
msg = "Error in DBSWriter.insertFiles\n"
msg += "No PNN associated to file"
#print "FAKING seName for now"
#seName="cmssrm.fnal.gov"
raise DBSWriterError(msg)
insertFiles.append(dbsfile)
# //Processing Jobs:
# // Insert the lists of sorted files into the appropriate
#// fileblocks
sumSize = 0
sumFiles = 0
tmpFiles = []
blockList = []
#First, get the block. See if the block already exists
try:
fileBlock = DBSWriterObjects.getDBSFileBlock(
self.dbs,
procDataset,
pnn)
fileBlock['files'] = []
#if not fileBlock in affectedBlocks:
# affectedBlocks.append(fileBlock)
except DbsException as ex:
msg = "Error in DBSWriter.insertFilesForDBSBuffer\n"
msg += "Cannot retrieve FileBlock for dataset:\n"
msg += " %s\n" % procDataset['Path']
msg += "%s\n" % formatEx(ex)
raise DBSWriterError(msg)
filesToCommit = []
for file in insertFiles:
# First see if the block is full
if self.manageFileBlock(fileBlock = fileBlock, maxFiles = maxFiles,
maxSize = maxSize, timeOut = timeOut, algos = ialgos,
filesToCommit = filesToCommit, procDataset = procDataset):
fileBlock['OpenForWriting'] = 0
if not fileBlock in affectedBlocks:
affectedBlocks.append(fileBlock)
# Then we need a new block
try:
fileBlock = DBSWriterObjects.getDBSFileBlock(
self.dbs,
procDataset,
pnn)
fileBlock['files'] = []
except DbsException as ex:
msg = "Error in DBSWriter.insertFilesForDBSBuffer\n"
msg += "Cannot retrieve FileBlock for dataset:\n"
msg += " %s\n" % procDataset['Path']
msg += "%s\n" % formatEx(ex)
raise DBSWriterError(msg)
fileBlock['files'].append(file['LogicalFileName'])
filesToCommit.append(file)
if len(filesToCommit) >= fileCommitLength:
# Only commit the files if there are more of them then the maximum length
try:
self.dbs.insertFiles(procDataset, filesToCommit, fileBlock)
filesToCommit = []
logging.debug("Inserted files: %s to FileBlock: %s" \
% ( ([ x['LogicalFileName'] for x in insertFiles ]),fileBlock['Name']))
except DbsException as ex:
msg = "Error in DBSWriter.insertFiles\n"
msg += "Cannot insert processed files:\n"
msg += " %s\n" % ([ x['LogicalFileName'] for x in insertFiles ],)
msg += "%s\n" % formatEx(ex)
raise DBSWriterError(msg)
if len(filesToCommit) > 0:
try:
self.dbs.insertFiles(procDataset, filesToCommit, fileBlock)
filesToCommit = []
logging.debug("Inserted files: %s to FileBlock: %s" \
% ( ([ x['LogicalFileName'] for x in insertFiles ]),fileBlock['Name']))
except DbsException as ex:
msg = "Error in DBSWriter.insertFiles\n"
msg += "Cannot insert processed files:\n"
msg += " %s\n" % ([ x['LogicalFileName'] for x in insertFiles ],)
msg += "%s\n" % formatEx(ex)
raise DBSWriterError(msg)
if not fileBlock in affectedBlocks:
affectedBlocks.append(fileBlock)
## Do bulk inserts now for DBS
#filesToCommit = []
#count = 0
#count2 = 0
#for file in insertFiles:
# count += 1
# #Try and close the box
# logging.error("Should have a file")
# logging.error(len(filesToCommit))
# count2 += len(filesToCommit)
# if self.manageFileBlock(fileBlock = fileBlock, maxFiles = maxFiles,
# maxSize = maxSize, timeOut = timeOut, algos = ialgos,
# filesToCommit = filesToCommit, procDataset = procDataset):
# fileBlock['OpenForWriting'] = '0'
# if not fileBlock in affectedBlocks:
# affectedBlocks.append(fileBlock)
#
#
#
# # Then we need a new block
# try:
# fileBlock = DBSWriterObjects.getDBSFileBlock(
# self.dbs,
# procDataset,
# seName)
# fileBlock['files'] = []
# except DbsException, ex:
# msg = "Error in DBSWriter.insertFilesForDBSBuffer\n"
# msg += "Cannot retrieve FileBlock for dataset:\n"
# msg += " %s\n" % procDataset['Path']
# msg += "%s\n" % formatEx(ex)
# raise DBSWriterError(msg)
# #At this point, we should commit the block as is
# fileBlock['files'].append(file['LogicalFileName'])
# if jobType == "MergeSpecial":
# for file in fileList:
# file['Block'] = fileBlock
# msg="calling: self.dbs.insertMergedFile(%s, %s)" % (str(file['ParentList']),str(file))
# logging.debug(msg)
# try:
# #
# #
# # NOTE To Anzar From Anzar (File cloning as in DBS API can be done here and then I can use Bulk insert on Merged files as well)
# self.dbs.insertMergedFile(file['ParentList'],
# file)
#
# except DbsException, ex:
# msg = "Error in DBSWriter.insertFiles\n"
# msg += "Cannot insert merged file:\n"
# msg += " %s\n" % file['LogicalFileName']
# msg += "%s\n" % formatEx(ex)
# raise DBSWriterError(msg)
# logging.debug("Inserted merged file: %s to FileBlock: %s"%(file['LogicalFileName'],fileBlock['Name']))
# else:
# filesToCommit.append(file)
# if len(filesToCommit) >= fileCommitLength:
# # Only commit the files if there are more of them then the maximum length
# try:
# logging.error("About to commit %i files" %(len(filesToCommit)))
# count2 += len(filesToCommit)
# self.dbs.insertFiles(procDataset, filesToCommit, fileBlock)
# filesToCommit = []
# logging.debug("Inserted files: %s to FileBlock: %s" \
# % ( ([ x['LogicalFileName'] for x in insertFiles ]),fileBlock['Name']))
#
# except DbsException, ex:
# msg = "Error in DBSWriter.insertFiles\n"
# msg += "Cannot insert processed files:\n"
# msg += " %s\n" % ([ x['LogicalFileName'] for x in insertFiles ],)
# msg += "%s\n" % formatEx(ex)
# raise DBSWriterError(msg)
#
#
#
#
## If we still have files to commit, commit them
#logging.error("Got to the end of the loop")
#logging.error(len(filesToCommit))
#logging.error(count2)
#if len(filesToCommit) > 0:
# try:
# logging.error("About to insert some files")
# self.dbs.insertFiles(procDataset, filesToCommit, fileBlock)
# filesToCommit = []
# logging.debug("Inserted files: %s to FileBlock: %s" \
# % ( ([ x['LogicalFileName'] for x in insertFiles ]),fileBlock['Name']))
#
# except DbsException, ex:
# msg = "Error in DBSWriter.insertFiles\n"
# msg += "Cannot insert processed files:\n"
# msg += " %s\n" % ([ x['LogicalFileName'] for x in insertFiles ],)
# msg += "%s\n" % formatEx(ex)
# raise DBSWriterError(msg)
if not fileBlock in affectedBlocks:
affectedBlocks.append(fileBlock)
return list(affectedBlocks)
def insertFiles(self, fwkJobRep, insertDetectorData = False):
"""
_insertFiles_
Process the files in the FwkJobReport instance and insert
them into the associated datasets
A list of affected fileblock names is returned both for merged
and unmerged fileblocks. Only merged blocks will have to be managed.
#for merged file
#blocks to facilitate management of those blocks.
#This list is not populated for processing jobs since we dont really
#care about the processing job blocks.
"""
insertLists = {}
orderedHashes = []
affectedBlocks = set()
if len(fwkJobRep.files)<=0:
msg = "Error in DBSWriter.insertFiles\n"
msg += "No files found in FrameWorkJobReport for:\n"
msg += "==> JobSpecId: %s"%fwkJobRep.jobSpecId
msg += " Workflow: %s"%fwkJobRep.workflowSpecId
raise DBSWriterError(msg)
for outFile in fwkJobRep.sortFiles():
# //
# // Convert each file into a DBS File object
#//
pnn = None
if "PNN" in outFile:
if outFile['PNN'] :
pnn = outFile['PNN']
logging.debug("PNN associated to file is: %s"%pnn)
## remove the fallback to site se-name if no SE is associated to File
## because it's likely that there is some stage out problem if there
## is no SEName associated to the file.
# if not seName:
# if fwkJobRep.siteDetails.has_key("se-name"):
# seName = fwkJobRep.siteDetails['se-name']
# seName = str(seName)
# logging.debug("site SEname: %s"%seName)
if not pnn:
msg = "Error in DBSWriter.insertFiles\n"
msg += "No PNN associated to files in FrameWorkJobReport for "
# msg += "No SEname found in FrameWorkJobReport for "
msg += "==> JobSpecId: %s"%fwkJobRep.jobSpecId
msg += " Workflow: %s"%fwkJobRep.workflowSpecId
raise DBSWriterError(msg)
try:
if ( insertDetectorData ):
dbsFiles = DBSWriterObjects.createDBSFiles(outFile,
fwkJobRep.jobType,
self.dbs)
else:
dbsFiles = DBSWriterObjects.createDBSFiles(outFile,
fwkJobRep.jobType)
except DbsException as ex:
msg = "Error in DBSWriter.insertFiles:\n"
msg += "Error creating DbsFile instances for file:\n"
msg += "%s\n" % outFile['LFN']
msg += "%s\n" % formatEx(ex)
raise DBSWriterError(msg)
if len(dbsFiles)<=0:
msg="No DbsFile instances created. Not enough info in the FrameWorkJobReport for"
msg += "==> JobSpecId: %s"%fwkJobRep.jobSpecId
msg += " Workflow: %s"%fwkJobRep.workflowSpecId
raise DBSWriterError(msg)
for f in dbsFiles:
datasetName = makeDBSDSName(f)
hashName = "%s-%s" % (pnn, datasetName)
if hashName not in insertLists:
insertLists[hashName] = _InsertFileList(pnn,
datasetName)
insertLists[hashName].append(f)
if not orderedHashes.count(hashName):
orderedHashes.append(hashName)
# //Processing Jobs:
# // Insert the lists of sorted files into the appropriate
#// fileblocks
for hash in orderedHashes:
fileList = insertLists[hash]
procDataset = fileList[0]['Dataset']
try:
fileBlock = DBSWriterObjects.getDBSFileBlock(
self.dbs,
procDataset,
fileList.pnn)
except DbsException as ex:
msg = "Error in DBSWriter.insertFiles\n"
msg += "Cannot retrieve FileBlock for dataset:\n"
msg += " %s\n" % procDataset
# msg += "In Storage Element:\n %s\n" % fileList.seName
msg += "In PNN:\n %s\n" % fileList.pnn
msg += "%s\n" % formatEx(ex)
raise DBSWriterError(msg)
if fwkJobRep.jobType == "Merge":
# //
# // Merge files
#//
for mergedFile in fileList:
mergedFile['Block'] = fileBlock
affectedBlocks.add(fileBlock['Name'])
msg="calling: self.dbs.insertMergedFile(%s, %s)" % (str(mergedFile['ParentList']),str(mergedFile))
logging.debug(msg)
try:
self.dbs.insertMergedFile(mergedFile['ParentList'],
mergedFile)
except DbsException as ex:
msg = "Error in DBSWriter.insertFiles\n"
msg += "Cannot insert merged file:\n"
msg += " %s\n" % mergedFile['LogicalFileName']
msg += "%s\n" % formatEx(ex)
raise DBSWriterError(msg)
logging.debug("Inserted merged file: %s to FileBlock: %s"%(mergedFile['LogicalFileName'],fileBlock['Name']))
else:
# //
# // Processing files
#//
affectedBlocks.add(fileBlock['Name'])
msg="calling: self.dbs.insertFiles(%s, %s, %s)" % (str(procDataset),str(list(fileList)),str(fileBlock))
logging.debug(msg)
try:
self.dbs.insertFiles(procDataset, list(fileList),
fileBlock)
except DbsException as ex:
msg = "Error in DBSWriter.insertFiles\n"
msg += "Cannot insert processed files:\n"
msg += " %s\n" % (
[ x['LogicalFileName'] for x in fileList ],
)
msg += "%s\n" % formatEx(ex)
raise DBSWriterError(msg)
logging.debug("Inserted files: %s to FileBlock: %s"%( ([ x['LogicalFileName'] for x in fileList ]),fileBlock['Name']))
return list(affectedBlocks)
def manageFileBlock(self, fileBlock, maxFiles = 100, maxSize = None,
timeOut = None, algos = [], filesToCommit = [],
procDataset = None):
"""
_manageFileBlock_
Check to see wether the fileblock with the provided name
is closeable based on number of files or total size.
If the block equals or exceeds wither the maxFiles or maxSize
parameters, close the block and return True, else do nothing and
return False
"""
# //
# // Check that the block exists, and is open before we close it
#//
fileblockName = fileBlock['Name']
blockInstance = self.dbs.listBlocks(block_name=fileblockName)
if len(blockInstance) > 1:
msg = "Multiple Blocks matching name: %s\n" % fileblockName
msg += "Unable to manage file block..."
raise DBSWriterError(msg)
if len(blockInstance) == 0:
msg = "Block name %s not found\n" % fileblockName
msg += "Cant manage a non-existent fileblock"
raise DBSWriterError(msg)
blockInstance = blockInstance[0]
isClosed = blockInstance.get('OpenForWriting', '1')
if isClosed != '1':
msg = "Block %s already closed" % fileblockName
logging.warning(msg)
# Now we need to commit files
if len(filesToCommit) > 0:
try:
self.dbs.insertFiles(procDataset, filesToCommit, fileBlock)
filesToCommit = []
except DbsException as ex:
msg = "Error in DBSWriter.insertFiles\n"
msg += "Cannot insert processed files:\n"
raise DBSWriterError(msg)
# Attempting to migrate to global
if self.globalDBSUrl:
self.dbs.dbsMigrateBlock(srcURL = self.args['url'],
dstURL = self.globalDBSUrl,
block_name = fileblockName,
srcVersion = self.version,
dstVersion = self.globalVersion,
)
#for algo in algos:
# self.globalDBS.insertAlgoInPD(dataset = get_path(fileblockName.split('#')[0]),
# algorithm = algo)
logging.info("Migrated block %s to global due to pre-closed status" %(fileblockName))
else:
logging.error("Should've migrated block %s because it was already closed, but didn't" % (fileblockName))
return True
# //
# // We have an open block, sum number of files and file sizes
#//
#fileCount = int(blockInstance.get('NumberOfFiles', 0))
fileCount = len(fileBlock['files'])
totalSize = float(blockInstance.get('BlockSize', 0))
msg = "Fileblock: %s\n ==> Size: %s Files: %s\n" % (
fileblockName, totalSize, fileCount)
logging.warning(msg)
# //
# // Test close block conditions
#//
closeBlock = False
if timeOut:
if int(time.time()) - int(blockInstance['CreationDate']) > timeOut:
closeBlock = True
msg = "Closing Block based on timeOut: %s" % fileblockName
logging.debug(msg)
if fileCount >= maxFiles:
closeBlock = True
msg = "Closing Block Based on files: %s" % fileblockName
logging.debug(msg)
if maxSize != None:
if totalSize >= maxSize:
closeBlock = True
msg = "Closing Block Based on size: %s" % fileblockName
logging.debug(msg)
if closeBlock:
# Now we need to commit files
if len(filesToCommit) > 0:
try:
self.dbs.insertFiles(procDataset, filesToCommit, fileBlock)
filesToCommit = []
#logging.debug("Inserted files: %s to FileBlock: %s" \
# % ( ([ x['LogicalFileName'] for x in insertFiles ]),fileBlock['Name']))
except DbsException as ex:
msg = "Error in DBSWriter.insertFiles\n"
msg += "Cannot insert processed files:\n"
#msg += " %s\n" % ([ x['LogicalFileName'] for x in insertFiles ],)
msg += "%s\n" % formatEx(ex)
raise DBSWriterError(msg)
# //
# // Close the block
#//
self.dbs.closeBlock(
DBSWriterObjects.createDBSFileBlock(fileblockName)
)
if self.globalDBSUrl:
self.dbs.dbsMigrateBlock(srcURL = self.args['url'],
dstURL = self.globalDBSUrl,
block_name = fileblockName,
srcVersion = self.version,
dstVersion = self.globalVersion
)
for algo in algos:
pass
#self.globalDBS.insertAlgoInPD(dataset = get_path(fileblockName.split('#')[0]),
# algorithm = algo)
logging.info("Migrated block %s to global" %(fileblockName))
else:
logging.error("Should've migrated block %s, but didn't" % (fileblockName))
return closeBlock
def migrateDatasetBlocks(self, inputDBSUrl, datasetPath, blocks):
"""
_migrateDatasetBlocks_
Migrate the list of fileblocks provided by blocks, belonging
to the dataset specified by the dataset path to this DBS instance
from the inputDBSUrl provided
- *inputDBSUrl* : URL for connection to input DBS
- *datasetPath* : Name of dataset in input DBS (must exist in input
DBS)
- *blocks* : list of block names to be migrated (must exist
in input DBS)
"""
if len(blocks) == 0:
msg = "FileBlocks not provided.\n"
msg += "You must provide the name of at least one fileblock\n"
msg += "to be migrated"
raise DBSWriterError(msg)
# //
# // Hook onto input DBSUrl and verify that the dataset & blocks
#// exist
reader = DBSReader(inputDBSUrl)
inputBlocks = reader.listFileBlocks(datasetPath)
for block in blocks:
# //
# // Test block exists at source
#//
if block not in inputBlocks:
msg = "Block name:\n ==> %s\n" % block
msg += "Not found in input dataset:\n ==> %s\n" % datasetPath
msg += "In DBS Instance:\n ==> %s\n" % inputDBSUrl
raise DBSWriterError(msg)
# //
# // Test block does not exist in target
#//
if self.reader.blockExists(block):
# //
# // block exists
#// If block is closed dont attempt transfer
if not self.reader.blockIsOpen(block):
msg = "Block already exists in target DBS and is closed:\n"
msg += " ==> %s\n" % block
msg += "Skipping Migration of that block"
logging.warning(msg)
continue
try:
xferData = reader.dbs.listDatasetContents(datasetPath, block)
except DbsException as ex:
msg = "Error in DBSWriter.migrateDatasetBlocks\n"
msg += "Could not read content of dataset:\n ==> %s\n" % (
datasetPath,)
msg += "Block name:\n ==> %s\n" % block
msg += "%s\n" % formatEx(ex)
raise DBSWriterError(msg)
xferData = _remapBlockParentage(datasetPath, xferData)
try:
self.dbs.insertDatasetContents(xferData)
except DbsException as ex:
msg = "Error in DBSWriter.migrateDatasetBlocks\n"
msg += "Could not write content of dataset:\n ==> %s\n" % (
datasetPath,)
msg += "Block name:\n ==> %s\n" % block
msg += "%s\n" % formatEx(ex)
raise DBSWriterError(msg)
del xferData
return
def importDatasetWithExistingParents(self, sourceDBS, sourceDatasetPath, targetDBS,
onlyClosed = True):
"""
_importDataset_
Import a dataset into the local scope DBS.
It complains if the parent dataset ar not there!!
- *sourceDBS* : URL for input DBS instance
- *sourceDatasetPath* : Dataset Path to be imported
- *targetDBS* : URL for DBS to have dataset imported to
"""
reader = DBSReader(sourceDBS)
inputBlocks = reader.getFileBlocksInfo(sourceDatasetPath, onlyClosed, locations = False)
for inputBlock in inputBlocks:
block = inputBlock['Name']
# //
# // Test block does not exist in target
#//
if self.reader.blockExists(block):
# //
# // block exists
#// If block is closed dont attempt transfer
if not str(inputBlock['OpenForWriting']) != '1':
msg = "Block already exists in target DBS and is closed:\n"
msg += " ==> %s\n" % block
msg += "Skipping Import of that block"
logging.warning(msg)
locations = reader.listFileBlockLocation(block)
# only empty file blocks can have no location
if not locations and str(inputBlock['NumberOfFiles']) != "0":
msg = "Error in DBSWriter.importDatasetWithExistingParents\n"
msg += "Block has no locations defined: %s" % block
raise DBSWriterError(msg)
logging.info("Update block locations to:")
for pnn in locations:
self.dbs.addReplicaToBlock(block,pnn)
logging.info(pnn)
continue
try:
xferData = reader.dbs.listDatasetContents(
sourceDatasetPath, block
)
except DbsException as ex:
msg = "Error in DBSWriter.importDatasetWithExistingParents\n"
msg += "Could not read content of dataset:\n ==> %s\n" % (
sourceDatasetPath,)
msg += "Block name:\n ==> %s\n" % block
msg += "%s\n" % formatEx(ex)
raise DBSWriterError(msg)
try:
self.dbs.insertDatasetContents(xferData)
except DbsException as ex:
msg = "Error in DBSWriter.importDatasetWithExistingParents\n"
msg += "Could not write content of dataset:\n ==> %s\n" % (
sourceDatasetPath,)
msg += "Block name:\n ==> %s\n" % block
msg += "%s\n" % formatEx(ex)
raise DBSWriterError(msg)
del xferData
locations = reader.listFileBlockLocation(block)
# only empty file blocks can have no location
if not locations and str(inputBlock['NumberOfFiles']) != "0":
msg = "Error in DBSWriter.importDatasetWithExistingParents\n"
msg += "Block has no locations defined: %s" % block
raise DBSWriterError(msg)
for pnn in locations:
self.dbs.addReplicaToBlock(block,pnn)
return
def importDataset(self, sourceDBS, sourceDatasetPath, targetDBS,
onlyClosed = True):
"""
_importDataset_
Import a dataset into the local scope DBS with full parentage hirerarchy
(at least not slow because branches info is dropped)
- *sourceDBS* : URL for input DBS instance
- *sourceDatasetPath* : Dataset Path to be imported
- *targetDBS* : URL for DBS to have dataset imported to
"""
reader = DBSReader(sourceDBS)
inputBlocks = reader.getFileBlocksInfo(sourceDatasetPath, onlyClosed, locations = False)
blkCounter=0
for inputBlock in inputBlocks:
block = inputBlock['Name']
# //
# // Test block does not exist in target
#//
blkCounter=blkCounter+1
msg="Importing block %s of %s: %s " % (blkCounter,len(inputBlocks),block)
logging.debug(msg)
if self.reader.blockExists(block):
# //
# // block exists
#// If block is closed dont attempt transfer
if str(inputBlock['OpenForWriting']) != '1':
msg = "Block already exists in target DBS and is closed:\n"
msg += " ==> %s\n" % block
msg += "Skipping Import of that block"
logging.warning(msg)
locations = reader.listFileBlockLocation(block)
# only empty file blocks can have no location
if not locations and str(inputBlock['NumberOfFiles']) != "0":
msg = "Error in DBSWriter.importDataset\n"
msg += "Block has no locations defined: %s" % block
raise DBSWriterError(msg)
logging.info("Update block locations to:")
for pnn in locations:
self.dbs.addReplicaToBlock(block,pnn)
logging.info(pnn)
continue
try:
self.dbs.migrateDatasetContents(sourceDBS, targetDBS, sourceDatasetPath, block_name=block, noParentsReadOnly = False)
except DbsException as ex:
msg = "Error in DBSWriter.importDataset\n"
msg += "Could not write content of dataset:\n ==> %s\n" % (
sourceDatasetPath,)
msg += "Block name:\n ==> %s\n" % block
msg += "%s\n" % formatEx(ex)
raise DBSWriterError(msg)
locations = reader.listFileBlockLocation(block)
# only empty file blocks can have no location
if not locations and str(inputBlock['NumberOfFiles']) != "0":
msg = "Error in DBSWriter.importDataset\n"
msg += "Block has no locations defined: %s" % block
raise DBSWriterError(msg)
for pnn in locations:
self.dbs.addReplicaToBlock(block,pnn)
return
def importDatasetWithoutParentage(self, sourceDBS, sourceDatasetPath, targetDBS,
onlyClosed = True):
"""
_importDataset_
Import a dataset into the local scope DBS with one level parentage,
however it has severe limitation on its use due to the "ReadOnly" concept.
- *sourceDBS* : URL for input DBS instance
- *sourceDatasetPath* : Dataset Path to be imported
- *targetDBS* : URL for DBS to have dataset imported to
"""
reader = DBSReader(sourceDBS)
inputBlocks = reader.getFileBlocksInfo(sourceDatasetPath, onlyClosed, locations = False)
for inputBlock in inputBlocks:
block = inputBlock['Name']
# //
# // Test block does not exist in target
#//
if self.reader.blockExists(block):
# //
# // block exists
#// If block is closed dont attempt transfer
if str(inputBlock['OpenForWriting']) != '1':
msg = "Block already exists in target DBS and is closed:\n"
msg += " ==> %s\n" % block
msg += "Skipping Import of that block"
logging.warning(msg)
locations = reader.listFileBlockLocation(block)
# only empty file blocks can have no location
if not locations and str(inputBlock['NumberOfFiles']) != "0":
msg = "Error in DBSWriter.importDatasetWithoutParentage\n"
msg += "Block has no locations defined: %s" % block
raise DBSWriterError(msg)
logging.info("Update block locations to:")
for pnn in locations:
self.dbs.addReplicaToBlock(block,pnn)
logging.info(pnn)
continue
try:
self.dbs.migrateDatasetContents(sourceDBS, targetDBS, sourceDatasetPath, block_name=block, noParentsReadOnly = True )
except DbsException as ex:
msg = "Error in DBSWriter.importDatasetWithoutParentage\n"
msg += "Could not write content of dataset:\n ==> %s\n" % (
sourceDatasetPath,)
msg += "Block name:\n ==> %s\n" % block
msg += "%s\n" % formatEx(ex)
raise DBSWriterError(msg)
locations = reader.listFileBlockLocation(block)
# only empty file blocks can have no location
if not locations and str(inputBlock['NumberOfFiles']) != "0":
msg = "Error in DBSWriter.importDatasetWithoutParentage\n"
msg += "Block has no locations defined: %s" % block
raise DBSWriterError(msg)
for pnn in locations:
self.dbs.addReplicaToBlock(block,pnn)
return
def getOutputDatasetsWithPSet(payloadNode):
"""
_getOutputDatasetsWithPSet_
Extract all the information about output datasets from the
payloadNode object provided, including the {{}} format PSet cfg
Returns a list of DatasetInfo objects including App details
from the node.
"""
result = []
for item in payloadNode._OutputDatasets:
resultEntry = DatasetInfo()
resultEntry.update(item)
resultEntry["ApplicationName"] = payloadNode.application['Executable']
resultEntry["ApplicationProject"] = payloadNode.application['Project']
resultEntry["ApplicationVersion"] = payloadNode.application['Version']
resultEntry["ApplicationFamily"] = item.get("OutputModuleName", "AppFamily")
try:
config = payloadNode.cfgInterface
psetStr = config.originalContent()
resultEntry['PSetContent'] = psetStr
except Exception as ex:
resultEntry['PSetContent'] = None
result.append(resultEntry)
return _sortDatasets(result)
from DBSAPI.dbsApiException import *
from DBSAPI.dbsOptions import DbsOptionParser
from DBSAPI.dbsApi import DbsApi
try:
optManager = DbsOptionParser()
(opts, args) = optManager.getOpt()
args = {}
# These dummy values are required to create the DbsApi object so let them here
args[
'url'] = 'http://cmsdbsprod.cern.ch/cms_dbs_prod_global/servlet/DBSServlet'
args['version'] = 'DBS_2_0_9'
args['mode'] = 'POST'
api = DbsApi(args)
if len(sys.argv) < 3:
print "USAGE: python %s <srcURL> <dstURL> block_name" % sys.argv[
0]
sys.exit(1)
srcURL = sys.argv[1]
dstURL = sys.argv[2]
block = sys.argv[3]
api.dbsMigrateBlock(srcURL, dstURL, block)
except DbsApiException, ex:
print "Caught API Exception %s: %s " % (ex.getClassName(),
ex.getErrorMessage())
if ex.getErrorCode() not in (None, ""):
print "DBS Exception Error Code: ", ex.getErrorCode()
print "Done"