Topic: Problem with data spikes on WU only

[hifi_hound]

I have been trying to solve a problem with multiple data spikes showing up in my daily Weather Underground graphs.  These seem to have started around Jan 31 from what I can see going back through my data.  At that time the data was being supplied by my Ecowitt GW1000.  The spikes look like they are happening across the board much of the time, meaning they appear in temperature, UV, solar rad, sometimes pressure, and rain.  I've seen some other threads about random UV spikes, but mine are happening in these other areas mentioned as well.  I've tried relocating the GW1000 to different outlets which are away from cycling refrigerators etc., but I can't seem to get rid of the problem.  I recently purchased the Ecowitt Wittboy, and transfered the WU duties over to the GW2000 and I am still getting the spikes.  I'm not sure what to do at this point.  My Ecowitt.net data along with Weathercloud and WeatherObservations data don't seem to show the same spikes or problems. Also, the WU data temperature lows don't seem to match the graphs either, with the low temperature be stated for the day, much lower than any readings in the graphs.  I just can't figure out why there are discrepancies, but I think it has to do with the data spikes/corruptions.  You can look at the historical data charts and see what I mean.  The Ecowitt site graphs don't show any problems.

https://www.wunderground.com/dashboard/pws/KCAEMERA6
https://www.ecowitt.net/home/share?authorize=Y6FJKF

Here is a sample from WU

 [ You are not allowed to view attachments ]

I'm running 2 stations at the moment, the Wittboy, and my original station which includes an Ecowitt WS68 mounted 30ft up, and a homemade FARS with an WH32_EP temp/humidity probe.  The WH32_EP and the FARS unit have been running for about a year, so it doesn't seem to be related to those. At the moment I have the GW2000 reporting to WU with WH32_EP providing the outdoor temp and the Wittboy providing the rest of the data.  I'm not really sure what the issue is since the problem is across 2 different stations and 2 receivers/upload units.  Also, my internet is gig fiber, and my router/wifi hasn't changed in over a year, so no recent changes there either.  Any help or suggestions would be appreciated.

[Gyvate]

at first glance it looks like a WU issue which is not in your control.

If you used FOSHKplugin, you could make related tests right away.
for FOSHKplugin see https://www.wxforum.net/index.php?topic=39625.0

You could install FOSHKplugin if you happen to have a RaspberryPi easily - or even install and run it in the Windows 10/11 WSL (Windows Subsystem for Linux).
FOSHKplugin documents the incoming and outgoing posts in logging files. Not just with a time stamp, but the full post including the payload.
FOSHKplugin would post the data to WU just as your console would do (using custom server and WU protocol).
So you could compare the data sent with the data shown for that time stamp at WU.

Of course, you wouldn't know if for whatever reason the standard WU posting of the console behaved out of the usual and not in tune with a related custom server setup.
In order to know this you would need to set up DNS spoofing or DNS redirection by installing e.g. DNSmasq redirecting the WU post to your FOSHKplugin, forward them to WU from there and then compare the values.

[hifi_hound]

Thanks for the reply! I briefly looked at FOSHKplugin awhile back, but haven't really looked at it lately.  Just curious, what makes you think this is a WU issue?

[Gyvate]

Thanks for the reply! I briefly looked at FOSHKplugin awhile back, but haven't really looked at it lately.  Just curious, what makes you think this is a WU issue?

your own statement   

My Ecowitt.net data along with Weathercloud and WeatherObservations data don't seem to show the same spikes or problems.

but to verify I proposed to use FOSHKplugin

[Jonesy]

Wow that's very weird!
I've just setup a Wittboy along side with my old fine-offset WH2950 unit/console. I've never seen spikes like that in the WU data. I previously had updates set at 5 minutes and now I've set them both to 1 minute intervals without problems. I set the Wittboy up on new site ID's for all the uploads (ecowitt, wu, weathercloud, wow) to see how it compares with my old unit. If as you say all looks good on the sites other than WU for your Wittboy/GW2000 then I can only think there is a problem on the WU server side. At the moment the only other thing I can think of which seems unlikely, is if you have multiple gateways/consoles sending reports to the same site WU ID.

[hifi_hound]

Yes it is strange.  I thought the same thing that somehow maybe I hadn't removed my WU credentials from my GW1000.  But I have checked multiple times just to be sure.  I also still have an Acurite 5 in 1 from probably 10 years ago still running, but everything is turned off on that unit as far as reporting to outside services.  It only shows up on the console and on myAcurite, but nowhere else.  I can't figure out anything that might have changed at the end of January either that could cause the spikes.  It started while the GW1000 was in service and now it's the GW2000 with the same issue.  Also 2 different sending units, other than the WH32_EP which was sending to both receivers.  I might try and create a new WU setup and see if it happens there as well.