meteohub.de

Well, I rebooted meteobridge, the log errors disappeared and the services Test button works now : My God ! Is there a Windows subsystem in there ?

1) After logging as root into Meteobridge / OpenWrt, doing a wget HTTP... and a wget HTTPS... it appears that the network stack does NOT support HTTPS, as compiled.
Is this specific to the MR3020 ? If yes, I can't remember having read that in the docs. Would it be different in a Raspberry Pi ...

Just bought a license for a TP-Link MR3020 latest version (no USB hub).
The MR3020 talks to a Davis Vantage Vue Prro2 and is in a LAN behind a 4G router. It sends requests every minute to populate a log file on a VPS I control.
I don't believe it might have any relation (Meteobridge certainly does ...

My meteobridge is on a Tp-Link 3020 behind a 4G router, and I access it from outside with an autossh session. This has been working well for a few months.
Recently I had to swith off and restart the router and the tp-link server. Now when I restart them that's what i get :
" Meteobridge Landing Page ...

Software version is Meteobridge 3.2, may 29 2017

I access OpenWRT on my TP-Link TL-MR3020 through Network/Expert Mode, to add 2 lines on crontab. They are saved fine. But on reboot the contents of crontab go back to the original version.
Any way to prevent that ? Would it be the same if I access OpenWRT by SSH ?

I assume we are speaking about the humidity sensor in the outdoor Integrated Sensor Suite.
If it's "not connected correctly" and throws a repeated boot error, how is it possible that the reported values are in a reasonable range (reported between 80% and 95% today) ?

TL-MR3020, no USB hub (I know it should have one), with Davis Vantage Pro2.

I get this message at boot :
logger (01.02.2018 07:23:50): transmitter 1 active as station type (#0) "ISS".
logger (01.02.2018 07:23:50): transmitter 2 active as station type (#7) "Soil".
logger (01.02.2018 07:23:50 ...

If I understand your question, in the simplest case, you tell your router to forward port 80 to the local IP address of your Meteobridge, something like " port 80 -> 192.168.1.120 port 80 ".
So you will access your Meteobridge with http://your-public-ip/.

If there are several HTTP servers behind ...

I should have remembered that one, of course it works, thanks for the reminder.

Not sure why, but if I send an HTTP request with the identifiers, it won't work : I have to use HTTPS for the URL to work.

I setup another domain without SSL nor authentication, so that I could get the meteobridge data.
I'd be curious to have an answer to my previous question.

Just a remark : when one modifies an HTTP event (I changed the domain name in the request) and when one pushes the [TEST] button, the previous ...

I have secured my personal domain and server with a combination of HTTPS and the basic Apache module mod_auth_basic, which means that to access the server, one has to provide the pre-defined identifiers through a dialog box sent by Apache.
Is there a mechanism to send the username and password with ...

Now all data are displayed, maybe because my station has been sending data for 24 hours.
Just a small glitch:
data is sent to wunderground.com, and on the dashboard the " Current Conditions " are Ok, the " graphs " and " table " values for the day are Ok, but the " Summary " values (max, min, avg ...

A new MR3020 did solve the problem, which means the previous router is dead.
Just a small glitch:
data is sent to wunderground.com, and on the dashboard the " Current Conditions " are Ok, the " graphs " and " table " values for the day are Ok, but the " Summary " values (max, min, avg) are empty ...

Since I can -sometimes- ping it, and it -sometimes- answer the SSH attempts, the network is properly configured.
And I don't see how it could be denied internet access, nothing having changed in the router.

What is suspect is that it works a little bit "sometimes" ! Could this MR3020 be defective ...