[Python] How to call a azure function that works on localhost and return 500 on azure portal?

I have a Azure Function created in VSCode with Python. The API contain the access to a AzureWebJob, when I tested it in the local host http://localhost:7071/api/FUNCTIONNAME?Param1=value1 the response is 200 and the JSON return is ok but, when I deploy the function to Azure the link https://APPSERVICENAME.azurewebsites.net/api/FUNCTIONNAME?code=APICODE==&Param1=value1 return 500 Internal Server Error.

I'm using Postman to test and replied the query on Chrome.

The code used is:

import logging
import azure.functions as func
from io import StringIO
import pandas as pd
from datetime import datetime as dt
import json

#filter the dataframe
def filter(pdf, keys, daterange):
jkeys=[]
for key in keys:
#filter by numeric key
if(reg.isdigit()):
jkeys.append(int(key))
#filtering by two types of keys
pdfA=pdf[pdf['key'].isin(keys)]
pdfS=pdf[pdf['numerickey'].isin(jkeys)]
#concat the DFs
pdf=pd.concat([pdfS,pdfA],axis=0)
#Filtering by a dateRange or single date
if(daterange.isdigit()):
pdf=pdf[pdf['Day']==int(daterange)]
return pdf
else:
#getting the two dates
date=daterange.split('-')
#filtering the rows
pdf=pdf[(pdf['Day']>=int(date[0])) & (pdf['Day']<=int(date[1]))]
return pdf

# function to validate the params
def param_validation(param: str,
param_name: str,
req: func.HttpRequest) -> str:

if not param:
try:
req_body = req.get_json()
except ValueError:

if param != None and '-' in param:
return func.HttpResponse(
"Set a date in the query",
status_code=400
)
else:
param = req_body.get(param_name)
return param
else:
return param

def main(req: func.HttpRequest,
inputBlob: func.InputStream) -> func.HttpResponse:
logging.info('Python HTTP trigger function processed a request.')

#get the params
daterange = req.params.get('daterange')
keys = req.params.get('keys')
name=req.params.get('name')

daterange = param_validation(daterange, 'daterange', req)
keys = param_validation(keys, 'keys', req)

#this condition is included just to test fast if works
if name:
return func.HttpResponse(f"Hello {name}!")
#This is to get the real data request function
if daterange and (keys):
try:

if daterange and keys:
# get the data from a csv in the datalake
data = pd.read_csv(StringIO(inputBlob.read().decode('utf-8')))
keys=keys.split(',')
dfFilt=filter(data,keys,daterange).fillna(0)

return func.HttpResponse(f"{json.dumps(dfFilt.to_dict(orient='record'))}")

else:
return func.HttpResponse(
"insert a valid daterange to the query",
status_code=400
)
except:
return func.HttpResponse(
"Params are wrong.",
status_code=500
)
else:
return func.HttpResponse(
"Params missed",
status_code=400
)



The purpose is to get a csv from datalake and return the rows filtered by 2 args provided.

I found more information about the error, the platform return a 137 python code. I deployed the last functional api with a stale csv (that is lighter, 60mb) and the api works good. But when I a set the new csv (300mb) the api starts to return the code 500 internal server error.

I think that is just this memory problem, because in the local host works perfectly.

are another way to get the data from datalake, filtering two columns using a azure function? Or its posible to fix this memory problem ?

Thanks!

Continue reading...