RPi CPU Temperature monitoring - simple but informative

I've got plenty of PiZero (most are $5 originals with a WiFi dongle hacked onto them)

I do use a couple of Wemos D1 minis as well to just receive MQTT generated by the Pi computers

I generally use the Pi for when I want a bit more than just switching stuff on or off

I find it MUCH easier to maintain, understand and change Node-RED code than Arduino C++ stuff :slight_smile:

4 Likes

That's a good point about getting power to the arduino's - although you'll have the same problem with the RPi's.

In a couple instances, I've had to pull wire and install another outlet (i.e. for a sensor in the attic, water monitor in the crawspace, or temp/humidity sensor over my porch), but I understand that's inconvenient or dangerous for DIY...

Wow, that's a great deal. I see wifi zero's are $15 now but still that beats a nano with wifi shield. Having said that, I still love a chance to do some C coding - something about programming a microcontroller in C is awesome fun... I used to be a PIC guy till arduino spoiled me :slight_smile:

@edje11 : Is your RPi monitoring system available?

Well, after thinking about it... I have decided to retire as much of my python scripting as possible over time, as I move things into node-red. Ignore the grown man crying behind the curtain... All the python I have was a real time investment... oh well. LOL.

1 Like

I've just this minute (OK about 15mins) installed Node Red onto my Pi zero W. Just got to think of some uses for it now. Probably ideal for experimenting away from my production version.

I use a pizerow as my mqtt broker and in no I have it checking on my Wemos’s telling me who’s alive and what version of espeasy is installed on them

1 Like

I found one odd ball issue, I put a image built on a Pi4 on my PiZero, node-red fails to start. The issue is nodejs for ARM v7 can't handle ARM v6 based Pi devices. I had to uninstall nodejs manually and then let the node-red install script reinstall nodejs to fix the issue. Unfortunately, I lost all my flows in the process, not sure why that happened, but it did. Maybe the way I did the sequence of steps caused the loss of the flows.

The answer is Node-Red Starts on RPi3+ but not on RPi Zero W

That will not have deleted the flows file, but it might be that node-red cannot find it for some reason. If the flows were important then immediately make a copy of your .node-red folder (so you do not accidentally delete anything) and show us what is in the .node-red folder by running
ls -l ~/.node-red
and show us the startup log you see from running
node-red-stop && node-red-start
Copy/paste please rather than a screenshot if possible.

by default they use flows_{hostname}.json - so if the pi0 has a different name it will start with a different file... the old one may indeed still be there and can be copied/renamed etc.

@dceejay
That is likely the issue, I do change the host names, via DHCP provided name, when a new image boots for the first time.

Binding an installation, in this case create flows, to a static host name seems odd to me. I have a long history with engineering enterprise deployment of operating system images, to virtual machines, containers, etc. So when solutions are bound to something that can change like in this case, we encouraged solution provides to avoid it.

Is this something that can be changed in the future?

Not sure if it will be changed but you can handle this yourself by providing a flow file name on the command line that starts node-red.

I think it may also be possible in settings.js (I forget)

Yes, the name of the flow file can be configured in settings.js, it is documented in the file, or you can set it in the command line.

1 Like

It gets lots easier if you enable projects because you will name the flow when starting a project.

Settings.js file will be the preferred method, since I am going to use ansible to install node-red and validate nodejs, etc.