ch2i / loragw-setup Goto Github PK

Running setup_config.py fails as router is None.

I fixed by changing line 96 from:

router = router.hostname

to

router = router.path

I have installed my gateway according to instructions. But when I called sudo journalctl -f -u loragw.service it only displayed

Jun 14 17:08:51 loragw-9883 loragw[613]: 17:08:51  INFO: Packet logger is disabled
Jun 14 17:08:51 loragw-9883 loragw[613]: 17:08:51  INFO: Flush output after statistics is disabled
Jun 14 17:08:51 loragw-9883 loragw[613]: 17:08:51  INFO: Flush after each line of output is disabled
Jun 14 17:08:51 loragw-9883 loragw[613]: 17:08:51  INFO: Watchdog is disabled
Jun 14 17:08:51 loragw-9883 loragw[613]: 17:08:51  INFO: Contact email configured to ""
Jun 14 17:08:51 loragw-9883 loragw[613]: 17:08:51  INFO: Description configured to ""
Jun 14 17:08:51 loragw-9883 loragw[613]: 17:08:51  INFO: [Transports] Initializing protocol for 2 servers
Jun 14 17:08:51 loragw-9883 loragw[613]: 17:08:51  ERROR: [TTN] Connection to server "" failed, retry in 30 seconds
Jun 14 17:09:35 loragw-9883 loragw[613]: 17:09:35  ERROR: [TTN] Connection to server "" failed, retry in 60 seconds
Jun 14 17:10:35 loragw-9883 loragw[613]: 17:10:35  ERROR: [TTN] Connection to server "" failed, retry in 120 seconds

The server address in /opt/loragw/global_conf.json was empty. "server_address": "",
I had to change it to "server_address": "router.eu.thethings.network",

shouldn't this work out of the box?

Hi,
I've been happily using your designs and solution for the last years, on and off, depending on my (mini) project focus.

Now, with the upcoming disabling of the The Things stack v2, I would like to upgrade the software on the Raspberry towards The Things Community Edition and also change Things Stack version.

I saw the steps in the link below, those are probably doable without changing the software. https://www.thethingsnetwork.org/docs/the-things-stack/migrate-to-v3/migrate-gateways/

But I suspect that fully migrating to the TTN Stack V3 will require code changes.
How would we go about to extend the life of your (still perfect working) gateway?

Thanks

I did a fresh install of repo https://github.com/ttn-zh/ic880a-gateway on my RPi with your ch2i backplane and the GW ist working fine with TTS V3. Unfortunately my two sensors BME280 and SI7021 aren't working anymore and all 4 LEDs are without function. Is there a chance to get it back to the state before the fresh install? What have I to do for that? Your repo for the old backplane ist deprecated and is not installable.

Hi,
1st of all, thanks for the great work you've done (and are still doing), the main function, working as a TTN Lorawan gateway, is working perfectly. I have the Pi0W + RAK831 board, PCB version 1.3 (small PCB). There is one issue though, the monitor doesn't properly start because the WS281X lib is not properly compiled. The file is in the expected location but is 0 bytes. A bit too short :) Below is the log for the monitor service.

What could I check/do to make it work? Compile something again differently/by itself?

Thanks,
Geert

Oct 03 18:48:42 loragw-ac6d systemd[1]: Started LoraGW monitoring service.
Oct 03 18:48:46 loragw-ac6d monitor[1438]: Traceback (most recent call last):
Oct 03 18:48:46 loragw-ac6d monitor[1438]:   File "/opt/loragw/monitor.py", line 25, in <module>
Oct 03 18:48:46 loragw-ac6d monitor[1438]:     from neopixel import *
Oct 03 18:48:46 loragw-ac6d monitor[1438]:   File "build/bdist.linux-armv6l/egg/neopixel.py", line 5, in <module>
Oct 03 18:48:46 loragw-ac6d monitor[1438]:   File "build/bdist.linux-armv6l/egg/_rpi_ws281x.py", line 7, in <module>
Oct 03 18:48:46 loragw-ac6d monitor[1438]:   File "build/bdist.linux-armv6l/egg/_rpi_ws281x.py", line 6, in __bootstrap__
Oct 03 18:48:46 loragw-ac6d monitor[1438]: ImportError: /root/.cache/Python-Eggs/rpi_ws281x-1.0.0-py2.7-linux-armv6l.egg-tmp/_rpi_ws281x.so: file too short
Oct 03 18:48:46 loragw-ac6d systemd[1]: monitor.service: Main process exited, code=exited, status=1/FAILURE
Oct 03 18:48:46 loragw-ac6d systemd[1]: monitor.service: Unit entered failed state.
Oct 03 18:48:47 loragw-ac6d systemd[1]: monitor.service: Failed with result 'exit-code'.
Oct 03 18:48:52 loragw-ac6d systemd[1]: monitor.service: Service hold-off time over, scheduling restart.
Oct 03 18:48:52 loragw-ac6d systemd[1]: Stopped LoraGW monitoring service.

Hello,
I have lora gw with ch2i V1.2 board, ic880A and a RPI 3B. During running 2_setup.sh there is question, which HW combination is used for the gateway (see line 91 of 2_setup.sh). I choose "CH2i ic880a". This will select GPIO5 as reset pin for ic880A board. But this is wrong. It has to be GPIO17 for this combination. Please can you correct script. I do not know, if the selection for the other HW combinations is correct.

It would be great if this could be used with resin.io much like many other configurations: https://github.com/kersing/ttn-resin-gateway-rpi

Hello,
in Python script monitor_gpio.py line 138 is commented out. With this selection the function to shutdown GW with the push-button does not work in my case. If I remove the comment character # it works. Please can you check.

Hello,
Any special thing needs to be done if I wan't to connect OLED in shield 1.5 ?
I have connected OLED to I2C-3 ... it smoked when I plug the RPi in :D ( 5V via USB power )

Thanks for the awesome setup procedure. I've successfully used it with a RAK831 board and RAK's official shield (converter board) for the Raspberry Pi 3. However the GPS module soldered onto the shield was not configured.

With manual work, I was able to get the GPS module running. The relevant steps are:

Serial port

On the Raspberry Pi 3, the serial port /dev/ttyAMA0 is used for the bluetooth module. The GPS module is found on /dev/ttyS0. The path /dev/serial0 is linked to that device as well. The Raspbian distribution configures it as a login shell, which is in conflict with the use for the GPS module. So it must be changed, either by removing console=serial0,115200 from /boot/cmdline.txt or by using raspbi-config:

• Interfacing Options
• Serial
• Would you like a login shell to be accessible over serial? No
• Would you like the serial port hardware to be enabled? Yes

Permissions for serial port

In order to access the serial port (without being root), the user must belong to the group dialout. The user pi belongs to it, the user loragw doesn't. So:

sudo usermod -a -G dialout loragw

Discovery

Even after these changes, the setup script did not properly discover the GPS module. It said:

Has hardware GPS:	False
Hardware GPS port:	/dev/ttyAMA0
Using fake GPS

I have no idea how the discovery works. Is the GPS port a default value? Anyhow, I've modified the global_conf.json manually:

        "fake_gps": false, 
        "gps": true,
	"gps_tty_path": "/dev/ttyS0"

After a reboot, the GPS module is now working. It would be cool if no manual work was needed and the setup scripts would automatically detect and configure the GPS module.

I'm trying to follow this deployment guide but I'm stuck a little, more confused though.

From what I gathered this deployment script does not count with using own Loara Server and does not install lora gateway bridge on the rpi, which would be more secure to let it talk to mqtt server directly ( with ssl and pass ).

Own installed loraserver will not give you: serv_gw_id and serv_gw_key

Hi Charles,
I wanted to give my gateway to a colleague and try to install the newest software. Unfortunately, the script does no more work and creates many errors (e.g. node-js but also some python errors.)

Would it be possible to adapt it? Or at least give a workaround that it still can be used. I like this small gateway...

Regards
Andreas

I promise this is the last issue I have :D But what should be on port 1688 ? The script did not put anything there. Its not open and there is no traffic on that port either....OLED therefore does not show LoraWan Data Status... any idea ?

Hey Charles!
I have used your awesome script up boot up my ttn gateway.. it's working well.
I have kept it On for around a week now.
For testing purposes, I've kept nodes transmitting at intervals of 20 secs.
The problem is that after a continuous transmission of a lot of packages for an hour or so, the data transmission stops abruptly. I can't see node data on the Gateway traffic on ttn, neither on my node-red setup which i have installed. TTN problem is because of the fair usage policy, but why a problem in my private network?
Restarting solves this problem for the moment. What could be the issue? Is it some memory problems?
Btw, am using the loraserver.io also with your gateway setup.

Hi, I was trying to read out both SNR and RSSI values to display on the OLED display from oled.py script and it was unsuccessful. I did add "lora_snr" variable in both global and UDPhandler method.

So I am wondering if I want to get the SNR value, is it to do with oled.py or other files?

Thanks!