def run(self):
log.info("Setting up directories for {}".format(self.swathDir))
swathDirName = os.path.basename(self.swathDir)
workspaceRoot = os.path.join(self.paths["processingDir"], swathDirName)
workingFileRoot = os.path.join(workspaceRoot, "working")
if not os.path.exists(workingFileRoot):
os.makedirs(workingFileRoot)
stateFileRoot = os.path.join(workspaceRoot, "state")
if not os.path.exists(stateFileRoot):
os.makedirs(stateFileRoot)
localTmpDir = os.path.join(workingFileRoot, "tmp")
if not os.path.exists(localTmpDir):
os.makedirs(localTmpDir)
outputFile = {
"swathDir": self.swathDir,
"workspaceRoot": workspaceRoot,
"workingFileRoot": workingFileRoot,
"stateFileRoot": stateFileRoot,
"localTmpDir": localTmpDir,
"demFilename": self.demFilename,
"arcsiReprojection": self.arcsiReprojection,
"outWktFilename": self.outWktFilename,
"projAbbv": self.projAbbv,
"metadataConfigFile": self.metadataConfigFile,
"metadataTemplate": self.metadataTemplate,
"maxCogProcesses": self.maxCogProcesses
}
with self.output().open("w") as outFile:
outFile.write(wc.getFormattedJson(outputFile))
def run(self):
getInputSwaths = {}
with self.input().open('r') as getInputSwathsInfo:
getInputSwaths = json.load(getInputSwathsInfo)
tasks = []
for swath in getInputSwaths["swaths"]:
task = SetupWorkDir(swathDir=swath["swathDir"],
paths=self.paths,
demFilename=self.demFilename,
arcsiReprojection=self.arcsiReprojection,
outWktFilename=self.outWktFilename,
projAbbv=self.projAbbv,
metadataConfigFile=self.metadataConfigFile,
metadataTemplate=self.metadataTemplate,
maxCogProcesses=self.maxCogProcesses)
tasks.append(task)
yield tasks
outputFile = {"swathSetups": []}
for task in tasks:
with task.output().open('r') as taskOutput:
swathSetup = json.load(taskOutput)
outputFile["swathSetups"].append(swathSetup)
with self.output().open("w") as outFile:
outFile.write(wc.getFormattedJson(outputFile))
def run(self):
basketDir = self.paths["basketDir"]
inputFiles = []
for inputFile in glob.glob(os.path.join(basketDir, "S1*")):
inputFiles.append(inputFile)
outputFile = {"basket": basketDir, "inputFiles": inputFiles}
with self.output().open("w") as outFile:
outFile.write(wc.getFormattedJson(outputFile))
def run(self):
basketDir = self.paths["basketDir"]
basketSubDirs = next(os.walk(basketDir))[1]
swaths = []
for subDir in basketSubDirs:
subDirProducts = []
swathDir = os.path.join(basketDir, subDir)
for product in glob.glob(os.path.join(swathDir, "S2*")):
subDirProducts.append(product)
if len(subDirProducts):
swath = {"swathDir": swathDir, "productPaths": subDirProducts}
swaths.append(swath)
outputFile = {"basket": basketDir, "swaths": swaths}
with self.output().open("w") as outFile:
outFile.write(wc.getFormattedJson(outputFile))
def run(self):
try:
outputFile = {
"productId": self.productName,
"sbatchScriptPath": self.sbatchScriptPath,
"jobId": None,
"submitTime": None
}
outputString = ""
if self.testProcessing:
randomJobId = random.randint(1000000, 9999999)
outputString = "JOBID USER STAT QUEUE FROM_HOST EXEC_HOST JOB_NAME SUBMIT_TIME"\
+str(randomJobId)+" test001 RUN short-serial jasmin-sci1 16*host290. my-job1 Nov 16 16:51"
else:
sbatchCmd = "sbatch {}".format(self.sbatchScriptPath)
log.info("Submitting job using command: %s", sbatchCmd)
output = subprocess.check_output(sbatchCmd,
stderr=subprocess.STDOUT,
shell=True)
outputString = output.decode("utf-8")
regex = '[0-9]{5,}' # job ID is at least 5 digits
match = re.search(regex, outputString)
self.jobId = match.group(0)
log.info(
"Successfully submitted lotus job <%s> for %s using sbatch script: %s",
self.jobId, self.productName, self.sbatchScriptPath)
outputFile["jobId"] = self.jobId
outputFile["submitTime"] = str(datetime.datetime.now())
with self.output().open('w') as out:
out.write(wc.getFormattedJson(outputFile))
except subprocess.CalledProcessError as e:
errStr = "command '{}' return with error (code {}): {}".format(
e.cmd, e.returncode, e.output)
log.error(errStr)
raise RuntimeError(errStr)
def run(self):
getInputs = {}
with self.input().open('r') as getInputsInfo:
getInputs = json.load(getInputsInfo)
tasks = []
for inputFile in getInputs["inputFiles"]:
task = SetupWorkDir(inputPath=inputFile,
paths=self.paths,
spatialConfig=self.spatialConfig,
removeSourceFile=True)
tasks.append(task)
yield tasks
outputFile = {"productSetups": []}
for task in tasks:
with task.output().open('r') as taskOutput:
productSetup = json.load(taskOutput)
outputFile["productSetups"].append(productSetup)
with self.output().open("w") as outFile:
outFile.write(wc.getFormattedJson(outputFile))
def run(self):
getInputSwaths = {}
with self.input()[0].open('r') as getInputSwathsInfo:
getInputSwaths = json.load(getInputSwathsInfo)
setupWorkDirs = {}
with self.input()[1].open('r') as setupWorkDirsInfo:
setupWorkDirs = json.load(setupWorkDirsInfo)
basketDir = self.paths["basketDir"]
with open(
os.path.join(self.paths["templatesDir"],
's2_serial_GenerateReport_job_template.bsub'),
'r') as t:
bsubTemplate = Template(t.read())
tasks = []
for swathSetup in setupWorkDirs["swathSetups"]:
productName = wc.getProductNameFromPath(swathSetup["swathDir"])
for swath in getInputSwaths["swaths"]:
if swath["swathDir"] == swathSetup["swathDir"]:
noOfGranules = len(swath["productPaths"])
break
arcsiReprojection = "--outWkt={} --projAbbv={}".format(
self.outWktFilename,
self.projAbbv) if self.arcsiReprojection else ""
metadataTemplate = ""
if self.metadataTemplate is not None:
metadataTemplate = "--metadataTemplate={}".format(
self.metadataTemplate)
arcsiCmdTemplate = ""
if self.arcsiCmdTemplate is not None:
arcsiCmdTemplate = "--arcsiCmdTemplate={}".format(
self.arcsiCmdTemplate)
reportFileName = "{}-{}.csv".format(
os.path.basename(self.paths["basketDir"]),
datetime.now().strftime("%Y%m%d%H%M"))
bsubParams = {
"maxRunTime": noOfGranules * self.hoursPerGranule,
"jobWorkingDir": swathSetup["workspaceRoot"],
"workingMount": swathSetup["workingFileRoot"],
"stateMount": swathSetup["stateFileRoot"],
"inputMount": swathSetup["swathDir"],
"staticMount": self.paths["staticDir"],
"outputMount": self.paths["outputDir"],
"s2ArdContainer": self.paths["singularityImgPath"],
"dem": self.demFilename,
"arcsiReprojection": arcsiReprojection,
"metadataConfigFile": self.metadataConfigFile,
"metadataTemplate": metadataTemplate,
"arcsiCmdTemplate": arcsiCmdTemplate,
"reportFileName": reportFileName,
"reportMount": self.paths["reportDir"],
"databaseMount": self.paths["databaseDir"]
}
bsub = bsubTemplate.substitute(bsubParams)
bsubScriptPath = os.path.join(
swathSetup["workspaceRoot"],
"submit_GenerateReport_job_for_{}.bsub".format(productName))
with open(bsubScriptPath, 'w') as bsubScriptFile:
bsubScriptFile.write(bsub)
tasks.append(
SubmitJob(paths=self.paths,
productName=productName,
bsubScriptPath=bsubScriptPath,
testProcessing=self.testProcessing))
yield tasks
outputFile = {"basket": basketDir, "submittedSwaths": []}
for task in tasks:
with task.output().open('r') as taskOutput:
submittedSwath = json.load(taskOutput)
outputFile["submittedSwaths"].append(submittedSwath)
with self.output().open("w") as outFile:
outFile.write(wc.getFormattedJson(outputFile))
def run(self):
setupWorkDirs = {}
with self.input().open('r') as setupWorkDirsInfo:
setupWorkDirs = json.load(setupWorkDirsInfo)
basketDir = self.paths["basketDir"]
with open(os.path.join(self.paths["templatesDir"], 's1_job_template.sbatch'), 'r') as t:
sbatchTemplate = Template(t.read())
reportFileName = "{}-{}.csv".format(os.path.basename(self.paths["basketDir"]), datetime.now().strftime("%Y%m%d%H%M"))
tasks = []
for productSetup in setupWorkDirs["productSetups"]:
productName = wc.getProductNameFromPath(productSetup["inputPath"])
path = Path(productSetup["inputPath"])
inputDir = path.parent
removeSourceFileFlag = "--removeInputFile" if self.removeSourceFile else ""
sbatchParams = {
"jobWorkingDir" : productSetup["workspaceRoot"],
"reportMount": self.paths["reportDir"],
"databaseMount": self.paths["databaseDir"],
"workingMount": productSetup["workingFileRoot"],
"stateMount": productSetup["stateFileRoot"],
"inputMount" :inputDir,
"staticMount" :self.paths["staticDir"],
"outputMount": self.paths["outputDir"],
"s1ArdContainer": self.paths["singularityImgPath"],
"productName": productName,
"snapConfigUtmProj": self.spatialConfig["snapConfigUtmProj"],
"snapConfigCentralMeridian": self.spatialConfig["snapConfigCentralMeridian"],
"snapConfigFalseNorthing": self.spatialConfig["snapConfigFalseNorthing"],
"snapRunArguments": self.spatialConfig["snapRunArguments"],
"sourceSrs": self.spatialConfig["sourceSrs"],
"targetSrs": self.spatialConfig["targetSrs"],
"filenameDemData": self.spatialConfig["filenameDemData"],
"filenameSrs": self.spatialConfig["filenameSrs"],
"demFilename": self.spatialConfig["demFilename"],
"demTitle": self.spatialConfig["demTitle"],
"metadataProjection": self.spatialConfig["metadataProjection"],
"metadataPlaceName": self.spatialConfig["metadataPlaceName"],
"metadataParentPlaceName": self.spatialConfig["metadataParentPlaceName"],
"removeSourceFileFlag": removeSourceFileFlag,
"reportFileName": reportFileName
}
bsub = sbatchTemplate.substitute(sbatchParams)
sbatchScriptPath = os.path.join(productSetup["workspaceRoot"], "process_s1_ard.sbatch")
with open(sbatchScriptPath, 'w') as sbatchScriptFile:
sbatchScriptFile.write(bsub)
task = SubmitJob(
paths = self.paths,
productName = productName,
sbatchScriptPath = sbatchScriptPath,
testProcessing = self.testProcessing
)
tasks.append(task)
yield tasks
outputFile = {
"basket": basketDir,
"submittedProducts": []
}
for task in tasks:
with task.output().open('r') as taskOutput:
submittedProduct = json.load(taskOutput)
outputFile["submittedProducts"].append(submittedProduct)
with self.output().open("w") as outFile:
outFile.write(wc.getFormattedJson(outputFile))
def run(self):
setupWorkDirs = {}
with self.input()[0].open('r') as setupWorkDirsInfo:
setupWorkDirs = json.load(setupWorkDirsInfo)
prepareArdProcessingJobs = {}
with self.input()[1].open('r') as submitProcessRawToArdJobsInfo:
prepareArdProcessingJobs = json.load(submitProcessRawToArdJobsInfo)
basketDir = self.paths["basketDir"]
with open(
os.path.join(self.paths["templatesDir"],
's2_mpi_GenerateReport_job_template.sbatch'),
'r') as t:
sbatchTemplate = Template(t.read())
tasks = []
for swathSetup in setupWorkDirs["swathSetups"]:
productName = wc.getProductNameFromPath(swathSetup["swathDir"])
for submittedSwath in prepareArdProcessingJobs["submittedSwaths"]:
if submittedSwath["productId"] == productName:
upstreamJobId = submittedSwath["jobId"]
arcsiReprojection = "--outWkt={} --projAbbv={}".format(
self.outWktFilename,
self.projAbbv) if self.arcsiReprojection else ""
metadataTemplate = ""
if self.metadataTemplate is not None:
metadataTemplate = "--metadataTemplate={}".format(
self.metadataTemplate)
arcsiCmdTemplate = ""
if self.arcsiCmdTemplate is not None:
arcsiCmdTemplate = "--arcsiCmdTemplate={}".format(
self.arcsiCmdTemplate)
reportFileName = "{}-{}.csv".format(
os.path.basename(self.paths["basketDir"]),
datetime.now().strftime("%Y%m%d%H%M"))
sbatchParams = {
"upstreamJobId": upstreamJobId,
"jobWorkingDir": swathSetup["workspaceRoot"],
"workingMount": swathSetup["workingFileRoot"],
"stateMount": swathSetup["stateFileRoot"],
"inputMount": swathSetup["swathDir"],
"staticMount": self.paths["staticDir"],
"outputMount": self.paths["outputDir"],
"s2ArdContainer": self.paths["singularityImgPath"],
"arcsiReprojection": arcsiReprojection,
"dem": self.demFilename,
"metadataConfigFile": self.metadataConfigFile,
"metadataTemplate": metadataTemplate,
"arcsiCmdTemplate": arcsiCmdTemplate,
"maxCogProcesses": self.maxCogProcesses,
"reportFileName": reportFileName,
"reportMount": self.paths["reportDir"],
"databaseMount": self.paths["databaseDir"]
}
sbatch = sbatchTemplate.substitute(sbatchParams)
sbatchScriptPath = os.path.join(
swathSetup["workspaceRoot"],
"submit_GenerateReport_job_for_{}.sbatch".format(productName))
with open(sbatchScriptPath, 'w') as sbatchScriptFile:
sbatchScriptFile.write(sbatch)
tasks.append(
SubmitJob(paths=self.paths,
productName=productName,
sbatchScriptPath=sbatchScriptPath,
testProcessing=self.testProcessing))
yield tasks
outputFile = {"basket": basketDir, "submittedSwaths": []}
for task in tasks:
with task.output().open('r') as taskOutput:
submittedSwath = json.load(taskOutput)
outputFile["submittedSwaths"].append(submittedSwath)
with self.output().open("w") as outFile:
outFile.write(wc.getFormattedJson(outputFile))
def run(self):
getInputSwaths = {}
with self.input()[0].open('r') as getInputSwathsInfo:
getInputSwaths = json.load(getInputSwathsInfo)
setupWorkDirs = {}
with self.input()[1].open('r') as setupWorkDirsInfo:
setupWorkDirs = json.load(setupWorkDirsInfo)
prepareArdProcessingJobs = {}
with self.input()[2].open('r') as submitPrepareArdProcessingJobsInfo:
prepareArdProcessingJobs = json.load(
submitPrepareArdProcessingJobsInfo)
basketDir = self.paths["basketDir"]
with open(
os.path.join(self.paths["templatesDir"],
's2_mpi_ProcessRawToArd_job_template.sbatch'),
'r') as t:
sbatchTemplate = Template(t.read())
tasks = []
for swathSetup in setupWorkDirs["swathSetups"]:
productName = wc.getProductNameFromPath(swathSetup["swathDir"])
for swath in getInputSwaths["swaths"]:
if swath["swathDir"] == swathSetup["swathDir"]:
noOfGranules = len(swath["productPaths"])
break
for submittedSwath in prepareArdProcessingJobs["submittedSwaths"]:
if submittedSwath["productId"] == productName:
upstreamJobId = submittedSwath["jobId"]
testProcessing = "--testProcessing" if self.testProcessing else ""
sbatchParams = {
"upstreamJobId": upstreamJobId,
"nodes": noOfGranules + 1,
"jobWorkingDir": swathSetup["workspaceRoot"],
"workingMount": swathSetup["workingFileRoot"],
"stateMount": swathSetup["stateFileRoot"],
"inputMount": swathSetup["swathDir"],
"staticMount": self.paths["staticDir"],
"singularityDir": self.paths["singularityDir"],
"arcsiContainer": self.paths["arcsiMpiBaseImg"],
"testProcessing": testProcessing
}
sbatch = sbatchTemplate.substitute(sbatchParams)
sbatchScriptPath = os.path.join(
swathSetup["workspaceRoot"],
"submit_ProcessRawToArd_job_for_{}.sbatch".format(productName))
with open(sbatchScriptPath, 'w') as sbatchScriptFile:
sbatchScriptFile.write(sbatch)
tasks.append(
SubmitJob(paths=self.paths,
productName=productName,
sbatchScriptPath=sbatchScriptPath,
testProcessing=self.testProcessing))
yield tasks
outputFile = {"basket": basketDir, "submittedSwaths": []}
for task in tasks:
with task.output().open('r') as taskOutput:
submittedSwath = json.load(taskOutput)
outputFile["submittedSwaths"].append(submittedSwath)
with self.output().open("w") as outFile:
outFile.write(wc.getFormattedJson(outputFile))
