def run():
args = parser_cl()
Base.metadata.create_all(engine)
session = Session()
root = get_xml_root(args.file)
# f = open('log', 'w')
for i, sms in enumerate(root):
json_data = xml_to_dict(sms)
sender = sms.get('contact_name').encode('utf-8') if sms.get(
'type') == '1' else 'Me'
# print('sender = %s' % sender.decode("utf-8") )
# print(type(sender))
json_data['sender'] = sender.decode('utf-8')
item = json_to_b(json_data)
session.add(item)
date = time.strftime("%D %H:%M", time.localtime(int(sms.get('date'))))
body = sms.get('body').encode('utf-8')
# line = '%s - %s : %s\n' % (date, sender, body)
# f.write(line)
session.commit()
# for sms in session.query(Backup).all():
# print(sms.body)
# print('Testing f to_file')
# to_file("I'm testing\ncheck")
# print('Testing find data')
ret = find_data(session.query(Backup).all)
print(ret)
def university_basic_populate():
Base.metadata.create_all(engine)
with open('../scrapers/university.json', 'r') as f:
university_data = json.load(f)
for uni in university_data:
session = Session()
university = University(uni["university_id"], uni["name"])
university.set_extra_info(uni["type"], uni["website"],
uni["oos_tuition"], uni["state_tuition"],
uni["survey_year"])
university.set_location(uni["longitude"], uni["latitude"],
uni["county_id"], uni["state"])
university.set_enrollment(uni["enrolled_men"], uni["enrolled_women"])
university.set_demographics(uni["demographics_asian"],
uni["demographics_white"],
uni["demographics_black"],
uni["demographics_hispanic"],
uni["demographics_other"])
session.add(university)
print("Added " + uni["name"])
session.commit()
session.close()
def add_twitch_clip(self, id):
ret = False
#print(id)
if (storageSystem.get_twitch_clip(id) != None):
return ret
session = Session()
clipObject = twitchmisc.getClipObjects([id])
if (len(clipObject) == 0 or clipObject[0] == None):
return False
#print (clipObject[0])
try:
session.add(clipObject[0])
session.commit()
#update_clips(all_clip_ids, all_clips)
for i in range(len(clipObject)):
if clipObject[i].id not in all_clip_ids:
session.expunge(clipObject[i])
all_clips.append(clipObject[i])
all_clip_ids.append(clipObject[i].id)
ret = True
except:
ret = False
finally:
session.close()
return ret
def add_to_liked_clips(self, clip_id, user_id):
ret = False
session = Session()
if (self.get_twitch_clip(clip_id)
is not None) and (self.user_exists(user_id) == True):
id = None
while id is None or session.query(LikedClips).filter(
LikedClips.id == id).first() is not None:
chars = digits + ascii_lowercase + ascii_uppercase
id = "".join([choice(chars) for i in range(25)])
new_relationship = LikedClips(id=id,
clip_id=clip_id,
user_id=user_id)
try:
session.add(new_relationship)
except:
raise
else:
ret = True
session.commit()
finally:
session.close()
return ret
else:
session.close()
return ret
def import_procedures(procedures):
print("Importing Procedures")
count = 0
session = Session()
for procedure in procedures:
source_id = procedure['subject']['reference'].split('/')[1]
patient_id = session.query(
Patient.id).where(Patient.source_id == source_id).one().id
try:
procedure_date = procedure['performedDateTime']
except KeyError:
procedure_date = procedure['performedPeriod']['start']
procedure_obj = Procedure(
source_id=procedure['id'],
patient_id=patient_id,
procedure_date=procedure_date,
type_code=procedure['code']['coding'][0]['code'],
type_code_system=procedure['code']['coding'][0]['system'])
session.add(procedure_obj)
count += 1
print(".", end='', flush=True) # Show progress
session.commit()
session.close()
print(f"\nImported {count} procedures")
def new_user(self, email, username, password):
session = Session()
if self.email_to_id(email) is not None:
return 1
id = None
while id is None or session.query(User).filter(
User.id == id).first() is not None:
chars = digits + ascii_lowercase + ascii_uppercase
id = "".join([choice(chars) for i in range(25)])
print("Using new user id: " + str(id))
new_user = User(email,
hashlib.sha512(password.encode("utf-8")).hexdigest(),
id, username)
try:
session.add(new_user)
session.commit()
session.close()
return 0
except:
session.close()
return 2
def create_user():
if (request.method == 'GET'):
return render_template('create_user.html')
elif (request.method == 'POST'):
name = request.form['name']
phone = request.form['phone']
# Remove everything except for digits
p = re.compile(r"[^0-9]")
phone = re.sub(p, '', phone)
if (len(phone) != 10):
response = {"error": "Phone number is wrong length"}
return jsonify(response), 500
# Check for a duplicate phone number
session = Session()
duplicate = session.query(User).filter(User.phone == phone).all()
if (len(duplicate) > 0):
response = {"error": "This phone number is already in the system"}
return jsonify(response), 500
# All tests have passed, so create and insert the user.
new_user = User(name, phone)
session.add(new_user)
session.commit()
session.close()
# After we have finished inserting the new user,
# redirect to show all users including the one we just inserted.
return redirect(url_for('users'))
class DBQuery:
Base.metadata.create_all(engine)
def __init__(self):
self.session = Session()
def __enter__(self):
return self
def __exit__(self, exc_type, exc_value, traceback):
self.session.close()
def get_items(self):
query = self.session.query(Item).all()
items = [item.columns() for item in query]
return items
def add_item(self, values):
self.session.add(values)
self.session.commit()
def update_item(self, item_id, values):
self.session.query(Item).filter(Item.id == item_id).update(values)
self.session.commit()
def delete_item(self, item_id):
self.session.query(Item).filter(Item.id == item_id).delete()
self.session.commit()
def handler(event, context):
authorized_user_types = [
UserType.ADMIN,
]
success, _ = check_auth(event['headers']['Authorization'],
authorized_user_types)
if not success:
return http_status.unauthorized()
body = json.loads(event["body"])
user_email = body.get('email')
if not user_email:
return http_status.bad_request()
chat_freq = 4
attributes = {
"custom:user_type": "MENTOR",
"custom:declared_chats_freq": str(chat_freq),
"custom:remaining_chats_freq": str(chat_freq)
}
admin_update_user_attributes(user_email, attributes)
admin_enable_user(user_email)
session = Session()
for i in range(chat_freq):
chat_type = ChatType["ONE_ON_ONE"]
chat = Chat(chat_type=chat_type,
chat_status=ChatStatus.PENDING,
senior_executive=user_email)
session.add(chat)
session.commit()
session.close()
return http_status.success()
def save(self):
s = Session()
if not self.id:
s.add(self)
s.commit()
s.expunge(self)
s.close()
def offerTransaction(rem, dest, offer):
destUser = User.get_by_email(dest)
account_2 = destUser.blockHash
remUser = User.get_by_email(rem)
nonce = web3.eth.getTransactionCount(remUser.blockHash)
amount = offer.precio
strOffer = "Pago por oferta: " + offer.nombre
float_amount = float(amount) / valorUDC
tx = {
'chainId': 3, # es 3 para Ropsten
'nonce': nonce,
'to': account_2,
'value': web3.toWei(float_amount, 'ether'),
'gas': 50000,
'gasPrice': web3.toWei(50, 'gwei'),
'data': bytes(strOffer, 'utf8')
}
signed_tx = web3.eth.account.signTransaction(tx, remUser.pk)
tx_hash = web3.eth.sendRawTransaction(signed_tx.rawTransaction)
s = Session()
dateTimeObj = datetime.now()
timestampStr = dateTimeObj.strftime("%d-%m-%Y (%H:%M:%S.%f)")
t = Transaccion(timestampStr, tx_hash, rem, destUser.organizacion, None,
amount, "", "")
s.add(t)
s.commit()
s.close()
def updateGridTaskStatus(gridTask):
#get GridWay status
(result, status, error)=drmaa_job_ps(gridTask.gwID)
#error control
if result != DRMAA_ERRNO_SUCCESS:
print >> sys.stderr, "updatingGridTaskStatus, drmaa_job_ps() failed for task " + gridTask.gwID + ". Error: %s" % (error)
gridTask.status = "DONE" #DONE does not imply success, just that it has finished
else:
if gridTask.status == "CLEAR": #ya se ha acabado, nothing to do here
return
elif status == DRMAA_PS_UNDETERMINED or status == DRMAA_PS_QUEUED_ACTIVE:
gridTask.status = "SUBMITTED"
elif status == DRMAA_PS_RUNNING:
#if needed, update task execution start
if gridTask.status == "SUBMITTED" or gridTask.status=="WAITING":
gridTask.executionStartDate = datetime.now()
print ("Task " + gridTask.gwID + "with type " + gridTask.type + " on host " + gridTask.host.hostname + " with ID=" + str(gridTask.host.id) + " has started its execution")
gridTask.status = "RUNNING"
elif status == DRMAA_PS_DONE or status == DRMAA_PS_FAILED:
gridTask.status = "DONE"
Session.add(gridTask)
def main(filename):
'''Funcón principal que nos permite crear el schema, tomar datos del csv y pasarlo a un .db
Parameter
--------
- filename : str
Nombre del dataset csv
Return
-------
- database : database from sqlite
'''
Base.metadata.create_all(engine) # Nos genera el schema
session = Session() # Generamos nuestra sesion
articles = pd.read_csv(filename)
for index, row in articles.iterrows(): # Comenzamos a iterar por todos elementos del archivo
logger.info('Loading article uid {} info DB'.format(row['uid']))
article = Article(row['uid'],
row['body'],
row['host'],
row['newspaper_uid'],
row['n_token_body'],
row['n_token_title'],
row['title'],
row['url'],)
session.add(article)
session.commit()
session.close()
def sendCoins(dest, amount, imgHash, urlProof):
destUser = User.get_by_email(dest)
account_2 = destUser.blockHash
nonce = web3.eth.getTransactionCount(test_address)
accion = Accion.getActionById(session['accionId'])
float_amount = float(amount) / valorUDC
bytesStr = "acc:" + accion.nombre + " img: " + imgHash
tx = {
'chainId': 3, # es 3 para Ropsten
'nonce': nonce,
'to': account_2,
'value': web3.toWei(float_amount, 'ether'),
'gas': 50000,
'gasPrice': web3.toWei(50, 'gwei'),
'data': bytes(bytesStr, 'utf8')
}
signed_tx = web3.eth.account.signTransaction(tx, private_key)
tx_hash = web3.eth.sendRawTransaction(signed_tx.rawTransaction)
s = Session()
dateTimeObj = datetime.now()
timestampStr = dateTimeObj.strftime("%d-%m-%Y (%H:%M:%S.%f)")
t = Transaccion(timestampStr, tx_hash, accion.empresa, dest,
accion.campanya_id, amount, imgHash, urlProof)
s.add(t)
s.commit()
query = s.query(Accion)
kpi = int(float(request.form['kpi']))
dictupdate = {Accion.kpi: Accion.kpi + kpi}
query.filter(Accion.id == accion.id).update(dictupdate,
synchronize_session=False)
s.commit()
s.close()
def updateStatus(self, gridTasks):
#===============================================================================
# print ("updatestatus: UPDATING INFRASTRUCTURE STATUS")
#
# print(" numbver of tasks to process: " + str(len(gridTasks)))
#===============================================================================
for host in self.hosts:
host.currentSlotCount = 0
for gridTask in gridTasks:
if gridTask.status == "RUNNING":
try:
gridTask.host.currentSlotCount +=1
except:
pass
#===============================================================
# print (" task running on host " + gridTask.host.hostname + "with id " + str(gridTask.host.id) +", that makes " + str(gridTask.host.currentSlotCount) + " tasks on the host")
#===============================================================
for host in self.hosts:
host.maxSlotCountThisTime = max(host.currentSlotCount, host.maxSlotCountThisTime)
if host.maxSlotCountThisTime > host.maxSlotCount:
host.maxSlotCount = host.maxSlotCountThisTime
Session.add(host)
def insert_attack(host, port):
Base.metadata.create_all(engine)
session = Session()
attack = Attack(host, port)
session.add(attack)
session.commit()
session.close()
def Add_Department():
dep_name = input("Enter Department Name: ")
new_dep = Department(dep_name=dep_name)
session = Session()
session.add(new_dep)
session.commit()
def get_or_create_solution(n_queens,solution):
session = Session()
sol = session.query(SolutionModel).filter_by(solution = ";".join(str(v) for v in solution)).first()
if sol:
return sol
else:
sol = SolutionModel(n_queens = n_queens,solution = ';'.join(str(v) for v in solution))
session.add(sol)
session.commit()
return sol
def updateInfoAfterExecutionInPilots(self, gridTask): self.updateInfoAfterExecution(gridTask) #queremos aumentar el numero de tareas otra vez (porque lo restamos al enviarla, y el updateInfo... lo vuelve a restar) # -si ha acabado OK, para que el numero de pendientes se quede como estaba # - si NO ha acabado Ok, para que se vuelva a incrementar, ya que lo restamos al enviarla self.remainingSamples += gridTask.realSamples Session.add(self)
def getTask():
newParameter = Session.query(Parameter).filter(Parameter.status=="WAITING").first()
if newParameter == None:
return template('{{id}}', id=-1)
newParameter.status="RUNNING"
newParameter.executionStartDate = datetime.now()
Session.add(newParameter)
Session.commit()
return template('{{id}}', id=newParameter.id)
def main(filename):
Base.metadata.create_all(engine)
session = Session()
articles = pd.read_csv(filename, encoding='utf-8')
for index, row in articles.iterrows():
logger.info('Loading article uid: {}'.format(row['uid']))
article = Article(row['uid'], row['body'], row['host'],
row['newspaper_uid'], row['n_tokens_body'],
row['n_tokens_title'], row['title'], row['url'])
session.add(article)
session.commit()
session.close()
def AddPost():
s = Session()
text = request.form['text']
date = datetime.strptime(request.form['date'], '%Y-%m-%d')
user = request.form['user']
title = request.form['title']
spoiler = request.form['spoiler']
p = Post(text, date, user, title, spoiler)
s.add(p)
s.commit()
s.close()
return jsonify()
def failedTask(idInput=-1):
#por si acaso
try:
id = int(idInput)
except:
return template('-1')
failedParameter = Session.query(Parameter).filter(Parameter.id==id).first()
failedParameter.status="WAITING"
failedParameter.executionStartDate = None
Session.add(failedParameter)
Session.commit()
return template('{{id}}', id=failedParameter.id)
def main(filename):
Base.metadata.create_all(Engine)
session = Session()
articles = pd.read_csv(filename)
for index, row in articles.iterrows():
logger.info(f"Loading article uid {row['uid']} into DB...")
article = Article(row['uid'], row['body'], row['host'],
row['newspaper_uid'], row['n_tokens_body'],
row['n_tokens_title'], row['title'], row['url'])
session.add(article)
session.commit()
session.close()
def add_university_image_link():
with open('../scrapers/university.json', 'r') as f:
university_data = json.load(f)
for uni in university_data:
session = Session()
university = session.query(University).filter(
University.id == uni["university_id"]).first()
university.set_image_link(uni["image_link"])
session.add(university)
session.commit()
print(university.image_link)
session.close()
def addToTable(self, FS, IS, VS, CO, TVOC, cTemp, fTemp, humidity,
pressure):
# adds collected data to local SQLite database using SQLAlchemy ORM
# Base, Session, and engine are all defined in base.py
# should be MySQL db of SensHost
Base.metadata.create_all(engine)
session = Session()
new_reading = Readings(strftime("%Y-%m-%d %H:%M:%S",
gmtime()), self.node.name, FS, IS, VS,
CO, TVOC, cTemp, fTemp, humidity, pressure)
session.add(new_reading)
session.commit()
session.close()
def main(filename):
Base.metadata.create_all(engine)
session = Session()
articles = pd.read_csv(filename, encoding='latin1')
for index, row in articles.iterrows():
logger.info('Loading article id {} into DB'.format(row['uid']))
article = Article(row['uid'], row['area'], row['gerente'],
row['fecha'], row['banda'], row['calificacion'],
row['salario'], row['seguro'])
session.add(article)
session.commit()
session.close()
def main(filename):
Base.metadata.create_all(engine)
session = Session()
articles = pd.read_csv(filename)
for index, row in articles.iterrows():
logger.info("Loading article into DB")
article = Article(row["uid"], row["body"], row["host"],
row["newspaper_uid"], row["n_tokens_title"],
row["title"], row["url"])
session.add(article)
session.commit()
session.close()
def main(filename):
Base.metadata.create_all(Engine) #Genera el SCHEMA
session = Session()
articles = pd.read_csv(filename)
for index, row in articles.iterrows():
logger.info('Loading article uid {} into DB'.format(row['uid']))
article = Article(row['uid'], row['body'], row['title'],
row['newspaper_uid'], row['n_token_body'],
row['n_token_title'], row['title'], row['url'])
session.add(article)
session.commit()
session.close()
def analyzeParametricJobfile(self):
tmpFile = "/tmp/tmpParameterFile"
process = monteraLocation + '/templateManager -c ' + self.parametricJobFile + " > " + tmpFile
parameters = self.runProcess(process)
numTasks = 0
for line in open(tmpFile, 'r').readlines():
print ("Parameter: " + line)
myParameter = Parameter(line)
Session.add(myParameter)
numTasks+=1
Session.commit()
def add_city_relationship():
with open('../scrapers/university.json', 'r') as f:
university_data = json.load(f)
for uni in university_data:
session = Session()
university = session.query(University).filter(
University.id == uni["university_id"]).first()
city = session.query(City).filter(City.id == uni["city_id"]).first()
university.set_city(city)
session.add(university)
session.commit()
print(city.city_name)
session.close()
def finishedTask(idInput=-1):
#por si acaso
try:
id = int(idInput)
except:
return template('-1')
finishedParameter = Session.query(Parameter).filter(Parameter.id==id).first()
finishedParameter.status="DONE"
finishedParameter.executionEndDate = datetime.now()
Session.add(finishedParameter)
Session.commit()
#TODO: en algun sitio habria que verificar los datos de salida antes del done
return template('{{id}}', id=finishedParameter.id)
def store(request):
title = request.form.get("title")
body = request.form.get("body")
resume = rf.resume(body)
Base.metadata.create_all(engine)
session = Session()
new_corpus = Corpus(title=title, body=body, resume=resume)
session.add(new_corpus)
session.commit()
session.close()
return redirect("http://127.0.0.1:5000/")
def handler(event, context):
# check authorization
authorized_user_types = [
UserType.ADMIN
]
success, _ = check_auth(event['headers']['Authorization'], authorized_user_types)
if not success:
return http_status.forbidden()
# validate body
body = json.loads(event["body"])
senior_executive = body.get('senior_executive')
chats_new = body.get('chats')
if not senior_executive or not chats_new:
return http_status.bad_request("invalid parameter(s): 'senior_executive, chats'")
session = Session()
for chat_new in chats_new:
if not chat_new.get('chat_type') or chat_new['chat_type'] not in ChatType.__members__:
session.close()
return http_status.bad_request("invalid parameter(s): 'chat_type'")
chat_type = ChatType[chat_new['chat_type']]
description = chat_new.get('description')
tags = chat_new.get('tags')
fixed_date = chat_new.get('fixed_date')
if chat_type in mandatory_date and not fixed_date:
session.rollback()
session.close()
return http_status.bad_request("missing body attribute { fixed_date } with chat_type { %s }" % (chat_type.name))
chat = Chat(
chat_type=chat_type, description=description,
chat_status=ChatStatus.PENDING, tags=tags,
senior_executive=senior_executive
)
admin_update_declared_chats_frequency(senior_executive, 1)
if fixed_date:
chat.fixed_date = datetime.fromtimestamp(fixed_date).replace(hour=0, minute=0,second=0, microsecond=0)
chat.chat_status = ChatStatus.ACTIVE
else:
admin_update_remaining_chats_frequency(senior_executive, 1)
session.add(chat)
session.commit()
session.close()
return http_status.success()
def updateAdd():
requestDetails = request.json
userID = requestDetails["userID"]
maskID = requestDetails["maskID"]
session = Session()
newOrder = Order(userID, maskID)
session.add(newOrder)
session.commit()
session.close()
return '', 200 # Successfully update the masks for the user
def main(filename):
Base.metadata.create_all(Engine)
session = Session()
articles = pd.read_csv(filename)
# uid,body,title,url,newspaper_uid,host,token_title,token_body
for index, row in articles.iterrows():
logger.info(f'Saving article {index}')
article = Article(row['uid'], row['body'], row['title'], row['url'],
row['newspaper_uid'], row['host'],
row['token_title'], row['token_body'])
session.add(article)
session.commit()
session.close()
def popular_banco():
session = Session()
emerson = Pessoa("Emerson")
juca = Pessoa("Juca")
natalia = Pessoa("Natalia")
# session.add(emerson)
# session.add(juca)
session.add(natalia)
natalia_casa = Telefone("(48) 3242-1119", natalia)
session.add(natalia_casa)
session.commit()
session.close()
def updateInfoAfterProfiling(self, gridTask):
#1.- leer la info correspondiente a esa aplicacion
#tener en cuenta que es una serie de parejas de valores
execution_file = base.tmpExecutionFolder + "execution_result_" + gridTask.gwID + ".xml"
try:
doc = xml.dom.minidom.parse(execution_file)
except:
print("failed when profiling host " + gridTask.host.hostname + ". File " + execution_file + " could not be found")
gridTask.host.updateInfoAfterFailedProfiling()
Session.add(gridTask.host)
return
executionInfoList = doc.getElementsByTagName('results')
executionResults=[]
for executionData in executionInfoList:
profileInfoList = executionData.getElementsByTagName('profile')
for profileInfo in profileInfoList:
samples = int(profileInfo.getElementsByTagName("samples")[0].firstChild.data) #TODO: remove "unicode" from TEXT
time = int(profileInfo.getElementsByTagName("time")[0].firstChild.data)
if samples == 0:
print ("There was an error on host " + gridTask.host.hostname + ". File " + execution_file + ", no samples have been executed")
gridTask.host.updateInfoAfterFailedProfiling()
Session.add(gridTask.host)
return
executionResults.append([samples, time])
#2.- procesar los resultados
#TODO: esto es muy posible que esté mal
numSimulations = len(executionResults)
if numSimulations == 0:
print ("There was an error on host " + gridTask.host.hostname + ". File " + execution_file + ", no simulations performed")
gridTask.host.updateInfoAfterFailedProfiling()
Session.add(gridTask.host)
return
avgSampleTime = 0
for i in range(numSimulations-1):
newSamples = executionResults[i+1][0] - executionResults[i][0]
newTime = executionResults[i+1][1] - executionResults[i][1]
if newTime <=0:
value = 0
continue
else:
value = newTime / newSamples ######TODO esta es la clave!!!
#weighted sample time, so later executions are more important than the first ones
#the reason is that they executed more samples, so their information is more valuable
if i==0 or avgSampleTime == 0:
avgSampleTime = value
else:
avgSampleTime = 0.6 * value + 0.4 * avgSampleTime
acumConstantTime = 0
#this has been modifiied to avoid zero values
for i in range(len(executionResults)):
acumConstantTime += max(0, executionResults[i][1] - executionResults[i][0] * avgSampleTime)
avgConstantTime = acumConstantTime / numSimulations
normalizedSampleEffort = avgSampleTime * gridTask.host.getWhetstones()
normalizedConstantEffort = avgConstantTime * gridTask.host.getWhetstones()
print("RESULTADO DEL PROFILING")
print ("task id: " + gridTask.gwID)
print("normalizado, this_sample_effort = " + str(normalizedSampleEffort))
print("normalizado, this_constant_effort = " + str(normalizedConstantEffort))
print("STATUS DEL PROFILING")
print("normalizado, sample_effort = " + str(self.profile.sampleEffort))
print("normalizado, constant_effort = " + str(self.profile.constantEffort))
print("")
if (normalizedSampleEffort < 0 ) or (normalizedConstantEffort < 0):
print ("A normalized effort below zero makes no sense, dismissing profile")
return
self.profile.updateInfoAfterProfiling(normalizedConstantEffort, normalizedSampleEffort)
Session.add(self.profile)
from GridTask import GridTask
import UserInterface
if __name__ == '__main__':
requirementsFile = "/soft/users/home/u5682/nfs/workspace/montera/FLUKA/fluka_example.mt"
print("Starting connection with database")
metadata = MetaData()
myDBDesign = DBDesign()
hostDB = myDBDesign.HostDBDesign(metadata)
applicationDB = myDBDesign.ApplicationDesign(metadata)
appProfileDB = myDBDesign.AppProfileDBDesign(metadata)
gridTaskDesignDB = myDBDesign.GridTaskDBDesign(metadata)
parameterDesignDB = myDBDesign.parameterDBDesign(metadata)
metadata.create_all(base.engine)
print("Database is correct")
print("")
myInfrastructure = Infrastructure.Infrastructure('lrms')
myInfrastructure.createHostProfilingTasks()
Session.add(myInfrastructure)
Session.commit()
Base.metadata.create_all(engine) session = Session() author_1 = Author(name='J.R.R. Tolkien', birth=date(1892, 1, 3)) author_2 = Author(name='J.K. Rowling', birth=date(1965, 7, 31)) author_3 = Author(name='Stephen King', birth=date(1947, 9, 21)) book_1 = Book(title='The Hobbit', published_in=date(1937, 9, 21), author_id=1) book_2 = Book(title='The Lord of the Rings', published_in=date(1954, 7, 29), author_id=1) book_3 = Book(title='Harry Potter', published_in=date(1997, 6, 26), author_id=2) book_4 = Book(title='Carrie', published_in=date(1974, 1, 1), author_id=3) book_5 = Book(title='Salem Lot', published_in=date(1975, 1, 1), author_id=3) book_6 = Book(title='The Shining', published_in=date(1977, 1, 1), author_id=3) book_7 = Book(title='Rage', published_in=date(1977, 1, 1), author_id=3) session.add(author_1) session.add(author_2) session.add(author_3) session.add(book_1) session.add(book_2) session.add(book_3) session.add(book_4) session.add(book_5) session.add(book_6) session.add(book_7) session.commit() session.close()
def createInfrastructureTasks(self, infrastructureTasks):
print ("---------------------")
print ("---------------------")
print ("---------------------")
print ("CREATE INFRASTRUCTURE TASKS")
hostsToProfile = []
hostList = obtainGWResources()
for hostInfo in hostList:
hostName = hostInfo.getElementsByTagName("HOSTNAME")[0].firstChild.data #TODO: remove "unicode" from TEXT
try:
foundArch = hostInfo.getElementsByTagName("ARCH")[0].firstChild.data
except:
foundArch=""
try:
foundCpuMHz = int(hostInfo.getElementsByTagName("CPU_MHZ")[0].firstChild.data)
except:
foundCpuMHz = 0
try:
foundLrms = hostInfo.getElementsByTagName("LRMS_NAME")[0].firstChild.data
except:
foundLrms = None
try:
freeNodeCount = int(hostInfo.getElementsByTagName("FREENODECOUNT")[0].firstChild.data)
except:
freeNodeCount = 0
if foundLrms != None:
if foundLrms == "jobmanager-pilot":
#solo tenemos en cuenta los pilots con al menos un slot disponible
if not freeNodeCount > 0:
continue
#if a certain LRMS is desired, remove the hosts with a different one
if self.lrms != None:
if foundLrms != self.lrms:
continue
#if host is unknown, create a profiling task
currentHost = self.getHost(hostName)
if currentHost == None:
newHost = Host(hostName, arch=foundArch, cpuMHz = foundCpuMHz, lrms=foundLrms)
self.hosts.append(newHost)
hostsToProfile.append(newHost)
#store new host on databae (faiulre resistance
Session.add(newHost)
#if information has changed, update host information
elif (currentHost.arch != foundArch) or (currentHost.cpuMHz != foundCpuMHz):
#TODO: pensar que hacer aqui. habria que eliminar el viejo o solo sobreescribir la información? Si se elimina el viejo, que pasa con las tareas ahí ejecutadas? No es trivial
currentHost.arch = foundArch
currentHost.cpuMHz = foundCpuMHz
if currentHost.lrms == None:
currentHost.lrms = foundLrms
hostsToProfile.append(currentHost)
Session.add(currentHost)
elif currentHost.shouldBeProfiled():
if currentHost.lrms == None:
currentHost.lrms = foundLrms
hostsToProfile.append(currentHost)
#print("Host profiling: submission of 1 tasks per host")
hostProfilingTasks = [ExecutionManager.createHostProfilingTask(host)
for host in hostsToProfile
for i in range(1)]
siteTasks = []
for task in hostProfilingTasks:
found=False
for gridTask in infrastructureTasks:
if gridTask.host.hostname == task.host.hostname:
found=True
break
if not found:
siteTasks.append(task)
#Esto es para el primer experimento de montera + gwpilot
#queremos tener pilots funcionando, así que los arranco con esto
if self.lrms=="jobmanager-pilot":
print ("creating fake profiling tasks")
existingFakeTasks = len([task for task in infrastructureTasks if task.host.hostname=="" and task.status != "PENDING"])
existingGoodPilots = len (self.getGoodHosts())
existingProfilingTasks = len(hostProfilingTasks)
#fakeTasksToCreate = base.maxRunningTasks - (existingFakeTasks + existingGoodPilots + existingProfilingTasks)
fakeTasksToCreate = base.maxRunningTasks - existingFakeTasks
print (" Desired tasks: " + str(base.maxRunningTasks))
print (" Existing fake tasks: " + str(existingFakeTasks))
print (" Existing good pilots: " + str(existingGoodPilots))
print (" created: " + str(fakeTasksToCreate))
emptyHost = FakeHost()
fakeHostProfilingTasks = [ExecutionManager.createWakeUpask(emptyHost)
for i in range(fakeTasksToCreate)]
siteTasks+=fakeHostProfilingTasks
return siteTasks
def updateInfoAfterProfiling(self, gridTask):
print ("Updating info after profiling site " + gridTask.host.hostname)
print (" Task info:")
print (" Task id: " + str(gridTask.id))
print (" GW ID: " + gridTask.gwID)
print (" desired host: " + gridTask.host.hostname)
print (" Host type: " + str(gridTask.host.__class__))
gridTask.status="CLEAR"
#1.- abrir el archivo correspondiente a esa task
execution_file = base.tmpExecutionFolder + "execution_result_" + gridTask.gwID + ".xml"
try:
doc = xml.dom.minidom.parse(execution_file)
except:
print("failed when profiling host " + gridTask.host.hostname + ". File " + execution_file + " could not be found")
gridTask.host.updateInfoAfterFailedProfiling()
Session.add(gridTask)
return
executionInfoList = doc.getElementsByTagName('execution_info')
for executionData in executionInfoList:
try:
gridTaskType = executionData.getElementsByTagName("type")[0].firstChild.data #TODO: remove "unicode" from TEXT
remoteHostName = executionData.getElementsByTagName("hostname")[0].firstChild.data
whetstones = float(executionData.getElementsByTagName("whetstones")[0].firstChild.data)
waitingTime = float(executionData.getElementsByTagName("execution_time")[0].firstChild.data)
dataSize = float(executionData.getElementsByTagName("data_size")[0].firstChild.data)
except:
print("failed when profiling host " + gridTask.host.hostname + ". File " + execution_file + " could not be found")
gridTask.host.updateInfoAfterFailedProfiling()
Session.add(gridTask)
return
#2.- procesar los resultados
if gridTaskType != "benchmark":
print ("ERROR when updating info from a site profiling")
print("Incorrect task type, readed " + gridTaskType + " and should be \"benchmark\"")
print (" considering the execution as failed")
gridTask.host.updateInfoAfterFailedProfiling()
Session.add(gridTask)
return
if remoteHostName != gridTask.host.hostname:
print ("ERROR when updating info from a site profiling")
print("Incorrect host name, readed" + remoteHostName + " and should be " + gridTask.host.hostname)
print (" considering the execution as failed")
gridTask.host.updateInfoAfterFailedProfiling()
Session.add(gridTask)
return
totalActiveTime = InformationManager.readTotalActiveTime(gridTask.gwID)
transferTime = totalActiveTime - waitingTime
queueTime = InformationManager.readQueueTime(gridTask.gwID)
if transferTime > 0:
bandwidth = dataSize / transferTime
else:
bandwidth = -1
#3.- suministrar esa info al host.
gridTask.host.updateInfoFromSuccessFulExecution(whetstones, queueTime, bandwidth)
#4.- eliminar archivos temporales
try:
#os.remove(execution_file)
print ("IN application.py, I would be deleting " + execution_file)
print ("Profiling file has been successfully deleted: " + execution_file)
except:
print ("Could not delete profiling file " + execution_file)
gridTask.endDate = datetime.now()
while True:
print('Add records:', '\n')
print('Tables list:', '\n')
print(tables, '\n')
user_response = input('Enter \'table #\' or \'x\' to exit: ')
if user_response.lower() == 'x' or user_response == '':
break
else:
table = tables.get(int(user_response), None)
if table is not None:
# Retrieve list of columns
columns = [col.name for col in table.__table__.columns if col.name != 'id']
print(columns)
# Initialize a new record
data = {}
for col in columns:
value = input('type a value for {0}: '.format(col))
data[col] = value
if input('Add record? y/n') == 'y':
record = table(**data)
session.add(record)
session.commit()
else:
print('Table not found', '\n')
def updateInfoAfterExecution(self, gridTask):
print("Updating info after exetcution of task " +gridTask.gwID + " on host " + gridTask.host.hostname + " (hostID " + str(gridTask.host.id) + ")")
gridTask.status="CLEAR"
#1.- abrir el archivo correspondiente a esa task
hostToUpdate = gridTask.host
execution_file = base.tmpExecutionFolder + "/execution_result_" + gridTask.gwID + ".xml"
#1.- abrir el archivo correspondiente a esa task
try:
doc = xml.dom.minidom.parse(execution_file)
except:
print("failed when updating info after execution. File " + execution_file + " could not be found")
hostToUpdate.updateInfoAfterFailedExecution()
Session.add(gridTask)
return
#si los archivos de salida deseados no existen, también la cuento como fallida
for outputFile in self.outputFiles.split(","):
#JOB_ID has to be replaced by gwID as it happens along the execution
splittedFile = outputFile.split('JOB_ID')
output=""
for pos in range(len(splittedFile)):
output += splittedFile[pos]
if pos < len(splittedFile) -1:
output+=gridTask.gwID
if not os.path.exists(base.tmpExecutionFolder + "/" + output):
print("failed when updating info after execution. output file " + base.tmpExecutionFolder + "/" + output + " could not be found")
hostToUpdate.updateInfoAfterFailedExecution()
Session.add(gridTask)
return
executionInfoList = doc.getElementsByTagName('execution_info')
gridTaskType = None
remoteHostName = None
executionTime = None
dataSize = None
realSamples = None
for executionData in executionInfoList:
try:
gridTaskType = executionData.getElementsByTagName("type")[0].firstChild.data #TODO: remove "unicode" from TEXT
remoteHostName = executionData.getElementsByTagName("hostname")[0].firstChild.data
executionTime = float(executionData.getElementsByTagName("execution_time")[0].firstChild.data)
dataSize = float(executionData.getElementsByTagName("data_size")[0].firstChild.data)
realSamples = int(executionData.getElementsByTagName("real_samples")[0].firstChild.data)
except:
print ("Error when reading execution file, exiting" )
Session.add(gridTask)
return
#2.- procesar los resultados
if gridTaskType != "execution":
print ("ERROR when updating info from an application execution")
print("Incorrect task type, expecting \"execution\"")
gridTask.status = "CLEAR"
Session.add(gridTask)
return
if remoteHostName != hostToUpdate.hostname:
print ("ERROR when updating info from a application execution")
print("Incorrect host name, expecting " + hostToUpdate.hostname)
gridTask.status = "CLEAR"
Session.add(gridTask)
return
if executionTime == 0:
print ("ERROR when updating info from an application execution")
print ("Execution time appears to be zero, and that's quite strange")
gridTask.status = "CLEAR"
hostToUpdate.updateInfoAfterFailedExecution()
Session.add(gridTask)
return
totalActiveTime = InformationManager.readTotalActiveTime(gridTask.gwID)
if totalActiveTime == -1:
print ("ERROR when updating info from an application execution")
print ("Could not read active time from GridWay log, considering that task failed")
gridTask.status = "CLEAR"
hostToUpdate.updateInfoAfterFailedExecution()
Session.add(gridTask)
return
queueTime = InformationManager.readQueueTime(gridTask.gwID)
if queueTime == -1:
print ("ERROR when updating info from an application execution")
print ("Could not read queue time from GridWay log, considering that task failed")
gridTask.status = "CLEAR"
hostToUpdate.updateInfoAfterFailedExecution()
Session.add(gridTask)
return
transferTime = totalActiveTime - executionTime
if transferTime > 0:
bandwidth = dataSize / transferTime
else:
bandwidth = -1
#3.- actualizar rendimiento del host
computationalEffort = self.profile.constantEffort + self.profile.sampleEffort * realSamples
whetstones = computationalEffort / executionTime
hostToUpdate.updateInfoFromSuccessFulExecution(whetstones, queueTime, bandwidth)
hostToUpdate.failedProfilings -=1
#4 actualizar estado de la tarea y la apliación.
gridTask.realSamples = realSamples
gridTask.status = "CLEAR"
self.remainingSamples -= realSamples
print("APPLICATION UPDATE: " + str(self.remainingSamples) + "/" + str(self.desiredSamples) + " left")
#5.- eliminar archivos temporales
try:
#os.remove(execution_file)
print ("In Application.py, I would be deletign" + execution_file)
print ("Execution file has been successfully deleted: " + execution_file)
except:
print ("Could not delete profiling file " + execution_file)
#6 update info on DB
gridTask.endDate = datetime.now()
Session.add(gridTask)
Session.add(self)
def createInfrastructureTasks(self, infrastructureTasks):
print ("-------------------")
print ("-------------------")
print ("createInfrastructureTasks- NewPilotInfrastructure")
# self.showHosts()
hostList = obtainGWResources()
hostsToProfile = []
print ("Analyzing resources ")
for hostInfo in hostList:
hostName = hostInfo.getElementsByTagName("HOSTNAME")[0].firstChild.data.strip().lower() #TODO: remove "unicode" from TEXT
whetstones=0
try:
foundArch = hostInfo.getElementsByTagName("ARCH")[0].firstChild.data
except:
foundArch=""
try:
foundCpuMHz = int(hostInfo.getElementsByTagName("CPU_MHZ")[0].firstChild.data)
except:
foundCpuMHz = 0
try:
foundLrms = hostInfo.getElementsByTagName("LRMS_NAME")[0].firstChild.data.strip().lower()
except:
foundLrms = None
print ("Could not find LRMS for host " + hostName + ", skipping it")
continue
try:
freeNodeCount = int(hostInfo.getElementsByTagName("FREENODECOUNT")[0].firstChild.data)
except:
freeNodeCount = 0
if foundLrms == "jobmanager-pilot":
#solo tenemos en cuenta los pilots con al menos un slot disponible
if not freeNodeCount > 0:
continue
username = os.getenv("USER")
genericStringArgs = hostInfo.getElementsByTagName("GENERIC_VAR_STR")
for node in genericStringArgs:
if node.attributes['NAME'].value =="PILOT_REAL_HOSTNAME":
workerNode = node.attributes['VALUE'].value.strip().lower()
if node.attributes['NAME'].value =="PILOT_REAL_RESOURCE":
site = node.attributes['VALUE'].value.strip().lower()
genericIntArgs = hostInfo.getElementsByTagName("GENERIC_VAR_INT")
for node in genericIntArgs:
if node.attributes['NAME'].value =="PILOT_" + username + "_VAR_5":
whetstones = int(node.attributes['VALUE'].value.strip().lower())
if whetstones > 65534:
whetstones = 0
# whetstones = 0
#if host is unknown, create a profiling task
currentHost = self.getHost(hostName)
if currentHost == None:
print ("Host/Pilot not found. hostname: " + hostName + ", LRMS: " + foundLrms)
if foundLrms == "jobmanager-pilot":
#he encontrado un pilot:
#primero busco e resource, y si no existe lo creo.
#luego creo un pilot que utilice ese resource
pilotResource = base.Session.query(PilotResource).filter(PilotResource.site == site, PilotResource.workerNode == workerNode).first()
if pilotResource == None:
print (" PilotResource was not found, creating a new one")
pilotResource = PilotResource(site, workerNode)
print (" Creating a new Pilot in NewPilotInfrastructure.createInfrastructureTasks")
newHost = Pilot(hostName, arch=foundArch, cpuMHz = foundCpuMHz, pilotResource = pilotResource, whetstones = whetstones)
self.pilots.append(newHost)
Session.add(newHost)
else:
print (" Creating a new Host in NewPilotInfrastructure.createInfrastructureTasks")
newHost = Host(hostName, arch=foundArch, cpuMHz = foundCpuMHz, lrms=foundLrms)
self.hosts.append(newHost)
Session.add(newHost)
#ESTO ES PARA HACER EL PROFILING DE LOS PILOT SI NO HAN PUBLICADO LOS WHETSTONES, SI NO NO HACE FALTA
#===============================================================
# if whetstones == 0 or whetstones > 65534:
# whetstones = 0
# print (" Host to profile: " + hostName + ": whetstone value not initialized ")
# hostsToProfile.append(newHost)
# #store new host on databae (faiulre resistance
# Session.add(newHost)
#===============================================================
#if information has changed, update host information
elif (currentHost.getWhetstones() != whetstones):
#va con un set porque es una operación más complicada, así que está encapsulada en esta funcion
currentHost.setWhetstones(whetstones)
Session.add(currentHost)
print ("Host: " + hostName + " UPDATED, new whetstones=" + str(whetstones))
elif currentHost.lrms == None:
currentHost.lrms = foundLrms
#pprofiling of new sites
hostProfilingTasks = [ExecutionManager.createHostProfilingTask(host)
for host in hostsToProfile
for i in range(base.profilingTasksPerHost)]
#estamos asumiento que todos los pilots publican la variable esa con su
#rendimiento, con lo que no hay que hacer el profiling de nada.
#AHORA, EN ESA NUEVA APROXIMACION, QUEREMOS TENER UNOS CUANTO SBENCHMARKS PARA IR ARRANCANDO PILOTS
print ("creating fake profiling tasks")
existingFakeTasks = len([task for task in infrastructureTasks if task.host.hostname=="" and task.status != "PENDING"])
existingGoodPilots = len (self.getGoodHosts())
existingProfilingTasks = len(hostProfilingTasks)
#fakeTasksToCreate = base.maxRunningTasks - (existingFakeTasks + existingGoodPilots + existingProfilingTasks)
fakeTasksToCreate = base.maxRunningTasks - existingFakeTasks
print (" Desired tasks: " + str(base.maxRunningTasks))
print (" Existing fake tasks: " + str(existingFakeTasks))
print (" Existing good pilots: " + str(existingGoodPilots))
print (" created: " + str(fakeTasksToCreate))
emptyHost = FakeHost()
fakeHostProfilingTasks = [ExecutionManager.createFakeHostProfilingTask(emptyHost)
for i in range(fakeTasksToCreate)]
hostProfilingTasks+=fakeHostProfilingTasks
return hostProfilingTasks
