bad sensor = connection refused? ** solved **

[crm0922]

Ok, my outdoor temp sensor died and now the sensor page says:

error connect socket 127.0.0.1:5555 :Connection refused

Sensors cannot be displayed because data logging process has been halted.

The data logging process has not been halted. Periodically, when the temp sensor comes back to life for some reason, everything goes back to normal and the sensor page shows up fine, etc.

How can I just tell meteohub to ignore the bad sensor or to use the indoor sensor until I can install the replacement outdoor sensor?

Chris

[sevenless]

I'd first try unplugging the TH sensor cable (the 'phone jack' style plug) inside the transmitter box, restart Meteohub and see if that removes the sensor from the list and resumes data logging.

If that doesn't work then try going to the Weather Station page and re-select the VP2 and hit save -- that might trick the Meteohub into re-polling the sensors to see what is available.

[crm0922]

I might have forgotten to mention that the station is 2400 miles away. However, I did some diagnostics last time I was there, and it is definitely a bad sensor. I just didn't have any way of getting a new one delivered before I had to leave.

I think I have tried reassigning the station type, but I'll try again tomorrow morning.

But is there something inside meteohub, like editing a file somewhere that can tell meteohub to use the indoor sensor instead and forget about the outdoor sensor?

The Davis VP2 base station is fine, it just shows dashes where the outdoor temp/hum is supposed to be. And WeatherLink downloads the information fine, it just marks the outdoor sensors grayed out as if it is unplugged.

Chris

[admin]

weather network uploads are halted when th0 is not there. Does that explain the behavior you see?

[crm0922]

Yes. So how can I bypass the bad sensor with the indoor sensor until I can get out there to replace the outdoor sensor?

Can't we tell it that the outdoor th0 is a duplicate of thb0 or whatever? Or just ignore the sensor somehow?

I still want my rain and wind data if possible. It's going to be 3 months until I can get out there to repair the temp sensor...

Chris

[admin]

We could try to change the file that checks for th0 being there, before uploading data to the networks.
Do you have ssh access to the unit? (port 22?)

to replace th0 by thb0 won't be easy. omitting th0 could work as described above. But I don't know what the weather networks will do with data records without outdoor temp.

[crm0922]

I have ssh access and I'm quite comfortable with anything unix, so lead the way, and I'll follow...

Omitting the check seems fine to me. Will that allow the sensor/dashboard/etc. pages to stop displaying the "Connection Refused" error?

Chris

[admin]

ok,
1) download this file: "http://www.meteohub.de/meteonet"
2) copy downloaded file to your meteohub's network share "/puplic/transfer".
2) log in via ssh, user root, password meteohub
3) stop meteonet daemon by "rcmeteonetd stop"
4) copy downloaded file by "cp /data/transfer/meteonet /home/meteohub/meteonet"
5) start meteonet daemon: "rcmeteonetd start"

Now weather network updates should work, even if th0 is not there.

[crm0922]

Great! Will all the other broken pages work now as well?

[crm0922]

Ok, I put in the patched file. The sensor page still says:

error connect socket 127.0.0.1:5555 :Connection refused


Sensors cannot be displayed because data logging process has been halted.

I'm not sure how this is ever going to work if the sensor page, dashboard, etc. are all throwing this error.

Can we stop the data collection from caring so much about a bad sensor? I looked in the raw data and nothing is getting collected when the sensor craps out.

I should also point out that the weather uploads still don't work...

Thanks for your help with this.

Chris

[admin]

Your problem looks like the data logger gets restarted again and again. This happens when it is going to be stopped by some reason.

What does "meteohub log" say?

Any chance for me to login to your meteohub on ports 22 (ssh) and 7777 (http)?

[crm0922]

The log looks like:

logger (05.10.2008 15:38:24): USB chip set: Vendor ID 10c4:ea60
logger (05.10.2008 15:38:25): rain collector size 0.01 inch per tick
logger (05.10.2008 15:38:25): transmitter 1 active as station type (#0) "ISS".
logger (05.10.2008 15:38:54): started Vantage (5).
logger (05.10.2008 15:38:54): USB chip set: Vendor ID 10c4:ea60
logger (05.10.2008 15:38:58): rain collector size 0.01 inch per tick
logger (05.10.2008 15:38:58): transmitter 1 active as station type (#0) "ISS".
logger (05.10.2008 15:39:24): started Vantage (5).
logger (05.10.2008 15:39:24): USB chip set: Vendor ID 10c4:ea60
logger (05.10.2008 15:39:25): rain collector size 0.01 inch per tick
logger (05.10.2008 15:39:25): transmitter 1 active as station type (#0) "ISS".
logger (05.10.2008 15:39:54): started Vantage (5).
logger (05.10.2008 15:39:54): USB chip set: Vendor ID 10c4:ea60
logger (05.10.2008 15:39:55): rain collector size 0.01 inch per tick
logger (05.10.2008 15:39:55): transmitter 1 active as station type (#0) "ISS".
logger (05.10.2008 15:40:24): started Vantage (5).
logger (05.10.2008 15:40:24): USB chip set: Vendor ID 10c4:ea60
logger (05.10.2008 15:40:25): rain collector size 0.01 inch per tick
logger (05.10.2008 15:40:25): transmitter 1 active as station type (#0) "ISS".
logger (05.10.2008 15:40:54): started Vantage (5).
logger (05.10.2008 15:40:54): USB chip set: Vendor ID 10c4:ea60
logger (05.10.2008 15:40:55): rain collector size 0.01 inch per tick
logger (05.10.2008 15:40:55): transmitter 1 active as station type (#0) "ISS".

Over and over and over....just like you suggested would be the case...

My station is on a satellite link that doesn't accept incoming connections, so I set up a WRT54G creating ssh tunnels to my PC at home (2400 mi away). So I can log in to the distant meteohub by SSH'ing to a port on my machine at home.

I could re-forward a port from my home connection outside, to the inside PC that hosts the tunnel for you, but it will be unbelievably slow. But it is useable.

How can I contact you with the info?

Chris

[admin]

info(at)meteohub.de

[admin]

when logging in via ssh, please try.

1) rcwmr928d stop (to stop data logging)
2) /home/meteohub/wmr928d -d -v -c /home/meteohub/wmr928d.conf
Please log what (2)outputs and send it to me.

3) rcwmr928d start (to restart data logging)

my guess is that you see the wmr928d child process stopping by a "segmentation violation" or so. If this is the case I would need access to drill down.

Please help me. What is special with your station? Just a missing th0 sensor? or anything else?

Another question: Is there any data logged at all? Does the "data inspection" page show any logged data for the last hours or days?

[crm0922]

Nothing is in the Data Inspection pages since the last and final failure of the temp sensor th0.

Sensor th0 was intermittent for a while, usually it would work from 10:30am to 7:30pm, I assumed this was because the ambient temperature was warmer and somehow bringing the sensor back to life.

A couple days ago it just never came back online at 10:30am.

However, when th0 starting working, all the meteohub pages acted normally, then went back to this Connection Failed stuff at night when the sensor was failing.

See the attached file from the output of the procedure you asked me to follow.

Only thb is reporting in this log, is that normal? I am almost 100% sure the ISS is still reporting rain and wind data to the console.

Thanks again,

Chris [file name=output.txt size=14580]http://www.meteohub.de/joomla/images/fb ... output.txt[/file]