GW2000 Extra Sensors

[jchurch]

Strange issue. CMX V4.0.1 Extra Sensors causes GW2000 to disconnect and stop sending data. This happens a few minutes to a couple hours after connecting to CMX. Data is from Customized setting in GW2000
Protocol: Ecowitt
Server: 192.168.1.16
Path: /station/ecowittextra
Port: 8998
Interval: 20

As soon as I disable Extra Sensors, connection is restored and I can disable Customized on the GW2000. It will run forever without this enabled. FYI, Using Davis Vantage as primary Station. No errors showing up in CMX.

Thanks for all input.;

[freddie]

Could you please enable debug and data logging, and send a log file capturing the disconnection. Log files are in the MXdiags folder of your installation. You may need to zip up the log file before attaching it to a reply to this topic. Log files are named with a series of numbers (actually date and time info) and have a .txt extension.

[water01]

Have you disabled all the sensors on the GW2000 you are not using, as sometimes if you don't it causes problems?

[jchurch]

Yes. All disabled. Running data capture now.

Thanks.

[jchurch]

Here is the data capture. It ran for 2 hours and a few minutes before the hub shut down and quit sending data. FYI, I ran the hub for 24 hours with no errors without CMX. All was updating correctly until failure.

Thanks for the assist. I am baffled.

20240610-081541.zip

[water01]

Well it looks like something is passing a clean shutdown CTRL+C to the executable and because of that it shuts down.

Have you a CRON job running that does a reboot / shutdown. Cannot think of a setting that would cause this so it must be an external source. Is the program running out of swap space and the Pi automatically shuts down?

Clutching at straws. It may be beneficial to switch debug logging on and run it to the point of shutdown and see what shows up. Data logging will not be of any use as the data all looks good.

[freddie]

From your log file, MX was still receiving data from Ecowitt just before you closed it down:

Code: Select all

2024-06-10 09:12:22.835 ProcessExtraData: Processing posted data
2024-06-10 09:12:22.835 ProcessExtraData: Payload = PASSKEY=<PassKey>&stationtype=GW2000B_V3.1.3&runtime=9&heap=95576&dateutc=2024-06-10+13:12:23&tempinf=74.66&humidityin=46&baromrelin=29.852&baromabsin=29.852&freq=915M&model=GW2000B&interval=20
2024-06-10 09:12:22.836 ApplyExtraData: Complete
2024-06-10 09:12:22.836 ProcessExtraData: Complete

You shut down MX 12 seconds after this data was received - which is shorter than the 20 second interval between Ecowitt updates.

Code: Select all

2024-06-10 09:12:34.160 Ctrl + C pressed
2024-06-10 09:12:34.161 Cumulus closing

[jchurch]

I shut it down when the hub lights went off and sensor data went away. I am going to rerun and leave it until I receive the 10 minute alert of no data transmitted to Ecowitt.Net.

Thanks for the quick response.

[jchurch]

Ok. Alll is working. Here is a screenshot of both GW2000 and CMX updating solar as expected.

Thanks again.

[water01]

Are you talking about the hub lights on your router, in which case is it not a problem with your router, such as a lease renewal time in your DHCP settings?

[dazza1223]

no he talking about the gateway hub the lights go off... i had this same prob i just gave up on them

[jchurch]

Here it is again failed. New logs. Also screen grab. Turn off CMX and everything comes back..

Thanks again.

CMX Failed.png

20240610-095944.zip

[freddie]

Sounds almost like the hub is crashing after sending x requests to CMX. That will be a problem with the hub, not MX, if that is the case. Interesting to hear you experienced the same issue dazza.

[SamiS]

It would be very interesting to know if the behaviour is same with CMX v3.x.

After all, when CMX configured as a custom server to the gw, CMX is just a webserver that responds to post request made by the gateway. So basically it should not be possible for CMX to cause GW to crash. But if there is a bug on the gw firmware regarding connection handling, and CMX does something differently than other webservers, anything is possible.

[dazza1223]

Sounds almost like the hub is crashing after sending x requests to CMX. That will be a problem with the hub, not MX, if that is the case. Interesting to hear you experienced the same issue dazza.

i got them to send me a new hub and it the same happened to that one to