wunderground API usage alert

[Ajderev]

Hello All

I started using the Domoticz in two days with my Evohome.
To see outside temp I added wunderground API.

I got two emails saying: Your wunderground API key (xxxxxxxxxx) exceeded its allotted usage today by making 502 calls in a day but the limit is 500.

Is there something I can do so I do not get the API disabled?
My Settings are show below:

1.JPG (98.92 KiB) Viewed 4585 times

2.JPG (127.56 KiB) Viewed 4585 times

I do not know how are the API calls made?
Is one API call for all sensor of the weather station or each sensor gets API call?

I will really appreciate some quick answer because I have only 1 Rain drop left.

Thaks in advance to anybody which spend a time to help me.

Best Regards
Stefan Ajderev

[pvm]

You don't have additional dashboards in addition to the standard domoticz one right?

[Ajderev]

No, I didn't add anything to the standard installation.
I'm running on Windows 10 PC

[DennisD]

I have had the same emails, apparently Domoticz was making 503 calls instead of the maximum of 500. Nothing has changed and i'm running the latest beta's on an rpi 3

[Ajderev]

So is this mean there is nothing I can do to limit the number of calls?
Should I disable wunderground API for the moment until it is fixed so it does not get disabled?

I'm an average user and Domoticz allowed me to do use it without any scripting nor Linux knowledge.
It is really great software I'm just discovering. I love how it works with my Evohome.

[Ajderev]

I just noticed in the log

2017-07-27 21:52:51.538 Wunderground: Worker started...
2017-07-27 21:54:07.408 Wunderground: Worker started...
2017-07-27 21:55:23.217 Wunderground: Worker started...
2017-07-27 21:56:39.081 Wunderground: Worker started...
2017-07-27 21:57:54.914 Wunderground: Worker started...

It is running every minute, not every 5 as I thought

Also, I have
2017-07-27 21:59:13.797 EventSystem - Python: Failed dynamic library load, install the latest libpython3.x library that is available for your platform.
Can you advise me how should I fix it in windows installation, please?

I downloaded and installed Download Python 3.6.2 for windows but still getting the error.

[Ajderev]

Ok, So I guess I need to disable it, for now, itril some solution came to light.

[Toulon7559]

24hours = 24*60minutes = 1440minutes/500calls ~ 1 call/3minutes
And all calls count, which apply your API-key.
No other solution than significant reduction of calling rate.
Anyway it makes no sense to call faster than once per 5 or 10 minutes, because the meteo-data at the source (= WUnderground) is not refreshed more often.

[Ajderev]

Hello Toulon7559,

Now if you tell me how to make it do a call once every 10 minutes it will be priceless

[tlpeter]

Isn't your system rebooting all the time? It will call WU at a service restart. Maybe the service restarted more than 500 times in one day.

[mivo]

Hello Toulon7559,

Now if you tell me how to make it do a call once every 10 minutes it will be priceless

Hi,

there is 10 min interval in C++ code and it works as expected on my system (beta 3.8143):

Code: Select all

2017-07-29 11:37:38.274  (Weather Underground LKPR) Temp + Humidity + Baro (WeatherUG LKPR)
2017-07-29 11:37:38.410  (Weather Underground LKPR) Wind (WeatherUG LKPR vitr)
2017-07-29 11:37:38.544  (Weather Underground LKPR) Rain (WeatherUG LKPR dest)
2017-07-29 11:47:38.592  (Weather Underground LKPR) Temp + Humidity + Baro (WeatherUG LKPR)
2017-07-29 11:47:38.727  (Weather Underground LKPR) Wind (WeatherUG LKPR vitr)
2017-07-29 11:47:38.862  (Weather Underground LKPR) Rain (WeatherUG LKPR dest)

It is question, why your Wunderground worker is restarting. Do you set something in "Data timeout" parameter ? It should be set to: disabled

[benbammens]

I have the same problem. It only happens when I have the dashboard open the whole day on my tablet.

[Toulon7559]

@Ajderev

Unfortunately not able to help you reducing the call-rate in Evohome, because no knowledge & no experience with that system.

My information on the number of calls is by experience.
At a certain time I had 2 lua-scripts uploading happily at 5 minute-rate to related WU-accounts and at the same time Domoticz making calls to get information from those 2 WUnderground-APIs: then you get quite soon an Alert from WUnderground .........
No (known) control on the call-rate by Domoticz, but the call-rate by the lua-scripts can be manipulated in the script.
Emperically I found a setting for the call-rate of the scripts of approx. 15 to 20 minutes, which in combination with the call-rate by Domoticz does not exceed the daily limit. Apparently, splitting the 2 accounts to 2 API-keys does not make difference.

very ROM deduction, but if I take 15 minutes as reference, it means that those 2 lua-scripts daily take approx. 2*24*4 calls = 192 calls, leaving a space of approx. 300 calls/day for Domoticz' requests, and for my calls to 2 accounts that is equivalent to 10min interval per call per account, confirming mivo's observation.
Then surprising that with calls from Domoticz to 1 WU-account you exceed the daily allowed volume.
Apparently we are missing something ..........

As experienced/observed, all accumulated calls to your WUnderground-API are counted, whether from Domoticz or from elsewhere.
Some aspects to check in that perspective?
1) Did you apply something like this to make a connection between Evohome and WUnderground? Would not be surpised if that requires extra calls to WUnderground.
2) (Perhaps far-fetched), what is your polling interval for Evohome: it should be >120s. (see Domoticz' Wiki on Evohome)

On the other hand, you also get an Alert-message in case of >10 calls per minute (in combination with free WU-membership), and therefore a few 'bursts' may trigger WUnderground for an alert (although I do not see how that could happen in your case).

[Ajderev]

Hello All

I do not have almost any scripting skills. I consider my self an average user
So no script making additional calls to WU.
I do not use ifttt.
Evohome is not calling WU.

I have the Domoticz on my desktop PC and it works fine for me and I do not know is the script for WU restarting but as long as I can tell the Domoticz do not restart.
I might need to try to do a Linux VM and install it on it and see how it goes there.

Thank you all for trying to help.

[Ajderev]

I installed Ubuntu 16.04 on VM and installed Domoticz and I still have the same issue.

I think the issue is that every time Evohome via Web API pulls data the Domoticz restart.

Aug 4 23:35:42 Domotixz domoticz[3648]: 2017-08-04 23:35:42.865 (Evohome) fetch data from server
Aug 4 23:35:42 Domotixz kernel: [ 4155.997270] domoticz[3659]: segfault at 8 ip 0000000000970675 sp 00007f6e1626a710 error 4 in domoticz[400000+94e000]
Aug 4 23:35:42 Domotixz systemd[1]: domoticz.service: Main process exited, code=killed, status=11/SEGV
Aug 4 23:35:42 Domotixz systemd[1]: domoticz.service: Unit entered failed state.
Aug 4 23:35:42 Domotixz systemd[1]: domoticz.service: Failed with result 'signal'.
Aug 4 23:36:42 Domotixz systemd[1]: domoticz.service: Service hold-off time over, scheduling restart.
Aug 4 23:36:42 Domotixz systemd[1]: Stopped domoticz_service.
Aug 4 23:36:42 Domotixz systemd[1]: Started domoticz_service.
Aug 4 23:36:42 Domotixz domoticz[3712]: 2017-08-04 23:36:42.988 Domoticz V3.8153 (c)2012-2017 GizMoCuz
Aug 4 23:36:42 Domotixz domoticz[3712]: 2017-08-04 23:36:42.988 Build Hash: 494fff71, Date: 2017-07-30 11:19:41

Is omeone else using Evohome via Web API axpirience the same issue ?

[vosmanz]

Since a couple of days i have a simular issue.
Every night I get an e-mail with the message:

Your wunderground API key (*key here*) exceeded its allotted usage today by making 20 calls within a minute but the limit is 10.
...
Details: exceededrate: maxrate: 20 (of 10)
Time: Thu, Aug 10, 2017 00:00 UTC

Not sure if it is related, but I recently updated to build V3.8153. I'm using the package on my Synology NAS. It has been running ok for a long time (temperature log goes back to nov. 2015) without any changes. Data timeout is disabled in the settings.

Anyone else with this issue, and does anyone know how to fix it?

[dws199]

I installed Ubuntu 16.04 on VM and installed Domoticz and I still have the same issue.

I think the issue is that every time Evohome via Web API pulls data the Domoticz restart.

Aug 4 23:35:42 Domotixz domoticz[3648]: 2017-08-04 23:35:42.865 (Evohome) fetch data from server
Aug 4 23:35:42 Domotixz kernel: [ 4155.997270] domoticz[3659]: segfault at 8 ip 0000000000970675 sp 00007f6e1626a710 error 4 in domoticz[400000+94e000]
Aug 4 23:35:42 Domotixz systemd[1]: domoticz.service: Main process exited, code=killed, status=11/SEGV
Aug 4 23:35:42 Domotixz systemd[1]: domoticz.service: Unit entered failed state.
Aug 4 23:35:42 Domotixz systemd[1]: domoticz.service: Failed with result 'signal'.
Aug 4 23:36:42 Domotixz systemd[1]: domoticz.service: Service hold-off time over, scheduling restart.
Aug 4 23:36:42 Domotixz systemd[1]: Stopped domoticz_service.
Aug 4 23:36:42 Domotixz systemd[1]: Started domoticz_service.
Aug 4 23:36:42 Domotixz domoticz[3712]: 2017-08-04 23:36:42.988 Domoticz V3.8153 (c)2012-2017 GizMoCuz
Aug 4 23:36:42 Domotixz domoticz[3712]: 2017-08-04 23:36:42.988 Build Hash: 494fff71, Date: 2017-07-30 11:19:41

Is omeone else using Evohome via Web API axpirience the same issue ?

Hi

I have exactly the same issue as you and my setup is very similar, running the latest version of Domoticz on a Windows 10 machine with Evohome Web API device. I'm receiving regular emails from Wunderground too so have disabled the service for now.

If it is evohome causing the restarts then the only workaround that I can think of is to setup another Domoticz instance instance (on a rpi) for example and have that as a slave device to the main Domoticz device. This slave could then be used purely for evohome devices.

Dave