Node-red not available on LAN when internet is down

Hi all,

Having a lot of internet outages and just wondering if there is an explanation for this phenomenon. My node-red instance is on my Cerbo within the OS Large package and the Cerbo is connected by ethernet to my router as is my monitor screen and a few other things. I would have thought that the Cerbo (so Node-red as well) and the screen would always be in communication but node -red is “stopped in safe mode” and the Remote console accessed from the screen can’t stay connected for 2 minutes.

Any ideas what I can do to access Node-red during these outages ?

By screen, you mean a PC on a LAN? Are you accessing by IP address or by name?

Do you use both interfaces on the Cerbo? Make sure the active one and no other is set to “internet access” in the network settings.
There should be no reason why you can’t access it independently of internet access.
Does trying via it’s IP work or are you using the mdns name?

You would need to dig into the node red logs and see why it is defaulting to safe mode.
Do you have any special nodes loaded?
Any flows that rely on internet access?

Are you sure the cerbo didn’t reboot?

Yes screen is a linux pc on LAN, remote console does connect but disconnects after a very short time. It is accessed by IP. Node-red is accessed by venus.local. It is visible but says “stopped in safe mode”, no inject nodes respond. i don’t know what both interfaces on cerbo are.

I suppose there are plenty of nodes that need internet, timers etc

Wifi and wired ethernet.

Do a test, disable all your flows and then disconnect your internet.
See if it stays up.

Disable all flows how ? Disconnect the ethernet cable ?

Click on the flow and choose the “disable” option. Then deploy.

Looks like the flows are working now, I had refreshed the page a dozen times but it seems it wanted me to do any edit just so that I could redeploy. However, remote console still goes disconnected after a very short time. It will reconnect but can’t get it to stay open. In settings ethernet is active, wifi not.

What I suggest. Enable ssh to the cerbo.
Go to /var/log
There you will see folders and files for nodered, you can check the current log to see what it is doing when this happens.
You can do the same with the venus log.

This will help narrow down what the issue might be.

If you disable nodered completely (doesn’t affect config) does the remote console behave normally on internet loss?

I have diasbled Node-red so we shall see what happens. Also I noticed in the the settings that node-red was “enabled safe mode” plus i keep getting “bms lost connection” errors.

Maybe consider reflashing your GX and starting clean. Export your flows and import them afterwards again (after loading manually added nodes from palette manager).

Eek ! :scream:

Flows disabled, still disconnects remote console. Maybe I’ll have to reflash…

What results do you get if you ping your Cerbo’s IP? 5-10 responses?

Where should I ping from ? It’s a bit of a pain at the moment with limited options, still no wifi

Any device on the local lan.
The gx also has an internal wifi ap, so you can join to that to get local console access (no internet though)

4 sent, 4 responses, 0 bytes lost, 0ms average time

So not an IP issue then.

I have accessed the node-red-venus log files but cannot get them copied onto the server so I can show you them. Maybe you would see a clue.

Zip them and attach the file to a post.

SSHing I have tried scp -r root@IP_ADDRESS:/var/log/node-red-venus ~/Documents hoping it would copy the log folder to the local laptop/Documents but..nah