ERROR: Request failed! Error code ENETUNREACH.

[RPJacobs]

After a couple of hours I get this error message.... devices are unavailable.

ERROR: Request failed! Error code ENETUNREACH. Check the URL in config.json and ensure that the URL can be reached from this system!

After a restart all is fine.

[jmichas]

Hi, are you using my version of the plugin? This was an error that happened in the old code that I had forked from, it was fixed.

[RPJacobs]

Yes I use your plugin. Equally I used the other plugin before and had a username password error. You fixed that one with the cookie.

[jmichas]

I will look through the code to see if I can figure out what that error code really means. Sorry, it is sometimes tough because I am not intimately familiar with existing code.
Is there a time frame this starts to occur? Like is it hours after a restart, days, weeks? Or just random?

[jmichas]

Actually, also, what version of homebridge and node are you running?
This has only been tested on HB 1.7 and node 18.16, so if youa re running a more recent version that may be the issue.

[RPJacobs]

It's a fresh install pi4B.

[jmichas]

Yeah, that's what I thought. I don't currently have an environment to test against the latest node/HB.
If a restart fixes the problem a workaround for now could be to just set your device to reboot in the middle of the night using cron. It is pretty simple to setup, you can google how to use cron. I used to have my pi reboot at 3am every night to fix the old problem before we fixed it properly in this repo. As soon as I can I will look into setting up some sort of VM using latest node to code and test against.

[RPJacobs]

Moved to a docker image with node.js 18 to test:

The (almost the same) error is still there but only for a minute and then it recovers..

[RPJacobs]

don't think it's a session error locking at the session log in z-way-server. The last session was created a day before the error....