NOAA reports stopped being saved / uploaded

[SUBYDAZZ]

I assume I may have to leave the ftplog going overnight to get the answers we may need?

[mcrossley]

Yes please.

I see from that log though that you have both "delete before upload" and "rename" enabled. As the rename is working fine, the delete before upload is largely redundant and you could consider disabling it to improve performance.

[SUBYDAZZ]

Thanks, I've changed that setting.

I'm beginning to think that the PC that Cumulus is on is causing some of the issues, seems to be very slow to do anything / very unresponsive. For example, lately when I get rain, it doesn't record in Cumulus MX straight away, but then all comes as one big lump some time later and shows an impossible rain rate during/after that event.

Anyway, here's my latest mxdiags and ftplog. Apologies that I had to split them, they were too big to upload...

[mcrossley]

FTP looks like it is behaving now.

As for your rainfall, I just looked at few log files...

20201116-055000 - rain data is being sent by your station once every minute
20201116-231814 - no rain data sent all in this log from 23:18 to 23:40

I do not really know how the WMR protocol works, but a working assumption - If the station loses contact with a sensor it does not send any data packets for the sensor?

I'd check your base station reception for the rain sensor.

[Tigro11]

good evening,
I wanted to report this error to the monthly noaa with davis station.
Do I need to change any particular settings?
Thanks
Valerio

[mcrossley]

What do you see under Settings|NOAA Settings?

[Tigro11]

[mcrossley]

Sorry, I didn't look beyond the URL in the "state" name.

You need to post your day file.

[Tigro11]

could you tell me the name of the file you need?

[mcrossley]

/data/dayfile.txt

Please also zip up your /MXdiags folder and attach that as well.

[Tigro11]

http://carlinometeo.altervista.org/dayfile.txt
http://carlinometeo.altervista.org/MXdiags.rar

[mcrossley]

Your day file is a bit of a mess.

At the start you have many days 1/11/19 to 13/11/19 that contain the same bad data. You need to delete those.

Then at line 443 dated 16/11/20 through to line 457 dated 30/11/20 (!!) you have similar bad data. The correct data for 16/11/20 starts again at line 458.
Delete those lines dated 16/11/20 through 30/11/20.

The diagnostic files show lots of problems connecting to the VP2 over the network. Do you have a Davis IP logger or a third party device?

I also noted that the console clock is not set correctly which is causing the whole logger contents to be downloaded at every start-up, and you are missing data at start-up because of the time mismatch.

[Tigro11]

yes on the davis vp2 console I mount the wifilogger2 module.
If possible, can you post me the configuration to be entered for a davis with ip address?
Thanks
Valerio

[mcrossley]

I find wifiloggers to be unreliable to be honest - I have puts lots of extra checks and retries in the code to try and cope with them but there is only so much I'm prepared to do.

You just put in the IP address in the config the same as for a Davis IP logger.

[PaulMy]

Hi Mark,
Sorry for going off-topic, and am not using the b3097 beta with WiFiLogger but just want to mention that I've been using/testing the WiFiLogger with CumulusMX for over 3 years with the builds before 3044 very unreliable. But since then Mark has done a lot of work on it and has become very stable for me (it does have some issues with the disconnect when also using all the WiFiLogger uploads). Wow, over 50 CMX builds! Thanks Mark. WiFiLogger also has gone through a number of updates during that time which I think has helped as well.

My b3097 beta testing is with Weatherlink Live.

Enjoy,
Paul