I am using an azure function app to get some weather data using open-meteo api, to do so I need to call their API from inside my function APP. Before deploying on azure, the http triggered function works without any issue and responses are in a matter of seconds. when I deploy It to an azure app it works the first time then times out the second time then works again from different laptops but times out after. I have tried multiple approaches: -Used a context manager to ensure the session is closed after use. -Implemented retries with exponential backoff using the retry_requests library -creased the timeout and memory allocation for your Azure Function. But nothing is working, below is the part of my code that gets the data from the api: def fetch_and_process_data(lat, lon, input_str): today_str, formatted_datetime = format_datetime(input_str) retry_session = retry(requests.Session(), retries=5, backoff_factor=0.5) openmeteo = openmeteo_requests.Client(session=retry_session) # Fetch weather data weather_url = "https://api.open-meteo.com/v1/forecast" weather_params = { "latitude": lat, "longitude": lon, "timezone": "Europe/London", "hourly": ["wind_speed_10m", "wind_direction_10m"], "wind_speed_unit": "ms", "start_date": today_str, "end_date": today_str } weather_responses = openmeteo.weather_api(weather_url, params=weather_params) weather_response = weather_responses[0] hourly_weather = weather_response.Hourly() weather_data = { "date": pd.date_range( start=pd.to_datetime(hourly_weather.Time(), unit="s", utc=True), end=pd.to_datetime(hourly_weather.TimeEnd(), unit="s", utc=True), freq=pd.Timedelta(seconds=hourly_weather.Interval()), inclusive="left" ), "wind_speed_10m": hourly_weather.Variables(0).ValuesAsNumpy(), "wind_direction_10m": hourly_weather.Variables(1).ValuesAsNumpy() } weather_dataframe = pd.DataFrame(data=weather_data) When used locally it works as expected but not when deployed to azure. What might be the issue? any help would be much appreciated Continue reading...
