Raspbian lite - pigpiod not accessible from node-red container

[craigh1015]

I installed pigpiod and started it up. By default it starts locally and the node-red container cannot see port 8888 of the PI. If I change the startup to global (-g) pigpiod fails to start.

[gcgarner]

Which version of rasbian are you using? If you installed it then i assume you are using rasbian lite

https://github.com/gcgarner/IOTstack/wiki/Node-RED#gpio Did the method described here not work for you?

[gcgarner]

I just started a brand new Raspian Lite, installed the pigpio as described, added the rc.local (rebooted) and mine works. My IP is 192.168.88.81

were you running it from the CLI?

[craigh1015]

I am running 2019-07-10 rasbian buster lite.

I was originally trying to run pigpiod through systemctl - I have disabled that now.

I made the change to /etc/rc.local after reboot I see pigpiod is running (vi ps) however it crashes.

I have now been successful running sudo pigpiod &

I am not sure why this is able to keep running - it will not survive a restart - but at least I can now proceed

[gcgarner]

Roughly how long does it take before it starts to crash?
Is it after you take any specific action? I would like to see if i can reproduce your results
after running sudo pigpiod & in the background does it bomb out at all?
for me running sudo systemctl status yeilds:

do you maybe have duplicates sudo systemctl status | grep "gpio" other that the grep

[craigh1015]

Using the default systemctl config pigpiod gets started with "-l" flag.
NodeRed then gives an "UNREACHABLE" error.

When I edit the installed pigpiod.service file or create a new one so I can start with "-g" flag then
systemctl start pigpiod fails to complete

I have now run into a new issue with the pi3 reading a DS18B20 temp sensor over 1wire so am going to start everything from scratch later this week or over the weekend.

[craigh1015]

I have been able to get my DS18B20's connected to node-red and did some further testing with the pigpio.

I have /usr/bin/pigpiod in /etc/rc.local after a reboot everything work fine initially then after a couple of minutes pigpio crashes and I get the following error in /dev/pigerr
2019-11-09 07:03:59 sigHandler: Unhandled signal 18, terminating

I do not appear to get this error if I run sudo pigpiod & after boot

[gcgarner]

I'm going to see if i can replicate on my spare 3B, on rasbian lite. Did you get the ds18b20 working in your docker version of nodered, if so which node did you use?

[gcgarner]

Test a 18b20 with node-red-contrib-ds18b20-sensor and it works fine on my 3b running docker. now to test the lite with your gpio issues

[craigh1015]

I am using node-red-contrib-ds18b20-sensor v1.3.6 for temp sensor. It is happily reading 3 sensors.

pigpio started from command line sudo pigpiod & appears to continue to be running fine.

[gcgarner]

Last night i setup a new Raspbian lite image on a 3B. Set it up and followed the rc.local method for getting the gpio working. So far I've had 10 hours of uptime on the GPIO without a crash.

I'm going to see what happens if I mess around with it's service so see if I can break and repair the pigpiod

[craigh1015]

My architecture has changed and I am now running my temp sensors on some remote ESP8266 boards.

I no longer have any pin usage on the pi.

Closing the issue as I am no longer effected by it.